摘要:

NOVEMBER, 1963 Vol. 88, No. 1052 THE ANALYST EDITORIAL Further Development of The Analyst SIX months ago the Society introduced a revised “Notice to Authors” and, simultaneously, a modified system for subjecting papers to the scrutiny of referees. At the same time we mentioned in an Editorial some alterations in the appearance of The Analyst that had taken place at the beginning of the year. Then, after outlining the changed responsibilities of the Editorial (formerly Publication) Committee, and the formation of the Publications Policy Committee, we said that many proposals of the AnaZyst Development Committee remained to be implemented. These changes (like the change in the paper stock on which the journal is printed) could not be introduced except at the beginning of a volume, but will take place in January, 1964.The most important of these proposals was undoubtedly one that the Proceedings of the Society, i.e., the reports of the activities of the Society, its Sections and its Groups, other “domestic” news pertaining to the Society, obituaries, and certain other items, should be published separately from the scientific matter at present making up most of The Analyst. The Publications Policy Committee has given further consideration to this proposal and has studied possible alternatives and matters of detail. Its report has been accepted by Council, which has decided that from January The Afialyst shall be devoted entirely to matters appropriate to the journal of a learned Society. It will contain Special Lectures, Review Papers, Original Papers (including Short Papers), Book Reviews, Communications and, when required, Editorials.Communications are a new feature being introduced in response to requests for a medium for the prompt publication of urgent material. It is not intended for simple claims to priority, but rather for the brief description of work that has made some progress and then has stopped for lack of time, or of resources, or even of inspiration. Publication has a twofold purpose: the ideas, although not fully worked out, may yet be valuable to workers on other problems, or other workers may be able to provide the necessary inspiration or resources to permit the completion of the original piece of research. Manuscripts must obviously be on urgent matters of some scientific importance (for time does not permit examination by referees), and must not exceed 300 words.They cannot include diagrams, although there will be no objection to formulae or tabular matter. Timing also precludes any question of appeal against the Editor’s refusal to include a particular Communication, although a manuscript so refused can always be submitted to the Editorial Committee for its decision after the usual examination by referees. In return The AnaZyst offers publication in a minimum time from receipt of 4 to 5 weeks, which is the length of the final stage of the production schedule. To this it may be necessary to add 30 days should a script arrive 1 day too late for a particular issue, making the maximum time 9 weeks. Such rapid publication can only be achieved for a small proportion of the material in each issue.The Proceedings of the Society for AnaZytical Chemistry will bring the rest of the traditional contents of The Analyst to members and non-members alike, and will also incorporate notices 823 Certain conditions must be fulfilled.824 PROCEEDINGS [Analyst, Vol. 88 of forthcoming meetings and other material that has hitherto appeared in the Bulktin. The Publications Policy Committee foresees the possibility of the Proceedings developing as a vehicle for other material worthy of the widest possible dissemination. Proceedings will therefore be treated from its inception as a fully-fledged journal and will be supplied in due time with a title page and index to each volume to facilitate the binding of this record of the Society’s activities. The cover of The Analyst has been re-designed slightly to lay greater emphasis on the fact that it is the journal of our Society. Also, in response to several requests, space will be found among the advertisements for duplicate copies of summaries of papers in a form in which they can readily be cut out for pasting on index cards, a service already available in other major analytical journals. Two other changes remain to be mentioned.

ISSN:0003-2654    

DOI:10.1039/AN9638800823

出版商:RSC    

年代:1963

数据来源: RSC

摘要:

824 PROCE.EDINGS [Analyst, Vol. 88 PROCEEDINGS OF THE SOCIETY FOR ANALYTICAL CHEMISTRY ORDINARY MEETINGS AN Ordinary Meeting of the Society was held at 6.30 p.m. on Wednesday, October 9th, 1963, in the Main Chemistry Lecture Theatre, Imperial College of Science and Technology, Imperial Institute Road, London, S.W.7. The Chair was taken by the President, Dr. D. C. Garratt, Hon.M.P.S., F.R.I.C. The subject of the meeting was “Thin-layer Chromatography” and the following papers were presented and discussed: “General Aspects,” by L. J. Morris, B.Sc., Ph.D. ; “Specific Separations on Impregnated Thin Layers,” by L. J. Morris, B.Sc., Ph.D. ; “Application of Thin-layer Chromatography to Inorganic Systems,” by K. Burton, B.Sc., D. Lyons, B.Sc., G. Nickless, B.Sc., Ph.D., and F. H.Pollard, D.Sc., Ph.D. ; “Effects of Alkyl and Alkenyl Substitution on the Chromatography of Quinols and Related Compounds,” by D.. McHale, B.Sc., Ph.D., A.R.I.C. ; “Separation of Steroids on Microslides,” by P. Oxley, M.A., B.Sc., A.R.I.C. ; “Thin-layer Chromatography of Lipids,” by B. W. Nichols, MSc., Ph.D. DEATHS WE record with regret the deaths of Leonard Balm f or t h Daryl Robert O’Dea William Stross. MIDLANDS SECTION AN Ordinary Meeting of the Section was held at 7 p.m. on Tuesday, October 15th, 1963, at the Nottingham and District Technical College, Burton Street, Nottingham. The Chair was taken by the Chairman of the Section, Mr. W. H. Stephenson, F.P.S., D.B.A., F.R.I.C. The following paper was presented and discussed : “Sugar in Foodstuffs-Some Newer Methods,” by J. L. Buchan, M.Sc., A.R.I.C. MICROCHEMISTRY GROUP THE forty-first London Discussion Meeting of the Group was held at 6.30 p.m. on Wednesday, October 16th, 1963, at “The Feathers,” Tudor Street, London, E.C.4. The Chair was taken by Mr. T. R. F. W. Fennell, B.A. A discussion on “The Determination of Boron” was opened by H. J. Cluley, M.Sc., Ph.D., F.R.I.C., and M. R. Hayes, A.R.I.C.

ISSN:0003-2654    

DOI:10.1039/AN9638800824

出版商:RSC    

年代:1963

数据来源: RSC

摘要:

November, 19631 GILLARD : CIRCULAR DICHROISM 825 Circular Dichroism A Review* BY R. D. GILLARD (Department of Chemistry, Imperial College of Science and Technology, London, S. W.7) Plane-polarised light may be resolved into Zaevo- and dextvo-circularly polarised light. When it is passed through an optically active medium, the amplitudes of these components are initially equal, but if the refractive indices of the medium are different for each component, there is rotation of the emer- gent light, which is now elliptically polarised. The phenomenon of circular dichroism (the differential absorption of Zaevo- and dextvo-circularly polarised light) is a function of wavelength and the shape of the molecules in the medium. This short review draws attention to a technique that may well prove invaluable in analysis.INTEREST in optical activity has recently revived.l The most commonly used property of optically active molecules in research applications has been the optical rotatory dispersion curve (the variation of specific rotation of plane polarised light with the wavelength of the light). Several valuable reviews of this technique are available, for example, those of Djerassi2 and K l ~ n e , ~ but so far it has not attracted attention as an analytical method, probably because the rotatory dispersion curves of single substances are not easy to interpret, and it is difficult, although not impossible, to draw quantitative conclusions from the rotatory dispersion curves of mixtures. Plane polarised light may be resolved into two circularly polarised components with equal amplitudes.Optical rotatory power is attributed to a difference in refractive indices of the optically active material for Zaevo (left) and dextro (right) circularly polarised light. If 91 (the refractive index for left circularly polarised light) is larger than T d (that for right circularly polarised light), the left circularly polarised light is delayed in traversing the medium, giving rise to a rotation of the plane-polarised resultant of the left and right circular components. In passing through absorption frequencies of an optically active molecule, the absorption of the left-handed component differs from that of the right-handed component. With the usual notation of E for molar extinction coefficient and with appropriate suffixes, circular dichroism at any wavelength is given by €1 - Ed, which we call A€.Cotton4 observed this in solutions of the tartrates of transition metals, and the phenomenon became known as the Cotton e f f e ~ t . ~ Since circular dichroism arises through electronic transitions, the spectra produced are similar to ultraviolet and visible spectra, but, for each circular dichroism band, A€, (€1 - Ed), may be either positive or negative, depending on the handedness of the molecule concerned. The requirements for the observation of a circular dichroism band are seen to be an optically active (asymmetric) molecule and an electronic absorption band. Any chromophore will give rise to circular dichroism bands; an example is the carbonyl (> C = 0) absorption (n -+ T*) at 3000 A. The circular dichroism of this band for a 16- ketosteroid is shown in Fig.1. MEASUREMENT The most obvious method of obtaining circular dichroism, A€, is to measure the differ- ence in optical density for Zaevo- and dextro-circularly polarised light ; apparatus has been described to permit this to be done.6 However, the method generally used involves the fact that, after plane polarised light has passed through an optically active medium in a region of electronic absorption, the circular components have been unequally absorbed, so that on recombination of the unequal components, elliptically polarised light is obtained instead of plane, polarised light. As the ellipticity is a function of the difference in absorption of Zaevo- and dextro-circularly polarised light, it is directly related to circular dichroism.The molecular ellipticity is given by- Most measurements of circular dichroism have been based on ellipticity determinations, as noted by Mitchell.' A recent commercial instrument, in which an oscillating crystal tech- niques is used, is also based on this well tried principle. [el = 3300 A€. * Reprints of this paper will be available shortly. For details, please see p. 902.826 GILLARD : CIRCULAR DICHROISM [Analyst, Vol. 88 Wavelength, mp Fig. 1. Circular dichroism for (n -+ T*) band of carbonyl group in a 16-kctostcroid CHEMICAL APPLICATIONS The principles implicit in the previous discussion may be summarised as: (a) Circular dichroism is a function of wavelength, A, with positions of maxima coincident with visible or ultraviolet absorption bands.However, two bands that are not resolved in the ordinary electronic absorption spectrum a.re often separate in circular dichroism. The selection rules for circular dichroism and electronic absorption are often different, which leads to useful differences in band shapes. (b) The sign of a circular dichroism peak depends on the actual shape (“handedness” or “absolute configuration”) of the molecule causing it. So far, circular dichroism has been studied in very few systems. Its uses seem likely to be in the fields detailed below: (i) Analytical studies of mixtures, either from reactions or from natural sources. This will be discussed later. (ii) Relative and absolute configurations of natural products such as steroids, terpenes, proteins and sugars. For example, the absolute configuration of the diterpenoid, cafestol, has been showng to be the opposite to that previously deduced from optical rotatory dis- persion.Several other erroneous configurations based on optical rotatory dispersion have also been corrected9 by studies of circular dichroism. I t has been foundlo that the circular dichroism of the 20-cyano-derivatives of 11-ketosteroids permitted the isomers to be distinguished and formulated, whereas neither the nuclear magnetic reson- ance nor infrared spectra were sufficiently distinctive to be useful. Similar applications are feasible in inorganic complex compounds. Since more circular dichroism bands appeared for the (-) trisoxalatocobaltateII1 anion, (-) [Co(C,O 4) 3]3-, than are expected for an ion with D, symmetry, the presence of some aquated ions of lower (C,) symmetry was postulated,ll with the formula [Co(C,O,) ,(C204H) (OH,)I2-.(iv) Relative and absolute configurations of octahedral co-ordination compounds. It has been deduced from the equivalence of the circular dichroism curves12 for the (-)546.1[C~(EDTA)]- and the (-546.1[Co(+PDTA)]- ions (where EDTA is ethylenedi- aminetetra-acetate and +PDTA is dextrorotatory propylenediaminetetra-acetate), that their relative configurations are the same. The absolute configuration of (--)546.1[C0- (+PDTA)]- is known (Fig. 2 ) , so that the absolute configuration of (-)546.1[C~(EDTA)j- is seen to be the same. (v) Because of the differences in selection rules for circular dichroism compared to electronic spectra (magnetic-dipole-allowed transitions appear strongly in circular dichroism) , distinctions may be made between spectroscopic bands.Confirmation was obtained in this (iii) Determination of structural features in complex organic molecules.November, 19631 GILLARD : CIRCULAR DICHROISM 827 way of the assignment13 of the lowest energy transition (IT, -+ lA,) in octahedral cobalt111 complexes, It was noticed in studies of optical rotatory dispersion that 2-a-iodocholestan- %one gave rise to a strong Cotton effect centred a t about 300mp, whereas no ultraviolet absorption band was observed in this region. Re-examination of the compound by Gillard (unpublished work) and Bose and co-workers14 showed that the ultraviolet band at 298 mp had an Emax value of less than 4, whereas the circular dichroism band had a value for (€1 - €d) of about 1, so that the magnetic dipole character of the transition responsible for the circular dichroism band is beyond doubt.0- \ CH3 Fig. 2. The absolute configuration of ( -)646.1[Co( +PDTA)]- (each curved chelate ring represents - CH,COO) AXALYTICAL USES Wherever optical activity occurs, as in nearly all natural products, circular dichroism is an obvious means of measuring both chemical and optical purity, since circular dichroism is a function of the concentration of the optically active species giving rise to it. For example, in the carbohydrate field, pure methyl-2-oxo-3,4-isopropylidene-/3-~-arabinoside shows a circular dichroism maximum of AE =- 1.8 at 312 nip.However, as in any spectrophoto- metric method, circular dichroism is useful only for chromophores absorbing between about 2000 and 7000 A. Concentrations required are about the same as those used for other spectrophotometric work, and the results are accurate to approximately 1 per cent. The amount of an optically active substance in an otherwise inactive mixture may be determined; applications to such materials as pine oil distillates and pyrethrum products are obvious. The occurence of the circular dichroism of a particular chromophore, such as carbonyl or nitro groups, within a narrow range of wavelengths offers possibilities for a “fingerprint” method of qualitative analysis for functional groups, particularly in synthetic products. So far, the characteristic circular dichroism of only a few functional groups is known, but more examples are being added rapidly.One might expect at least the chromophores listed below to find application : for a-amino-acids, the N-phthaloyl15 and thiourea16 derivatives ; for cyclo-alkenes, the osmate esters1’ ; for a-substituted carboxylic acids, the acylthiourea derivativesls ; and for a-hydroxyacids the dithiocarbonate derivatives.16 Other chromophores that have been used in optical rotatory dispersion include thiones,19 disulphides,20 N-nitroso- amides21 and nitrites.22 presents the circular dichroism curves for several cobalamins. These curves show much more marked differences than the absorption spectra; for example, for cyanocobalamin and aquocobalamin, which have similar absorption spectra, the circular dichroism bands at 500 mp are, respectively, negative and positive, and the signs of several other bands are also different for the two compounds.(References are to optical rotatory dispersion work.) The cobalamin system presents many analytical problems ; an interesting recent828 GILLARD CIRCULAR DICHROISM [Analyst, Vol. 88 CONCLUSION Little analytical work involving the use of circular dichroism has appeared to date, though mixtures of 1 l-ketosteroids and 20-ketosteroids have been analysed by circular di~hroism.~* However, circular dichroism has many potential applications, particularly in the field of natural products, for which the technique may prove useful for handling mixtures now amenable only to complicated treatments.1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. REFERENCES Mason, S. F., Quart. Rev., 1963, 17, 20. Djerassi, C., “Optical Rotatory Dispersion,” McGraw-Hill Book Co. Inc., New York, 1960. Klyne, W., “Stereochemical Correlations,” R.I.C. Monograph No. 4, London, 1962. Cotton, A., Compt. Rend., 1895, 120, 1044. Mitchell, S., “The Cotton Effect,” G. Bell & Sons Ltcl., London, 1933. -, J . Sci. Instrum.., 1957, 34, 89. -, Nature, 1950, 166, 434. Grosjean, M., Lacam, A., and Legrand, M., Bull. SOC. Chim. France, 1959, 1495. Scott, A. I., Sim, G. A., Ferguson, G., Young, D. W., and McCapra, F., J . Amer. Chem. SOC. Bertin, D., and Nedelec, L., Bull. SOC. Chim. France, 1963, 406. McCaffery, A. J., and Mason, S. F., Proc. Chern. SOC., 1962, 388. Gillard, R. D., Nature, 1963, 198, 580. -, J . Chem. SOC., 1963, 2092. Bose, A. K., Manhas, M. S., Cambie, R. C., and Mander, L. N., J . Amer. Chem. SOC., 1962, 84, 3201. Djerassi, C., Lund, E., Bunnenberg, E., and Sheehan, J. C., J . Org. Chern., 1961, 26, 4509. Sjoberg, B., Fredga, A., and Djerassi, C., J . Amer. Chem. SOC., 1959, 81, 5002, Bunnenberg, E., and Djerassi, C., Ibid., 1960, 82, 5953. Djerassi, C., Undheim, K., and Weidler, A. M., Acta Chem. Scavzd., 1962, 16, 1147. Djerassi, C., and Herbst, D., J . Org. Chem., 1961, 26, 4675. Djerassi, C., Fredga, A., and Sjoberg, B., Acta Chem. Scand., 1961, 15, 417. Djerassi, C., Lund, E., Bunnenberg, E., and Sjoberg, B., J . Amer. Chem. SOC., 1961, 83, 2307. Djerassi, C., Harrison, 1. T., Zagneetko, O., and Nussbaum, A. L., J . Org. Chem., 1962, 27, 1173. Legrand, M., and Viennet, R., Bull. SOG. Chim. France, 1962, 1435. Lacam, A,, and Viennet, R., Ibid., 1961, 1974. Received July 9th, 1963 1962, 84, 3197.

ISSN:0003-2654    

DOI:10.1039/AN9638800825

出版商:RSC    

年代:1963

数据来源: RSC

摘要:

November, 19631 CURRY : INFORMATION RETRIEVAL IN THE ANALYTICAL LABORATORY 829 Information Retrieval in the Analytical Laboratory A Review" BY D. R. CURRY (Overseas Geological Surveys (Department of Technical Co-oWation), 64-78, Gray's Inn Road, London, W.C. 1) The potential applications of modern information-retrieval techniques in A summary is given of the simple Attention is drawn to published the analytical laboratory are outlined. manual punched-card systems available. material on the various aspects of the subject. EVERY analytical laboratory deals with information, both as a receiver and as a creator. The process of analysis gives a continuing stream of new information about samples being examined. It is desirable that this information should be stored accessibly so that the effort originally expended in getting the analytical result is not wasted.A t the same time the analytical chemist has, as the starting material for many of his investigations, the ever increasing scientific literature. In both these aspects the techniques of information retrieval may be useful. Information retrieval in this context is taken to mean manual systems as opposed to automatic data-processing techniques requiring the use of electronic computers. Save in exceptional circumstances the use of electromechanical card-sorters is seldom justified. They achieve little that cannot be done manually and may frequently take longer. In this paper the requirements of information retrieval are discussed under three headings- (a) (b) (c) Literature. Laboratory records of work done; Reference data for instrumentai methods ; As the analytical chemist is personally concerned in these problems, it is most strongly urged that study of the techniques available will prove worthwhile.Purely office routines are properly the concern of an 0. and M. Officer, but manipulation of scientific facts should not be delegated to unqualified personnel. THE PROBLEM OF LABORATORY RECORDS- Most analytical laboratories, whether routine or research, have some method of recording samples, in which a costing procedure may also be incorporated. The system used should be able to provide the head of the laboratory with up-to-date information on the work in hand. A t the same time records of past work done should not be left to memory, since the staff may change or retire.It should give all relevant details on previous examinations of particular types of sample; and also, for administrative purposes, show the numbers of these types of sample being analysed (i.e. how the flow of work is varying). INSTRUMENTAL METHODS AND LITERATURE- In many instrumental methods the recorded characteristics of a sample are assessed by comparison with similar records of known materials. At the same time the collection of standard records should be accessible via the various materials or classes of materials. In this way one should be able to find the characteristics that a series of organic homologues may have in common. Many analytical laboratories are only small parts of larger organisations and their libraries may not, either through limitations of staff or knowledge of subject, be able to provide the analytical chemist with an adequate bibliographic service.Analytical information may not be found exclusively in the primary analytical literature or in abstracts, as some relevant material is bound to appear in the general literature of the field in which the analyst is working, e.g. petroleum, metallurgy or semi-conductors. The complexities of cross referencing often dissuade the analytical chemist from maintaining his own conventional card index. At the For details, please see p. 902 * Reprints of this paper will be available shortly.830 CURRY: INFORMATION RETRIEVAL IN THE ANALYTICAL LABORATORY [ArtdySi!, Vol. 88 same time the analytical chemist may partly remember the reference of a particular item of interest (e.g., author or journal) and an index should also answer problems from this approach.In all the instances mentioned above, records and information accessible in the laboratory and under the control of chemical staff are needed. Before considering detailed methods for tackling these problems, a brief outline is given of the various types of manual punched cards. MANUAL PUNCHED CARDS EDGE-PUNCHED CARDS- Edge-punched cards are the simplest of manual punched cards, and their form is familiar to most peop1e.l to 8 The cards have a series of holes along the edges, so that a slot may be formed by punching away the small amount of card between the hole and the edge. When a pack of cards is aligned and a sorting needle inserted through one of the holes, those cards that have been slotted in that particular position will fall away.The holes may be assigned various meanings, and a given set or “field” of holes grouped together to represent letters or figures. It is also possible to have more than one row of holes along part of the edge, to increase the number of selections available. The cards may be sorted by hand as indicated or in some systems by simple frames holding a series of needles, so that all the slots in one edge of a series of cards can be sorted simultaneously. Systems have also been devised for punching plain cards and sorting them on an array of barsg BODY-PUNCHED CARDS- Instead of slots at the edge, however, slots are punched to join pairs of holes in the body of the card.Sorting has to be carried out in a special cradle that permits all the slots to be examined simultaneously. The theory of body-punched cards is the same as that of edge-punch cards. There is an unpunched area on the card for written information.lO FEATURE CARDS- Feature cards are known by various names-coincidence cards, Batten cards or more popularly “Peek-a-boo” cards. The principle on which these operate is the reverse of edge punched item cards and is illustrated in Fig. 1. A feature card is prepared for each feature or concept considered relevant to part of a collection of items (such as abstracts). Each item is assigned a serial number, and every feature card has a number of squares, each corres- ponding to a particular numbered item. When a feature corresponds to a particular item a hole is punched in that item’s square in the relevant feature card.When a series of feature cards is superimposed, any co-incident holes represent items possessing all the features selected. Feature cards are punched for items, whereas edge-punched item cards are slotted for features. COMPARISON OF ABOVE SYSTEMS- For a small collection of items not requiring complex indexing, the edge-punched cards offer economy and convenience. They have an important use, since they also reveal any concurrence of features. For example, if a series of cards is selected from a collection by sorting for holes A, C and D, and all the selected cards are also slotted in position B, this is immediately obvious as another common feature. Feature cards are most valuable when a higher degree of cross referencing of a large number of features is essential.Speed of sorting is greater than with either of the other two systems, provided that the number of items sought is small, as the items themselves have to be in a subsidiary numerical file. The feature-card system is most adaptable to changes of emphasis as new subjects of interest develop. It is usual with edge- or body-punched cards to use custom-printed cards, so that the user does not have to refer to a book for the meanings of the various holes. Although unused holes should be available in the original design of the card their use will be limited, whereas it i s a simple matter to introduce new feature cards at any time.November, 19631 CURRY : INFORMATION RETRIEVAL IN THE ANALYTICAL LABORATORY 831 I I I marble, carbonatlte 5 I S I 5 131 45 55 6S \ 7 J ES 19s 05 I S '25 135 145 55 65 1 75 ES 95 05' 6 16 1 6 / 3 6 46 Y 6 6 / 7 6 % / % M 16 I 6 ) Y I Y % 6 6 / 7 6 1 E l % M 7 I 7 01 37 47 57 b7 I 7 7 87 I 9 7 07 I 7 21 \ 37 147 I 7 67 i 77 i 07 I 9 7 07 Fig.1. Illustration of feature-card system12: ( a ) , (b) and (c) are parts of feature cards from a sample index representing source, type of material and elements determined, respectively. Each hole in a card represents the numbers of samples having that par- ticular feature. ( d ) shows the position when card ( a ) is superimposed on card (c) (i.e., samples 2, 90 and 162 originated from Nyasaland and were analysed for niobium). Simi- larly (e) illustrates card (c) when placed on top of card ( b ) , showing 90 and 158 as common holes.If all the cards are put together the result is shown by (f), only sample 90 being a carbonate rock from Nyasaland analysed for niobium. (Published by permission of the Controller, H.M. Stationery Office)832 CURRY: INFORMATION RETRIEVAL IN THE ANALYTICAL LABORATORY [Analyst, Vol. 88 LABORATORY RECORDS Although there are various references to efficient laboratory organisation, for example see Kent- Jones,ll little literature giving practical details of achieving such control has been published. One the earliest comprehensive systems used an edge-punched master card for each sample analysedl and for a routine laboratory, a simple system has been advocated.2 In the United Kingdom one of the edge-punched cards in use for some years at the Macaulay Institute for Soil Research, Aberdeen, has been discus~ed.~ There is an example of the use of body-punched cards for record keeping (although not specifically chemical) ,lo and recently a machine has been designedg at the Royal Armament Research and Development Establish- ment, Fort Halstead, Sevenoaks, Kent for converting plain record cards for edge-notched sorting.The application of feature cards to laboratory records has also been reported.12 REFERENCE DATA FOR INSTRUMENTAL METHODS Before considering reference data in detail, it should be remembered that the quality of data retrieved can only be as good as the input. All authors, referees and editors of primary literature presenting data should give attention to this matter, which has been discussed by the Director of the Office of Critical Tables.13 X-RAY DIFFRACTION- Chronologically, the first reference data system available on a commercial basis was the “X-Ray Diffraction Data Index” published in 1942 by the American Society for Testing Materials.In this system the three strongest lines in terms of inter-planar spacing of the powder diffraction pattern are used to select a card giving the full details of the pattern and the reference to the source of information. When the series was initiated, plain cards were used and filed numerically, but with the use of edge-punched cards, advocated by matt hew^,^ the three strongest lines and the chemical composition could be indexed. Practical use, however, established that the cumulative index book was most efficient, except for special searches meriting the use of I.B.M.cards. More recently, a proposal by Dr. Matthews to use a feature card index has been adopted.l* Whereas the edge-punched cards were supplied neither punched nor marked for punching, the set of 150 feature cards covering 5,698 sub- stances are supplied punched, and it is stated that “as new sections are added to the data file the feature cards can be returned for the additional data to be entered.” The 10,000 position feature cards should therefore facilitate searching the index for identification purposes. INFRARED SPECTRA- In the United Kingdom there are two common systems for indexing infrared spectra. The Sadtlerl5 system is in book form and has various indexes covering formula, name of compound and also the “Spec-finder” for identification of unknown spectra.In this pro- cedure the strongest bands in twelve sections of an unknown spectrum are listed and compared with a numerical list automatically prepared by computer. A separate section of the Sadtler index deals with commercial products. The Butterworth-D.M.S.lG system was described by Thompson5 and involves the use of edge-punched cards, which are available punched or marked ready for punching. The strongest bands in the spectrum are indexed and also various structural features of the compound. The body of the card also gives bibliographic details and physical properties in addition to the full spectrum. In 1960 a second series of cards was started, in which the same indexing procedures for inorganic spectra was used.Concurrently with the spectral cards, a series of literature abstract cards is supplied; these are similar edge-punched cards indexing author or authors, year of publication and the topics covered (theories, apparatus, etc.). Recently, a feature-card index to the spectral series has been issued (DMS-I-Cards) ; 21 1 of these deal with an extended range of the holes in the original series, and the capacity of each card is 5000 items. A valuable collective index1’ of available spectra has recently been compiled and will be kept up to date by annual supplements.November, 19631 CURRY : INFORMATION RETRIEVAL IN THE ANALYTICAL LABORATORY 833 DIFFERENTIAL THERMAL ANALYSIS- The latest recruit to punched card data indexes is the Scifax DTA data index issued in 1962.18 This is based on a simple edge-punched card, the body of which contains details of the differential thermal analysis record and bibliographic references.One edge is supplied punched with principal peaks, the remaining edges being left free for the user to punch as required, although for minerals it is suggested that the Hey classification be used. LITERATURE SYSTEMS The universal decimal classification for analytical chemistry,lg class 543, may serve its purpose for arranging books, but a single glance should reveal its inadequacy for indexing complex papers in the literature. The problems associated with comprehensive indexing of analytical literature have received more attention than other aspects of information re- trieval previously considered.Three papers on edge punched cards6 ,’ ~8 give various examples of the different ways information can be coded by using numbers and letters. It is possible to index authors’ names, substances, subjects and dates, thus permitting a reasonable amount of cross referencing. Recently in American literature a co-ordinate indexing system has been introduced.*O This is based on the feature-card principle, although for economy the method suggested is the ‘Uniterm’ card system. In my opinion this is a false economy, since the manual effort of entering numbers on cards and recognising correlations far outweighs the additional cost of feature-card equipment if the capacity of the system is likely to exceed 500 items.A feature- card system specifically applied to inorganic analytical literature and having single sided Analytical Abstracts as its foundation has been described. This largely obviates the need for extensive coding of subjects, and makes the index more useful to the analytical chemist in the laboratory.12 CONCLUSION Information retrieval techniques have many applications in the analytical laboratory. It is vital to determine the probable scope required before deciding on the optimum system. Factors to be considered include the type of information available, the information to be retrieved, reasons for requiring the information and the method of seeking it. If a collection of spectra is not to be used to determine the spectrum of compound X, it is unnecessary to include a formula index.When the probable size and nature of material to be included in an index have been determined, the various mechanical means of achieving an efficient form can be considered. Small systems of transient interest and those to which reference will be made infrequently and from a particular aspect are probably most easily derived by using edge-punched cards. Feature cards are adaptable and most efficient when the population of the system is measured in thousands. Anyone considering setting up a system of any size for information retrieval will find study of two books helpfu1.21*22 For those wishing to keep abreast of this rapidly growing subject there is the Journal of Chemical Documentation issued by the Division of Chemical Literature of the American Chemical Society.A comprehensive review of the state of the subject in the United States has been published,23 and, in the general field of information retrieval, A SLIB Proceedings and the series issued by the National Science Foundation “Current Research and Development in Scientific Documentation” are of interest. I thank Overseas Geological Surveys for permission to publish this paper. The upper capacity limit of a manual system depends again on use. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. REFERENCES Hale, A. H., and Stillman, J. W., Anal. Chem., 1952, 24, 143. Naimark, G. M., and Prindle, R. F., Ibid., 1954, 26, 645. Muir, J. W., and Hardie, G. G. M., J . Soil Science, 1962, 13, 249. Matthews, F. W., Ana2. Chem., 1949, 21, 1172. Thompson, H. W., J . Chem. SOC., 1955, 4501.Cox, G. J., Bailey, C. F., and Casey, R. S., Chem. Cr Eng. News, 1945, 23, 1623. Krieger, K. A., J . Chem. Educ., 1949, 26, 163. Breger, I. A., Econ. Geol., 1958, 53, 325. Loneragan, R. I., 0. & M . Bulletin. 1960, 15, 125. Renwick, A., and Flinter, B. H., Overseas Geology and Mineral Resources, 1958,7, 36. 6.834 CURRY: INFORMATION RETRIEVAL IN THE ANALYTICAL LABORATORY [Analyst, Vol. 88 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. Kent-Jones, D. W., Chem. G. Ind., 1962, 1937. Curry, D. R., and Moore, P. J., Overseas Gaology and Mineral Resources, 1963, 9, 61. Waddington, G., “The Confidence Factor-What is fit to Store?” Paper presented a t the Sym- posium on Storage and Retrieval of Analytical Data, ASTM/SAS. Pittsburgh, March, 1963. Mat. Res. & Stand., 1962, 643 and 842. Available from Sadtler Research Laboratories, Philadelphia, U.S.A., or Heyden & Sons, Ltd., 64 Vivian Avenue, London, N.W.4. Available from Butterworth & Co. (Publishers) Ltd., 4-5 Bell Yard, London, W.C.2. “Molecular Formula List of Compounds, Names and References to Published Infrared Spectra,” Available from Cleaver-Hume Press Ltd., 31 Wrights Lane, London, W.8. British Standard 1000 C: 1963. Cushing, R., Chem. Eng., 1963, 73. “Punched Cards, Their Applications to Science and Industry,” Casey, R. S., Perry, J. W., Kent, A , , and Berry, M., Editors, Second Edition, Reinhold Publishing Co., New York, 1958. Scheele, M., “Punch Card Methods in Research and Documentation,” Translated by Holmstrom, J. E., Interscience Publishers Inc., New York and-London. 1962. ‘‘Documentation, Indexing and Retrieval of Scientific InfO~atlOn,” Senate Document 86th Congress, Second Session, No. 113, US. Government Printing Office, Washington, D.C.. 1961. A .S.T.M. Sfiecial Technical Publication No. 331. Received April 22nd, 1963

ISSN:0003-2654    

DOI:10.1039/AN9638800829

出版商:RSC    

年代:1963

数据来源: RSC

摘要:

November, 19631 ANALYTICAL METHODS COMMITTEE 835 Analytical Methods Committee REPORT PREPARED BY THE ADDITIVES I N ANIMAL FEEDING STUFFS SUB-COMMITTEE THE Analytical Methods Committee has received the following report from its Additives in Animal Feeding Stuffs Sub-committee. The Report has been approved by the Analytical Methods Committee and its publication has been authorised by the Council. REPORT In 1958 The Society for Analytical Chemistry received a request from the Scientific Sub-committee of the Standing Advisory Committee, Fertiliser and Feeding Stuffs Act, 1936, to the Ministry of Agriculture, Fisheries and Food for advice on the availability of suitable methods of analysis for the various types of additives commonly included in animal and poultry feeding stuffs. The matter was referred to the Analytical Methods Committee of the Society, which appointed the Additives in Animal Feeding Stuffs Sub-committee to study this problem.The additives involved fall into six principal groups-antibiotics, synthetic hormones, minerals, prophylactics, and water-soluble and oil-soluble vitamins. On the basis of this classification, six panels were set up by the Sub-committee to investigate suitable methods of analysis for the selected groups of additives. The growing practice of adding to animal rations not only vitamins and nutritional supplements, but also drugs and medicaments for the stimulation of growth or control of disease has set new problems in analysis and manufacturing control. A particular feature of this development has been the need to bring together two hitherto unrelated fields of analytical activity-that of the pharmaceutical industry on the one hand and of the animal feeding stuffs industry on the other.Not only has the supplemented feed (as fed to the animal) been considered, but also the higher-potency preparations supplied as diet supplements (pre-mixes) , either for intimate admixture by the feeding-stuffs manufacturers or for individual administration by the user. Although analytical methods exist (published or unpublished) for most, if not all, of the substances per se used as additives, in few instances has it been established that such methods could be applied to the determination of these additives when present in feeds. To a lesser extent, this also applies to the higher-potency supplements.Broadly speaking, there are three factors to be taken into account when applying an analytical method to feeding stuffs; these are- (;) the sensitivity of the method (because the additive in question may be present at low concentration) ; (ii) the interference from the ingredients of the basic feed; (iii) the interference from other additives (in some instances, the identity of these may The representative sampling of feeding stuffs is of considerable importance in the analysis of the products because of the low concentrations of additives usually encountered. This aspect, however, does not strictly come within the purview of the Sub-committee’s work, and so, except in Part 5 of this Report, which deals with the determination of oil-soluble vitamins, no recommendations have been made. In order to assess the precision of any proposed method, each Panel prepared special samples, in addition to using marketed com- pound feeding stuffs, for collaborative investigations, because experience has shown that extreme care is required in the admixture of additives, and that the degree of fineness of milling can be critical in ensuring the production of a uniform mixed sample for analytical purposes. This report is divided into six parts, each part dealing with the work of one of the Panels, together with their recommended methods for determining the additives allocated to the particular panel. not be known to the analyst),

ISSN:0003-2654    

DOI:10.1039/AN9638800835

出版商:RSC    

年代:1963

数据来源: RSC

摘要:

836 ANALYTICAL METHODS COMMITTEE: REPORT OF THE ADDITIVES [Analyst, VOl. 88 PART 1. REPORT OF THE ANTIBIOTICS PANEL The Determination of Penicillin, Chlortetracycline and Oxytetracycline in Diet Supplements and Compound Feeding Stuffs INTRODUCTION THE Antibiotics Panel was set up under the chairmanship of Mr. S. A. Price, and its member- ship was: Mr. A. J. Cavell, Mrs. J. Gammon, Mr. 0. Hughes, Mr. W. P. Jones, Mr. G. Sykes (deputy Miss F. N. Mulholland) and Mr. S. Varsanyi, with Miss A. M. Parry as Secretary; Mr. A. H. Sexton, Mr. J. S. Simpson and Mr. J. H. Taylor also served on the Panel. The Panel was appointed to consider methods for determining antibiotics. Only three antibiotics-penicillin, chlortetracycline and oxytetracycline-are permitted by law to be added to feeding stuffs in Great Britain.For this purpose they are available as diet supplements, with potencies ranging from about 1 to about 250 g of antibiotic per lb, for incorporation in the feeding stuff at the rate of a few pounds per ton, and as other much less concentrated preparations for incoporation at the rate of several hundredweights per ton. Procedures for determining the three antibiotics in both high-potency and low-potency diet supplements, as well as in compound feeding stuffs, have been devised. EXPERIMENTAL AND RESULTS HIGH-POTENCY DIET SUPPLEMENTS- In investigations of the extraction and assay of supplements the Panel was guided by information from the manufacturers of the products. Early collaborative tests confirmed that chemical as well as microbiological methods were applicable to these relatively high- potency products.The chemical methods recommended by the Panel for each of the three anitbiotics in their respective supplements are described in Appendixes I, I1 and 111, and the results obtained by the collaborating laboratories in a final collaborative test with these methods are summarised in Tables I, I1 and 111. Only one figure for each laboratory is included for a given antibiotic; in some instances this was the result of one determination and in others two or more. TABLE I RESULTS OF FINAL COLLABORATIVE TEST OF T H E DETERMINATION OF PENICILLIN IN A DIET SUPPLEMENT BY THE RECOMMENDED CHEMICAL METHOD Sample contained approximately 1 g of procaine penicillin peclb. Laboratory A .. . . . . .. .... B .. . . . . .. . . .. c . . . . .. . . .. .. D .. . . .. .. .. . . E . . . . . . .. .. . . F .. . . . . . . .. . . G . . . . .. . . . . .. Mean . . . . .. .. . . .. Standard deviation . . .. .. . . Coefficient of variation . . .. .. Procaine penicillin, g per lb 1.15 1.03 1.02 - 1-04 1.14 0.92 1.05 0.085 8-1 yo From the inter-laboratory variances of these results it was considered that the methods for the assay of high-potency diet supplements are satisfactory, and that, if several labora- tories examine samples of this type containing penicillin or chlortetracycline in concen- trations of about 1 or 3.6 g per lb., 19 out of 20 of the results might be expected to fall within 20 per cent. of the means; with Supplements containing about 5 g of oxytetracycline per Ib, 19 out of 20 of the results might be expected to fall within 10 per cent.of the mean.November, 19631 I N ANIMAL FEEDING STUFFS SUB-COMMITTEE, PART 2 837 LOW-POTENCY DIET SUPPLEMENTS- The methods recommended below for compound feeding stuffs are generally suitable for these supplements. COMPOUND FEEDING STUFFS- The problems associated with the assays of compound feeding stuffs are much more complex. First, the concentrations of the antibiotics (1 to 20g per ton) are so low that chemical methods are inapplicable and it is necessary to use microbiological procedures. Second, the composition of feeding stuffs is so variable and the range of additives used is so TABLE I1 RESULTS OF FINAL COLLABORATIVE TEST OF THE DETERMINATION OF CHLORTETRACYCLINE I N A DIET SUPPLEMENT BY THE RECOMMENDED CHEMICAL METHOD Sample contained approximately 3.5 g of chlortetracycline hydrochloride per lb.Laboratory x . . . . . . . . B . . . . . . . . c .. .. . . .. D . . . . , . . . E . . .. . . . . F . . . . . . .. G . . .. . . .. Mean . . . . . . Standard deviation . . Coefficient of variation . . Chlortetracycline hydrochloride, g per lb . . .. 3.37 . . . . 3.30 . . . . 3.10 . . . . 3-53 . . . . 3.52 . . .. 2.75 . . .. 3-26 . . .. 0.296 - . . . . . . .. 9*1yo TABLE I11 RESULTS OF FINAL COLLABORATIVE TEST OF THE DETERMINATION OF OXYTETRACYCLINE I N A DIET SUPPLEMENT BY THE RECOMMENDED CHEMICAL METHOD Sample contained approximately 5 g of oxytetracycline per lb. Laboratory A . . .. .. .. B . . . . . . . . c . . . . . . . . D . . . ... . . E . . . . . . . . I; . . . . .. . . G . . .. .. . . Mean . . . . . . Standard deviation . . Coefficient of variation . . . . .. ,. .. . . .. . . * . . . .. . . .. . . .. . . .. . . .. Oxytetracycline (free base), g per lb 4.76 4.81 4-73 4.83 4.83 4.73 4-78 0.176 - 3.7% extensive that a method found satisfactory for one feeding stuff will not necessarily be sat isfactory for another. A t the beginning of the work the Panel agreed on the use of plate, rather than tube, methods for the microbiological assays, and the use of large plates rather than Petri dishes was preferred. It was also decided that the extraction methods to be used should be those that had been found satisfactory for the respective supplements by the antibiotic manu- facturers themselves.For the early investigations, these extraction methods and the organisms to be used were defined, but the details of the microbiological techniques were left to the choice of the individual analysts. The results of these trials were encouraging with penicillin and oxytetracycline, but with chlortetracycline it was apparent that the procedure had to be investigated more fully if accurate and precise results were to be obtained. Experiments were therefore undertaken to find the optimum conditions for plate assays, particularly as applied to chlortetracycline.838 ANALYTICAL METHODS COMMITTEE REPORT OF THE ADDITIVES [Analyst, VOl. 88 Among the factors examined were the constituents of the assay medium and its depth, the cavity size, the strain of test organism and whether or not the inoculum should be vegetative or a spore suspension, and the incubation temperature.Inocula and media were exchanged, and a detailed examination was made of the size and clarity of the inhibition zones in relation to the results obtained. It was evident from this work that the methods for all three anti- biotics should be specified in considerable detail. With some samples it is desirable that an unsupplemented sample of the feeding stuff under test should be available as a blank sample; it can then be extracted by exactly the same procedure as is applied to the test sample, and the extract used as a diluent for the standard, i.e., to give a “modified standard.” As an unsupplemented sample is rarely available in practice, the Panel also investigated methods for destroying or removing the antibiotic from the sample or the sample extract; the procedures found satisfactory for this purpose are included in the methods recommended for the assay of compound feeding stuffs described in Appendixes IV, V and VI.The methods drafted by the Panel specify in unambiguous detail the procedure to be followed in the extraction stage, but at the same time are sufficiently flexible to be carried out in any laboratory equipped for, and experienced in, this type of work. Throughout the collaborative studies, the Panel has been impressed by the importance of experience, not only in microbiological assays in general, but in the assay of each particular antibiotic. It is assumed also that anyone undertaking such assays is familiar with the statistical design and calculations involved. The results obtained by the collaborating laboratories in a final collaborative test with these methods are summarised in Tables IV, V and VI.TABLE ITIT RESULTS OF ASSAYS OF PENICILLIN IN COMPOUND FEEDING STUFFS MEASURED AGAINST VARIOUS STANDARDS Key to Standnrds-Standard l-Unmodified. Standard 2-Modified with unsupplemented poultry meal extract. Standard 3-Modified with supplemented poultry meal extract Standard ”Modified with unsupplemented pig meal extract. Standard 5--Modified with supplemented pig meal extract treated treated to remove the antibiotic. to remove the antibiotic. Penicillin in poultry meal Penicillin in pig meal Laboratory Standard 1, g per ton h . . . . . . . . B . . . . . . .. c . . . . * . . . D .. . . . . . . E .. . . .. . . F .. . . . . . . Mean . . .. . . Standard deviation . . Coefficient of variation. . 5.55 6.10 5 40 5.12 5.19 5.63 5.50 0.35 6.46 ”/o Standard 2, g per ton 5.50 6.10 5.29 5-34 5.59 5-51 5.55 0.29 5.23% Standard 3, g per ton 5-2 1 5-50 5.70 5.15 5.49 5.4 1 0.23 4.25 yo - Standard 1, g per ton 5-05 5.57 5.10 5-05 5 0 7 5-36 Standard 4, g per ton 5-28 5.90 5.30 5-58 5.12 5.52 5-20 5.45 0-22 0.28 4.17% 5.14% Standard 5, g per ton 4.90 5.60 5-50 5.17 5-23 5.28 0.28 5.31 ”/o - At first sight it would appear from the Tables that the results outlined with unmodified standards do not differ materially from those obtained with modified standards. In some laboratories however, the simpler procedure, in which unmodified standards were used, occasionally resulted in invalid assays, because the dose - response curves were non-parallel.Provided that there is no such deviation from parallelism between the slopes, the simpler procedure, in which the standard is diluted with buffer solution rather than with unsupple- mented or treated feed extract, may be used. Some collaborating laboratories found evidence of non-parallelism of the response curves when unmodified standards were used; the means and standard deviations of results so obtained have not, therefore, been included in the Tables. It must be emphasised that in using these methods it is assumed that the identity of the antibiotic is known and that only one is present. The determination of two or moreNovember, 19631 IN ANIMAL FEEDING STUFFS SUB-COMMITTEE, PART 1 839 antibiotics present together in a feeding stuff is likely to present special problems to the analyst, and the presence of certain prophylactics and other drugs may complicate otherwise satisfactory microbiological methods.These are problems that could well occupy attention in the future, but their complete solution will probably depend on more detailed fundamental work as well as on collaborative investigation. From the inter-laboratory variances shown in Tables IV, V and VI it may be stated that, if several laboratories experienced in microbiological assays of this type all assay the same samples of poultry or pig meals containing 5 g of penicillin or 10 g of chlortetracycline or oxytetracycline per ton, 19 out of 20 results may be expected to be within about 10 per cent.of the mean for penicillin, 20 per cent. of the mean for chlortetracycline and 25 per cent. of the mean for ox yt et racycline. TABLE V RESULTS OF ASSAYS OF CHLORTETRACYCLINE IN COMPOUND FEEDING STUFFS MEASURED AGAINST VARIOUS STANDARDS For Key to Standards-See Table I V Chlortetracycline in poultry meal Chlortetracycline in pig meal Laboratory Standard 1, g per ton ,4 . . .. . . . . 8-51 B . . . . . . . . 10.00 c . . .. .. . . 8.34 D . . . . . . . . 9-84 E . . . . . . . . 7-97 F . . .. . . . . 7.67 Mean . . . . . . 8.72 Standard deviation . . 0-97 Coefficient of variation. . 11-27; Standard 2, g per ton 9.95 8.99 10.01 9.05 8-91 9.38 0-55 5-87 yo - Standard 3, g per ton 9.90 9.20 8.9 1 9-39 8.97 7-24 8-92 0.86 9.64% Standard 1, g per ton 7.41 9.30 9-27 9-27 10.02 8-47 8.96 0.90 10.06 yo Standard 4, Standard 5, 9.77 8.90 - 8-60 9.35 9.35 9.64 8.74 9.75 10-03 8.90 8.33 9.48 8-99 0.3 7 0.6 1 3.91% 6*79yo g per ton g per ton TABLE VI RESULTS OF ASSAYS OF OXYTETRACYCLINE IN COMPOUND FEEDING STUFFS MEASURED AGAINST VARIOUS STANDARDS For Key to Sfandards-See Table IV Oxytetracycline in poultry meal Oxytetracycline in pig meal Laboratory Standard 1, Standard 2, g per ton g per ton A .. .. . . . . 7-89 8-98 B . . . . . . . . 9-15 11.75 c . . .. . . . . 9-53 9.37 D . . . . . . . . 7.66 9-29 E . . . . . . . . 9.21 9.66 F . . .. . . . . 11.05 11.40 Mean . . . . . . 9.08 10.07 Standard deviation . . 1-23 1-18 Coefficient of variation. . 13-57; ll*76y0 Standard 3: g per ton 8-71 11.45 9-42 8.69 9.80 10.70 9.79 1.10 11.28y0 L- 7 Standard 1, Standard 4, 7-85 8-76 9.10 10.85 9.32 9.31 8-39 10.0 8.80 9.18 9-73 9.87 8.86 9.66 0.6’7 0.74 7.61”/, 7.66% g per ton g per ton 7 -~ Standard 5, g per ton 8.78 11-10 9.10 9.17 9.72 9.80 9-61 0.83 8.6% ACKNOWLEDGMENTS The Panel is grateful to the organisations in whose laboratories the work was carried out, and in particular to Messrs.Boots Pure Drug Co. Ltd., Cyanamid of Great Britain Ltd., Distillers Company Ltd., Glaxo Laboratories IAd., Pfizer Ltd. and Vitamins Ltd. for generously supplying materials for the collaborative tests.840 ANALYTICAL METHODS COMMITTEE : REPORT OF THE ADDITIVES [Analyst, Vol. 88 Appendix I CHEMICAL DETERMIXATION OF PENICILLIN I N DIET SUPPLEMENTS PRINCIPLE OF METHOD- The supplement is extracted with water, and the procaine penicillin in the extract is determined by the iodimetric method of the “British Pharmacopoeia 1958”l for total peni- cillins in benzylpenicillin (see Note 1).APPLICABILITY AND RANGE- per lb. REAGENTS- The method is applicable to diet supplements containing not less than 1 g of penicillin Water-Use distilled or de-ionised water. The other reagents required are described in the “British Pharmacopoeia 1958.”1 PROCEDURE EXTRACTION OF SAMPLE- (a) For supplements containing 1 to 5 g of penicillin per lb-Weigh accurately about 25 g of sample, and transfer quantitatively to a small glass mortar. Mix with a few millilitres of water, and grind to form a smooth paste. Transfer the mixture quantitatively to a tared 250-ml conical flask, rinsing in with water, and add sufficient water to the mixture to give a total weight of 175 g. Mix thoroughly, and set aside for 5 minutes, with occasional shaking.Filter the supernatant liquid through a fluted 24-cm Whatman No. 42 filter-paper, discard the first 20 ml of filtrate, and then collect 60 ml. Determine the procaine penicillin in the filtrate by the method described below (see Note 2). (b) For sufiplements containing 16 g of penicillin per lb-Extract the sample as described above under (a), but use about 12 g of sample and make the weight of the mixture up to 162 g with water. (c) For supplements containing 224 g of penicillin per Ib-Weigh accurately about 2.5 g of sample, and transfer quantitatively to a glass mortar. Mix with a few millilitres of water, and grind to form a smooth paste.Transfer the mixture quantitatively to a 500-ml calibrated flask, dilute to about 400 ml with water, and set aside for 5 minutes, with occasional shaking. Dilute the mixture to the mark, mix well, immediately filter through a fluted 24-cm Whatman No. 42 filter-paper, discard the first 20 ml of filtrate, and then collect 60 ml. Determine the procaine penicillin in the filtrate by the method described below (see Note 2). DETERMINATION OF PROCAINE PENICILLIN- Determine the procaine penicillin in a 10-ml portion of the filtrate from the extraction of the sample by the iodimetric method for total penicillins exactly as described in the “British Pharmacopoeia 1958” under “Benzylpenicillin,” p. 89 (see Note 3).RESULTS- Express the results as grams of procaine penicillin per lb of supplement. NOTES 1 . Since this report was prepared the “British Pharmacopoeia 1963”2 has been published, and in this new pharmazopoeia a method is included for the determination of total penicillins in procaine penicillin. The B.P., 1963, method is similar to that recommended by the Panel, except that the procaine is first removed from solution by precipitation with sodium silicotungstate ; as the procaine has been removed, i t is unnecessary to apply the correction factor of 1.04 (see Note 3). 2. The analysis should be completed with the minimum of delay, because the aqueous penicillin solution slowly decomposes. 3. It is nezessary, in order to determine the factor applicable in each laboratory, to standardise the volumetric reagents by using a standard preparation of benzylpenicillin sodium and to convert the fastor obtained to a figure applicable to procaine penicillin by multiplying by the appropriate molecular-weight ratio ; the molecular weight of benzylpenicillin sodium is 356.4 and that for procaine penicillin (monohydrate) is 588.7, so that the ratio of molecular weights is 1 t o 1.652.It is generally recognised that when the iodimetric method is applied to procaine penicillin it gives results that are 4 per cent. low; it is recommended, therefore, that the factor be multiplied by 1.04 to correct for this.November, 19631 IN ANIMAL FEEDING STUFFS SUB-COMMITTEE, PART 1 Appendix I1 CHEMICAL DETERMINATION OF CHLORTETRACYCLINE IN DIET SUPPLEMENTS PRINCIPLE OF METHOD- The method is based on that of Chiccarelli, Woolford and Tr~mbitas.~ 841 REAGENTS- Water-Use distilled or de-ionised water.Hydrochloric acid, 5 N. Dilute sodium hydroxide solution-Dilute 8 ml of 5 N sodium hydroxide to 100 ml with water. Sodium metabisulphite solution-A 10 per cent. w/v solution in water. The solution must be freshly prepared. Phosphate bz@er solution (pH 7.5)-Dissolve 178 g of dipotassium hydrogen ortho- phosphate, K,HPO,, and 22 g of potassium dihydrogen orthophosphate, KH,PO,, in 1 litre of water. Stock standard chlortetracycline solution-Transfer exactly 100 mg of chlortetracycline hydrochloride B.P. to a 100-ml calibrated flask, dilute to the mark at 20" C with water, and mix well. Store the solution in an amber-glass bottle at 5" to 8" C; under these conditions the solution is stable for one week.Working standard chlortetracycline solution-Dilute 5.0 ml of stock standard chlortetra- cycline solution to 100 ml at 20" C with water, and mix well; Prepare the solution immediately before use. Filter the solution before use. 1 ml = 0-05 mg of chlortetracycline hydrochloride. PROCEDURE EXTRACTION OF SAMPLE- Weigh accurately a portion of the sample expected to contain approximately 5mg of chlortetracycline hydrochloride (see Note 1) into a 100-ml calibrated flask. Add about 70 ml of water and 4 ml of 5 N hydrochloric acid, shake the mixture for 10 minutes, dilute to the mark with water, and mix thoroughly (see Note 2). Filter a portion of the mixture through Celite filter aid, discarding the first 20 to 30 ml of filtrate, collect the clear filtrate, and determine the chlortetracycline in the filtrate by the method described below. DETERMINATION OF CHLORTETRACYCLINE- Transfer by pipette two 10-ml portions of the filtrate from the extraction of the sample into separate 50-ml calibrated flasks; these form the sample and sample blank solutions.Transfer by pipette two 10-ml portions of the working standard chlortetracycline solution into further separate 50-ml calibrated flasks; these form the standard and standard blank solutions. To the solutions representing the sample and the standard add, in order, 12 ml of 5 N hydrochloric acid, 15 ml of phosphate buffer solution (pH 7 3 ) , 2 ml of sodium metabisulphite solution and 3ml of dilute sodium hydroxide solution, and suspend the flasks in a bath of boiling water for exactly 7 minutes, swirling the contents occasionally (see Note 3).To the solutions representing the sample blank and the standard blank add 15 ml of phosphate buffer solution (pH 7-5), 2 ml of sodium metabisulphite solution and 3 ml of dilute sodium hydroxide solution, and suspend the flask in a bath of boiling water, swirling the contents occasionally (see Note 3). After exactly 5 minutes, add 12 ml of 5 N hydrochloric acid, and heat for an additional 2 minutes. Immediately after the completion of the heat treatment, cool the four flasks in running water, dilute the contents of each flask to the mark at 20" C with water, and mix well.842 ANALYTICAL METHODS COMMITTEE REPORT OF THE ADDITIVES [Analyst, VOl.88 If a cloudiness develops in the solutions at this point, spin the sample and sample blank Measure the optical densities, at 445 mp, of the solutions representing the standard and solutions in a centrifuge until clear. the sample against their respective blank solutions with a suitable spectrophotometer. CALCULATION- The amount of chlortetracycline hydrochloride in the sample is given by the expression- Chlortetracycline hydrochloride content = (A,,, sample) x 0.01 x (dilution of sample) x 453-6 (A445 standard) x (weight of sample) x 1000 g per lb, where A,,5 = optical density at 445 mp and 0.01 = final concentration of the standard in mg per ml. RESULTS- Express the results as grams of chlortetracycline hydrochloride per lb of supplement.NOTES 1 . The sample taken must not weigh less than 100 mg. If high-potency material is t o be assayed and a weight of 100 mg contains more than 5 mg of chlortetracycline hydrochloride, take a sample weighing 100 mg or more, and make appropriate dilutions to give a concentration of 0.05 mg per ml. Increase thc amount of 5 N hydrochloric acid used in direct proportion to the dilution. 2. The concentration of this solution is approximately 0.05 mg of chlortetracycline hydrochloride per ml. 3. It is essential that the water is boiling throughout the entire heating period. Appendix 111 CHEMICAL DETERMINATION OF OXY’TETRACYCLINE I S DIET SCPPLEMEYTS PRINCIPLE OF METHOD- tetracycline with ferric chloride in the presence of 0-01 N hydrochloric acid. The method is based on the measurement of the colour produced by reaction of oxy- APPLICABILITY- iron, interfere with the method.be removed before the analysis is undertaken. Ingredients such as phosphates, fluoride, thiocyanates, etc., that combine with ferric These interfering substances must therefore be absent or RANGE- For supplements containing not less than 5 g of oxytetracycline per lb. REAGENTS- Water-Use distilled or de-ionised water. Dilute acetic acid-A 10 per cent. v/v solution of glacial acetic acid in water. Concentrated hydrochloric acid, sp.gr. 1-16 to 1.18. Hydrochloric acid, 0.01 N-Adjust to pH 2.0 if necessary. Ferric chEoride soZution--A 0.05 per cent. w/v solution of ferric chloride, FeCl,.GH,O, in 0.01 N hydrochloric acid. Standard oxytetracycline solzttion-Dissolve sufficient oxytetracycline hydrochloride in 0.01 N hydrochloric acid, and dilute with the same acid to give a solution containing 400 pg of oxytetracycline per ml.November, 19631 I N ANIM..IL FEEDING STUFFS SUB-COMMITTEE, PART 1 843 PROCEDURE EXTRACTION OF SAMPLE- (a) For supplemenis containing 5 g of ox-ytetracycline per lb-Weigh accurately about 3 g of the sample into a 250-ml beaker, add 50 ml of acetic acid, and set the mixture aside for 20 minutes, stirring gently (see Note 1).Filter the solution quantitatively through a Buchner funnel, more than once if necessary, until the filtrate is absolutely clear. Wash the filter-paper with 10 ml of water, combine the filtrate and washings, and adjust the clear mixture to pH 2.0 0.05 (measured with a pH meter) with a few drops of concentrated hydrochloric acid.Transfer the solution quantitatively to a 100-ml calibrated flask, rinsing in with 0.01 N hydrochloric acid, and dilute to the mark at 20" C with the same acid (see Note 2). Determine the oxytetracycline hydrochloride in the solution by the method described below. (b) For supplements containing 10 g of oxytetracycline per Zb-Extract the sample as described under ( a ) , but use about 2.5 g of sample. (c) For supplements containing 25 g of oxytetracycline per lb-Extract the sample as described under ( a ) , but use about 1 g of sample. DETERMINATION OF OXYTETRACYCLINE- Transfer a 5-ml portion of the solution from the extraction of sample to each of two clean test-tubes. Add to one tube 15 ml of 0.01 N hydrochloric acid and to the other tube 5 ml of 0.01 N hydrochloric acid and 10 ml of ferric chloride solution, mix, and set aside at 20" to 25" C for 20 minutes.Measure the optical density of each solution in turn at 490mp., in a 1-cm cell with a suitable spectrophotometer. Prepare, in 0.01 N hydrochloric acid, dilutions of standard oxytetracycline solution covering the range 0 to 400 pg of oxytetracycline per ml, and treat 5-ml portions of each dilution as described above for the solution from the extraction of the sample. Construct a graph relating the optical densities to the number of micrograms of oxytetracycline in each dilution. By reference to this calibration graph, determine the amount of oxytetracycline in the solution from the extraction of the sample, and hence calculate the amount of oxytetracycline in the sample.RESULTS- Express the result as grams of oxytetracycline (free base) per lb of supplement. NOTES 1 . h magnetic stirrer is suitable. 2. If the material being assayed contains 6 g of oxytetracycline per lb, the concentration in this solution is approximately 300 pg per ml. Appendix IV MICROBIOLOGICAL ASSAY OF PENICILLIN I N DIET SUPPLEMENTS AND COMPOUND FEEDING STUFFS Normal bacteriological procedures and precautions must be adopted in preparation of cultures, sterilisation of media and glassware, etc. REAGENTS- Assay medium- Peptone . . . . .. .. .. . . 5.0g Lab. Lemco . . . . .. . . . . 3.0g Agar . . . . . . .. . . . . 15.0g Glass-distilled water . . . . . . . . to 1 litre The recommended procedure for preparing the medium is to dissolve the peptone and the Lab.Lemco in the water, adjust the solution to pH 7.0, stir in the agar, and transfer the mixture to an autoclave. Steam for 30 minutes, and then heat at 115" C for 20 minutes. Transfer 200- to 250-ml amounts (see Note 1) to 12-oz screw-capped bottles, and sterilise at 115" C for 20 minutes.844 ANALYTICAL METHODS COMMITTEE : REPORT OF THE ADDITIVES [Anahst, VOl. 88 Acetone - phosphate bufer so2ution (pH 7-5)-Dissolve 9.78 g of disodium hydrogen orthophosphate, Na,HPO,, and 1-85 g of potassium dihydrogen orthophosphate, KH,PO,, in 200 ml of hot distilled water, cool, add 250 ml of acetone, and dilute to 1 litre with distilled water. Phosphate bufer solution (PH 7.O)-Dissolve 5.0 g of dipotassium hydrogen ortho- phosphate, K,HPO,, and 3-9 g of potassium dihydrogen orthophosphate, KH,PO,, in distilled water, and dilute to 5 litres with distilled water.The solution may be stored; if so, it should be sterilised. ORGANISM- strain, e.g., ATCC 6633. two sterile media listed below- The organism used in the assay is Bacillus subtilis (NCIB 8236), or any other suitable Maintenace of culture-Maintain cultures of the organism on slopes of either of the Peptone . . .. .. . . . . . . 6.0g Casein hydrolysate (enzymic) . . . . . . 4-Og Marmite . . .. .. . * .. . . 3-Og Lab. Lemco . . .. .. . . . . 1.5g Dextrose . . .. .. , . ,. . . b o g Agar . . . . . . . . .. , . 15.0g Glass-distilled water . . . . . . . . to 1 litre Adjust to pH 6.5 to 6.6, and sterilise by heating at 115" C for 10 minutes in 10-ml bottles.07 Difco Penassay Seed Agar Sterilise by heating at 115" C for 10 minutes in 10-ml bottles. Preparation of spore suspension-Transfer 200 ml of one of the media to a Roux bottle, and sterilise by heating at 115" C for 10 minutes. Cool, inoculate with the culture, and incubate at approximately 30" C for 7 to 8 days. Wash the growth from the surface of the medium with 150ml of sterile glass-distilled water, spin the suspension in a centrifuge, re-suspend the deposit in sterile glass-distilled water, and dilute to match approximately opacity 4 on Brown's opacity tubes. Fill 1-02 screw-capped bottles with the suspension, pasteurise at 70" to 75" C for 30 minutes, and store in a refrigerator. PROCEDURE PREPAR-4TION OF PLATES- Melt the contents of a bottle of assay medium by steaming for approximately 1 hour, and cool to about 50" C in a water bath at this temperature.Inoculate the melted medium with the spore suspension of B. subtilis-dilute the sus- pension, if necessary, and use an inoculum density found by previous experiment to give zone diameters of suitable size (see Note 2)-mix well, and pour on to a 12-inch x 12-inch sterilised plate (see Note 3). The plate must be supported on a level surface while the medium is being poured. Leave the plate at room temperature with the lid raised until the medium has set, and store in a refrigerator for at least 1 hour. Cut 64 cups in the solidified medium with the aid of a template; use a No. 5 cork-borer (8 mm in diameter) or nearest available size.PREPARATION OF STANDARD SOLUTIONS- Dissolve 50 mg of benzylpenicillin sodium of known potency in phosphate buffer solution (pH 7.0), and dilute to 500 ml with the buffer solution (see Note 4). Take a portion of the solution containing as much antibiotic activity as is estimated to be present in the test sample (see Note 5 ) , dilute to 100 ml witlt acetone - phosphate bufer solution (PH 7-5), and add 20g of the blank (unsupplemented) feed (see Note 6). Shake for 1 hour, and then sfiin in a centrifuge, orjlter. Take a portion of the solution, dilute, if necessary, with acetone - phosphate buffer solution (pH 7.5) to give a solution containing 1.0 i.u. of penicillin per ml, and from this make a further dilution in acetone - phosphate buffer solution (pH 7-5) to give a solution containing 0-2 i.u.of penicillin per ml.November, 19631 IN ANIMAL FEEDING STUFFS SUB-COMMITTEE, PART 1 845 PREPARATION OF TEST SOLUTIONS- Sup$Zements-Weigh a 1-g sample, add 250 ml of acetone - phosphate buffer solution (pH 7 6 ) , and shake vigorously. Spin in a centrifuge, or filter, and dilute the solution with acetone - phosphate buffer solution (pH 7.5) in accordance with the expected potency to the levels required in the assay, namely, approximately 1.0 and 0.2 i.u. of penicillin per ml. Feeding stufs-Weigh a 20-g sample, add 100 ml of acetone - phosphate buffer solution (pH 7 6 ) , and shake for 1 hour. Spin in a centrifuge, or filter, and dilute the solution with acetone - phosphate buffer solution (pH 7.5) according to the expected potency to the levels required in the assay, namely, approximately 1-0 and 0.2 i.u. of penicillin per ml.ASSAY- By means of a statistically satisfactory assay design, such as suggested by Brownlee et aZ.,4 Lees and T ~ o t i l l , ~ Price and Boucher6 or Simpson and Lees,’ compare the test and standard solutions in a cup-plate assay by using the previously prepared plates. Plate out the solutions according to the chosen assay design; deliver a uniform volume of 0.05 to 0.1 ml into each cup of the prepared plates. Set aside at room temperature for 1 hour to allow pre-diffusion, and then place in an incubator at approximately 30” C. After incubation for 18 to 20 hours, measure the diameters of the inhibition zones with finely pointed vernier callipers or a projection device. Check the parallelism of the test and standard responses, and, if satisfactory, calculate the potency, P, of the test solution from the equation- where T , = the total of the responses to the high dose of test solution, T, = the total of the responses to the low dose of test solution, S , = the total of the responses to the high dose (1.0 i.u.per ml) of standard solution, and S , = the total of the responses to the low dose (0.2 i.u. per ml) of standard solution Calculate the potency of the sample from the dilution employed. RESULTS- ton of compound feeding stuff. Express the results as grams of procaine penicillin per pound of diet supplement or per NOTES 1. The 1-olume of the medium should be such that when poured into a 12-inch x 12-inch plate the depth of medium is 0.1 inch.2 . Zone diameters reported by the Panel members ranged from 15 t o 28 mm for a dose level The increase in response 3. Sterilisation may be achieved by heating in an autoclave or by swabbing with acid alcohol. 1. For the assay of diet supplements, the instructions in italic type can be omitted. They may also bc omitted in the assay of compound feeding stuffs if, by previous experiment, it has been found that the test response is parallel to that of an “unmodified standard.” 5 . Take a 0.6-ml portion for a feeding stuff estimated to contain 5 g of procaine penicillin per ton. 6. If unsupplemented feed is not available, a suitable preparation can be made by steaming of 1.0 i.u. per ml and from 10 t o 20 mm for a dose level of 0-2 i.u.per ml. for this five-fold increase in dose ranged from 5 t o 10 mm. a portion of the test sample for 1 hour or heating in an autoclave a t 120* C for 15 minutes.846 ANALYTICAL METHODS COMMITTEE : REPORT OF THE ADDITIVES [Analyst, Vol. 88 Appendix V MICROBIOLOGICAL ASSAY OF CHLORTETRACYC,LINE IS DIET SUPPLEMENTS AXD COMPOUXD FEEDING STUFFS Normal bacteriological procedures and precautions milst be adopted in preparation of cultures, sterilisation of media and glassware, etc. REAGENTS- Assay medium- Yeast extract . . . . . . . . . . 1-og Ammonium nitrate . . . . . . . . 5.og Sodium dihydrogen ort hophosphat e (NaH2P0,.2H,0), B.P. grade . . . . 5*0g Dextrose . . .. , . . . . . . . 5-Og Agar . . .. . . . . . . . . 15.0g Glass-distilled water .. . . . . . . to 1 litre The recommended procedure for preparing the medium is to dissolve the yeast extract, ammonium nitrate and sodium dihydrogen orthophosphate in the water, adjust the solution to pH 7.5, stir in the agar, and transfer the mixture to an autoclave. Steam for 30 minutes, and then heat at 115" C for 20 minutes. Remove from the autoclave, add the deptrose and filter clear through a prepared paper-pulp pad about 0.5 inch thick. Cool, and adjust the solution to pH 7.0. Transfer 200- to 250-ml amounts (see Note 1) to 12-oz screw-capped bottles, and sterilise at 115" C for 20 minutes. Aqueous acid - acetone mixture-Mix 40 ml of hydrochloric acid, sp.gr. 1.16 to 1.18, with 1300 ml of acetone, and dilute to 2 litres with distilled water. Sodium hydroxide, N.Phosphate bufer solution (PH 4.5)-Dissolve 13.6 g of potassium dihydrogen ortho- phosphate, KH2P0,, in distilled water, and dilute to 1 litre with distilled water. Hydrochloric acid, N. Sodium chloride solution-A 5.85 per cent. w/v solution in distilled water. The final pH of the medium is 6-7. ORGANISM- The organism used in the assay is Bacilhs cereus ECIB 8849 or 9231 (ATCC 11778). , Maintenance of culture-Maintain cultures of the organism on slopes of "Oxoid" Blood Agar Base ("Oxoid" Nutrient Broth + 1.5 per cent. of agar) that has been sterilised at 115" C for 10 minutes in 10-ml screw-capped bottles. Preparation of spore sus+ension-Prepare the following medium- Protone (Difco) . . . . . . . . . . 5.0g Manganese sulphate (MnS0,$.4H20) . . . . 0.001 g Agar .. . . . . * . . . . . 2o.og Glass-distilled water . . . . . . . . to 1 litre Transfer 200-ml amounts of the medium to Roux bottles, and sterilise by heating at 120" C for 15 minutes. Wash the growth from the surface of the medium with 150 ml of sterile glass-distilled water, and spin the suspension in a centrifuge; wash the deposit with three successive 150-ml portions of sterile glass-distilled water, spinning the suspension in a centrifuge each time. Re-suspend the growth in sterile glass-distilled water, and dilute to match opacity 1, or less, on Brown's opecity tubes. Cool, inoculate with the culture, and incubate at 30" C for 5 to 7 days. Store the suspension in a refrigerator. PROCEDURE PREPARATION OF PLATES- and cool to about 50" C in a water bath at this temperature.Melt the contents of a bottle of assay medium by steaming for approximately 1 hour,November, 19631 I N ANIMAL FEEDING STUFFS SUB-COMMITTEE, PART 1 845 Inoculate the melted medium with the spore suspension of B. cereus-dilute the suspen- sion, if necessary, and use an inoculum density found by previous experiment to give zone diameters of suitable size (see Note 2), mix well, and pour on to a 12-inch x 12-inch sterilised plate (see Note 3). The plate must be supported on a level surface while the medium is being poured. Leave the plate at room temperature with the lid raised until the medium has set, and store in a refrigerator for at least 1 hour. Cut 64 cups in the solidified medium with the aid of a template; use a No. 5 cork-borer (8 mm in diameter) or nearest available size.PREPARATION OF STANDARD SOLUTIONS- Dissolve 50 mg of chlortetracycline hydrochloride of known potency in phosphate buffer solution (pH 4 4 , and dilute to 500 ml with the buffer solution (see Note 4). Measure 25 ml of solution A (see below under “Preparation of Test Solutions”) into a 50-ml beaker, add 2-5 ml of N sodium hydroxide, boil gently for 15 minutes, and cool. Add 2-5 ml of N hydrochloric acid, transfer the solution quantitatively to a 50-ml calibrated $ask, add a portion of the standard chlortetracycline solution containing as much antibiotic activity as is estimated to be present in the test sample, and dilute to the mark with Phosphate bu#er solution (PH 4.5). Dilute a portion with phosphate buffer solution (pH 4-5) to give a solution contain- ing 0.25 pg of chlortetracycline per ml, and from this make a further dilution in phosphate buffer solution (pH 4.5) to give a solution containing 0.05 pg of chlortetracycline per ml.PREPARATION OF TEST SOLUTIONS- Supplements-Weigh a 1-g sample, add sufficient aqueous acid - acetone mixture to give a volume of 250m1, shake for 1 hour, and filter. Dilute the filtrate with phosphate buffer solution (pH 4-5) according to the expected potency to the levels required in the assay, namely, approximately 0.25 and 0.05 pg of chlortetracycline per ml. Feeding stufs-Weigh a 20-g sample, add 100 ml of aqueous acid - acetone solution, shake for 1 hour, and spin in a centrifuge. Titrate a 5-ml portion of the supernatant liquid with N sodium hydroxide, with methyl orange solution as indicator.Transfer a 25-mI portion of the supernatant liquid to a 100-ml calibrated flask, add a volume of N sodium hydroxide equivalent to five times the volume required in the titration, and dilute to the mark with phosphate buffer solution (pH 46). To 10 ml of Solution A, add 1 ml of sodium chloride solution, and dilute with phosphate buffer solution (pH 4.5) according to the expected potency to the levels required in the assay, namely, approximately 0.25 and 0-05 pg of chlortetracycline per ml, (see Note 5). ASSAY- By means of a statistically satisfactory assay design, such as suggested by Brownlee et a,?.: Lees and T ~ o t i l l , ~ Price and Bouchere or Simpson and Lees,’ compare the test and standard solutions in a cup-plate assay by using the previously prepared plates.Plate out the solutions according to the chosen assay design; deliver a uniform volume of 0.05 to 0.1 ml into each cup of the prepared plates. Set aside at room temperature for 1 hour to allow pre-diffusion, and then place in an incubator at approximately 30” C. After incubation for 18 to 20 hours, measure the diameters of the inhibition zones with finely pointed vernier callipers or a projection device. Check the parallelism of the test and standard responses, and, if satisfactory, calculate the potency, P, of the test solution from the equation- (This is Solution A.) where T2 = the total of the responses to the high dose of test solution, T , = the total of the responses to the low dose of test solution, S2 = the total of the responses to the high dose (0.25 pg per ml) of standard solution S, = the total of the responses to the low dose (0.05 pg per ml) of standard solution.and Calculate the potency of the sample from the dilution employed.848 ANALYTICAL METHODS COMMITTEE: REPORT OF THE ADDITIVES [Analyst, vol. 88 RE s u LT s- supplement or per ton of compound feeding stuff. Express the results as grams of chlortetracycline hydrochloride per pound of diet NOTES 1. The volume of medium should be such that when poured into a 12-inch x 12-inch plate the 2. Zone diameters reported by the Panel members ranged from 14 to 30 mm for dose levels of The increase in response 3. Sterilisation may be achieved by heating in an autoclave or by swabbing with acid alcohol.4. For the assay of diet supplements the instructions in italic ty$e can be omitted. They may also be omitted in the assay of compound feeding stuffs if, by previous experiment, i t has been found that the test response is parallel to that of an unmodified standard. 5. If, for example, the sample contains 10 g per ton, or 9.84 pg per g, dilute to 20 ml. depth of medium is 0.1 inch. 0.25 pg per ml and from 12 to 25 mm for dose levels of 0.05 pg per ml. for this five-fold increase in dose ranged from 2.2 to 9 mm. Appendix VI MICROBIOLOGICAL ASSAY OF OXYTETRACYCLINE IN DIET SUPPLEMENTS AND COMPOUND FEEDING STUFFS Normal bacteriological procedures and precautions must be adopted in preparation of cultures, sterilisation of media and glassware, etc. REAGENTS- Assay medium- Yeast extract .. .. .. .. . . b o g Ammonium nitrate . . . . .. . . 5.0g (NaH2P0,.2H,0), B.P. grade .. . . 5*0g Dextrose . . .. ,. .. .. , . 5-Og Agar .. . . .. . . .. . . 150g Sodium dihydrogen orthophosphate Glass-distilled water . . .. .. . . to llitre The recommended procedure for preparing the medium is to dissolve the yeast extract, ammonium nitrate and sodium dihydrogen orthophosphate in the water, adjust the solution t o pH 7.5, stir in the agar, and transfer the mixture to an autoclave. Steam for 30 minutes, and then heat at 115” C for 20 minutes. Remove from the autoclave, add the dextrose, and filter clear through a prepared paper-pulp pad about 0.5 inch thick. Cool, and adjust the solution to pH 7.0. Transfer 200- to 250-ml amounts (see Note 1) to 12-02 screw-capped bottles, and sterilise at 115” C for 20 minutes.Acid methanol sohtion, 2 per cent. v/v-Dilute 20 ml of hydrochloric acid, sp.gr. 1.16 to 3-18, to 1 litre with methanol. Sodium hydroxide, N. Hydrochloric acid, N. Phosphate bufer solution (@H 4.5)-Dissolve 13-6 g of potassium dihydrogen ortho- Sil Flo-Obtainable from Haller and Phillips Ltd., 68-70 Goswell Road, London, E.C.l. Sodium carbonate-Na,CO,. 10H,O. B.P. quality. The final pH of the medium is 6.7. phosphate, KH,PO,, in distilled water and dilute to 1 litre with distilled water. ORGANISM- The organism used in the assay is Bacillus cereus NCIB 8849 or 9231 (ATCC 11778). Maintenance of culture-Maintain cultures of the organism on slopes of “Oxoid” Blood Agar Base (“Oxoid” Nutrient Broth + 1.5 per cent.of agar) that has been sterilised at 115” C for 10 minutes in 10-ml screw-capped bottles.November, 19631 IN ANIMAL FEEDING STUFFS SUB-COMMITTEE, PART 1 Preparation of spore suspension-Prepare the following medium- Protone (Difco) . , * . . . . . . . 5.0g Manganese sulphate (MnS0,.4H20) . . . . 0.001 g Agar . . . . . . . . . . . . 20.0g Glass-distilled water . . . . . . . . to llitre 849 Transfer 200-ml amounts of the medium to Roux bottles, and sterilise by heating at 120" C for 15 minutes. Wash the growth from the surface of the medium with 150 ml of sterile glass-distilled water, and spin the suspension in a centrifuge; wash the deposit with three successive 150-ml portions of sterile glass-distilled water, spinning the suspension in a centrifuge each time.Re-suspend the deposit in sterile glass-distilled water, and dilute to match opacity 1, or less, on Brown's opacity tubes. PROCEDURE Cool, inoculate with the culture, and incubate a t 30" C for 5 to 7 days. Store the suspension in a refrigerator. PREPARATION OF THE PLATES- Melt the contents of a bottle of assay medium by steaming for about 1 hour, and cool to about 50" C in a water bath at this temperature. Inoculate the melted medium with the spore suspension of B. cereus-dilute the suspen- sion, if necessary, and use an inoculum density found by previous experiment to give zone diameters of suitable size (see Note 2)-mix well, and pour on to a 12-inch x 12-inch sterilised plate (see Note 3). The plate must be supported on a level surface while the medium is being poured.Leave the plate at room temperature with the lid raised until the medium has set, and store in a refrigerator for at least 1 hour. Cut 64 cups in the solidified medium with the aid of a template; use a S o . 5 cork-borer (8 mm in diameter) or nearest available size. PREPARATION OF STANDARD SOLUTIONS- Dissolve 50 mg of oxytetracycline hydrochloride of known potency in phosphate buffer solution (pH 449, dilute to 500 ml with the buffer solution and prepare modified standards by either method (i) or (ii) described below (see Note 4). (i) Standard solution modi$ed with inactivated supplemented feed-Prepare an adsorption column, 9 to 10 cm in length, composed of a mixture of equal parts of sodium carbonate and Sil Flo, in a chromatography tube, 3 cm in diameter.Pack the column dry, and tamp down before use. Transfer approximately 50 ml of Solution A (see below under "Preparation of Test Solutions") to the top of the dry adsorption column, allow the solution to pass through the column, and finally apply gentle suction. Pass the solution through the column once only. Adjust the pH of the eluate to pH 4-5 with N hydrochloric acid, and transfer 20ml of the solution quantitatively to a 50-ml calibrated flask. Add a portion of the standard oxytetra- cycline solution, prepared as described above, containing as much antibiotic activity as is estimated to be present in 4.0 g of the test sample, and dilute to the mark with phosphate buffer solution (pH 4-5). Dilute a portion with phosphate buffer solution (pH 4.5) to give a solution containing 0-8 pg of oxytetracycline per ml, and from this make a further dilution in phosphate buffer solution (pH 4-5) to give a solution containing 0-2 pg of oxytetracycline per ml.(ii) Standard solution modi$ed with unsupplemented feed-Weigh a 20-g sample of the unsupplemented feeding stuff, add 100 ml of acid methanol, shake for 1 hour, and spin in a centrifuge. Adjust the supernatant liquid to pH 4-5 with N sodium hydroxide, and transfer 20 ml of the solution quantitatively to a 50-ml calibrated flask. Add a suitable portion of the standard oxytetracycline solution, and dilute as described above under (i). PREPARATION OF TEST SOLUTIONS- Supplements-Weigh a 1-g sample, add sufficient acid methanol to give a volume of 250m1, shake for 1 hour, and filter.Dilute the filtrate with phosphate buffer solution (pH 4-5) according to the expected potency to the levels required in the assay, namely, approximately 0.8 and 0.2 pg of oxytetracycline per ml.850 ANALYTICAL METHODS COMMITTEE [Analyst, Vol. 88 Feeding shfls-weigh a 20-g sample, add 100ml of acid methanol, shake for 1 hour, and spin in a centrifuge. Transfer 10 ml of Solution A to a 25-ml calibrated flask, adjust the pH to 4.5 with N sodium hydroxide, and dilute to the mark with phosphate buffer solution (pH 4.5). If necessary, dilute portions with phosphate buffer solution (pH 4.5) according to the expected potency to the levels required in the assay, namely, approximately 0.8 and 0.2 pg of oxy- tetracycline per ml. (This is Solution A.) ASSAY- plates and a statistically satisfactory assay design.of 0-05 to 0.1 ml into each cup of the prepared plates. 1 hour to allow pre-diffusion, and then place in an incubator at approximately 30" C. finely pointed vernier callipers or a projection device. the potency, P, of the test solution from the equation- Compare the test and standard solutions in a cup-plate assay; use the previously prepared Plate out the solutions according to the chosen assay design; deliver a uniform volume Set aside at room temperature for After incubation for 18 to 20 hours, measure the diameters of the inhibition zones with Check the parallelism of the test and standard responses and, if satisfactory, calculate where T , = the total of the responses to the high dose of test solution, T , = the total of the responses to the low dose of test solution, S , = the total of the responses to the high dose (0.8 pg per ml) of standard solution S, = the total of the responses to the low dose (0.2 pg per ml) of standard solution. and Calculate the potency of the sample from the dilution employed. RESULTS- or per ton of compound feeding stuff. Express the results as grams of oxytetracycline (base) per pound of diet supplement NOTES 1. The volume of medium should be such that when poured into a 12-inch x 12-inch plate the 2. Zone diameters reported by Panel members ranged from 16 to 26 mm for dose levels of 0.8 p g The increase in response for this four-fold 3. Sterilisation may be achieved by heating in an autoclave or by swabbing with acid alcohol. 4. For the assay of supplements a standard diluted further with the buffer solution to give solutions containing 0.8 and 0-2 p g of oxytetracycline per ml may be used instead of one modified with a feeding stuff as described below. Likewise, with feeding stuffs if, by previous experiment, it has been found that the test response is parallel to that of an unmodified standard, a modified standard need not be used. depth of medium is 0.1 inch. per ml and from 11 to 18 mm for dose levels of 0.2 p g per ml. increase in dose ranged from 3.2 to 10 mm. REFERENCES 1. 2. 3. 4. 5. 6. 7. British Pharmacopoeia, 1958, p. 89. British Pharmacopoeia, 1963, p. 656. Chiccarelli, F. S., Woolford, M. H., and Trombitas, R. W., J . Ass. 08. Agric. Chem., 1957, 40, 922. Brownlee, K. A., Delves, C. S., Dorman, M., Green, C. A., Johnson, J. D. .4., and Smith, N., Lees, K. A., and Tootill, J. P. R., Analyst, 1955, 80, 95, 110 and 531. Price, S. A., and Boucher, K. A., Analyst, 1954, 79, 150. Simpson, J. S., and Lees, K. A., Analyst, 1956, 81, 562. J . Gen. Microbiol., 1948, 2, 40.

ISSN:0003-2654    

DOI:10.1039/AN9638800836

出版商:RSC    

年代:1963

数据来源: RSC

摘要:

November, 19631 CHIRNSIDE, CLULEY, POWELL AND PROFFITT 851 The Determination of Small Amounts of Iron and Chromium in Sapphire and Ruby for Maser Applications BY R. C. CHIRNSIDE, H. J. CLULEY, R. J. POWELL AND P. M. C. PROFFITT (The General Electric Company Limited, Central Research Laboratories, Hirst Research Centre, Wembley, England) It has been required to know the amount and distribution of para- magnetic ions in synthetic sapphire and ruby crystals used for maser devices. The extremely hard and chemically inert nature of the sample materials creates special analytical problems. Methods, both chemical and spectrographic, have been developed for determining 0.002 to 0.2 per cent. of iron and of 0.01 to 0-5 per cent. of chromium. The methods are applicable to small samples (> 20 mg) so that material from different regions of a crystal can be analysed to assess homogeneity.In the spectrographic method, the crushed sample is mixed with graphite powder containing cobalt oxide as internal standard, and a d.c. arc total- combustion technique is used to determine chromium and iron. The chemical methods involve spectrophotometric determination of chromium with di- phenylcarbazide and of iron with bathophenanthroline, both applied after decomposition of the crushed sample with a sodium carbonate - sodium tetra-borate flux; for determining iron at low levels rigorous purification of reagents i s essential to achieve suitably low blank values. The chemical methods have higher precision, but the spectrographic method is more rapid and is preferred for routine work.Results by the various methods are presented. SAPPHIRES and rubies have long been prized as gemstones. Chemically and crystallographi- cally they are essentially identical, consisting of single-crystal cc-alumina or corundum. Both owe their colour to the presence of other metallic elements as impurities; in ruby, for example, the characteristic red colour is imparted by chromium which, in the form of oxide, is incor- porated in the aluminium oxide structure. In the absence of impurity elements sapphire is a colourless or “white” crystal. Blue, red (ruby) and “white” sapphire have all been made for many years in the labora- tory. Originally the intention was presumably to make available gemstones at a price much lower than the natural variety, but at a later date sapphire and ruby were made to supply industrial needs. Among the most important of these has been material for jewel bearings, manufactured for instruments, watches, etc.This use arises from the fact that corundum is extremely hard and second only to diamond on the conventional Moh’s scale of hardness. Recently many new industrial applications have arisen for most of which the colourless or “white” sapphire is the material required. The most successful and most common method for the production of these synthetic crystals is the flame-fusion process developed by Verneuil about 1904.192 A large single crystal or boule, probably weighing 50 to 70 g, is the first product of the Verneuil process, and this boule can then be cut and worked to give material of the dimensions and shape required for the intended industrial use.For production of the coloured crystals the appro- priate impurity is added to the aluminium oxide powder that serves as the “raw material’’ of the flame-fusion process ; however, much of this additive, together with other accidental impnrities, is driven off at the high temperature attained in the process (about 2000” C). Some degree of inhomogeneity of impurity distribution may in fact occur in the boule single crystal. A recent and most important application of synthetic ruby is in the preparation of maser crystals. The maser is a device for microwave amplification by stimulated emission of radiation, and it has been found that the chemical and crystallographic nature of ruby852 CHIRNSIDE et d.: DETERMINATION OF SMALL AMOUNTS [AnajYSt, Vol.88 is particularly suited to a special version of this device known as the laser, in which the micro- waves are replaced by optical radiation in the visible or near-visible range. One of the problems in maser studies on ruby, and in related work on “white” sapphire, is to investigate the relationship between the microwave spectral data and the nature, amount and distribution of impurities in the crystals. It is desirable for this purpose to have detailed analytical information concerning the concentration and distribution of added and accidental impurities down to the lowest detectable level. Bearing in mind the refractory nature of sapphire and ruby, particularly in the form of a massive piece of single crystal, the analytical problem is considerable.The materials are difficult to decompose except by certain fusion methods, for which the sample must be in a finely crushed form. The extreme hardness of the materials creates special problems in the preparation of an uncontaminated finely crushed sample, particularly for the determination down to low levels of an impurity such as iron. In the sections below are described the analytical investigations by means of which methods have been established for determining chromium, an added impurity, and iron, generally a fortuitous impurity, in single-crystal material. These methods, both chemical and spectrographic, relate to the determination of the average or total amount of chromium or iron in the whole of the sample selected for examination. In addition, an account is given of appropriate methods of sampling used to make a preliminary assessment of the distribution of chromium in ruby boule. THE ANALYTICAL PROBLEM CHROMIUM IN RUBY- Ruby crystals of different chromium contents, representing the range of colours normally used for gems or jewel bearings, are available, and it has been possible to explore the range of chromium contents in which maser action is most prominent.This has shown the need, in work on maser crystals, for analytical methods to cover the range 0.01 to 0.5 per cent. of chromium. To permit analysis of small samples, e.g., in distribution studies, it was required that the method should be applicable to sample weights of about 10 mg, i.e., corresponding to the determination of 1 to 50pg of chromium.Suitable methods of measurement, both chemical and spectrographic, are, of course, available for these amounts of chromium. For the chemical determination the method adopted was decomposition of the crushed sample by fusion and subsequent development of the chromium - diphenylcarbazide colour, which lends itself to precise spectrophotometric measurement. While corrections for the blank value are essential in this method, the amounts of chromium likely to be introduced from reagents, etc., are not sufficiently large to endanger the accuracy of the determination. The real difficulty is associated with the sampling and preparation of material of such considerable hardness and of the geometrical shape of the Verneuil boule; particular care is needed to avoid the introduction of extraneous chromium during the crushing process.IRON IN SAPPHIRE AND RUBY- Most of the analytical interest in iron has been concerned with the determination of the amounts present as an adventitious impurity in “white” sapphire and in ruby; these amounts, which are small compared with the chromium content of ruby, fall in the range 0.002 to 0-01 per cent. Consideration was again given to methods applicable to small weights of sample, of about 20 mg, corresponding to the determination of 0.4 to 2 pg of iron. Some determina- tions have been required on experimental materials having deliberate additions of iron, but even here the levels (up to 0-2 per cent.) have been lower than those mentioned in connection with chromium.In the determination of iron, the problems of sample preparation are more severe than for the chromium determination. The risk of contamination with the element to be deter- mined is much greater, and the levels to be deterrnined are lower. For chemical determination of iron, reagents of sufficient sensitivity are available, but further problems can arise. The necessary fusion stage provides an opportunity for an exchange of iron between melt and crucible material; in addition, reagents of normal analytical purity may bring into the deter- mination amounts of iron which in total may be as much or more than the amount of iron it is required to measure. On the other hand, a direct spectrographic technique enables the analyst to avoid these chemical difficulties, though not those of the sampling process. Both chemical and spectrographic methods have again been used.November, 19631 OF IRON AND CHROMIUM IN SAPPHIRE AND RUBY 853 THE DETERMINATION OF CHROMIUM IN RUBY SAMPLING- One satisfactory method involves the initial reduction of the ruby sample to a coarse powder by thermal shock.Pieces of boule, held in platinum-tipped tongs, are heated t o redness in a flame and dropped into water. The resulting coarse powder is then further crushed in a mortar made from either steel or tungsten of high purity; the crushed material is sifted through a 60-mesh silk or nylon screen for the spectrographic method or through a 120-mesh screen for the chemical method; oversize material is again crushed and sifted.The powdered sample is treated with hydrofluoric and nitric acids to remove any tungsten derived from the mortar, or is boiled for a few minutes with aqua regia if a steel mortar has been used. Finally the acid-washed powder is washed three times by decantation with de-ionised water and dried at 110” C for 30 minutes. When the determination of both iron and chromium is required on the same prepared sample, use of the tungsten mortar is essential (see section on “Determination of Iron”- “Sampling”). When knowledge about the local concentration of chromium is required a different method of sampling is employed. The boule is trepanned at each sampling point, to a depth of 3 to 4 mm, with a tool made from 16 s.w.g. steel tube. This tool is attached to a Mullard ultrasonic 60-watt drill unit; boron carbide (300 mesh) is used as the abrasive. Each cylindrical piece cut out by this technique weighs 7 to 10 mg.Each piece is crushed in a steel percussion mortar, and the resulting powder is sifted, acid-cleaned, washed and dried as described above. CHEMICAL METHOD The method adopted involves fusion of the prepared ruby sample in platinum with a sodium carbonate - sodium tetraborate mixture; this is in our experience the most suitable method of attack on sapphire and ruby materials. The determination is completed spectro- photometrically by a conventional diphenylcarbazide method. REAGENTS- potassium dichromate in water, and dilute to 1 litre. Standard potassium dichromate stock solution-Dissolve 0.707 g of analytical-reagent grade 1 ml = 250 pg of chromium.Standard potassium dichromate working solution-Prepare when required for use by diluting 1 ml of the stock solution to 100 ml with water. 1 ml = 2.5 pg of chromium. Sodium carbonate - sodium tetraboratejux-Mix intimately 10 g of analytical-reagent grade anhydrous sodium carbonate with 5 g of anhydrous sodium tetraborate (prepared by ignition of the hydrated analytical-reagent grade salt). Aluminium potassium sulphate solution-Dissolve 1 g oflanalytical-reagent grade hydrated aluminium potassium sulphate in 100 ml of water. 1 ml = approximately 1 mg of alumina. Sulphuric acid, 0*4~-Dilute 400ml of N sulphuric acid with water to 1 litre. Sul~hur dioxide solution-Prepare a saturated solution of the gas in water. Potassium permanganate, approximately 0.1 N-Dissolve 0.32 g of analytical-reagent grade Sodium azide soldion-Dissolve 1 g of sodium azide in 100 ml of water.Diphenylcarbaxide solution-Dissolve 0.25 g of analytical-reagent grade diphenylcarbazide potassium permanganate in 100ml of water. in 100 ml of analytical-reagent grade acetone. CALIBRATION- In each of six 100-ml beakers place 0-1 g of flux, 10 ml of aluminium potassium sulphate solution and 25 ml of 0.4 N sulphuric acid. To the contents of the separate beakers add from a burette amounts of 0, 1, 2, 3, 4 and 5 ml of the working standard dichromate solution (1 ml = 2.5 pg of chromium). Treat each solution as described below. Prepare fresh each day.854 CHIRNSIDE et at!.: DETERMINATION OF SMALL AMOUNTS [Analyst, VOl. 88 Add 2 drops of sulphur dioxide solution to reduce the dichromate, and boil the solution vigorously for 1 to 2 minutes to remove excess of sulphur dioxide; then add 3 drops of 0.1 N potassium permanganate, and continue boiling for 5 minutes to re-oxidise the chromium to the sexavalent state.Cool somewhat, then add sodium azide solution dropwise to de-colorise the solution. Avoid using a large excess; 5 to 10 drops is usually adequate. Boil the solution for 1 to 2 minutes to remove hydrazoic acid, cool thoroughly, then transfer to a 50-ml calibrated flask. Add 5 ml of diphenylcarbazide solution, dilute to the mark with water, and mix the solution. Allow 5 to 10 minutes to elapse for full colour development, and then measure the optical density of each solution at 540 mp in a 4-cm cell, with water in the reference cell.Deduct the reading for the solution containing no added chromium from the readings for the other solutions. From these results prepare a calibration graph or calculate the mean factor relating corrected optical density to micrograms of chromium. PROCEDURE- On a semi-micro balance weigh out 2 to 10 mg of the prepared sample, depending on the expected chromium content. (For chromium contents up to 0.1 per cent. use approximately 10 mg of sample, for contents of 0.1 to 0.2 per cent. use approximately 5 mg and for contents of 0.2 to 0.5 per cent. use approximately 2 mg of sample.) Transfer the weighed sample portion to a small platinum crucible of about 5 ml capacity, add 0.1 g of flux, and mix throughly with a glass rod.Heat the covered crucible over a Meker burner, gently at first, then strongly for 10 minutes. Swirl the crucible occasionally to aid attack of the sample. Cool, then transfer the crucible and lid to a 100-ml beaker, add 25 ml of 0.4 N sulphuric acid, and heat on a hot-plate until the melt has dissolved; complete dissolution should be obtained. Remove the crucible and lid, rinse them, add the rinsings to the solution, and heat the solution to boiling. Add 3 drops of 0.1 N potassium permanganate, and continue boiling for 5 minutes. Complete the procedure for development and measurement of the chromium colour as described under “Calibration.” Run a blank determination in parallel with the sample determination, and correct for any chromium found. Finally, calculate the chromium content of the sample. SPECTROGRAPHIC METHOD For the spectrographic determination of chromium, the prepared sample is mixed with a carbon-powder spectrographic buffer containing cobalt as the internal standard.The mixture is placed in a cupped graphite electrode, and the chromium is determined by a d.c. arc total-combustion technique. Standardisation is effected by comparison with mixtures prepared from ruby in which the chromium has been determined chemically. The same spectrographic buffer and arc conditions are also used in the determination of iron; if required, chromium and iron can be determined simultaneously, provided that the method of sample preparation has been appropriate for determining iron. For convenience, details of the spectrographic procedure relating to the determination of iron are given separately under “Determination of Iron.’’ In experiments on the spectrographic method, particular attention was paid to assessing the effect of particle size of the sample and to observing the variation with time of the evapora- tion of chromium and cobalt (and also iron).No difference in evaporation behaviour was observed between samples and standards. The use of a sample of different particle size showed that material of less than 60 mesh was satisfactory, no difference being observable with more finely crushed material. APPARATUS- zontal slit adjacent to the plate. circuit voltage at least 120 volts. S$ectrogra$h-Hilger large-quartz spectrograph with a mask containing a 5-mm hori- Excitation source-A d.c.arc source capable of providing a current of 15 amps; open Slit height-2 mm. Slit width-0.015 mm.November, 19631 of the arc on the collimating lens of the spectrograph. OF IRON AND CHROMIUM I N SAPPHIRE AND RUBY 855 External optics-A 30-cm quartz lens, placed adjacent to the slit so as to form an image Wavelength range-3400 to 2450 A. PEate calibrator-A rotating sector disc with a step-to-step ratio of 2 to 1. Densitometer-Hilger non-recording densitometer. Electrodes-Upper electrode (negative) : a 1 inch length of 6.35-mm diameter Ringsdorff Photographic plates-Kodak B. 10. RW-0 graphite rod. Lower electrode (positive) : Ringsdorff RW0078 electrode. REAGENTS- chromium content as determined by the chemical method. cribed under “Sampling.” rods, with 0.28 g of cobalt oxide in an agate mortar. Crushed ruby bode-Ruby boule crushed as described under “Sampling” and of known Crushed sapphire bode--“White” (i.e., chromium-free) sapphire boule crushed as des- Spectrographic bufer-Mix 9.72 g of carbon powder, prepared by turning JMIB carbon Developer-Dilute 1 part of Ilford 1D2 stock solution with 2 parts of water. PREPARATION OF CHROMIUM STANDARDS- Prepare in duplicate a series of standards, covering the range 0.005 t o 0.25 per cent. Weigh out 5.0 mg of crushed, chemically analysed ruby boule, or appropriate portions Mix with of chromium, as described below. of chemically analysed ruby and “white” sapphire to give a total of 5.0mg. 5-0mg of spectrographic buffer in an agate mortar, and transfer to an electrode.Record the spectra of the standards as described below. PROCEDURE- Weigh out 5.0 mg of the prepared sample, mix in an agate mortar with 5.0 mg of spectro- graphic buffer, and transfer to an electrode. Place a pair of electrodes in the arc stand with the filled one in the lower (positive) arm. Touch the electrodes together, switch on the current, separate the electrodes to give a 4-mm gap, and adjust the control rheostat to give a current of 10 amps. Run the arc for 2 minutes, maintaining an electrode separation of 4 mm. To record a calibrating spectrum on the same plate, remove the mask, increase the slit height to 12 mm, place the sector disc in position, and photograph a 5-amp iron arc for 60 seconds. Develop the plate for 3 minutes at 20” C, fix, wash and dry.Measure the transmission of the chromium line at 267243A and the cobalt line a t 2650.27 A, and of the spectral background adjacent to each line. CALCULATION OF RESULTS- Construct a plate calibration curve from the transmission readings from the calibrating spectrum. Convert the transmission readings for the chromium and cobalt lines to relative intensity, and correct the line intensities for background intensities. Calculate the ratio of the chromium and cobalt line intensities, and convert this to percentage chromium by reference to a working curve. The working curve is prepared on log - log paper from the spectra of the standards. RE s u LTS When ruby samples were analysed for chromium both chemically and spectrographically, satisfactory agreement between the methods was obtained.This is illustrated in Table I, which shows values obtained by both methods on samples covering a range of chromium contents.856 [Analyst, Vol. 88 A typical experiment to assess the distribution of chromium in a “flat” ruby boule is illustrated in Fig. 1. The samples for analysis were removed from the flat face of the boule, at the points shown, by the trepanning technique previously described under “Sampling” ; CHIRNSIDE et al.: DETERMINATION OF SMALL AMOUNTS [-- Front view t Side view Fig. 1. Distribution of chromium in a flat ruby boule. Percentage of chromium found at sampling points: A, 0.021 ; R, 0.021 ; C, 0.021; D, 0.017; E, 0.016; F, 0.016 the chromium contents were determined chemically. As shown in Fig. 1, the results obtained covered the range 0.016 to 0-021 per cent.of chromium, although the alumina powder from which the boule was grown contained an addition corresponding to 0.1 per cent. of chromium. This ihstrates the large loss of additive that can occur in the flame-fusion process. TABLE I COMPARISON OF CHEMICAL AND SPECTROGRAPHIC RESULTS FOR CHROMIUM I N RUBY Chromium content found by- r A -l Sample number Chemical method, yo Spectrographic method, yo 26777 ,. . . 0.015 0.016 26707 . . .. 0.058 0.053 27163 .. .. 0.10 0.11 27154 .. . . 0.20 0.19 THE DETERMINATION OF IRON IN SAPPHIRE-AND RUBY SAMPLING- Extreme care is required to avoid the introduction of extraneous iron during the prepara- tion of the sample. Heat the sample piece, held in platinum-tipped tongs, to red heat in a flame, and then drop into “pure” water to shatter the sample by thermal shock.Pour off the water, dry the sample at 110” C, and then transfer to a mortar made from tungsten of high purity and provided with a well-fitting tungsten rod as a pestle. Fit a polythene cap to the upper end of the pestle, and tap this gently with a hammer to crush the sample. Sift the fine material through a 60-mesh silk or nylon screen for the spectrographic method or through a 120-mesh screen for the chemical method, Return the oversize material to the mortar, and again crush so that all the material passes through the screen. Transfer the sifted powder to a clean platinum crucible, add 2 ml of hydrofluoric acid and 2 ml of nitric acid, and warm the mixture for about 5 minutes to dissolve any tungsten acquired in the crushing.Dilute with water, allow the powder to settle, and carefully decant off the liquid. Wash the powder with water by decantation several times, dry at 110” C, and store the dried powder in a clean capped specimen-tube. The method described below was found to be satisfactory. SPECTROGRAPHIC METHOD In this particular context, a spectrographic determination offers great advantages over a chemical method. The relatively high blank values for iron resulting from reagents of normal purity (and the consequent need in a chemical method to purify the reagents) areNovember, 19631 OF IRON AND CHROMIUM I N SAPPHIRE -4ND RUBY 857 eliminated by the use of a direct spectrographic technique. It was established by experiment that the presence of chromium had no effect on the iron spectrum, so that the same technique was applicable to the determination of iron in colourless sapphire and in ruby.Although it has been shown that the iron content of ruby and sapphire is generally in the region of 0.005 per cent., the method described below has been devised for iron contents up to 0.2 per cent.; this has permitted the analysis of experimental material containing deliberately added iron. As stated earlier, the spectrographic buffer and arc conditions used are the same as for the chromium determination, so that iron and chromium can be determined simultaneously if required. APPARATUS AND REAGENTS- The apparatus, spectrographic buffer, electrodes, photographic plates and developer are as described for the determination of chromium. Other reagents required specifically for the determination of iron are listed below.Crushed bode-Ruby or sapphire bode of low iron content, crushed as described under “Sampling.” Aluminium ammonium sulphate, hydrated-Analytical-reagent grade. Ammonium ferric sulphate, hydrated-Analytical-reagent grade. Ammonia solution, s9.p. 0-88-Analytical-reagent grade. Wash solution-A 2 per cent. w/v solution of ammonium nitrate made just alkaline to methyl red by adding ammonia solution. PREPARATION OF STOCK ALUMINA - IRON MIXTURES FOR IRON STANDARDS- Prepare two stock mixtures of alumina, containing 0.1 and 1 per cent. of iron, as described below. For the 0-1 per cent. mixture, weigh out and transfer to a 250-ml beaker 848g of aluminium ammonium sulphate and 0-0086 g of ammonium ferric sulphate.For the 1 per cent. mixture weigh out and transfer to a second beaker 8.80 g of aluminium ammonium sulphate and 0.0861 g of ammonium ferric sulphate. Add 2 or 3 drops of methyl red indicator, heat the solution to boiling, and add ammonia solution until the indicator just turns yellow. Set the solution aside on a water-bath for 10 minutes. Filter off the precipitate on a Whatman No. 41 or equivalent filter-paper, and wash the precipitate with hot wash solution. Transfer the paper and precipitate to a silica crucible, dry, and then ignite at a low heat until all carbonaceous matter has been removed. Finally, heat the crucible and contents in a muffle furnace at 1000” C for 1 hour. Cool, transfer the mixture to an agate mortar, and grind to a fine powder.PREPARATION OF IRON STANDARDS- as described below. specified in Table 11, and mix in an agate mortar. mortar, and transfer the mixture to an electrode. TABLE I1 WEIGHTS OF ALUMINA - IRON MIXTURES FOR SPECTROGRAPHIC Dissolve the salts in water, and dilute to about 150ml. Prepare in duplicate a series of standards, covering the range up to 0.2 per cent. of iron, Weigh out 5.0 mg of spectrographic buffer, add the weight of alumina - iron mixture Add 5-0 mg of crushed boule, mix in the Record the spectra of the standards as described below. IRON STANDARDS Iron content of standard, Blank 0.006 0.02 0.06 0.20 Weight of alumina - iron mixture added - % 0.30 mg of O.lyo mixture 1-00 mg of O.lyo mixture 0.30 mg of 1% mixture 1.00 mg of 1% mixture PROCEDURE- graphic buffer, and transfer to an electrode.Weigh out 5-0 mg of the prepared sample, mix in an agate mortar with 5-0 mg of spectro-858 the same plate, exactly as described for chromium. fix, wash, and dry. line at 2590*59w, and of the spectral background adjacent to each line. CALCULATION OF RESULTS- Construct a plate calibration curve from the transmission readings for the calibrating spectrum. Convert the transmission readings for the iron and cobalt lines to relative intensity and correct the line intensities for background intensities. Calculate the ratio of iron and cobalt line intensities and convert this to percentage iron by reference to a working curve. The working curve is prepared on log - log paper from the spectra of the standards.For this purpose deduce from the spectra, by the method of standard addition^,^ the iron content of the crushed boule used in the standards, and correct the nominal iron content of each standard accordingly. CHEMICAL METHOD As described above, the preparation of standards in the spectrographic technique involves the mixing of solid materials, a crushed sapphire or ruby being “doped” with alumina - iron mixtures of known iron content. Excellent results over a range of iron contents, e.g., with materials containing deliberate additions of iron, were obtained by the use of these synthetic standards. However, a possible doubt attached to the spectrographic determination of iron at the lowest levels, e.g., in the purest undoped materials, as this necessarily involved an extrapolation of the working curve. For this and other reasons, a considerable effort was put into the development of a chemical method suitable for determining iron at the lowest levels.Spectrophotometric determination with bath~phenanthroline~ was used ; this is the most sensitive and selective reagent available for iron, and there was no difficulty in achieving the required sensitivity. In the development of the method, the two aspects requiring most consideration (apart from the sampling problem already discussed) were the method of decomposition and the need to attain adequately low blank values. These two points are discussed in more detail below. CHIRNSIDE et al. : DETERMINATION OF SMALL AMOUNTS [Analyst, Vol. 88 Excite the sample, and record the spectrum: also record a calibrating iron spectrum on Develop the plate for 3 minutes at 20” C, Measure the transmission of the iron lines at 2611-87 and 2621.68 A and of the cobalt DECOMPOSITION- As in the determination of chromium, the most suitable method of decomposition involves a fusion process for attack of the sapphire or ruby.Although there are no special problems with chromium, the choice of flux and of crucible material are all-important in the deter- mination of iron. In particular, there is a need to minimise the possibility of exchange of iron between crucible and melt [and of attack on the crucible), and it is desirable to achieve complete decomposition in the shortest possible time. After an investigation of various fluxes and crucible materials, the process finally adopted involved “sintering” in platinum at 1050” C with a small portion of flux consisting of anhydrous sodium carbonate and an- hydrous sodium tetraborate in the proportion 2 to 1 by weight. Only 50 mg of flux and 20 mg of sample were used, and the mix was heated in a platinum crucible in an electric muffle furnace.The “sinter” process was introduced by Finn and Klekotka of the National Bureau of Standards in 1930,5 and this technique has been in use in these laboratories for many years for decomposing glass and refractory materials. In the present context, the sinter technique has several advantages- (;) the process is quick; complete attack of the sample is obtained within 10 minutes’ heating ; (ii) the small amount of flux reduces the blank value arising from iron introduced from this source; <;;i) both constituents of the flux are readily re-crystallised and therefore potentially amenable to purification ; (iv) while it is known that fusions in platinum can give rise to exchange of iron between melt and crucible, the use of a sinter rather than a fluid melt minimises the extent of this exchange process.November, 19631 859 It has nevertheless been thought an essential precaution to do one or more blank fusions in the crucible immediately before and after each determination of iron.If results at the customary blank level are obtained, this provides an assurance that exchange of iron with the crucible has not occurred in the determination. As is our normal practice when deter- mining small amounts of iron, a platinum crucible was reserved solely for the work on sapphire and ruby.BLANK VALUES- In trace analysis, the contribution from reagents and apparatus of the element to be determined should ideally be appreciably less than the amount in the sample. It may be desirable, as in our work, to adopt an analytical process permitting the use of reagents that either are available in a high state of purity or readily lend themselves to purification. In the method devised for determining iron, the reagents used are: sodium carbonate - sodium tetraborate flux, hydrochloric acid, a combined reducing and buffer solution containing hydroxylammonium chloride and sodium acetate, bathophenanthroline solution and chloro- form.In initial work on this method, only the reducing - buffer solution was purified (by development and extraction of the iron - bathophenanthroline colour). The blank values were at the level of 0-7 to 0.8 pg of iron, i.e., similar to and sometimes greater than the amount of iron in a sample. Subsequently, the blank value was substantially reduced by decreasing the amount of bathophenanthroline used, by employing distilled hydrochloric acid and by purifying the constituents of the flux. Consistent blank values at the level of 0.2 pg of iron were then obtained, an amount equivalent to only a half of the smallest amount of iron determined in a sample. For purification of the constituents of the flux, use was made of the reagent phenyl- 2-pyridyl ketoxime, which has the unusual property of forming an extractable complex with ferrous iron in strongly alkaline s o l ~ t i o n s .~ ,7 Separate strong aqueous solutions of sodium carbonate and of sodium tetraborate were prepared, iron was extracted as the phenyl-2-pyridyl ketoxime complex, and the sodium carbonate and tetraborate were re-crystallised from solution. Purification in this manner was more successful with sodium tetraborate than with sodium carbonate, a possible explanation being the presence of iron in an insoluble form in the sodium carbonate. Better results with the sodium carbonate were achieved by introducing a filtration step, with titanium to serve as a “collector,” before the extraction. Experimental details of the chemical method for determining iron in sapphire and ruby are given below.PREPARATION AND PURIFICATION OF REAGENTS- Water-Use distilled water further purified by passage through mixed-bed ion-exchange resin. Hydrochloric acid-Purify by isopiestic distillations as described below. Clean a 10-inch vacuum desiccator with hydrochloric acid, dry the desiccator, and remove any lubricant. Into the base of the desiccator pour 500ml of analytical-reagent grade hydrochloric acid, sp.gr. 1.18. To a polythene or polystyrene container add 100 ml of water, and support the container on glass rods above the level of the acid; close the desiccator with the lid. At daily intervals withdraw a small aliquot from the solution in the plastic container, making use of the hole in the desiccator lid, and by titration determine the strength of acid.Continue the distillation until acid of 8 N concentration is obtained (2 to 3 days). Store the distilled acid in a glass-stoppered bottle. Bathophenanthroline solution-Dissolve 0-0668 g of bathophenanthroline (4,7-diphenyl- 1,lO-phenanthroline) in 100 ml of analytical-reagent grade ethanol, and dilute to 200 ml with water. OF IRON AND CHROMIUM IN SAPPHIRE AND RUBY Store in a polythene container. Store in a glass-stoppered bottle. Chloroform-Analytical-reagent grade. Reducing - bufler solution-Dissolve 20 g of analytical-reagent grade sodium acetate trihydrate and 2.0 g of analytical-reagent grade hydroxylammonium chloride in water, transfer to a separating funnel, and dilute to 200 ml. Add 2 ml of bathophenanthroline solution, mix, and set aside for 1 minute.Add 10ml of chloroform, shake vigorously for 1 minute, allow the layers to settle, and run off the chloroform layer. Repeat the extraction, adding further bathophenanthroline solution, if necessary, until a colourless extract is obtained. Store the purified solution in a glass-stoppered bottle.860 CHIRNSIDE et at?.: DETERMINATION OF SMALL AMOUNTS [Analyst, Vol. 88 Sodium carbonate-Dissolve 10 g of analytical-reagent grade anhydrous sodium carbonate in 50 ml of hot water, and add a solution containing the equivalent of about 5 mg of titanium dioxide (prepared by dissolving 25 mg of potassium titanyl oxalate in a few drops of sulphuric acid (1 $- I), heating to fumes, then cooling and diluting with about 5 ml of water). Digest the solution at about 70" C for 30 minutes, and then filter through a coarse paper, without washing the precipitate.Re-heat the filtrate to about 70" C, add 0.2 g of sodium dithionite and 5 ml of phenyl-2-pyridyl ketoxime solution (0.2 g dissolved in 100 ml of 0.1 N hydro- chloric acid), and keep the solution hot for about 5 minutes. Cool the solution, transfer to a separating funnel, and extract with successive 10-ml portions of chloroform until a colourless extract is obtained (2 or 3 extractions should suffice). Run the aqueous solution into a beaker, heat to 50"C, add 50ml of ethanol, and cool. Pour off the alcohol layer, again heat to 50" C, add 25 ml of ethanol, cool, and discard the alcohol layer; if necessary, repeat the heating, the addition of 25 ml of ethanol and the cooling to obtain crystallisation of sodium carbonate from the aqueous solution. Filter off the crystals on a No.4 grade sintered- glass crucible, and wash once with ethanol. Transfer the crystals to a clean platinum crucible, dry at 110" C then heat at 250" C to convert the precipitated monohydrate into anhydrous sodium carbonate. Sodium tetraborate-Dissolve 20 g of analytical-reagent grade sodium tetraborate in 100 ml of hot water. Add 0.2 g of sodium dithionite and 5 ml of phenyl-2-pyridyl ketoxime solution, and maintain the solution at about 70" C for 5 minutes. Cool the solution to about 40" C, transfer to a separating funnel, and extract the solution with two successive 10-ml portions of chloroform, rejecting the extracts; complete the extractions quickly so as to minimise risk of crystallisation of the sodium tetraborate.Pour the aqueous solution into a beaker, and cool the solution, stirring during the cooling to prevent the crystals adhering to the walls of the beaker. Filter off the crystals on a No. 4 sintered-glass crucible, and wash once with ethanol. Transfer the crystals to a clean platinum crucible, dry at 110" C, and then heat at 300" C to dehydrate the sodium tetraborate. Flux-Mix purified sodium carbonate and purified sodium tetraborate in the proportions 2 to 1 by weight. Standard iron stock solution-Dissolve 0.100 g of pure iron in a few millilitres of dilute sulphuric acid (1 + 6), and dilute with water to 1 litre. 1 ml = 100 pg of iron. Store in a screw-capped bottle. Store in a screw-capped bottle.Store in a screw-capped bottle. Standard iron working solution-Prepare when required for use by diluting 1 ml of the stock solution to 100 ml with water. CALIBRATION- To each of four dry 50-ml separating funnels add 2 ml of water, 0.2 ml of hydrochloric acid and 50mg of flux. To the solutions in three of the funnels add 0.5, 1.0- and 2-0-mI portions of the working standard iron solution, corresponding to 0.5, 1.0 and 2.0 pg of iron. Dilute all four solutions to 5 ml with water, and treat each as described below. Add 5 ml of reducing - buffer solution (the amounts of reagents used give a pH of 4.7 to 4.9 at this stage). Add 2 ml of bathophenanthroline solution, mix and set aside for 1 minute. Add from a pipette 10 ml of chloroform, shake vigorously for 1 minute, and allow the layers to settle.Run off the chloroform layer through a small roll of filter-paper in the stem of the funnel (to absorb any entrained water) into a 4-cm cell. With chloroform in the reference cell, measure the optical density of the extract at 535mp; use a suitable mask if the volume of solution is insufficient to fill the light path of the spectrophotometer. From the optical densities of the solutions with added iron subtract the optical density of the solution with no added iron. From these results prepare a calibration graph or calculate the mean factor relating corrected optical density to micrograms of iron. PROCEDURE- of flux. a t 1050" C, taking care to handle only the outside of the crucible. remove the crucible, and allow it to cool. 1 ml = 1 pg of iron.To a semi-micro platinum crucible, specially reserved for the determination, add 50 mg Cover the crucible and with platinum-tipped tongs transfer it to a muffle furnace Heat for 10 minutes,Kovember, 19631 861 Add 2 ml of water and 0-2 ml of hydrochloric acid, and gently heat the covered crucible on a hot-plate for a few minutes to dissolve the fused mass. Allow the crucible to cool, transfer the solution to a dry 50-ml separating funnel, and dilute with water to 5 ml, Add 5 ml of the reducing - buffer solution and 2 ml of bathophenanthroline solution, set aside for 1 minute, and then proceed with the extraction and spectrophotometric measurement of the extract as described under “Calibration.” If the amount of iron found is at the normal level of the blank value, proceed with a determination.If an excessive amount of iron is obtained, or if a crucible is being used for the first time, repeat the blank fusion as above until the amount of iron found is at a suitably low level. (With reagents purified as described, the amount of iron customarily found in blank fusions was 0.2 & 0.05 pg of iron.) For the determination, add 50mg of flux and 20mg of the prepared sample to the crucible, and mix with a clean glass rod. Cover the crucible, and heat for 10 minutes at 1050” C in a muffle furnace. Allow the crucible to cool, add 2 ml of water and 0.2 ml of hydrochloric acid, and gently heat the covered crucible on a hot-plate until dissolution of the sintered mass is complete; heating for 15 to 20 minutes may be required.Cool the cruci- ble, transfer the solution to a separating funnel, and proceed with the development, extraction and measurement of the iron colour as previously described. Calculate the total amount of iron found. Without delay carry out a blank fusion in the crucible and determine the amount of iron obtained. If this amount is at the customary blank level, calculate the mean blank value from the amounts of iron found for the blank fusions made immediately before and after the determination on the sample. Subtract this mean blank value from the total amount of iron found in the determination on the sample, and hence calculate the percentage of iron in the sample. Because of the small a.mounts of iron to be determined, constant precautions are needed to avoid contamination of solutions and apparatus from airborne dust, etc.OF IRON AND CHROMIUM IN SAPPHIRE AND RUBY Calculate the amount of iron found. RESULTS To illustrate levels of the blank value and the actual amounts of iron commonly deter- mined, a chemical determination is shown in detail in Table 111. As will be seen from Table 111, the blank fusions before and after the determination on the sample gave amounts of iron at the customary blank level of 0.2 pg of iron, thus providing an assurance that no significant exchange of iron between crucible and melt had occurred in the intervening determination. The material analysed in this particular determination was sapphire grown in these laboratories. TABLE I11 DETAILS OF A DETERMINATION OF IRON IN “WHITE” SAPPHIRE Operation Optical density Iron found, CLg of extract Blank fusion before determination .. . . 0.033 0.23 Determination on 20-mg sample . . . . . . 0.080 0-55 Blank fusion after determination . . . . . . 0.028 0.19 Mean blank value = 0.21 pg of iron Iron in sample = 0.65 - 0.21 = 0.34 pg = 0.0020,a In Table IV are shown results obtained for iron by both chemical and spectrographic methods on a range of samples. The main interest in determining iron has been concerned with sapphire and ruby in which iron is present only as an adventitious impurity, and most of the results given in Table IV relate to this type of material (iron content generally 0.002 to 0.005 per cent.). Satisfactory agreement between the two methods was obtained at these levels and the chemical method, although relatively lengthy, has thus provided independent evidence that the spectrographic technique is applicable at these low levels of iron content.An independent check by neutron-activation methods on two of the samples listed in Table IV, was arranged through Dr. C. A. Parker of the Admiralty Materials Laboratory. Results for iron were obtained by direct gamma-ray spectrometry on the irradiated sample and also by a more rigorous method in which the iron was separated chemically before be- ing counted. The respective results by these two techniques were 0.004 and 0.001 per cent.862 [Analyst, Vol. 88 of iron on sapphire No. 28171 (for comparison with the figures given in Table IV, i.e., mean values of 0.002 per cent. chemically and 0.002 per cent.spectrographically) and 0-002 and 0.001 per cent. on ruby No. 27154 (for comparison with figures of 0.004 per cent. chemically and 0.0045 per cent. spectrographically in Table IV). These results by activation methods, though not in precise agreement with our own values, do help to establish the general level of iron content in the sapphire and ruby materials. It should be added that in the determination of chromium (a much less exacting problem), excellent agreement was obtained between results at the Admiralty Materials Laboratory and those reported in Table I. On ruby No. 27154 results of 0.20 and 0.20 per cent. of chromium were obtained by gamma-ray spectrometry on the irradiated sample, compared with values of 0.20 and 0.19 per cent. of chromium by chemical and spectrographic methods, respectively.To illustrate determination at higher levels, Table IV includes some results on “doped” materials, i.e., with iron deliberately added during their preparation. Again, satisfactory CHIRNSIDE et al. : DETERMINATION OF SMALL AMOUNTS TABLE IV COMPARISON OF RESULTS BY CHEMICAL AND SPECTROGRAPHIC METHODS FOR IRON IN SAPPHIRE AND RUBY Iron content found by- - -7 Sample Sample number spectrographic method, chemical method, % O / /O “Pure” materds- Sapphire . . .. .. 28171 0.0025, 0.002, 0.002, 0*0015 0.002, 0.002 Sapphire . . .. .. 27309 0.004, 0.004 0.003, 0.003, 0.002 Sapphire . . .. .. 27308 0.003, 0.003, 0.005 0-004, 0.003, 0.003 Ruby .. .. . . 26707 0.004, 0.004, 0.004, 0.006 0.004 Ruby .. .. .. 27154 0.0045 (mean value) 0.004 ,%faterials “doped” with ivon- Ruby .... .. 26777 Corundum . . .. 25163 0.10 0.09, 0.09 0.19, 0.22 0-18 agreement between chemical and spectrographic methods was obtained. At these higher levels, the problems of the chemical determination of iron are greatly simplified; purification of reagents becomes unnecessary and a less sensitive reagent, e.g., thioglycollic acid, can be used if required. DISCUSSION In the analysis of sapphire and ruby by chemical or spectrographic methods, the process of sample preparation is a critical part of the procedure, particularly in the determination of iron. In the investigation of various sampling techniques, spectrographic examination has been of value in indicating the nature and degree of any contamination that might arise a t the various stages in the sampling process.We are satisfied that the techniques adopted will yield prepared sample material essentially uncontaminated with the element to be determined. In the preceding descriptions of the analytical methods, the chemical method has been given first in the section on chromium, and the spectrographic method first in the section on iron; the order chosen is both logical and chronological. For chromium in ruby, the chemical method was developed first, and chemically analysed ruby was subsequently used as a means of calibrating the spectrographic method for chromium. With the determination of iron, the obvious difficulties of a chemical method encouraged the initial development of a spectro- graphic method; this in turn led to the development of a chemical method as a means of assessing the validity of the spectrographic method at low levels of iron content.The work has shown the spectrographic technique to be applicable over an appreciable range of chromium and iron contents, and this technique has been the preferred method for routine work. Once the calibrations have been completed the procedure is rapid; iron and chromium can be determined simultaneously, and the sample preparation is a little less exacting in that material of less than 60 mesh is required instead of less than 120 mesh for the chemical methods.November, 19631 863 The chemical method for iron has served its immediate purpose of verifying the validity of the spectrographic technique at low levels of iron content and of helping to establish the levels of adventitious iron present in sapphire and ruby.I t requires considerable care and is relatively time-consuming, and is now used only for the occasional confirmatory test, On the other hand, the chemical method for chromium has been of continuing value, because it has a higher precision than the spectrographic method. It has been used, for example, in establishing the existence of small differences in chromium contents in distribution studies of the type illustrated in Fig. 1. The analytical techniques described in this paper have fulfilled exploratory requirements to establish the levels of chromium in ruby and of iron in sapphire and ruby, and to provide some information on their distribution. Ultimately, the ideal analytical technique would be a non-destructive test capable of giving point-to-point information on the distribution of “impurities” in the actual crystal to be used in a maser device. Preliminary work on the determination of chromium suggests that both X-ray fluorescence and optical absorption spectrophotometry may be of some value in this connection; with X-ray fluorescence, how- ever, limitations may arise on sensitivity and from the fact that only the composition of material near the surface is assessed. When the work described in this paper was essentially complete, a method for analysing sapphire and ruby maser crystals, involving a rather different approach from our own, was reported by Dod~on.~ In this method up to 20mg of the crushed sample is decomposed by fusion for 2 to 2+ hours with about 1 g of potassium hydroxide in a zirconium crucible. Chromium is determined spectrophotometrically with diphenylcarbazide on an aliquot of the solution of the melt. On a separate aliquot iron is determined with 2,2’-dipyridyl; because of the pale colours obtained even with concentrated solutions, the iron colour is developed either in a small volume of solution in the depression of a white tile or on a small disc of filter-paper impregnated with the reagent, and the iron is assessed by visual comparison with standards similarly prepared. We thank Mr. H. B. Clarke and the Superintendent of the Admiralty Materials Labora- tory, Holton Heath, for the work on determination of iron and chromium by radioactivation, and for permission to quote the results obtained. REFERENCES OF IRON AND CHROMIUM IN SAPPHIRE AND RUBY 1. 2. 3. 4. 6. 6. 7. 8. 9. Dodson, E. M., Anal. Chem., 1962, 34, 966. Verneuil, M. A,, Ann. Chim. Phys., 1904, 3, 20. Brown, K. W., Chirnside, R. C., Dauncey, L. A., and Rooksby, H. P., G.E.C. Journal, 1944, Duffendack, 0. S., and Wolfe, R. A., Ind. Eng. Chem., Anal. Ed., 1938, 10, 161. Smith, G. F., McCurdy, W. H., jun., and Diehl, H., Analyst, 1952, 77, 418. Finn, A. N., and Klekotka, J. F., J . Bes. Nat. Bur. Stand., 1930, 4, 809. Trusell, F., and Diehl, H., Anal. Chem., 1959, 31, 1978. Cluley, H. J., and Newman, E. J., Analyst, 1963, 88, 3. Irving, H., and Cox, J. J., Ibid. 1958, 83, 526. 13 (2), 2. Received July 15th, 1963

ISSN:0003-2654    

DOI:10.1039/AN9638800851

出版商:RSC    

年代:1963

数据来源: RSC

摘要:

864 TRACE MATERIALS (COLOURS) COMMITTEE : EXTRACTION AND [Analyst, Vol. 88 The Extraction and Identification of Permitted Food Colouring Matters with Special Reference to the Changes Undergone During Processing and Subsequent Storage FIRST REPORT OF THE TRACE MATERIALS (COLOURS) COMMITTEE SET UP BY THE BRITISH FOOD MANUFACTURING IN’DUSTRIES RESEARCH ASSOCIATION THE Trace Materials (Colours) Committee of the British Food Manufacturing Industries Research Association was set up in April 1959 under the Chairmanship of Dr. A. A. Houghton with the following terms of reference- “The Committee exists to examine any problems which may arise in the use of coal- tar colouring matters in foodstuffs as a result of the Colouring Matters in Foods Regulations. The principal interest is in the effects of processing the colours in foodstuffs and of storage of the finished products.Particularly in the formation of new coloured products and the changing of the proportion of subsidiary dyes. Inevitably this involves the consideration of the best processes for isolating colours from foods, identifying them and estimating their concentration. Changes in the percentage of subsidiary dyes are important in view of the standards now being laid down by the British Standards Institution.” The present working membership of the Committee is-Mr. A. G. Sansome (Edward Sharp and Sons Ltd.), Chairman; Mr. N. R. Jones (B.F.M.I.R.A.), Secretary; Dr. D. Dickinson, assisted by Mr. T. W. Raven (Fruit and Vegetable Canning and Quick Freezing Research Association), Messrs. K.J. Gardner (Mars Ltd.), R. A. Knight (British Baking Industries Research Association), F. J. Lipscomb (John Mackintosh and Sons Ltd.), R. C. Spalding (Kent County Council Public Analysts Department) and E. F. Williams, assisted by H. C. Hornsey (J. Sainsbury Ltd.). Dr. A. A. Houghton (Mars Ltd.) and Dr. A. G. Lipscomb (John Mackintosh and Sons Ltd.) previously served on the Committee as Chairmen, Changes undergone by food colours during processing and subsequent storage are important from two major aspects: (1) Fading or undesirable changes in shade may be such as to affect the acceptability to the consumer if the desired appearance cannot be achieved or is lost during storage. (2) The Colouring Matters in Foods Regulations permit only certain colours to be used.If the colours undergo changes resulting in the production of new coloured substances, it might be concluded that non-permitted colours had been added to the food in the first place. Also, should the situation eventually arise wherein food colours are required to conform to the appropriate British Standards (as is envisaged), it might happen that foodstuffs would be examined to ascertain whether or not the colouring matters in them were acceptable. In particular, the proportion of subsidiary dyes might be determined, and, therefore, changes in the proportions of the main dyes and any subsidiaries are of interest, as well as the appear- ance of dyes not present in the original colouring matter. To obtain this information, it was considered vital that the methods used should remove all the colour from the foodstuff, and permit it to be recovered for subsequent examination.It was considered to be particularly important that the procedures involved should avoid any severe heat treatment or the use of strong acids or alkalis, as any changes in the dyes during extraction would invalidate the conclusions drawn. The traditional wool-dyeing technique was thought to be undesirable, as repeated boiling is required to obtain complete recovery of the dyes.November, 19631 IDEKTIFICATION OF PERMITTED FOOD COLOURING MATTERS 865 METHODS Only the methods that have been employed in the work of the Committee will be described All methods may be divided into three separate parts- here, and only water-soluble dyes have been investigated.(a) (b) (c) transfer of the dyes from the foodstuff into solution, separation of the dye from extraneous substances and separation of individual dyes and identification. These parts can be regarded as unit operations that can be combined together to suit The Committee has examined the following products as representative of different types : So far, these products have been coloured with the food colours listed below, but for a few particular products and particular colouring matters. boiled sweets, toffees, fondant, madeira cake, fresh sausages and fish paste. of these products the whole range of colours was not used- Tartrazine Red 10B Orange G Sunset Yellou- Red 2G Yellow 2G Amaranth Ponceau NIX Blue VRS Ponceau 4K Ponceau SX Indigo Carmine Carmoisine Red 6B Erythrosine Green S Black PN In addition, Tartrazine was processed in canned peas.The choice of dyestuffs was conditioned by the setting up of specifications by the British Standards Institution (B.S.I.), which gave standards of purity for the dyestuffs and suggested suitable solvent systems for separating subsidiary dyes from the principal colouring matters by paper chromatography. The dyes used were all of good commercial foodstuffs quality, and, where B.S. I. specifications had been published, they conformed to these specifications. TRANSFER OF DYES FROM THE FOODSTCFFS IXTO SOLUTION Methods for extracting dyes from foods are described in the Association of Public Analysts (A.P.A.) Handbook1 and in other places. These methods have proved quite satisfactory in practice, and it is necessary to give only a brief indication of the method used for most products.Boiled Sweets and fondants-Direct dissolution in water was sufficient. Tofees- A solution in water was made just alkaline with ammonia solution and filtered by suction, a large excess of filter-aid (e.g., Hyflo Supercel) being used. Madeira cake-Air-dried, and de-fatted with light petroleum before extracting the colour by soaking for several hours or overnight with cold 50 per cent. aqueous alcohol containing 1 per cent of ammonia solution, spgr. 0.88. Solutions were clarified by centrifugation, and the extractions were repeated when necessary. The final extracts were combined and concentrated. Sausages-Extracted in ammoniacal 50 per cent. aqueous alcohol (20g of sausages in 14 ml of water, 25 ml of ethanol and 1 ml of ammonia solution, sp.gr. 0.88) by setting the mixture aside in the cold for half an hour. The solution was then filtered and concentrated. Fish paste-Extracted in ammoniacal 50 per cent. acetone (20 g of fish paste in 6 ml of water, 20 ml of acetone and 1 drop of ammonia solution, sp.gr. 0.88). After having been stirred for 10 minutes the solution was spun in a centrifuge and the acetone boiled off. Further extractions of the residues were often required to remove the whole of the colour. Fat was then removed by shaking with light petroleum. Canned Peas-The samples were macerated in their own liquor. Colour was extracted directly from this macerate by a solvent. In all these methods heating was kept to a minimum to avoid changes in the dyestuffs.In particular, the solutions should be concentrated under reduced pressure. Some of the methods used eliminated the need for concentration by evaporation. SEPARATION OF DYES FROM EXTRANEOUS SUBSTANCES The principles involved were those of solvent extraction, ion exchange and column chromatography. None of these methods is specific for synthetic dyes, nor is the wool- dyeing technique. I t is always desirable to purify the first extract, either by repeating the process used or by using a different principle. The latter would seem most desirable as the impurities carried over in, for example, solvent extraction may not be carried over in a different process, such as ion exchange.866 SOLVENT EXTRACTION- solvents investigated.conditions and removed from the solvent by weak ammonia solution. conditions were those for canned peas- TRACE MATERIALS (COLOURS) COMMITTEE : EXTRACTION .4ND [Analyst, Vol. St3 The higher alcohols, particularly isobutanol and n-butanol, were the most useful of the Dyes were normally extracted from aqueous solution under acid Typical working Macerate the peas with their own liquor. To 25 ml in a 50-ml stoppered cylinder add 2 ml of 25 per cent. w/v sulphuric acid. Warm to 50 to 60" C in a water-bath, add 15 ml of isobutanol, and shake vigorously. Repeat until extracts contain no further colours. The colour can be removed from this extract by direct chromatography on an alumina column or can be extracted into dilute ammonia solution. In the latter instance, ammonium sulphate will be present so that further purification by another method is desirable.Similar methods were used successfully on sausages and madeira cake (prepared as described above) and, by using n-butanol, on boiled sweets and fish paste. A method involving solvent extraction after formation of a complex with a quaternary ammonium compound was described by Drevon and Laur.2 This method had the dis- advantage that the complex needed to be broken down by the use of strong acids and high temperatures. A modified method, however, has been found successful, although it has not yet been tried with all the colours examined. In this modified method a cetylpyridinium salt is used to form the complex, which is then broken down in the cold by careful neutralisation with a dilute solution of Teepol, e g ., the procedure for toffees- Make the coloured syrup alkaline with ammonia solution, add Supercel, and filter; 25 m1 of syrup should contain approximately 0.2 mg of colour. Adjust the pH to 9, add 10 m1 of 0.1 per cent. cetylpyridinium chloride solution, and extract with 10 ml of chloroform. Repeat with a second 10-ml portion of chloroform, which should leave the aqueous phase colourless. Shake the chloroform extracts with successive 5-ml portions of 0.1 per cent. Teepol solution, rejecting the aqueous fractions until some colour is extracted. Retain this fraction in contact with the chloroform, and add 1 per cent. Teepol solution dropwise, shaking between additions, until no colour remains in the chloroform layer. Reject the chloroform.The aqueous layer contains a precipitate. Remove this precipitate by extracting twice with 10 ml of light petroleum (boiling-range, 40" to 60" C). Concentrate the clear aqueous dye solution under reduced pressure. With some colours, e.g., Yellow 2G, the colour complex formed a solid mass at the water- This could be overcome by gentle extraction with n-butanol instead Decant off the isobutanol extract. Combine the extracts. chloroform interface. of chloroform. ION EXCHAXGE- In the ion-exchange methods employed, aminoethylcellulose in the form of Whatman's AE.50 floc or powder or AE.30 paper was used. This is a weakly basic anion-exchange material that absorbs acid dyestuffs from acid solutions (at a pH below 4 to 5). After the aminoethylcellulose had been thoroughly washed, the colour was stripped with dilute ammonia solution. Hence this material acted in a manner similar to that of wool, but required no heating.It was very useful for absorbing the colour from large volumes of solution, thus achieving concentration without the use of heat. The floc and powder were used in the form of short columns (say 5 cm long, 1 to 1.5 cm in diameter), the powder being mixed with about 3 parts of Kieselguhr to improve its porosity. Typical working instructions were those for boiled sweets- Dissolve about 20 g of sweets in cold water, and, if not already acid, acidify with acetic acid; pass the solution through a I-inch column of Whatman's AE.50 floc. Wash the floc well with extremely dilute acetic acid, and then with water. Strip the colour from the floc with a little dilute ammonia solution (the strength required varies with the dye from about 1 to 10 per cent.*), collecting the coloured portions only.A few millilitres only should be required, but the ammonia can be removed by aspiration or during concentration under reduced pressure. * All concentrations of ammonia solution given in percentages refer to volumes of ammonia solution, sp.gr. 0.88, in 100 volumes of water.November, 19631 IDENTIFICATION OF PERMITTED FOOD COLOURING MATTERS 867 What- man’s AE.30 paper was used to purify the colour extract from toffees in the following manner after concentration under reduced pressure as described above- Take up the residue in a few millilitres of water, and make just acid to litmus with 1 per cent.hydrochloric acid. Pass the solution through a double disc of Whatman’s AE.30 paper in a Gooch crucible under suction. Pass the filtrate repeatedly until all the dye is absorbed. Wash the filtrate with 20 ml of cold water, and strip it by passing a minimum volume of 10 per cent. ammonia solution through it. Similar methods were also used for fondant, toffees, fish paste, and madeira cake. CHROMATOGRAPHY- Solutions of extracted dyes in isobutanol or n-butanol were sometimes purified by chromatography on an alumina column, e.g., the solvent method described for canned peas can be continued- Pass the combined extracts through a $-inch bore polythene tube packed with dry alumina. Cut the column to isolate the coloured bands, and stir each fraction in a small beaker with 5 to 10 ml of 30 per cent.v/v ethanol. After the alumina has been allowed to settle, the supernatant liquors can be applied directly to chromatography paper. APPLICABILITY OF METHODS- All the methods were not carried out on all products containing the whole range of dyes. Alternative methods were employed only when the first method tried did not give an appar- ently complete recovery of the colour. For some dyes difficulty was experienced in obtaining a complete extraction into solvent, a complete take up on ion-exchange material or a complete removal from that material, but it should be emphasised that for a simple identification, not requiring complete recovery, any of these methods should be satisfactory. With fresh sausages, methods involving the use of columns were unsatisfactory because the columns became blocked by colloidal proteins; this difficulty was not experienced with fish paste, presumably because much of the protein had been denatured by heat.Solvent methods were normally satisfactory for fresh sausages. Ion-exchange methods were found satisfactory and very convenient with simple systems such as plain boiled sweets and fondants. With the exception of certain colours they were also found satisfactory with products, such as toffees and fish paste, containing cooked protein. The colours that caused difficulties were Yellow 2G, Blue VRS, Black PN and Erythrosine. Erythrosine was extracted satisfactorily from an acid solution by ether. Blue VRS was found to be much more readily extracted into n-butanol than into isobutanol, and its partition was found to be strongly affected by pH, too acid a solution giving a poor extraction.Table I shows the results of solvent and cetylpyridinium methods as applied to these difficult dyes in toffees and fish paste. TABLE I SOLVENT AND CETYLPYRIDINIUM COMPLEX METHODS APPLIED TO TOFFEES AND FISH PASTE Toffees Fish paste A A I -l r > Cetylpyridinium Cetylpyridinium Colour Solvents complex Solvents complex Blue VRS . . . . Unsatisfactory Satisfactory Satisfactory with Satisfactory Unsatisfactory with Satisfactory Erythrosine . . Not tried Satisfactory Unsatisfactory Not tried Black PN . . . . Satisfactory Satisfactory Unsatisfactory Unsatisfactory Yellow 2G . . Not tried Satisfactory Satisfactory Not tried n-butanol isobu tanol Once an initial separation of the colour had been made, other methods, unsatisfactory for the initial separation, were usually found quite satisfactory for further purification, suggest- ing that the problem was probably one of interfering impurities such as soluble proteins, amino acids, etc.TABLE I1 CHANGES OCCURKING DURING PROCESSING AND STOKAGE n.c.p.-no change on processing n.c.s.-no further change on storage C.1.numbers refer to the 1956 colour index Fish yastc Madeira cakc Fresh sausages n.c.p. Colour Original dyc No subsidiaries Boiled sweets Fondants Toff W S n.c.p. - n.c.p. Tartrazine CI 19140 n.c.y. n.c.p. Traces of mauve and brown after 15 months’ storage 3 subsidiaries 0*42%, 0-32%, 0-42 % 2.2% of Fast Red E present Sunset Yellow F.C.F.CI 15985 n.c.p. n.c.p. - n. c . p. n.c.p. n.c.p. n.c.p. n.c.p. n.c.p. - New red produced. Fast Red E partly faded Yellow- brown n.c.p. colour produced Amaranth CI 16185 Ponceau 4K CI 16255 Orange-red and yellow subsidiaries, total 0.9% - Subsidiaries increased Ycllow subsidiary n.c.p. greatly increased. Pink produced. Yellow increased after 14 months’ storage as intense as main colour n.c.p. 0.4% red subsidiary Red and purple on processing Yellow and orange on n.c.p. processing. Additional yellow, orange, pink and brown after 14 months’ storage Mauve subsidiary n.c.p. intensified. Blue subsidiary lost. Carmoisine CI 14720 n.c.p. n.c.p. Red 10B CI 17200 Mauve, blue and pink subsidiaries, total 0.9% Additional blue on processing - Mauve subsidiary intensified.Orange- brown formed. Red formed. n.c.s. 13 months Some subsidiaries Orange-brown 3ormed. lost, orange produced n.c.s. 13 months after 9 months’ storage - n.c.p. Original colour Trace of pink after almost gone. Several 9 months’ storage yellows formed. Mauve and traces of pink after 13 months’ storage Red 2G C I 18050 0.2% pink subsidiary n.c.p. n.c.s. 12 months n.c.p.TABLE I1 (conntd.) Colour Original dye Boiled sweets Ponceau MX Red n.c.p. CI 16150 subsidiary n.c.s. 9 months Fondants Toffee Fish paste Madeira cake Yellow formed. n.c.p. Yellow lost after 9 months’ storage Fresh sausages Subsidiary lost n.c.p. n.c.p. n.c.s. 9 months n.c.s. 9 months PonceauSX No Trace of pink CI 1470 subsidiaries n.c.s. 8 months Trace of 3 n.c.p.pinks n.c.s. 8 months n.c.s. 8 months n.c.p. n.c.s. 8 months n.c.p. n.c.p. Red 6B No n.c.p. CI 18056 subsidiaries n.c.s. 9 months n.c.p. Trace of red. n.c.s. 9 months n.c.s. 9 months Brownish yellow Trace of red formed. Main colour lost after 6 months’ storage. Yellow, orange, and more brownish- yellow formed All colour lost on processing Orange G CI 16230 Yellow and orange subsidiaries Trace of new colour found. Subsidiaries lost after 9 months Red and new orange found. Subsidiaries lost after 9 months n..c. p . Subsidianes lost after 9 months n.. c. p . Subsidianes lost and new faint orange formed after 9 months n.c.p. n.c.p. Yellow 2G 2 yellow CI 18965 subsidiaries Traces of new colour formed Subsidiaries lost n.c.s. 8 months Trace of red n.c.p. n.c.s. 6 months n.c.p.Yellowish Erythrosine At least CI 45430 6 subsidiaries Red and orange. Possibly original subsidiaries intensified n.c.p. Some subsidiaries lost after 7 months n.c.p. Some subsidiaries lost after 7 months n.c.p. n.c.p. Faint brown after 7 months Blue VRS 4 subsidiaries CI 42045 Mauve, blue, blue-green Indigo- 4 blue Carmine subsidiaries CI 13015 Trace of mauve Trace of yellow n.c.s. 7 months Trace of red n.c.p. n.c.s. 6 months n.c.p. Faint yellow. Yellow lost after 7 months n.c.p. except almost complete fading n.c.p. n.c.p. except almost complete fading Yellow Green S 2 green CI 44090 subsidiaries n.c.p. n.c.p. n.c.p. n.c.p. n.c.p. n.c.p. Yellow Orange and yellow formed. 1 red subsidiary lost Orange Black PN 2red, 1 orange CI 28440 1 grey-blue subsidiaries n.c.p.[Analyst, Vol.88 These methods are adequately described in the A.P.A. Handbook, but the following additional notes should prove helpful. 870 TRACE MATERIALS (COLOURS) COMMITTEE : EXTRACTION AND SEPARATION AND IDENTIFICATION OF INDIVIDUAL DYES The final process in every investigation made use of paper chromatography. SOLVENTS FOR PAPER CHROMATOGRAPHY- The following solvents were used most frequently- n-ButaPzol - water - glacial acetic acid-(20 + 12 + 5) by volume. Ethyl acetate - Pyridine - water-(11 + 5 + 4) by volume. Ethyl methyl ketone - acetone - water-(7 - 3 + 3) by volume. (B.S.I. method for determining subsidiaries in Amaranth,3 Ponceau 4R4 and Red lOB.5) Ethyl methyl ketone - acetone - water - ammonia solution, sp.gr. 0.88-(700 + 300 + 300 + 2 ) by volume. (B.S.I.method for determining subsidiaries in Tartrazine,6 Sunset Yellow,' Carmoisine,s Red 2Gg and Yellow 2G.1°) Trisodium citrate dihydrate - water - ammonia solution, sp.gr. 0.88-Trisodium citrate dihydrate (2 g) in 95 ml of water and 5 ml of ammonia solution. (B.S.I. method for deter- mining subsidiaries in Orange Gll and Ponceau SX.12) The B.S.I. solvents were used when appropriate, as the work described here was designed partly t o look for changes in proportions of subsidiaries. However, it was often found desirable to extend the time for development when a more complete picture of the subsidiaries present was required ; by using several different solvent mixtures, subsidiaries were often separated that would not have been found in a single solvent.GENERAL COMMENTS ON PAPER CHROMATOGRAPHY- The following general observations, some of which have already been made in the A.P.A. Handbook, are emphasised. (i) R, value alone is not a safe criterion for identification, even when several different solvents are used. R, values are affected by such factors as temperature, length of run, paper variations (thickness, texture, machine direction, batch-to-batch variations), presence of impurities and other less easily defined conditions. (ii) The use of standard dyes as markers is essential, but even with this precaution, impurities extracted with the colours, etc., may cause differences between the R, values of the markers and extracted dyes. If no separation occurs in any solvent system the colours may be assumed t o be identical. At the same time spots of standard dye and extracted dye should be run separately at similar con- centrations.This is necessary because the standard dye itself may split into more than one component and give misleading results. (iv) Certain colours will produce two or more spots of identical composition when run in some solvents if the concentration on the paper is too high (e.g., Tartrazine in A.P.A. Solvent 5). RESULTS The effects of processing the various colours so far examined are shown in Table 11. No storage tests were carried out on fresh sausages or madeira cake, because both these products have only a short shelf-life. Uncoloured material was also processed to ensure that yellow and brown spots on the chromatograms did not originate entirely from ingredients other than the dyestuffs.Fading of colours on processing or storage is not reported here as it is almost universal, and no attempt has been made to investigate colourless substances produced by breakdown of the dyes. Certain other changes have been reported in the literature. Dickinson and Raved3 reported that Ponceau 4R was reduced to a yellow colour by hydrogen sulphide or sulphur compounds liberated during the processing and storage of certain foodstuffs, and that Erythrosine lost iodine to produce fluorescein when canned cherries coloured with Erythrosine were stored in unlacquered (Solvent No. 5 of the A.P.A. Handbook.) (iii) The only safe method is to superimpose standard dye and extracted dye.November, 19631 IDENTIFICATION OF PERMITTED FOOD COLOURING MATTERS 87 1 Ruiz and Laroche15 found that Black PN was reduced in some sugar confectionery to give an orange colour that was not a known food dye. They also found that during extraction (on wool) this dye was partly converted to another orange colour that was not a known food dye. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 16. REFERENCES “Separation and Identification of Food Colours Permitted by the Colouring Matter in Food Regula- Drevon, B., and Laur, J., Ann. FaZsif., 1959, 52, 155. British Standard 3341 : 1961. British Standard 3342 : 1961. British Standard 3610 : 1963. British Standard 3211 : 1960. British Standard 3340: 1961. British Standard 3343 : 1961. British Standard 361 1 : 1963. British Standard 3614 : 1963. British Standard 3612 : 1963. British Standard 3613 : 1963. Dickinson, D., and Raven, T. W., paper submitted to the Food Additives and Contaminants tions, 1957,” The Association of Public Analysts, London, 1960. Sub-committee by the Campden Research Station, November, 1961. Riiiz, I. S. L., and Laroche, C., Ann. FuZsif. Ex$. Chim., 1960, 53, 681. , , J . Sci. Food Agric., 1962, 13, 650. -- Received January 17th, 1963

ISSN:0003-2654    

DOI:10.1039/AN9638800864

出版商:RSC    

年代:1963

数据来源: RSC

摘要:

872 YANUKA et al.: ISOLATION AND SEPARATION OF DYES [Analyst, Vol. 88 The Isolation and Separation of Dyes from Foodstuffs by Column Chromatography BY Y. YANUKA, Y. SHALON, E. WEISSENBERG AND I. NIR-GROSFELI) (Institute fov Standardization and Control of Pharmaceuticals, Ministry of Health, Jerusalem, Isvael) A method is described for isolating from foodstuffs, and then separating Two The by column chromatography, twelve synthetic dyes used in Israel. adsorbents were used: alumina, a t different pH values, and silica gel. eluents used were water, n-butanol and methanol. IN a previous publication1 a paper chromatographic method was described for separating and identifying a group of twelve synthetic water-soluble food dyes in use in Israel. The procedure described can be applied directly to commercial dyes.Dyes to be determined in food must be extracted and then purified before they can be identified, since the impurities associated with food often cause considerable difficulty in identifying the dyes, by either paper chromatography, spectrophotometry or any other means of identification. Wool has been used for extracting coal-tar dyes from f o o d - ~ t u f f s , ~ ~ ~ ~ ~ and by suitable modification of this method, basic dyes could be separated from acidic ones.596 Helbergs suggested the use of solutions of different pH values for eluting the adsorbed dyes from the wool. This procedure involved the use of strong acids and heat, conditions liable to alter the chemical structure of the dyes. Other procedures for isolating and separating dyes include solvent extraction at different pH values‘ and column chr~matography.~ y 9 Kaolin, alumina and silica gel have been used as adsorbents in chromatography.l0y11J2 91391* The ninth edition of the “Official Methods of Analysis” of The Association of Official Agricultural Chemists, describes a method for separating dyes by using a powdered-cellulose column and a series of sodium chloride solutions as e1~ents.l~ This technique, however, was not always found to give the desired separation.16 A procedure for extracting dyes from foods and separating them by column chromato- graphy is described in this paper. By using silica gel and alumina adjusted to different pH values, and with water, butanol and methanol as solvents, the dyes tested could be separated.The adsorbents acted as partition, adsorption or ion-exchange columns, depending on the conditions used. The dyes were obtained chemically pure, and could subsequently be identified with ease. METHOD REAGENTS- All reagents used were of analytical-reagent grade. Hydrochloric acid, 2 N and N. Ammonia solution, sp.gr. 0-88, and N. Ammonia water-Dilute I ml of ammonia solution, sp.gr. 0.88, with 99 ml of water. Methanol, redistilled. Methanol, 90 per cent. v / v . Ammoniacal methanol-Dilute 1 ml of ammonia solution, sp.gr. 0.88, with 99 ml of n-B.utanol, dry. n-Butanol-Saturate with 2 N hydrochloric acid by shaking the two reagents together Chloroform, redistilled. Acetone, redistilled. Sodium hydroxide solution, 5 N. Sodium acetate solutions, M, 0-5 M and 0.1.M aqueous. methanol. for a few minutes.Kovember, 19631 FROM FOODSTUFFS BY COLUMN CHROMATOGRAPHY 873 DYES- Ponceau 4R Red 6B Tart razine Green S Carmoisine Amaranth Red FB Ponceau SX Naphthol Yellow S Blue VRS Indigotine Sunset Yellow FC,F ADSORBENTS- Silica gel-M.F.C. Materials for Chromatography, Hopkin & Williams Ltd. Basic alumina-B.D.H. Chromatography grade. Acid alumina-Place about 500 g of the basic alumina in a 5-litre Erlenmeyer flask, and cover with N hydrochloric acid to about 10 cm above the surface of the adsorbent. Shake the Erlenmeyer flask occasionally for about 1 hour, and set aside overnight. Decant the acid. Cover the alumina with a further portion of PI: hydrochloric acid. Shake again occasionally for an hour, and decant the acid.Wash the alumina thoroughly several times with distilled water until the pH of the washings reaches 3 to 3.5. Filter off the alumina on a Buchner funnel, transfer to a flat tray, and set aside to dry at room temperature. Neutral alumina-Place about 100 g of the acid alumina in a 1-litre Erlenmeyer flask, and cover with N ammonia solution to about 3cm above the surface of the adsorbent. Shake the Erlenmeyer flask for about an hour, set aside overnight, and decant. Cover the alumina with a further portion of IG ammonia solution, and shake again for an hour. Decant and wash the alumina with warm distilled water until free of chloride, and then filter it off on a Buchner funnel. Dry the alumina first at 110" C, and then activate it a t 300" C for 3 hours. PREPARATION OF COLUMNS- Columns of 20-cm length and 2-cm diameter were used.glass-wool at the bottom of the column. height of about 5 cm, and cover with another piece of wood. required solvent by applying slight pressure. PROCEDURE- Extraction of the dyes from foodstufs-The procedure for extracting dyes added to food stuffs differs according to the type of food, and for this purpose the foodstuffs may be divided into two main categories. Place a piece of cotton- or Fill loosely with the appropriate adsorbent to a Wet the adsorbent with the I. Foodstuffs soluble in, or miscible with water: (a) Soft and alcoholic drinks, sweets, jellies, etc. ( b ) Milk. Foodstuffs sparingly soluble, or insoluble, in water: (a) Non-fatty foodstuffs, like cereals or jams. ( b ) Foodstuffs containing fat, e.g., butter, sausage or cheese, and non-fatty food contain- ing natural pigments, e.g., paprika.I (a) Dissolve the foodstuff in a suitable amount of water at room temperature, or on a heated water-bath. Adsorb the solution on an acid-alumina column, and wash the column with 100 to 200 ml of warm water (between 50" and 70" C) ; liquid preparations may be adsorbed directly on the acid-alumina column. Elute the adsorbed dyes with ammonia water, and evaporate the eluate to dryness on a water-bath. I ( b ) Add three volumes of acetone to one volume of milk, shake and allow the precipitate to settle. Collect the precipitate on filter-paper, and wash with small amounts of acetone until the dye is completely extracted. Apply the acetone solution onto an acid-alumina column and elute with warm water and then ammonia water as described under I (a).I1 (a) Crush or cut the foodstuffs into small pieces and grind them in a mortar with 2s hydrochloric acid to a homogeneous paste. Extract the paste with 10-ml portions of butanol saturated with 2 N hydrochloric acid, and decant the liquid into a separating funnel; three or four portions of acid - butanol mixture will extract most of the dyes. Only Green S and Blue VRS will remain mainly in the paste and pass only partly into the butanol. Extract these two dyes from butanol with ten to twelve 5-ml portions of 2 N hydrochloric acid. IT.874 YANUKA et d.: ISOLATION AND SEPARATION OF DYES [Analyst, VOl. 88 Adsorb the butanol solution containing all the dyes (except Green S and Blue VRS) on an acid-alumina column ; wash and elute as described under I (a).Add the acid extract containing Green S and Blue VRS to the original paste. Triturate the paste well, and filter. Wash the precipitate with 2 N hydrochloric acid until all traces of the dyes are removed from the paste. Adjust the pH of the acidic solution to approximately 1 with 5 N sodium hydroxide, and adsorb on to a silica-gel column. Elute the dyes from the silica-gel column with ammoniacal methanol, and immediately evaporate to dryness on a water-bath. I1 ( b ) Crush or cut the foodstuffs into small pieces, and transfer them to an Erlenmeyer flask. Extract the mass in the Erlenmeyer flask with 20-ml portions of chloroform until the latter is colourless, while the flask is being shaken mechanically.Add 20 ml of N hydrochloric acid to the food mass, and extract again in the Erlenmeyer flask with 20-ml portions of chloroform. Remove the chloroform each time by decanting into a separating funnel. Return the hydro- chloric acid solution each time to the food mass in the Erlenmeyer flask, and continue the extraction until the chloroform layer is colourless. Discard the interfering natural pigments and any unlisted dyes present that are soluble in chloroform. Extract the acid solution and the food mass with 10-ml portions of butanol until all the dyes have been extracted from the food. If Green S and Blue VRS dyes are suspected to be present in the sample tested, extract the combined butanol solution with ten to twelve 5-ml portions of 2 N hydrochloric acid, and add the acid extracts to the original acid solution.Adjust the pH of the acidic solution to approximately 1 with 5 N sodium hydroxide, and proceed as outlined under I1 (a). If the dyes Green S and Blue VRS are not expected to be present, combine the butanol extracts and adsorb on an acid-alumina coloumn, wash, and elute as described under I (a). CHROMATOGRAPHIC SEPARATION OF THE DYES SEPARATION OF THE DYES GREEN S AND BLUE VRS- Dissolve the dyes obtained by one of the procedures outlined above (not more than 1 mg of each dye should be present in the sample) in 10 to 20 ml of distilled water, and pour the solution onto a silica-gel column. All the dyes except Green S and Blue VRS move quickly through the column. Blue VRS remains strongly adsorbed at the top of the column, while Green S moves slowly during the washing: continue until this dye reaches to about 0-5 cm from the bottom of the column.To separate all the dyes from Green S and Blue VRS, 20 to 30 ml of water are usually sufficient. However, should separation not be complete, a second silica-gel column may be inserted below the first. (This step may be omitted when Green S and Blue VRS have been extracted and separated from the food by procedure I1 (a) or I1 (b).) The dyes Green S and Blue VRS are eluted from the silica gel with ammoniacal methanol, and the solution is immediately evaporated to dryness on a water-bath. To separate dyes Green S and Blue VRS from each other, dissolve the residue in aqueous butanol (1 part water plus 10 parts butanol), and adsorb the solution on a basic-alumina column.Wash the column with 10 ml of aqueous butanol. Blue VRS is eluted first, followed by Green S. Adsorb each dye solution separately on an acid-alumina column, and elute each with ammonia water. The dyes eluted earlier from the silica-gel column with water (all 12 dyes except Green S and Blue VRS) are adsorbed on an acid-alumina column and eluted with ammonia water. Evaporate the solution to dryness on a water-bath, and dissolve the residue in 1.5 ml of water and then add 10 ml of butanol. Adsorb the solution on a basic-alumina column, and wash the column with butanol saturated with water. The dyes on the column separate into two distinct bands, an upper one, Fraction A (containing Ponceau 4R, Amaranth, Red 6B, Tartrazine, and Indigotine), and a lower one, Fraction B (containing Carmoisine, Red FB, Ponceau SX, Naphthol Yellow S, and Sunset Yellow FCF).Continue eluting the column with aqueous butanol until the lower Fraction B is completely eluted. Wash the column with 5-ml portions of distilled water. SEPARATION OF THE DYES IN FRACTION A-- Wash the column containing the remaining Fraction A with three 5-ml portions of methanol to remove the butanol, and then continue washing with 90 per cent. methanol. Indigotide, and then Red 6R are now eluted; each dye is adsorbed on a separate acid-aluminaNovember, 19631 FROM FOODSTUFFS BY COLUMN CHROMATOGRAPHY 875 column. Remove the remaining dyes, Ponceau 4R, Amaranth and Tartrazine, from the first column, by eluting with 10 to 15 ml of ammonia water.Evaporate to dryness. Dissolve the residue in water, and adsorb on an acid-alumina column. Elute with 0-5 M sodium acetate solution. Three distinct bands appear on the column. The first to be eluted is Tartrazine, and then Ponceau 4R; Amaranth remains adsorbed on the column. Dilute the acetate solutions containing Tartrazine and Ponceau 4R with water, (one part of acetate solution plus 10 parts of water), and adsorb again on acid alumina. Wash the last three acid-alumina columns with distilled water to remove the excess of sodium acetate before elution of the dyes. Finally, elute all the five dyes, now separated on individual columns, with ammonia water. SEPARATIOX OF THE DYES IN FRACTION B- Adsorb the aqueous butanol solution containing the dyes of Fraction B on an acid- alumina column, and wash with water to remove the solvent. Elute the dyes with ammonia water, and evaporate the solution to dryness on a water- bath.Dissolve the residue in 10 ml of methanol] and adsorb the solution on a basic-alumina column. Elute Sunset Yellow FCF by washing with 90 per cent. methanol, and elute the remaining dyes from the column with 0.1 M sodium acetate. Adsorb the acetate solution on neutral alumina. Wash the column with about 40 ml of 0.1 M sodium acetate, which will elute Naphthol Yellow S. Continue the washing with four or five 10-ml portions of M sodium acetate, which elutes Carmoisine and Ponceau SX together leaving Red FB at the top of the column. To separate Carmoisine from Ponceau SX, dilute the 1 M acetate solution with about 200 ml of distilled water, and adsorb again on acid alumina.Wash the column with water, and extract the dyes with ammonia water. Evaporate the solution to dryness on a water- bath. Dissolve the residue first in 1.5 ml of water, and then add 10 ml of butanol. Chromato- graphically separate the solution on a basic-alumina column, by using aqueous butanol (1.5 parts water $Zus 10 parts butanol) as solvent. Carmoisine separates from the column. Adsorb the eluted dyes Carmoisine, Naphthol Yellow S and Sunset Yellow FCF on acid- alumina columns, wash with water, and elute with ammonia water. Wash the column containing Red FR with water, and then elute the dye with ammonia water. Remove Ponceau SX directly from its column with ammonia water without previous washing with water.The isolated food colours can now be identified by paper chr0matography.l DISCUSSION OF METHOD Most of the synthetic dyes used for colouring foods are azo-compounds, containing various acidic and basic groups of different strength. These functional groups, their type, position and number in the dye-molecule, permit separation of the dyes by column chromato- graphy. The two adsorbents used (silica gel and alumina, in their various forms described above), together with the three eluents (water, butanol and methanol), provide a wide range of adsorption, partition and ion-exchange conditions. Silica gel adsorbs only dyes with basic groups from aqueous solution. Acid alumina strongly adsorbs dyes with acid groups, and permits removal of extraneous I t is possible to separate the dyes into groups or individual colours by using alumina of By the method described] about 100 pg of a dye may be recovered from up to 500 g of In starchy gelatinous food a minimum of about 500 pg of dye must be present in The dyes are separated and isolated without application of either heat or strong reagents.The extraction of small concentrations of dyes from food required the use of large amounts By passing diluted solutions of dyes through appropriate columns, the dyes Thus a dye solution could be concentrated without material. different pH values. food. 100 g of a sample. This will be advantageous when working with compounds liable to decompose. of solvents. were adsorbed and the solvents recovered.prolonged heating.876 YANUKA, SHALON, WEISSENBERG AND NIR-GROSFELD [AndySt, vO1.88 The use of small columns allowed the separated dyes to be obtained on single columns. In this way it was possible to proceed with the isolation and identification of some dyes (including those which remain strongly adsorbed on the column) before the complete separa- tion of all the dyes had been accomplished. We thank Miss Dinah Ezran for her technical assistance and Dr. Erich J. Diamant for his help in preparing this manuscript. 1. 2. 3. 4. 6. 7 . 8. 9. 10. 11. 12. 13. 14. 15. 0. 16. REFERENCES Yanuka, Y., Shalon, E., Weissenberg, E., and Nir-Grosfeld, I., Analyst, 1962, 87, 791. Balavoine, P., Trav. Chim. Alim. et Hyg., 1930, 21, 28. Helberg, E., Mitt. Lebensmitt. Hyg., 1946, 37, 408. Roleff, H., 2. anal. Chem., 1949, 129, 190. Fujii, S., Bull. Nut. Hyg. Lab., Tokyo, 1955, 73, 335. Giovanni, C., Ann. Sper. agr., 1959, 13, 545. Eisenbrand, Von J., Dtsch. Lebensmitt Bdsch., 1954, 50, parts 10/11. Kuggli, P., and Jensen, P., Helv. Chim. Acta, 1935, 18, 624. -~ , Ibid., 1936, 19, 64. Motiier, M., Mitt. Lebensmitt. Hyg., 1952, 43, 118. Mottier, M., and Potterat, M., Ibid., 1952, 43, 123. Potterat, M., and Mottier, M., Ibid., 1953, 44, 192. Mottier, M., and Potterat, M., Ibid., 1953, 44, 293. -__ , Anal. Chim. Acta, 1955, 13, 46. “Offikial Methods of Analysis,” Ninth Edition, The Association of Ofi‘icial A4gricultural Chemists, “Separation and Identification of Food Colours Permitted by the Colouring Matters in Food Received February 1 lth, 1963 Washington, D.C., 1960. Regulations 1957,” The A4ssociation of Public -halysts, London, 1960.

ISSN:0003-2654    

DOI:10.1039/AN9638800872

出版商:RSC    

年代:1963

数据来源: RSC

摘要:

November, 19631 BENNETT AND MARSHALL 877 The Separation and Determination of Chromium Sesquioxide in Chrome Ores and Chrome-bearing Refractories BY H. BENNETT AND K. MARSHALL (The British Ceramic Research Association, Queens Road, Penkhull, Stoke-ow Trent) Dichromate in solutions obtained from chrome ores and refractories by fusion can be almost completely removed by using liquid ion-exchange “resins” under slightly acid sulphate conditions. The chromium can then be quantitatively recovered from the organic phase by stripping with potassium hydroxide solution. After reduction, the chromium sesquioxide can then be determined spectrophotometrically as the complex with ethylenediamine- tetra-acetic acid to an accuracy within about 0-2 per cent. ; this is more than adequate for routine control analysis.The aqueous phase is available for the determination of other constituents. THE analysis of chrome ores and chrome-bearing refractories has been the subject of con- siderable study in the past few years. The current methods for analysing these materials are tedious, and the results often show considerable discrepancies between laboratories. Con- sequently, a more rapid and accurate method is needed and also one more suitable for routine control testing, when some accuracy may be sacrificed for the sake of increased speed. Bryant and Hardwick’s methodl has been tested by the Chemical Analysis Committee of the British Ceramic Research Association and has proved to be a satisfactory procedure for the direct determination of chromium sesquioxide in these materials.Also, the same Committee has evaluated and published, in a private communication to the industry, a method for the accurate and direct determination of total iron. There is still a need for a method whereby the same solution of the sample can be used for determining both chromium and the other constituents. This is important, as the decomposition of the material is difficult and time consuming, and, therefore, it is undesirable to develop a series of methods for various constituents each involving a separate decomposition. One of the main difficulties presented by this type of material is that the large amount of chromium present complicates the determination of many of the other constituents, and it is therefore desirable to separate this element before proceeding with the rest of the analysis.This paper represents a first step in a research programme, in that it describes a method for the prior separation and determination of the chromium to an accuracy adequate for routine control analyses. In the past, chromium has been separated in this class of materials by one of three methods : volatilisation as chromyl chloride ; electrolysis with the use of a mercury cathode ; or by solvent extraction with isobutyl methyl ketone. ’Each procedure has its disadvantages. Volatilisation as chromyl chloride involves the decomposition of the sample with perchloric acid, which is not invariably successful, and, in addition, complete removal of the chromium is rarely achieved. Separation via a mercury cathode simultaneously removes the iron and manganese, and also there is a danger that some of the silica may be precipitated.Further, the subsequent cleansing of the mercury may give rise to difficulties in some laboratories. Extraction with isobutyl methyl ketone is the most useful of these techniques, but is invariably incomplete owing to the reduction of some of the chromate by the organic phase in the presence of the necessary hydrochloric acid. Often the amount of the residual chromium is tolerable, but occasionally the amount can be high enough to introduce significant errors. Recently, liquid ion-exchange “resins” have become available, and these appeared to offer the possibility of removing chromate from slightly acid solutions. Green’s work2 on the removal of iron, titanium and zirconium suggested that the removal of chromium would be almost specific.Stripping the chromate back from the “resin” could then be accomplished by using alkali. LIQUID ION-EXCHANGE “RESINS”- The “resins” used in this work were manufactured by the Rohm and Haas Company of America and were obtained from their agents in this country, Charles Lennig and Co. Ltd.,878 [Analyst, Vol. 88 26-28 Bedford Row, London, W.C.l. They are Amberlite LA-1 and LA-2, liquid secondary amines insoluble in water and having a high molecular weight. The “resins” are soluble in a large range of organic solvents, and that in which they are generally supplied is stated to be toluene. EXPERIMENTAL Preliminary trials showed that when a slightly acid solution of potassium dichromate in sulphuric acid was shaken with an approximately equal volume of “resin” solution (4 per cent.in chloroform) almost all the dichromate ion was transferred to the organic phase. When this organic phase was shaken with N potassium hydroxide the chromate ion was stripped from the “resin” and returned to the aqueous phase. The solutions used in subsequent experiments were made to simulate those that would be obtained after fusion of a 0.5-g sample in sodium carbonate and boric acid and dissolution of the melt in sulphuric acid. A blank solution slightly acidified with sulphuric acid and containing 7 g of sodium carbonate and 2.5 g of boric acid per aliquot was used. The strength of the “resin” solution was increased from that used by Green to 10 per cent.by volume in the chosen solvent. EFFECT OF ACIDITY- The pH of the blank solution used for these experiments was found to be 2.4. Separations with “resin” LA-1 were carried out by using in the first extraction 25 ml of the “resin” solution and with 0-, lo-, 20- and 30-ml additions of dilute sulphuric acid (1 + 9). The competition of the sulphate ion reduced the efficiency of the extraction, some di- chromate being left in the aqueous phase. Two further extractions, each with 10 ml of “resin” solution were therefore carried out resulting in a colourless aqueous phase. This phase was then washed twice with 25 ml of chloroform to remove traces of “resin.” The aqueous phase was then tested for residual chromate and also for total chromium (see Table I).BENNETT AND MARSHALL : SEPARATION AND DETERMINATION The effects of acidity, type of “resin” and solvent were investigated. TABLE I Potassium dichromate equivalent to 125 mg of chromium sesquioxide was taken for each determination Sulphuric acid Chromate found in Total chromium found (.1 + 9) added, aqueous phase, in aqueous phase, ml mg of Cr,O, mg of Cr,O, 0 0.42 0.45 10 0.00 1.19 20 0.00 1.49 30 0.00 1.77 EFFECT OF ACID CONCENTRATION ON THE EXTRACTION OF CHROMATE The results in Table I show that more complete removal of the chromate is achieved at higher concentrations of acid, but is offset by increasing reduction of the sexavalent chromium. This reduction may be due to impurities in the “resin,” e.g., ferrous iron, organic reducing agent, or both. Stripping the “resin” before use to remove the ferrous iron, if present, was ineffective, and attempts to remove the organic reducing agents by treating the “resin” with permanganate revealed that the permanganate was reduced almost indefinitely ; hence it was concluded that the “resin” itself was being oxidised.To minimise this reduction and still obtain a complete removal of chromate, no acid was added until after the first two extractions, then 10 ml of sulphuric acid (1 + 9) were added before the third extraction. This technique proved effective. The pH of the solution had an initial value of 2.4; after the first extraction this value increased to pH 4.2 and after the second extraction to pH 4.6. After addition of the sulphuric acid the pH dropped to about 1.3 and thereafter increased only slightly.This would seem to indicate a definite loss of sulphuric acid into the “resin.” CHOICE OF “RESIN”- Amberlite LA-1 and LA-2 “resins” are similar in type, but, of the two, LA-2 is the more basic, so that, provided the sulphate was not preferentially extracted, it could be expected to be the more effective.November, 19631 OF CHROMIURI SESQUIOXIDE IK CHROME ORES 879 Again, total chro- mium and residual chromate were determined in the aqueous phases. In both instances the chromate content was negligible and the total chromium sesquioxide was 0.060 mg with LA-1 and 0-045mg with LA-2. The latter figure is equivalent to 0-009 per cent. of chromium sesquioxide if a 0.5-g sample of material had been taken. LA-2 was chosen for the method, as it also showed less tendency to emulsify.CHOICE OF SOLVENT- Chloroform had been used exclusively up to this stage of the work, and as a slight reduc- tion still occurred it was thought possible that a change of solvent might be beneficial. “Resin” having a concentration of 10 per cent. by volume was used throughout. Total residual chromium sesquioxide with LA-2 was 0-015 mg. This was a considerable improve- ment over the results obtained with chloroform, but the use of toluene, which is lighter than the aqueous phase, introduces manipulative complications. If it had proved possible to determine the chromium sufficiently accurately for a referee method, toluene would have been the solvent chosen, but, as the final method for chromium is of routine accuracy, chloro- form appears to be at least adequate.Carbon tetrachloride was next tried, but was even less effective than chloroform, the residual chromium sesquioxide being 0.1 1 mg. Mixtures of carbon tetrachloride or chloroform with toluene (so as to obtain a heavier organic phase) were not so effective, probably because separation of the phases was much slower. POSSIBLE EXTRACTION OF OTHER CONSTITUENTS- Silica-It was thought that silica might be partially extracted under these conditions. With the equivalent of 4.95 per cent. of silica present before the extraction, recoveries were 4.99 per cent. with Amberlite LA-1 being used and 4-90 per cent. with LA-2. These errors are probably not greater than the experimental error of the spectrophotometric method of determination. Ferric oxide-Under the conditions of acidity used here, Green found that extraction of iron took place, but the results of experiments showed that when a blank fusion solution (see “Experimental” p.878) was present no loss of iron occurred. Thus it may be that it is the concentration of the sulphate or competing ion that is the important factor. However, Green found that it was extracted in the presence of hydrogen peroxide, and, as this could provide the basis for a subsequent separation and determination, attempts were made to extract titanium from the solutions by this procedure. In all instances recoveries were found to be only about 50 per cent. of the amount present, so that again competition from the sulphate appears to restrict extract ion. Manganese oxide-No extraction of manganese oxide appears to occur, whether this is added in the manganous state or as permanganate.STRIPPING THE “RESIN” PHASE- A single extraction with N potassium hydroxide appeared to strip the “resin” completely, but a second extraction was made with the same reagent to ensure that no significant amount of chromium was left in the slightly cloudy emulsion in the organic phase. DETERMINATION OF THE CHROMIUM SESQUIOXIDE- Several methods for the determination of chromium were examined. Most of the accurate methods are based on its determination as chromate, but even by using such a tech- nique as Bryant and Hardwick’s, which involves an auxiliary oxidation, some chromate always appears to be reduced by the slight traces of “resin,” and results were always low.Other methods for determining the chromic ion were attempted, including gravimetric tech- niques with 8-hydroxyquinoline and benzoic acid, but low results were recorded. A volumetric method based on the formation of a complex with ethylenediaminetetra- acetic acid (EDTA) and back titration with zinc and with dithizone as indicator gave sur- prisingly sharp end-points in the strong purple solution, particularly if this is screened with Naphthol Green B to a dark blue with a slight tinge of green. But it became clear from the consistently low results that the amount of EDTA required to ensure complete reaction would mean excessively large back-titrations. Separations with each “resin” were carried out as indicated above. As the “resin” is already dissolved in toluene this was the obvious first choice.Titania-No extraction of titania was observed.880 [Analyst, Vol. 88 The presence of large amounts of sulphate ion rules out precipitation as barium or lead chromate. Pi-ibil and Klubalova’s spectrophotometric method, as described by Wel~her,~ proved to be successful. The chromate is reduced to the chromic ion with sodium sulphite and the chromium complexed with a large excess of EDTA at a pH between 2 and 4: The violet colour of the complex is measured at 550 mp. Solutions containing various amounts of chromate, and a portion of the blank fusion solution were extracted into the “resin” and the combined organic phase stripped with potassium hydroxide. The chromium111 recovered from the acidified aqueous solutions was then determined spectrophotometrically.The results, expressed as percentage of chromium sesquioxide equivalent to the chromiumII1, were- BENNETT AND MARSHALL : SEPARATION AND DETERMINATION Amount added .. 10 10 10 25 25 40 40 50 50 50 Amount recovered . . 10.0 10.0 9.8 24.9 24.9 40.1 39-9 50.0 50.0 49.8 It will be seen that the results have a slight negative bias, but the technique of calibrating the instrument on solutions of potassium dichromate would appear to produce answers accurate enough for routine control. However, Pi-ibil states that the sulphate ion causes the results to be slightly low (an error of up to 1 per cent.), and as the solutions contain sulphate extracted into the “resin” it seems legitimate to calibrate the instrument on solutions that have been extracted into the “resin” in the presence of a blank solution.Table I1 shows the results on a range of samples, the results having been calibrated against pure solutions and solutions passed through the “resin.” TABLE I1 DETERMINATION OF CHROMIUM SESQUIOXIDE IN DIFFERENT MATERIALS CONTAINING CHROMIUM The results are the mean of duplicate determinations Accepted value Material for Cr,O, content, % Mag-chrome brick, A . . . . . . 11-90 11-26 Chrome-mag brick, B . . . . . . 23.79 29-54 Turkish chrome-ore . . . . . . 39.44 Philippine chrome-ore . . . . . . 34-04 Grecian chrome-ore (B.C.S. 308) . . 41.50 Cr,O, content found after calibration against- A f > solutions extracted pure solutions, into “resin,” 11.8 11.9 11.1 11.3 23.6 23.8 29.2 29.5 39-0 39-2 33.9 34-2 41.1 41.3 % % An accuracy to within 0.5 per cent.of chromium sesquioxide is required in routine control work. The results obtained by the proposed procedure, after calibration against blank fusion solutions extracted into the “resin,” had a maximum error of k0.2 per cent. ; this is not significantly different from the error in the results obtained by Bryant and Hardwick’s procedure. The method takes a little longer to complete, but has the considerable advantage that other determinations can be made on the aqueous phase without recourse to a further fusion. Work is continuing on this possibility, with particular reference to the determination of lime, which has been a stumbling block in the past. 31 ETH 0 D “RESIN” SOLUTION- chloroform and mixing.DECOMPOSITION OF THE SAMPLE- Accurately weigh 0.500 g of the finely ground sample, dried at 110” C, into a platinum crucible. Add 7 g of anhydrous sodium carbonate and 2.5 g of boric acid, and mix thoroughly. Heat over a bunsen or Meker flame, slowly raising the temperature until the mixture begins to melt; keep at this temperature until melting is complete. Then raise the tempera- ture slowly and steadily to the full heat of the flame (about 950” C). After about 5 minutes at this temperature, swirl the contents of the crucible every 2 minutes, making sure that the particles of the sample on the side of the crucible come into contact with hot molten flux. Prepare a (1 + 9) solution of “resin” by adding 50 ml of Amberlite LA-2 to 450 ml ofNovember, 19631 OF CHROMIUAI SESQUIOXIDE IN CHROME ORES 88 1 If swirling is begun too early, it is difficult to detach unfused particles from the side of the Continue to heat and swirl at this temperature for 25 to 45 minutes depending on Completeness of decomposition can be checked visually by the absence of unfused par- The type and position of the burner should be chosen so as Cool, place the crucible and lid in a 250-ml beaker containing 85 ml of water and 15 ml Remove the crucible Cool the solution.Add diluted ammonia solution (1 + 1) until the first formation of a permanent precipitate, crucible. the nature of the sample. ticles at the bottom of the melt. to maintain oxidising conditions throughout. of diluted sulphuric acid (1 + l), and warm until extraction is complete.and lid from the beaker, washing them with the minimum volume of water. and then re-dissolve in 1 or 2 drops of diluted sulphuric acid (1 + 1). EXTRACTION OF CHROMATE- Transfer the solution to a 500-ml separating funnel, A ; the volume at this stage should not exceed 150 ml. Add 25 ml of dilute resin solution (I + 9), shake for 1 minute, and allow to separate. Transfer the organic phase to a second separating funnel, B, washing the stem of funnel A with chloroform. Repeat the extraction with 10ml of dilute resin solution, and again transfer the organic phase to funnel B. Add 10 ml of dilute sulphuric acid (1 + 9) to the aqueous phase, and repeat the extrac- tion with a further 10ml of dilute “resin” solution, again transferring the organic phase to funnel B.Remove the traces of “resin” from the aqueous phase by two extractions with 20-ml portions of chloroform, the chloroform layers being added to the combined “resin” solution in funnel B. Reserve the aqueous phase and washings from funnel A for any other determina- tions, if required. To funnel B add 50 ml of potassium hydroxide solution (approximately N), and shake for 1 minute. Allow the layers to separate, and transfer the organic layer to the now empty funnel A. Add to this funnel a further 50 ml of the potassium hydroxide solution, and again extract. Discard the organic layer, and combine the two aqueous phases. Remove traces of “resin” by shaking twice with 10-ml portions of chloroform, discarding the chloroform layers. Transfer the alkaline aqueous phase to a 450-ml beaker, add hydrochloric acid, spgr.1-18, until the solution becomes acid, and then add 6 drops in excess. Boil the solution to remove the chloroform and to reduce the volume to about 200ml. Transfer the solution to a 250-ml calibrated flask, and dilute to 250 ml. DETERMINATION OF CHROMIUM SESQGIOXIDE- If the solution appears cloudy, filter 40 to 50 ml through a dry Whatman KO. 42 filter- paper, discarding the first few millilitres. Transfer a 20-ml portion of this solution to a 250-ml beaker, add 20 drops of hydro- chloric acid, sp.gr. 1-18, and then 10 ml of a 5 per cent. solution of sodium sulphite, whik stirring. Cool to room temperature, add 10 ml of 5 per cent. EDTA solution, and then add ammonia solution, sp.gr. 0-88, until the first appearance of a permanent precipitate. Dissolve this precipitate by adding 20 drops of diluted acetic acid (1 + l ) , and dilute to 200 ml. Roil for 10 to 15 minutes, cool, and dilute to 250 ml in a calibrated flask. Measure the optical density of the solution against water at 550 mp in 4-cm cells. Determine the chromium sesquioxide content of the solution by reference t o a calibration graph prepared either from pure solutions of potassium dichromate or by passing dichromate solutions mixed with a blank fusion solution through the separation process. We thank the Director of Research of the British Ceramic Research Association, for permission to publish this paper. Dilute to about 50ml with water, and boil for about 5 minutes. REFEKENCES 1. 3. Green, H., Mefalluvgza, 1962, 65, 305; 1962, 66, 52. 3. Bryant, F. J., and Hardwick, P. J., Alzalyst, 1950, 75, 12. Welcher, F. J ., “The Analytical Uses of Ethylenediaminetetra-acetic Acid,” D. Van Nostrand Received A l q 17th, 1963 Co. Inc., Princeton, N.J., New York, ‘Toronto and London, 1957.

ISSN:0003-2654    

DOI:10.1039/AN9638800877

出版商:RSC    

年代:1963

数据来源: RSC