摘要:

JANUARY, I953 THE ANALYST Vol. 78, No. 922 Editorial REVIEW PAPERS IN analytical chemistry, as in every other advancing science, a stage is inevitably reached at which it is essential, if further progress is to be made in a particular branch, to survey critically the field already covered and to sift, sort and arrange the accumulated mass of facts and theories collected and recorded during, possibly, many years. Only by means of such a survey can these facts and theories be allotted their value for present needs and their potentialities for future progress. To make a survey of this kind is by no means easy. I t demands a wide theoretical and practical knowledge of the subject, a critical and orderly habit of mind and some practice in the use of words and their nice arrangement on paper.To be of sufficient value to readers of The -4naZyst a survey or review paper of this kind must be critical of its own contents; the sphere of applicability, and the usual working range and reproducibility of any method mentioned need to be given; it must contain all the information that an analyst confronted with a practical problem would wish to have. In other words, a review of this type should be conceived in the laboratory rather than in the library. The writer should be an authority on his subject-one well qualified by experience to act as guide to the student and inexperienced worker, and one who will be listened to with respect by his equals in knowledge. The Publication Committee is fully aware of these requirements and in plans for future reviews will keep them constantly in mind.As a pattern for future review papers suitable for the pages of The Analyst, there will be found in this issue a survey of the methods used in the micro-determination of proteins in blood plasma. The determination of protein-more precisely, of total nitrogen-on a macro scale is itself of a sufficiently sedate age to have become a respectable member of the group of standard methods, yet it is still the subject of occasional research and even of some resultant polemic. This is no doubt due in part to the complications that arise from the presence in most biological material of “non-protein” nitrogen, but the most serious difficulty arises directly from the immense variety of amino-acid patterns in the proteins elaborated by plants and animals. Those who estimate proteins by determining nitrogen are making a single factor do what a single factor cannot possibly do with accuracy. To these inherent methodological difficulties new ones are added when the analyst is compelled to work with milligrams rather than grams of sample. This is the situation for the biochemist or chemical pathologist who has to determine total or fractionated plasma protein. The review in this issue is addressed primarily to them, but we have reason to believe that its subject matter, as well as the approach and treatment, will commend it to a wider audience,

ISSN:0003-2654    

DOI:10.1039/AN9537800001

出版商:RSC    

年代:1953

数据来源: RSC

摘要:

2 PROCEEDINGS [Vol. 78 PROCEEDINGS OF THE SOCIETY OF PUBLIC ANALYSTS AND OTHER ANALYTICAL CHEMISTS AN Ordinary Meeting of the Society was held at 7 p.m. on Wednesday, October Ist, 1952, in the Meeting Room of the Chemical Society, Burlington House, London, W.l. The Chair was taken by the President, Dr. J. R. Nicholls, C.B.E., F.R.I.C. The following papers were presented and discussed : “Spectroscopic Properties of Vitamin A,. Application to the Assay of Cod-Liver Oil,” by H. R. Cama, B.A., M.Sc., Ph.D., and Professor R. A. Morton, Ph.D., D.Sc., F.R.S., F.R.I.C. ; “The Estimation of Carbonyl Com- pounds by Semicarbazide and Hydroxylamine with Special Reference to Fatty Acid Oxidation Products,” by A. J. Feuell, B.Sc., A.R.I.C., and J. H. Skellon, M.Sc., Ph.D., F.R.I.C. ; “Simultaneous Determination of Pentose and Hexose,” by W.R. Fernell, BSc., and H. K. King, M.A., Ph.D., F.R.I.C. AN Ordinary Meeting of the Society, organised by the Biological Methods Group, was held at 7 p.m. on Wednesday, November 5th, 1952, in the Meeting Room of the Chemical Society, Burlington House, London, W.l. The Chair was taken by the President, Dr. J. R. Nicholls, C.B.E., F.R.I.C. The subject of the meeting was “The Bio-assay of Vitamins with Special Reference to Microbiological Methods,” and the following papers were presented and discussed : “Vitamin Assays-The Relative Roles of Chemical, Biological and Microbiological Tests,” by L. H. Harris, Sc.D., D.Sc., Ph.D., F.R.I.C.; “Selection of Methods for Routine Assays for Members of the Vitamin B Complex,” by H.Pritchard, M.Sc., F.R.I.C.; “The Efficient Planning of Microbiological Assays, Particularly Assays of Vitamin B,,,” by E. C. Wood, B.Sc., Ph.D., A.R.C.S., F.R.I.C. NEW MEMBERS Everard George Adams, Ph.C. (Lond.), M.P.S. ; Per Olof Bethge, M.Sc. (Sweden) ; Peter Maxwell Grayson Broughton, B.Sc. (Lond.), A.R.I.C. ; Philip Arnold Farmer, M.Sc. (Lond.), D.I.C., A.R.I.C. ; Gordon Gough, B.A. (Cantab.) ; Philip Sydney Hall; Harry Brereton Heath, M.B.E., B.Pharm. (Lond.), Ph.C. ; Frank Rowland Holmes; Riaz Ahmed Khan, MSc. (Aligarh Univ.), Ph.D. (Lond.), A.R.I.C. ; Vazir Ahmed Kureshy, B.A. (Osmania Univ.), M.Sc. (Muslim Univ.) ; James Herbert Evan Marshall, M.A. (Cantab.), F.R.I.C. ; James Martin Ogilvie; Wilfred Thomas Roberts, B.Sc., Dip.Ed. (Wales), A.R.I.C.; John Geoffrey Washington, BSc. (Lond.). Brian Robert Arnold; Sahib Dayal Bhatia, B.Sc. (Allahabad), MSc. (Agra), A.R.I.C. ; William Henry Bull, B.Sc. (Edin.) ; Leslie Davies, B.Sc. (Lond.), A.R.I.C. ; Michael Trevor Davies, B.Sc. (Lond.), A.R.I.C. ; Sydney Albert Elks, M.A. (Oxon.), A.R.I.C. ; Harry John Finch, B.A. (Cantab.) ; Miss Edna Patricia Greenburgh; John Richard Heron, B.Sc., A.R.C.S. ; James Ivor Morgan Jones, D.Sc. (Wales), F.R.I.C. ; Albert Arthur Kendall; John Brian Mullin, B.Sc. (Liv.) ; Norman Arthur Chetwynd Pryce, B.Sc. (Wales), F.R.I.C. ; Eric Vernon Reid, A.R.I.C.; Harry Slack, D.C.M., BSc. (Vict.), F.R.I.C.; John Errol Still, B.Sc. (Lond.); Basil James Walby, B.Sc. (Lond.) ; Victor Trevor Walkley. ELECTED OCTOBER 1ST, 1952 ELECTED NOVEMBER 5TH, 1952 DEATHS Robert Edward Leo Davies. Harry Gordon Reeves. We regret to record the deaths of SCOTTISH SECTION AN Ordinary Meeting of the Section was held at 7.15, p.m. on Wednesday, November 5th, 1952, in Room 246, The Royal Technical College, Glasgow, C.l. A lecture on “Quantitative Microscopy in Relation to Plant Tissue” was given by Francis Fish, B.Pharm., Ph.C., followed by a demonstration of the equipment and materials used in quantitative microscopy.

ISSN:0003-2654    

DOI:10.1039/AN9537800002

出版商:RSC    

年代:1953

数据来源: RSC

摘要:

3 Obituary SIR JACK DRUMMOND ~ A C K CmiL DRUMMONI) was born in 1891. He was educated a t King’s College School and hast London College, taking his degree in 1913. After acting as Research Assistant at King’s College, Londoii, and serving for a short time in the Government Laboratory, he went to tlie Cancer Hospital Research Institute, where in 1918 he became Director of Biochemical Research. In the following year hc was appointed Reader in Chemical Physiology at University College, London, and was made the first Professor of Biochemistry in 1922. At that time such Chairs were few, and Drummond concentrated on the mode of action of vitamins and other food essentials. His school of research workers specialised on a variety of nutritional problems and became well-known in this field.Biological tests were essential for observing the effects of vitamins, and the writer remembers the occasion when the late H. E. Armstrong greeted Druminond with the query, “Well, Drummond, when are you going to give up ratting and do some chemistry?” Various papers in T h e LLJnaly~t show the interest that Drummond took in the early cheniical tests developed for determining vitamins and the researches he made to separate the minor constitucnts of thc unsaponifiable matter of oils with which vitamins are associated. In this connection lie was one of the first to use chromatographic techniques; his batteries of adsorption columns were impressive for their size and the amounts of material dealt with. Although worlting on the newer knowledge of nutrition, he was convinced that man had developed a natural aptitude or instinct for selecting food that would provide the necessary adjuncts.His researches into the food habits of tlie past were collected in “The Englishman’s Food,” piiblishetl in 1939 in collaboration with Miss Wilbraham, whom he married in 1940. A t thtt out I I J - U ~ o f war i i i 1939 [ i t \ offered his services to the Ministry of Food arid was (2iief Adviser 011 lTood Contamination until the following year, when he was appointed Scientific Adviser, a post which he held until 1946. He was concerned with feeding a people at a time of shortage, when there were great restrictions on the availability and choice of imports. The problem of emergency feeding after raids was a specially difficult one.Throughout he always insisted that special attention must be paid to the diets of children, adolescents, expectant and nursing mothers, the sick, and heavy industrial workers. His work was rewarded by a knighthood in 1044, and in the same year he was elected a Fellow of the Royal Society. He resigned his post with the Ministry of Food in 1946 to become Director of Research to Boots Pure Drug Co., Ltd. Between 1914 and 1948 Drummoiid published about 200 scientific communications either alone or jointly. About one-third appeared in The Biochemical . Jownal, the remainder being widely dispersed; ten appeared in The Analyst and about double that number in the Journal of the Society of Chemical Industry. They include references to almost all the then known vitamins and his contributions to nutritional science are noteworthy. Drummond served on the Council of the Society, on the Council of the Iiistitute of Chemistry, for which he acted as examiner for six years, and was Chairman of the London Section of the Institute while the writer was Secretary. He was a member of the Inter- departmental Committee on Food Standards from its formation in 1942. In 1947 this Committee was reconstituted as the Minister’s Food Standards Committee and Sir Jack retained his seat on the Committee until other commitments compelled him to resign in 1950. Drummond was always a young man, always a good companion arid always ready to be of assistance to others. The tragic news of his death and that o f his wife and daughter came as a great shock. J. R. NICHOLLS

ISSN:0003-2654    

DOI:10.1039/AN9537800003

出版商:RSC    

年代:1953

数据来源: RSC

摘要:

4 SALT MICRO-ANALTTICAL METHODS [Vol. 7s Micro-Analytical Methods for Proteins in Blood Plasma A Critical Review BY HAllO1,T) R. SALT SUMMAR\- 01; cos'ri<s.rs Introduction. Estimation of total protein in plasma or serum. By physical methods. By nitrogen content. By colour reactions for molecular groiipings. Electrophoretic separation. Differential precipitation by alcohol. Fractionation by salting-out. Fibrinogen by coagulation. Cryoglobulins by cold precipitation. Globulins by viscosity measurements. Metal-combining globulin by absorptiometric titration. Albumin by combination with haemin. Acid-soluble mucoproteins. Specific protein fractions by immunoche~nical precipitation. The differential determination of plasma protcins. The separation and determination of certain proteins by their special properties.Normal values. 1IEcm;T advances in our knowledge of the proteins of human blood plasma have shown the value of analysing plasma samples in detail and thereby delineating the various protein patterns that occur in health or in disease. Normally the plasma contains large amounts of albumins, globulins and fibrinogen, smaller amounts of lipoproteins and mucoproteins, and many antibodies, hormones and enzymes. Disease states are often accompanied by quantitative changes in the amounts or proportions of these plasma proteins, and some diseases are characterised by the appearance of abnormal proteins. Our more intimate knowledge has come about largely through the application of elaborate new techniques for the separation, identification and estimation of the plasma protein com- ponents.Ultracentrifuga1,l electrophoretic2 y3 and physico-chemical preparative procedures4 p5 have each helped to elucidate the problems, whilst other techniques, appropriately related to physiological or immunological mechanisms,6 have been applied to the study of those proteins that possess specific biological properties. The elaboration of these varied techniques continues to be the subject of important fundamental research. In the clinical laboratory, where the problems of human disease most urgently confront the chemist, detailed analyses for plasma proteins demand simpler and less costly methods than those proper to research. The complex nature of the proteins themselves and the necessary limitations of technical practice lead inevitably to an acceptance of analytical methods that are in greater or less degree empirical, and this has fostered, in turn, the develop- ment of many different procedures.Descriptions of these and discussions of their merits are widely distributed ; they form an extensive and perplexing literature. Most of the methods published up to 1047 were reviewed by Kirk7 in a discussion on the chemical determination of proteins. Other analytical considerations have been included by Edsall,8 reviewing the fractionation of plasma protein, and by Gutman9 in a survey of plasma proteins in disease. In this review only the simpler procedures acceptable to the clinical chemist will be discussed. Special attention will be directed to those model-n refinements of technique that best demonstrate the whole pattern of the plasma proteins and reveal any such particular abnormality as may be expected in any particular disease state. First, however, the problem of estimating the plasma total protein (or any one protein) must receive adequate consideration.Jan., 19631 FOR PROTEINS IS BLOOD PLASMA.A CRITICAL REVIEW 5 ESTIMATION OF TOTAL PROTEIN I N PLASMA OR SERUM B Y PHYSICAL METHODS- The effects of proteins on certain physical properties of their solutions provide simple methods for their estimation which avoid the need for first isolating them. Such methods are applicable to blood plasma, in which variations of physical properties are correlated closely with quantitative changes in protein content and but little affected by fluctuations in other constituents.The methods are especially useful when large numbers of observations are to be made, but they demand careful standardisation and the maintenance of exact techniques. Procedures based on specific gravity measurements have been described by Kagan,lo who timed the fall of a drop of serum through a fixed distance in oil, and by Lowry and Hunter,ll who noted the flotation of a drop of serum in an organic liquid mixture having its specific gravity graded throughout its height. A more recent technique, which has been widely used and studied, is that of Van Slyke, Hiller, Phillips, Hamilton, Dole, Archibald and Il:der,12 who compute the plasma total protein from the specific gravity determined by flotation of drops of plasma in solutions of copper sulphate. Sunderman13 has recommended that total protein be estimated by measurement of the refyactiue index of the plasma.The method is less sensitive than those based on specific gravity and may be inaccurate when plasma lipids are high. Other physical methods have been applied to the protein after its precipitation from solution. In the method of Looney and Walsh,14 the protein was precipitated from diluted serum by means of salicylsulphonic acid in the presence of ghatti gum, and the optical turbidity of the suspension was measured in a photelometer. This technique has been refined by Salt15 and described along with a similar method for serum globulins. Light dispersion methods of this kind are both accurate and sensitive, provided that the particle-size of the dispersed proteins are standardised by exact control of reaction conditions, of which the most important is the controlling effect of a protective colloid.Separation of proteins from solution, with subsequent purification, drying and weighing, was put forward as a pavimetric method by Robinson and Hogden.16 Unlike what has happened for most analytical methods, the gravimetric procedure has not been generally accepted as the most reliable standard reference method, owing to the difficulties of purifying and drying the protein precipitates. If these difficulties can be surmounted, the method has merit for reference purposes. I t is free from errors due to variations in protein nitrogen content (which also affect the accuracy of the Kjeldahl procedure, more often used as a reference method) but it requires large quantities of material.BY NITKOGEN CONTEST-- Determination of total nitrogen in purified protein fractions (or in cruder material with a suitable correction for non-protein nitrogen) has been adopted generally as the standard method to which others can be referred. Brand, Kassell and Saidell7 determined the percentages of nitrogen in separated fractions of normal human plasma by the micro-Dumas combustion method with the following results: albumin 15.95, P-globulin 14-84, y-globulin 16-03, fibrinogen 16.90. In the Kjeldahl technique as applied to plasma total protein, a constant protein-nitrogen content of 16.0 per cent. (factor 6.25) is usually assumed. The acceptability of this assumption has been confirmed by Hiller, Plazin and Van Slyke,l* who recently reviewed the experimental conditions variously recommended for Kjeldahl determinations of protein nitrogen.These authors concluded that mercury, which was first introduced as a catalyst only two years after Kjeldahl's original publication in 1883, is the only satisfactory accelerator of the digestion process. Their micro-Kj eldahl procedure makes use of sulphuric acid, potassium sulphate arid mercuric sulphate to effect complete digestion within 30 minutes of the clarification of tlie solution. Before distillation of the ammonia, zinc dust is added to reduce mercuric oxide to mercury, as otherwise a considerable fraction of the ammonia is bound by the mercuric oxide (precipitated by the added sodium hydroxide) and cannot be liberated by boiling.A convenient new form of all-glass apparatus for steam distillation has been described by Markham. l9 Despite criticisms1* of selenium as a catalyst, the colorimetric method of Campbell and Haniia,2" comprising micro-digestion and direct nesslerisation, has been found convenient and satisfactory in clinical work.6 SALT: MICRO-ANALYTICAL METIIO1)S [Vol. TS Several colour reactions, specific for characteristic groups occurring in all proteins, have led to colorimetric micro-methods in which preliminary destruction of the proteins is unnecessary. The most successful methods have been based on the constant occurrence of tyrosine or arginine, or of the “twin” peptide linkages found in the amino-acid chains of the protein molecule.The tyrosine reactio:t--In the method of Greenberg,’l the phenol reagent of Folin and Ciocalteu was used to give a blue colour with the tyrosine moiety of the proteins, and the colour was compared with that from a pure tyrosine standard. As tryptophan also gives some colour, the method is not specific, and an empirical relationship between the amounts of each protein and the colour intensities produced has to be established. )Yokes and Still22 investigated the method and found that it gave results lower than those given by Kjeldahl nitrogen determinations and that the values increased with time of storage of the sera before analysis. Several improvements to Greenberg’s technique were subsequently described and various procedures were developed including a preliminary heating of the protein in sodium hydroxide solution before proceeding with the colour reaction.This simple modifica- tion overcame the difficulties mentioned, although the method still remained empirical. I have found the technique of Minot and Keller32 to be both the simplest and the most reliable. An interesting further development of the reaction between proteins and the phenol reagent has been reviewed by Lowry, Rosebrough, Farr and Kanda11,24 and applied by them to a method for estimating as little as 0.2 pg of protein. By a preliminary reaction for 1 0 minutes at room temperature, the protein is held in an alkaline copper solution. The phenol reagent is then added and a blue colour appears in 30 minutes and, with serum proteins, has an intensity seven times that produced when the preliminary reaction is omitted.Although the procedure suffers from the disadvantage that the intensity of the colour varies with different proteins and is not strictly proportional to the concentration of protein, the greater sensitivity achieved makes it especially useful for the determination of small amounts of protein of fairly standard composition. The arginine reaction-The Sakaguchi reaction for arginine has been developed quantita- tively by Albanese, Saur and I r b ~ . ~ ~ In this micro-method, the plasma sample is diluted with 10 per cent. sodium hydroxide solution; water and ethanolic cc-naphthol are then added, and finally 0.06 N sodium hypochlorite and 20 per cent. urea solutions. Keyser26 found it necessary to increase the hypochlorite concentration to 0.15 2c’ to achieve satisfactory estimations of total protein, and he experienced other difficulties when attempting to estimate albumin and globulin separately.I have not found any special advantage in this nietliod; moreover, it cannot be said that the arginine content of the various plasma proteins is niore constant than the tyrosine content. The ninhydrin reactioir-The colour reaction2’ between cc-amino-acid groups and ninhydrin has been applied by Kunkel and Ward28 to the determination of protein, especially the sinall quantities obtain able fractionally from serum after immunochemical precipitation. A1 t hougl1 non-specific, inasmuch as peptides, amino-acids, amines and ammonia also interact , and although variable, in that different proteins give different intensities of colour, this simple method is highly sensitive and especially useful for minute quantities of particular proteins, for as little as 3 pg of albumin can be determined accurately by it.T?ze biziret reactiout-In my experience, the biuret reaction, in its modern form, has proved to be the most reliable absorptiometric method for direct determinations of protein in clinical work. All plasma proteins give the colour response at approximately the same intensity and the reaction is unaffected by small amounts of ammonium ions, which may remain after ammonium sulphate has been used for fractionation. Various biuret reagents and practical procedures have been devised by different analysts during the past four decades. The reaction itself has been re-examined recently by Mehl, Pacovska and Winzler,29 who show it to be similar to, but not identical with, that used by Lowry et ~ ~ 1 .2 ~ in their preliminary treatment of protein, before colour reaction with the phenol reagent, by the method noted above. A biuret technique has been adapted for visual colorimetry by making use of the Lovibontl comparator. This simple application gives an approximate determination of protein, but the method is not without its difficulties. and sekeral improvements have been introduced by Love and I i m i ~ d e n ~ ~ in a recent revision of the technique. BY COLOUR REACTIONS FOR MOLECULAR GROUPINGS- Some of these were studied by Wokes andJan., 19531 FOR PROTEINS IS BLOOD PLASMA.A CRITICAL REVIEW 7 For accurate protein determination, refined absorptiometry of the biuret colour is essential. 'This is achieved by use of the Ilford No. 605 colour filter in the photo-electric instrument described by Salt,15 as slightly modified to accommodate optical cells of 13 mm optical length. Extensive use of the method of Kingsley31 has in the past provided excellent results, but this has now been relinquished in favour of a newer procedure based on the studies of Weich~elbaum.~~ In recent years many analysts have sought to improve the biuret technique by establishing optimal reaction conditions and to simplify it by devising a reagent containing the copper and alkali together in one solution. A later technique of KingsleyS was criticised by Gornall, Bardawill and Davida and was further modified by who chose to apply the biuret reaction in highly alkaline medium and to use a solution of potassium permanganate as an artificial colour standard.Levin arid B r a ~ e r ~ ~ have also devised a 11 ighly alkaline biuret reagent containing both sodiuni and ammonium hydroxides. They found that ammoniiim hydroxide by itself failed to bring about the reaction. In contrast, Weichselbaum32 has put forward a biuret reagent h ~ i n g a high copper content, stabilised with sodium potassium tartrate and potassium iodide, and containing a small amount of sodium hydroxide. The method was devised for visual colorimetry, but was also described in a footnote32 in a suitably modified form for photo-electric ahsorptiometry.Experience with these several reagents has led me to adopt the weak117 alkaline type of reaction, despite the claima that optimal conditions are only established by the use of a higher concentration of sodium hydroxide. Weichselbaum's absorptiometric reagent has been effectively applied by Wolfson, Cohn, Calvary and Ichiba3' to the determination of serum protein and protein fractions in an analytical scheme that is described below (p. 9). The reagent is invariable, keeps indefinitely and rarely shows any turbidity even with highly lipaemic sera. If a turbidity does appear, the coloured solution can be clarified by shaking with ether and centrifugation. This procedure is more effective than the method of correcting for turbidity described by Keyser and \Taughn3s in connection with a biuret reaction of ;t different type.30 939 T H E DIFFEREXTIAL DETERMIXAl3OX OF PLASMA I'lIOTEISS In many types of disease the blood plasma is found to contain normal proteins in dis- proportionate amounts, and refined methods oi separation are therefore needed before exact analytical assessment of the several fractions can be made.Many methods of separation have been devised and comparative studies have been undertaken to (letermine in detail the protein patterns characteristic of health and of disease. E:LECTIIOPIIORETIC SEPARATION- In recent years, the moving-boundary method of electrophoretic :~nnlysis, based on tiit. principles laid down by Ti~elius,~O has become a standard method by which to evaluate other procedures for plasma protein fractionation.In the electrophoretic technique, the proteins are separated on the basis of their differential mobility in an electric field, and the determinations are made refractometrically. Other modes of differential deterniinatioii have been compared with the electrophoretic method by Edsall,8 Gutman!) and by Marracli and H o ~ h , ~ l with conclusions that generally testify to the superiority of electrophoresis. The moving-boundary electrophoretic procedure is, however, too elaborate and costly for routine use in clinical analysis, so that it will not be considered further here. elegant new technique for micro-electrophoresis of proteins on filter-paper, wl iicli has recently become available, is attractively simple and deserves detailed consideration. In principle, a strip of filter-paper is soaked in barbitone buffer at a pH of 8.6, a small volumc (as little as 0-01 ml) of serum is applied at a point on the paper, and a direct current (at, say, 100 volts) is passed for a sufficient time (about I8 hours) through the paper.After the period of electrophoresis, the paper is dried and the proteins are stained with a suitable dye, where- upon it is seen that the albumin fraction has migrated furthest along the paper strip towards the anode, and the ul-, u2-, 18- and y-globulins have migrated through progressively shorter distances, in that order. The somewhat elaborate apparatus of Cremer and T i ~ e l i u s , ~ ~ and the simpler forms described by Grassmann, Hannig and Knede1,43 Goa44 and Kunkel and T i s e l i ~ s , ~ ~ all contain the paper strip in a horizontal position.L a t r ~ e r ~ ~ has described an apparatus tl7iCt holds the paper sloping downwards at an angle of 24", while Durrum47 and Flynn and de Mayoj*'8 SALT : MICRO-ANALYTICAL METHODS [Vol. 78 have achieved protein separation on folded paper strips held in a steeply inclined position. Staining of the proteins has been accomplished by means of azocarmine,4Q J % ~ ~ bromo- phenol blue,42s44@s47s48 or Amidoschwarz 10B (Naphthalene Black 12B, 200).83s48 Ultimate differential determination of the proteins was completed by elution of the dye from cut portions of the paper when bromophenol blue was used, or by an elegant method of direct photometry of the coloured portions of the paper when the black dyea or azocarmine50 was used.An ingenious process of “retentiometry” was devised by Wieland and Wirth51 in connection with their azocarmine method. I have obtained good results, especially when using Whatman No. 100 filter-paper in the horizontal or slightly sloping positions, in apparatus similar to that of Grassman et U Z . , ~ and without the complication of any cooling device. The determinations have been completed successfully by using the acid-aqueous bromophenol blue staining method and subsequent washing with 0.5 per cent. aqueous acetic acid,& and also by staining with naphthalene black48 and estimating the proteins by direct photometry.& Micro-electrophoresis on paper, besides being simple, provides a means for the absolute separation of the several protein components from minute amounts of serum, and also an opportunity for analysing the fractions for components other than the proteins.For a general account of the possibilities of this technique the comprehensive paper of Kunkel and Tiselius& is especially recommended. DIFFERENTIAL PRECIPITATION BY ALCOHOL- The large-scale procedures, devised by E. J. Cohn and his colleagues (see review by Edsalls) for the separation of protein and other components of plasma, by means of various concentrations of ethanol at low temperatures with controlled pH and ionic strength, have now been applied to smaller volumes of plasma by Cohn et d5 Their new scheme, “Method 10,;’ has been further adapted by Lever, Gurd, Uroma, Brown, Barnes, Schmid and S c h ~ l t z ~ ~ to provide detailed analyses of 5.0-ml samples of plasma for their contents of various proteins, protein-bound lipids and carbohydrates.Some special equipment is required, but the scheme has the advantage that the biochemical properties of the separated proteins are unaltered. An application of this scheme of analysis to series of pathological plasma samples would be a valuable extension of this wwk and likely to produce important new knowledge. In a similar, but simpler, manner Pillemer and Hutchinson= effected separation of serum albumin and globulin at 0” C by means of 42.5 per cent. methanol at a pH of 6.8. Whereas these authors completed the determinations by Kjeldahl digestion, Christensen” demonstrated that the biuret reaction could be applied to the methanolic filtrate containing albumin.Salt15 has described a dispersimetric method for the direct determination of globulin in diluted serum, wherein the globulins are precipitated by 45 per cent. methanol at a pH of 6.6 in the presence of ghatti gum in 1 hour at 37°C. Under these conditions a uniform colloidal suspension of denatured globulin is formed, having considerable light-dispersing power especially for the violet waveband, so that a high degree of sensitivity is achieved. FRACTIONATIOK BY SALTING-OUT- In clinical chemistry, salting-out techniques have been both widely used and vigorously criticised. Often they have been modified or elaborated inorder to yield protein fractions closely in agreement with those obtained by electrophoretic procedures. Most interest has been centred around the globulin components of the serum, some of which are now known to have special biochemical functions in vizo.Ammonium sulphate-The classical method of separating plasma proteins by salting-out with ammonium sulphate is now regarded as unreliable except for the precipitation, by 0.34 saturation, of the y-globulin fraction in a fairly pure ~ t a t e . ~ The difficulties of nitrogen estimation, as a result of using ammonium sulphate precipitation of the protein, have been avoided in determinations of serum y-globulin by the use of the biuret r e a ~ t i o n , ~ ~ , ~ ~ or by a turbidimetric techniq~e.~6~~~ These simple methods are of practical value, despite failure to obtain exact correlation between amounts of precipitated protein determined t urbidime t ric ally and y -globulin isolated electrophoretic ally , either quali t a t ively58 or q~antitatively.~~ 969 Sodium suZphate-The classical method of Howeso for separating serum proteins into albumin and three globulin fractions by means of sodium sulphate at final concentrations of 21-5, 17.4 and 13.5 per cent.has been widely used until recent times, when it became evidentJan., 19533 FOR PROTEINS I N BLOOD PLASMA. A CRITICAL REVIEW !I that there were discrepancies between the fractions so obtained and those given by electro- phoresis. For example, Peterman, Young and Hogness61 were able to show that albumin values obtained by the Howe technique were in close agreement with the sum of the albumin and a-globulin concentrations determined electrophoretically.Later investi- g a t o r ~ ~ ~ , ~ ~ , ~ ~ , 6 ~ , ~ 6 , 6 7 have obtained results which, although riot altogether in agreement, have indicated that a better separation of serum proteins can be achieved by salting-out witl; sodium sulphate at the higher concentrations of 26-0, 1 9 4 and 15.0 per cent., the resulting iractions being then similar to the albumin and the three main globulins as separated 11) electrophoresis. *4s a precipitant, sodium sulphate has not been entirely satisfactory in my laboratory, because long and difficult filtrations in the warm are necessary and some inaccuracy can aristi from adsorption of proteins on the filter-paper.6s The separation of globulin from albumiii by means of sodium sulphate can be done, however, without adsorption error by the technique of Kingsley,G9 in which ether is also gently shaken into the mixture and separation of the.globulin effected in the centrifuge. Sodium sulfihite-Other neutral salts have been used for fractionation of serum proteins, notably sodium sulphite at reagent concentrations of 21.0, 18.0 and 15-0 per cent., as described by Campbell and Hanna.'O Although the 21-0 per cent. reagent is now regarded as insufficiently concentrated to precipitate all of the globulins from serum, the scheme of Campbell and Hanna gave fractions comparable with those prepared by the Howe technique, and without the need of working in the warm. The clinical usefulness of sulphite fractionation has been demonstrated by Salt7* for the serum proteins of patients with chronic rheumatic disorders.Covzbinatioizs o j saltiizg-ozd @vocedures-In 1948, Cohn and W0lfson7~ introduced the use of sodium sulphite at the higher concentration of 28.0 per cent. for precipitation of ali globulins from serum. This method was further studied73 and was finally elaborated to a form37 that provided rapid estimations of serum albumin and of the globulins in threc fractions. For the differentiation of the globulins, use was made of sodium sulphate at ;L reagent concentration of 234 per cent. and of an ammonium sulphate - sodium chloridc solution. This last reagent was devised as 19.3 per cent. ammonium sulphate in 4.0 per cent. sodium chloride, but was later modified to include only 3.0 per cent. of sodium chloride, as noted by de la Huerga and Popper.56 I have made extensive trials of the procedure of Wolfson et aL37 with satisfactory i-esult,c.It should be noted, however, that the biuret reagent described by the authors of the method is stated to contain 90 g of Rochelle salt, whereas I prefer only 30 g in 2 litres of reagent. The lower quantity appears to be more satisfactory, and it gives results that agrec with the original biuret procedure of Wei~hselbaum.~~ The separations of precipitated proteins are all quickly made in the centrifuge, ether being used as an aid in the sodium sulphate fractionation, and in the sodium sulphite fractionation ether containing 1.0 per cent. of Span 20. Several other non-ionic surface-active agents are equally effective, Empilan AQlOO being the one I normally use.Although this simple and useful scheme is recommended, it must be recognised that the separation of serum proteins into characteristic fractions, each in a state of purity, is inipossiblc by any kind of salting-out process. 'The comparative studies of Jager, Schwartz, Smith, Xickerson and Hr0wn7~ revealed that the protein obtained in filtrates from four procedures for the salting-out of total globulin in no instance consisted solely of albumin, as shown elec trophoretically. Despite this, evidence for the reliability of the methods of Wolfson et aZ.37 for determining the albumin and the albumin plus a-globulin fractions was recorded, and similar support was provided by I-evin, Oberholzer and Whitehead75 in respect of these €ractioiis and of y-globulin. Studies were also made by Popper, de la Huerga, Franklin, Bean, Faul, Routh and Schaffner7'j in order to evaluate the chemical methods of Wolfson et aZ.,37 the electrophoretic process being used for reference.The results led to a recommendation of the chemical methods for albumin, total globulins and y-globulin, but to only a cautious acceptance oE the determinations of K- and ,8globulins. These two fractioiis are both obtained by thc method of difference, and this, together with the fact that the globulins have lipid and carbohydrate moieties combined with them, may account for the larger experimental errors. I t may be recalled that 0 t h e r s ~ 6 , ~ ~ who used sodium sulphate for fractionation, also recorded poor results for the determinations of a- and ,&globulins.10 SALT : MICRO-ANALI’TICAL METHODS [Vol.‘78 When the lipid content of serum is greatly increased, special difficulties are encountered in both separating and estimating the protein fractions. Popj Ak and McCarthy77 have shown that complete saturation of lipaemic or of normal sera with magnesium sulphate successfully separates the globulins from albumin. after a few trials, questioned this claim, but Jager et aZ.7* established its validity. The main disadvantages of the magnesium sulphate procedure are that 24 hours are required for salting-out and that the biuret reaction is inapplicable, as the alkaline reagent gives a precipitate of magnesium hydroxide in the reaction mixture. I t is preferable to remove the lipids from lipaemic sera before analysing for proteins. The removal of lipids is done most successfully by shaking the serum with ether, freezing solid, then thawing ,and centrifuging; the whole process is repeated five times, each with fresh ether, as described by Popjiik and M~Carthy.~7 The cleared serum is then analysed without difficulty, either by the chemical method37 or by micro-electrophoresis on filter-paper.& Martin and THE SEFARATIOS AXD DETERMISATION OF CERTAIN PROTEINS BY THEIR SPECIAL PROPERTIES FIRRISOGEN BY COAGULATIOK- The classical method78 for the separation of fibrinogen from plasma, in which the addition of an excess of calcium ions activates the enzymic precipitation of fibrin, is an early example of a procedure that makes use of a special property of a particular protein.Subsequently, salting-out with 12.5 per cent.sodium sulphite7O or buffered sodium ~ulphate7~ was devised in order to include determinations of fibrinogen along with other plasma protein fractions in cwmprehensive schemes for differential analysis. Several of these methods were critically reviewed by Morrison,80 who then proceeded to a careful study of the clotting process brought about by the addition of a thrombin preparation derived from blood plasma. He defined the optimal conditions for coagulation of fibrinogen, such that complete separation of fibrin was effected, with but minimal occlusion of other proteins. Morrison’s method for fibrinogen has been usefully adapted by Ratnoff and Menziesl for the analysis of small clinical samples of plasma. Another methods2 that has been used successfully for the determination of plasma ‘tibrinogen depends upon the unique precipitability of fibrinogen when a solution of protamine and subsequently one of calcium acetate are added to dilute plasma.CRYOGLOBULINS BY COLD PRECIPITATIOK- Morrison’s studies80 on the coagulation of fibrinogen led also to the recognition, in normal plasma, of minute amounts of a protein that forms a reversible gel when a solution of it is cooled to 0” C. Similar proteins possessing the property of separating from solution when plasma is cooled had previously been discovered occasionally in the blood of persons with certain diseases. The earlier literature has been reviewed by Lerner and Watson,83 who described in detail a cold-precipitable globulin. The term “cryoglobulin” was proposed for this type of protein and further researches were recorded by Lerner and Greenberg.84 By a suitably devised technique, Lerner, Barnum and Wats0n8~ showed the presence of small amounts of serum cryoglobulin in many different disease states and its absence from the serum of normal controls.Subsequent contributions to our knowledge of the cryoglobulins have been reviewed and new data recorded in several recent publications.86~87~88 Evidently the cryoglobulins comprise a group of proteins with variable properties, but having one common property, zliz., that of spontaneous precipitation from solution at temperatures below 37” C, with the formation of a gel, an amorphous deposit or a crystalline precipitate. In all, the cryo- globulin can be separated from the cooled plasma and estimated, directly if the amount is large or by difference if small. When gel formation occurs in the plasma, the cryoglobulin may be absent from the serum, owing to occlusion with fibrin during the blood-clotting process.Hence, both serum and plasma samples should be examined for the presence of cryoglobulin by the methods indicated.85 787 GLOBULINS BY VISCOSITY MEASUREMEXTS-- high viscosity. Closely related to the occurrence of cryoglobulins is the existence of globulins of unusually Descriptions of this rare abnormality have been published,sg sgo and theJ;tn., 19531 FOR TJHOTI<INS I N BLOOD PLASMA. A CRITICAL REVIEW 11 teclmique of investigation , by measurement of serum viscosity at various temperatures, has been d e ~ c r i b e d .~ ~ ~ ~ ~ By the method of Woodmansey and 1&’ilsong2 for the measurenaent of viscosity, I have confirmed the value of that technique in three instances, but as the enhanced viscosity in these rare cases is related to the nature of the globulins rather than t o their absolute quantity, the method is chiefly of qualitative value. Small increases in the viscosity of serum or plasma may be found in a wide variety of disease states, owing mainly to increased concentrations of the normal globulins or of fibrinogen. The usefulness of single viscosity measurements is illustrated by results of the recent publications of Salt71 and of Houston, Whittington, Cowan and H a r k n e ~ s . ~ ~ Obviously, it is impossible to establish strict quantitative relationships, as abnormal proteins may also be present and others, especially albumin, may be depleted by disease.L a w r e n ~ e ~ * , ~ ~ has used viscometry for the assessment of the whole plasma and of tlie plasma proteins obtained in solution in four fractions after separation by chemical means, but does not advise conversion of the measured values to viscosity units or to protein concentrations. The interesting technique of Foster and Biguriag6 depends on the production of an increase i n the viscosity of serum by the addition of formaldehyde. This increase is due to interaction with the globulin moiety only, and the measured increase in viscosity may thus be related to the globulin content of the serum. ~IT<TAI.-COMBTXING GLOl3ITLIX BY ABSOKPTIOMETKIC TITRATION- A &-globulin constantly present in the plasma and possessing unique biochemical properties has been described by Surgenor, Koechlin and Strongg7 This protein, although present normally in only small amounts, is responsible for the transport of inorganic iron in the body and can be estimated readily by means of its iron-binding capacity.In the technique described by Rath and Finch,gs the existing serum iron is determined absorptio- metrically by means of the dipyridyl reactiong9 and the residual iron-binding capacity of the serum by absorptiometric titration with an acid solution of ferrous ammonium sulphate. The sum of the two values gives the total iron-binding capacity of the serum and, as this property is due entirely to the metal-combining /3,-globulin, its specific estimation is achieved.,ALRUMIN BY COMBINATION WITH HAEMIN- Iron in organic combination is not normally present in the plasma, but if haemin (ferri- protoporphyrin IX chloride) is added to plasma or serum, a coloured compound is formed specifically with the albumin fraction. The compound methaemalbumin has been studied by Rosenfeld and Surgenorloo and a method for the determination of albumin, based on its haemin-binding property, has been proposed. The method involves titration with haemin at a pH of 7.3, the titration being followed absorptiometrically at 403 mp. More complete details have now appeared,lol and the method has been used by Lever et for the determination of albumin in serum or, preferablv, in a serum fraction containing all the albumin together with only about 20 per cent.of the other proteins. ACID-SOLUBLE MUCOPROTEIKS- Protein-bound carbohydrate is present in the serum, being distributed generally on tlie globulins, and more especially throughout the a-globulins. These compounds, which have been variously designated as seromucoid, glycoprotein, mucoprotein and so on, are still only vaguely defined, although abnormalities in their concentration occur in certain diseases and these are of considerable importance. A distinct advance was made by Winder, Devor, Mehl and Smyth,l02 who used percliloric acid solution to precipitate the main plasma proteins and give an acid filtrate containing a mucoprotein fraction. The mucoprotein was then precipitated by means of phosphotungstic acid, the protein being estimated by the biuret reaction or from its tyrosine or nitrogen content.Alternatively, the characteristic carbohydrate fraction of the mucoprotein was determined by the colour reaction it gives with orcinol in sulphuric acid solution. This reaction has been studied by Friedmannlm and compared with similar reactions in which carbazole or skatole were used. Her results confirm the view that the hexose moiety of serum mucoprotein consists of mannose and galactose.12 SALT : MICRO-ANALYTICAL METHODS [I‘ol. 7s Greenspan, Lehman, Graff and Schoenbachlo4 have published careful details of their scheme for the assessment of serum “polysaccharide” (in arbitrary units), and for the cleter- mination of the carbohydrate of ethanol-precipitated serum protein and of separated miico- protein by condensation with tryptophan in the presence of sulphuric acid.In certain disease states, distinct abnormalities were revealed, including that of the ratio of hexose to protein (biuret value), which was frequently disproportionate, so indicating qualitative as well as quantitative changes in the mucoprotein fraction. Graff, Greenspan, Lehman and Holechek105 have adopted the anthrone reaction for the estimation of hexoses in protein and in the serum mucoprotein fraction. By this method a green colour is developed without the need for elaborately controlled reaction conditions, and all interference by non-specific reaction products is eliminated when measurements are made in the region of 620mp. I favour the use of the anthrone reaction, especially as it gives results agreeing closely with those by the tryptophan reagent.Besides mannose and galactose, glucosamine is also present in serum mucoprotein, although it is not estimated by any of the reactions above mentioned. West, Clarke and KennedylOG devised a method for hydrolysing diluted serum by boiling it with AT hydrochloric acid for 5 hours. The glucosamine in the neutralised filtrate is then determined absorptiometrically at 525 mp, after reaction with acetylacetone and condensation with 9-dimethylaminobenzaldehyde in hydrochloric acid. SPECIFIC I’KOTEIS FRACTIONS BY IMMUSOCHEMIC‘AL PRECIPITATIOS---- Since it became possible to prepare albumin and certain globulins of high purity, they have been available for the preparation of specific precipitating anti-sera from animals after suitable immunisation procedures. These anti-sera possess the property of precipitating specifically the homologous antigen from the plasma being analysed ; the amount of protein so obtained can then be determined by any suitably sensitive technique.In the method of Chow,107 plasma albumin is precipitated by this means and determined either turbidimetrically or by micro-Kjeldahl estimation of the nitrogen in the precipitate. Later, Chow and his colleaguesloS confirmed the value of the turbidimetric method, which avoids the necessity for separating the immunochemically precipitated protein and washing it before analysis. Moreover, the procedure was found to give results closely in agreement with those of electrophoretic analysis and superior to those obtained by salt-fractionation.Kunkel and Ward28 have also described a determination of serum albumin by precipitation with a specific anti-serum and estimation of the precipitated protein absorptiometrically after applying the ninhydrin reaction. In a similar way, Bendich and KabatlOQ used an anti-serum to precipitate y-globulin from plasma or serum samples and completed the deter- mination by nitrogen analysis of the precipitated protein. Provided that cross-reactions can be minimised between anti-sera and antigens closely related to the specific protein to be determined, immunochemical methods of serum protein fractionation are of great value. This is especially true when the amounts of serum available are minute, or when the particular protein is present in amounts too small for determination by any other technique. An immunochemical method for demonstrating the presence of Bence- Jones protein in serum, and for determining its concentration, has been described by Moore, Kabat and Gutman.llo In this procedure, the preparation of a potent anti- serum is difficult, but the method appears to be the only successful one for estimating this particular protein in serum.NORMAL VALUES A critical approach to the many techniques for separating and determining plasma or serum proteins and an appraisal of their several merits and difficulties clearly show that it is impossible to define exact limits for normality. Some analytical empiricism is inevitable and every result has to be related to the method by which it has been derived.Many of the publications already mentioned include valuable data about normal con- centrations of plasma protein components and tabulated values are presented in some of t h e ~ e . ~ , ~ , ~ ~ $2 The appropriate sections of a reviewlll of the chemical composition of blood plasma and serum also include values for electrophoretic and ethanol-separated protein fractions in normal plasma ; a recent publication112 on paper electrophoresis provides a useful comparison of the results produced by that technique with those by the original electrophoretic method of Tiselius.Jan., 19531 FOR PROTEINS I N BLOOD PLASMA. ‘4 CRITICAL REVIEW 13 REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 1s. 16. 17.18. 19. 20. 21. 22. 23. 24. 25. 26. 87. 28. 29. 30. 31. 32. 33. 34. 36. 36. 37. 38. 39. 40. 41. 42. 43. 44. 4.7. 46. 47. 48. 49. 50. 51. 62. 53. 54. ;IS. 57. 58. 59. 60. 61. 62. 63. 64. 65. 66. 67. -- .m. Pederson, I<. O., “Ultracentrifugal Studies of Serum and Serum Fractions,” Almqvist and Wiksells, Luetscher, J . A., jun., Yhysiol. Rev., 1947, 27, 621. Svensson, H., Advanc. Protein Chem., 1948, 4, 251. Mulford, D. J., A n n . Reu. Physiol., 1947, 9, 327. Cohn, E. J., Gurd, 1;. I<. K., Surgenor, D. M., Barnes, €3. A., Brown, K. K., Derouaux, G., Gillespie, J . M., Kahnt, F. W., Lever, W. F., Liu, C. H., Mittelman, I)., Mouton, R. IT., Schmid, K., and Uroma, E., J . ilmev. Chew. Soc., 1950, 72, 465. Treffers, H. P., ,4dvanc. Pyotein Chem., 1944, 1, 69. Kirk, P.L., Ibid., 1947, 3, 139. Edsall, J . T., Ibid., 1947, 3, 383. Gutman, A. H., Ibid., 1948, 4, 155. Kagan, B. M., J . Clin. Invest., 1938, 17, 369 and 373. Lowry, 0. H., and Hunter, T. H., J . Biol. Chem., 1945, 159, 465. Van Slyke, TI. D., Hiller, A., Phillips, H. A., Hamilton, f’. EJ., Dole, V. P., Archibald, R. M., and Sunderman, I?. W., Ibid., 1944, 153, 139. Looney, J. M., and Walsh, A. I., Ibid., 1939, 130, 635. Salt, H. B., J . I-ab. Clin. Med., 1950, 35, 976. Robinson, H. W‘., and Hogden, C. G., J . Hiol. Chew., 1941, 140, 853. Brand, E., Kassell, B., and Saidel, I,. J ., J . Clz%. Imvst., 1944, 23, 437. Hiller, A., Plazin, J., and Van Slyke, D. I)., ./. B i d . Chetu., 1948, 176, 1401. Markham, H., Riochn. J . , 1942, 36, 790. Campbell, W. K., and Hanna, M.I., J . Hiol. Chew., 1937, 119, 1. Greenberg, I). &I., Ibid., 1929, 82, 545. Wokes, F., and Still, 13. M., Biochpm. J.. 1942, 36, i97. Minot, A. S., and Keller, M., J . Lab. Clin. Med., 1935-36, 21, 743. Lowry, 0. H., Rosebrough, N. J., Farr, A. I,., and Randall, R. J., J . B i d . Chew,, 1951, 193, 265. Albanese, A. A., Saur, B., and Irby, V., J . Lab. Clin. Med., 1947, 32, 296. Keyser, J. W., Biockenz. J., 1948, 43, 488. Moore, S., and Stein, W. H., J . Biol. Cheni., 1948, 176, 367. Kunkel, H. G., and Ward, S. M., Ibid., 1950, 182, 597. Mehl, J . W., Pacovska, E., and Winzler, R. J., Ibid., 1949, 177, 13. Love, E. B., and Ramsden, F., J . Med. Lab. Terhnol., 1952, 10, 10. Kingsley, G. R., J . Biol. Chem., 1939, 131, 197. Weichselbaum, T. E., Amer.J . Clin. Path., 1946, 16, 40 (Tech. Sect.). Kingsley, G. R., J . Lab. Clin. Med., 1942, 27, 840. Gornall, A. G., Bardawill, C. J.. and David, M. M., J . Hiol. Chem., 1949, 177, 751. Kibrick, A. C., $1. Lab. Clin. Med., 1949, 34, 1171. Levin, R., and Brauer, R. W., Ibid., 1051, 38, 474. Wolfson, W. Q., Colin, C., Calvary, E., and Ichiba, F., Amw. J . Clin. Puth., 1948, 18, 723. Keyser, J. W., and Vaughn, J., Biochem. J., 1949, 44, xxii. Keyser, J. W., Ibid., 1949, 44, xxiii. Tiselius, A., Ihid., 1937, 31, 1464. Marrack, J . H., and I-Ioch, H., J . Clin. Path., 1949, 2, 161. Cremer, H., and Tiselius, A., Biochem. Z . , 1950, 320, 273. Grassmann, W., Hannig, K., and Knedel, M., Dtsch. med. Wschr., 1951, 76, 333. &a, J., Scand. J . Clin. Lab. Inuest., 1951, 3, 236. Kunkel, H.G., and ’Tiselius, A., J , Gen. Physiol., 1951, 35, 89. Latner, A. I,., Biochenz. J., 1952, 51, xii. Durrum, E. L., J . Amev. Chem. SOC., 1950. 72, 2943. Flynn, F. V., and de Mayo, P., Lancet, 1951, ii, 235. Turba, F., and Enenkel, H. J., Naturw’ssenschujten, 1950, 37, 93. Eisenreich, F., and Eder, M., Klin. Wschv., 1951, 29, 60. Wieland, T., and Wirth, L., <4ngew. Chem., 1950, 62, 473. Lever, W. F., Gurd, I;. H. N., Uroma, E., Brown, R. K., Barnes, B. A., Schmid, K., and Schultz, Pillemer, L., and Hutchinson, M. C., J . Biol. Chem., 1945, 158, 299. Christensen, H. N., J . Lab. Clin. Med., 1946, 31, 916. Jager, B. V., and Nickerson. M., J . Biol. Chem., 1948, 173, 683. de la Huerga, J., and Popper, H., J . Lab. Clin. Med., 1950, 35, 459. Levin, B., Oberholzer, V.G., and Whitehead, T. P., J . Clin. Path., 1950. 3, 284. de la Huerga, J., Popper, H., Franklin, M., and Routh, J. I., J . Lab. Clin. Med., 1950, 35, 466. Ricketts, W. E., Sterling, K., and Levine, R. S., Ibid., 1951, 38, 153. Howe, P. E., J . Bid. Chem., 1921, 49, 93, 109 and 115. Peterman, M. L., Young, N. F., and Hogness, K. R., Ibid., 1947, 169, 379. Majoor, C. L. H., Ibid., 1947, 169, 583. Milne, J., Ibid., 1947, 169, 595. Kibrick, A. C., and Blonstein, M., Ibid., 1948, 176, 983. Martin, N. H., and Morris, R., J . Clin. Path., 1949, 2, 64. Martin, N. H., Morris, R., and Smith, M., Ibid., 1950, 3, 266. Baker, R. W. R., and Merrivale, W. H. H., Scand. J . Clin. Lab. Invest., 1951, 3, 273. Uppsala, 1945. Eder, H. A . , Ibid., 1950, 183, 331. E.L., J . C l i ~ Invest., 1951, 30, 99.14 66. 69. 70. 71. 72. 73. 74. 75. 76. 77. 78. 79. 80. 81. 82. 83. 84. 86. 86. 87. 88. 89. 90. 91. 92. 93. 94. 95. 96. 97. 98. 99. 100. 101. 102. 103. 104. 106. 106. 107. 108. 109. 110. 111. 112. SALT [Vol. ‘is Robinson, H. W., Price, J. W., and Hogden, C. G., J . Bid. Chem., 1937, 120, 481. Kingsley, G. R., Ibid., 1940, 133, 731. Campbell, W. K., and Hanna, M. I., Ibid., 1937, 119, 9 and 15. Salt, H. R., Ann. Rheum. Dis., 1951, 10, 46. Cohn, C., and Wolfson, W. Q., J . Lab. Clin. Med., 1948, 33, 367. Wolfson, W. Q., Cohn, C., Calvary, E., and Thomas, E. M., Ibid., 1948, 33, 1276. Jager, B. V., Schwartz, T. B., Smith, E. L., Nickerson, M., and Brown, D. M., Ibid., 1950, 35, 76. Levin, B., Oberholzer, V. G., and Whitehead, T.P., J . Clin. Path., 1950, 3, 260. Popper, H , de la Huerga, J ., Franklin, M., Bean, W. B., Paul, W. I)., Routh, J. I., and Scbffner, F., Popjhk, G., and McCarthy, E. F., Biochem. J . , 1946, 40, 789. Cullen, G. E., and Van Slyke, D. I)., J . Biol. Cheulz., 1920, 41, 587. Hill, R. M., and Trevorrow, V., J . Lab. Clin. Med., 1941, 26, 1838. Morrison, P. R., J . Awer. Chem. Soc., 1947, 69, 2723. Katnoff, 0. D., and Menzie, C., J . Lab. Clin. Med., 1951, 37, 316. Mylon, E., Winternitz, M. C., and de Siito-Naigy, G. J., .J. Bid. C h e w , 1942, 143, 21. Lerner, A. B., and Watson, C. J., Atnev. J . Med. Sci., 1947, 214, 410. Lerner, A. B., and Greenberg, G. R., J . Biol. Chem., 1946, 162, 429. Ixrner, A. B., Barnum, C. P., and Watson, C. J., Anier. J . Med. Sci., 1947, 214, 116. Hill, R. M., lhinlop, S. G., and Mulligan, R. M., .I. Lab. Clin. Med., 1949, 34, 1057. Harr, D. P., Header, G. G., and Wheeler, C. H., Ann. Intern. Med., 1950, 32, 6. Rorvik, K., Acta Med. Scand., 1950, 137, 390. Shapiro, S., lioss, V., and Moore, D. H., J . Clin. Invest., 1043, 22, 137. Lucey, H. C., Leigh, E., Hoch, H., Marrack, J. R., and Johns, 13. G. S., Brit. J . Ex?. Pccth., 1930, Waldenstrom, J., Acta Med. Scand., 1944, 117, 216. Woodmansey, A., and Wilson, J. V., Ann. Rheum. &is., 1948, 7, 236. Houston, J., Whittington, R. B., Cowan, I. C., and Harkness, J., J . Clin. Invest., 1949, 28, $32. Lawrence, J. S., Ann. Rheum. Dis., 1949, 8, 209. -, J. Clin. Path., 1950, 3, 332. Foster, S., and Biguria, F., J . Lab. Clin. Med., 1943, 28, 1634. Surgenor, D. M., Koechlin, B. A., and Strong, L. E., J . Clin. irmml., 1949, 28, 73. Rath, C. E., and Finch, C. A., Ibid., 1949, 28, 79. Kitzes, G., Elvehjem, C . -4., and Schuette, H. -4., J . Riol. Chew%., 1944, 155, 653. Rosenfeld, M., and Surgenor, D. M., Ibid., 1950, 183, 663. , , Ibid., 1952, 199, 911. Winzler, R. J., Devor, A. W., Mehl, J. W., and Smyth, I. XI., J . Clin. Invest., 1948, 27, 609. Friedmann, R., Biochem. J . , 1949, 44, 117. Greenspan, E. M., Lehman, I., Graff, M. M., and Schoenbach, E. B., Cancer, 1951, 4, 972. Graff, M. M., Greenspan, E. M., Lehman, I. R., and Holechek, J. J., J . Lab. Clin. Med., 1951, West, R., Clarke, D. H., and Kennedy, E. M., 1. Clin. Invest., 1938, 17, 173. Chow, B. F., J . Biol. Chem., 1947, 167, 757. Chow, B. F., Homburger, F., DeBiase, S., and Peterman, M. L., J . Lab. Clin. Mod., 194$, 33, Bendich, A., and Kabat, E. A., Ibid., 1949, 34, 1066. Moore, D. H., Kabat, E. A., and Gutman, A. B., J . Clin. Invest., 1943, 22, 67. Krebs, H. A., Ann. Rev. Biochem., 1950, 19, 409. Koiw, E., Wallenius, G., and Gronwall, A., Scand. J . Clin. Lab. Invest., 1952, 4, 47. Amev. J . Clin. Path., 1950, 20, 630. 31, 380. -- 37, 736. 1052. THE BIOCHEMICAL LABORATORY THE ROYAL INFIRMARY WORCESTER August 28th, 19.52

ISSN:0003-2654    

DOI:10.1039/AN9537800004

出版商:RSC    

年代:1953

数据来源: RSC

摘要:

Jan., 19531 MORRIS AND JONES 15 A Plate-Assa y Technique for Biotin, Nicotinic Acid and Pantothenic Acid BY S. MORRIS AND A. ,JONES (Presented at the meeting of the BioZogical Methods Group on Friday, June lst, 1951) Methods have been devised for the assay of biotin, nicotinic acid and pantothenic acid by the plate-assay technique. During the course of the investigatory work, parallel assays by the tube technique were made. The methods proposed show that, within certain limits, the plate assay can be used. The results are significantly lower than those by the tube assay, though for such practical purposes as plant control and laboratory control of pharmaceutical products the loss of accuracy is unimportant compared with the gain in time. The plate assay has been used for nicotinic and pantothenic acids in vitamin tablets with results in good agreement with the tube assay.THE increasing use of the vitamin-B group in pharmaceutical products and the need for scientific control in the extraction of the B complex from materials like yeast and liver has made rapid and accurate assay methods for the B vitamins a necessity. It is desirable that these assay methods should make use of the least possible number of variants in basal media and assay organisms. During the last few years, plate-assay methods have been published for vitamin B,, the B6 complex and i n o s i t 0 1 . l ~ ~ ~ ~ ~ ~ These methods are found by statistical analysis to be at least as precise as any method previously described with the added advantage of speed and ease of manipulation. This paper deals with plate-assay methods for nicotinic acid, biotin and pantothenic acid.Not many details of plate-assay methods for these members of the B-vitamin group have so far been published. Bacharach5 first suggested that the cup-plate or plate method used in antibiotic estimation might be adapted to the assay of vitamins and amino-acids. Bacharach and Cuthbertson' next described the application of this method to thiamine (aneurine) estimations and to standard solutions of nicotinic acid and biotin, as did Genghof, Partridge and Carpenter6 and Williain~.~ The yeast, Saccharomyces car2sbergensi.s 4228, has already been used for the assay of the B, complex3 and of ino~itol.~ But in view of the high close levels for biotin, 0.08 to 0.64 pg per ml, and pantothenic acid, 0-5 to 10 pg per ml, and because nicotinic acid is not an essential growth factor for S.carlsbergensis, we used Lactobadus arabinosus for all three assays, METHODS MAINTENANCE OF CULTUFCE- L. avabinosus 17/5 is maintained by fortnightly subculture on yeast water - glucose agar. ASSAY MEDIA-- For the assay of pantothenic acid and nicotinic acid, Connor Johnson'sg modification of Wright and Skegg~'~ medium is used. For biotin, two media can be used, the modified Wright m d Skeggs'8 medium or Barton-Wright 'sl0 modification of Snell and Wright 'sll medium. For the assay of biotin, particular attention must be paid to ensuring that the acid casein hydrolysate is free from biotin. The hydrolysate we used, which was made in these laboratories, was treated, after neutralisation, thrice with charcoal, a t pH 5.5 to 6.0, 5 5 of hydrolysate being boiled with 2 g of charcoal each time.The number of treatments required may depend on the source of the hydrolysate and the grade of charcoal used. The media are distributed in boiling-tubes, 20 ml per tube, and sterilised by steaming for 20 minutes. Until required for assay the tubes are stored at room tcmpersturc, but for not longer than a month. When used after storage for periods longer than a month, the The media used are shown in Table I.I6 results are frequently irregular. in amounts of 300 to 500ml. MORRIS AND JONES: A PLATE-ASSAY TECHNIQUE FOR BIOTIN, [Vol. 78 Bulk pouring has also been used; the media being stored PREPAKATIOK OF THE INOCULUM- For the assay of nicotinic and pantothenic acids, the nicotinic acid basal medium of Wright and Skeggsg is used containing nicotinic acid and pantothenic acid but no agar.For the assay of biotin, a nicotinic acid basal mediumlo is used with the addition of @2pg per ml of nicotinic acid and 8mpg of biotin per 10 ml. In all the tubes 10 ml of inoculum medium are used. The medium is inoculated from a stock culture of L. arabinosw and incubated at 37" C for 18 to 20 hours. After centrifugation, the organisms are washed twice with sterile saline TABLE I Casein 113 tlrolysate . . r,-Cystinc . . .. nL-Tryptopha 11 . . Glucose . . . . --lmmoniuni sulphate . . Sodium acetate, hydrated Sodium chloritlc . . Adeninc . . . . (hanine . . . . IJracil . . . . . . Xanthine .. . . Inorganic salt solution A Inorganic salt solution I3 Thiamine . . . . Calcium r)-pantc)tlwnatc Riboflavin . . . . jb.Aminobenzoic acid . . Biotin . . . . .. Xicotinic acid . . . . Glass-distilled water to pH adjusted to .. .4gar . . . . .. Peptme (pl1utolysr.d) -/- Pyridoxine . . .. .. .. .. .. .. .. .. .. .. .. .. .. . . a . .. .. .. .. . . .. . . . . . . . . .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. BASAL MEDIA Pantothenic acid, Wright and Sktggs ( I ) 5.0 g 0.2 h' 0.2 g 20.0 ,g 33.2 g* -- _ . 0.005 g 0.005 h' 0.005 h' 0.005 h' 5.0 ml 5.0 ml 100.0 ml 1.0 mg 1.0 mg _- 200.0 p g 100.0 pg 1.0 Crg 1.0 mg 1000.0 ml 6.7-6.8 20.0 g Xicotinic acid, Wright and Slreggs ( I ) 5 . 0 g 0.2 g 0.2 g 20.0 g 33.2 g ___ 0.005 g 0.005 g 0.005 0" 0.005 g 5.0 ml 5.0 ml 1.0 nig 1.0 m g 1.0 mg - 200.0 pg 100*0 pg 1.0 Pg - 1000.0 ml 6.7-6.8 20.0 g Biotin r Wright and Skeggs (1) 5.0 g 0.2 g 0.2 h' 20.0 g __ 33.2 g -- 0.005 g 0.005 g 0.005 g 0.005 5.0 ml 5.0 ml 1.0 mg 1.0 mg 1.0 mg -- 200.0 pg 100.0 pg - 1.0 mg 1000.0 ml 6.7-6.8 20.0 g 1 Snell and Wright (2) 5-0 g 0.1 g 0.2 g 10.0 g 10.0 g 5.0 g 0.01 g 0.01 g 0.01 g 0.01 g 5.0 In1 5.0 ml 3.0 g - 100.0 pg 200.0 pg 200.0 pg 200.0 pg 100.0 pg _- 400-0 pg 1000.0 ml 6.7-6.8 20.0 g * 10 g of sodium acetate can be used in place of 33.2 g, but since the other media contain the higher quantity, it has usually been added to thc pantothenic acid medium, for easy preparation of all three media.t Barton-W'right's book,1° p. 65. SOTSS-( 1) Connor Johnson's modification.(2) Barton-Wright's modification. and finally suspended in saline. A heavier concentration of organisms is necessary for the nicotinic acid and pantothenic acid assays. For these, the organisms are re-suspended in sufficient sterile saline to make the opacity of the suspension, if diluted 1 to 2, correspond to tube No. 5 of the Wellcome series of standard opacity tubes. For biotin, the opacity of' the suspension, undiluted, is made to correspond to tube No. 2 or No. 3. PREPARATION OF THE PLATES- Immediately before use the required number of tubes are heated in boiling water. In the biotin assay, four tubes are taken for the standard and four for each test sample; The tubes are maintained at 48" to 50" C and 1 ml After thorough mixing, the contents of each tube When cold, holes are cut in each plate,.in the other assays five tubes are used. of inoculum suspension is added to each. are poured into Petri dishes and allowed to cool. four for the biotin assay and five for the nicotinic acid and pantothenic acid assays.Jan., 1953; NICOTIXIC ACID AXD PANTOTHENIC ACID 17 PREPARATION OF THE SAMPLE- Riotif?-In preliminary extraction experiments, 3 N sulphuric and 6 N hydrochloric acids were used. A mean value of 0.84 pg per g with the former, and 0.89 pg per g with the latter was found for yeast 22. These figures are well within the -&15 per cent. limit usually accepted for microbiological assays. As 2 N sulphuric acid released as much biotin from yeast as 6 N hydrochloric acid, it was used in the extraction. One and a half grains of dried yeast are added to 15 ml of acid and autoclaved for 2 hours at 15 lb pressure.When cold, 2 ml of 2.5 M sodium acetate arc added, the pH value adjusted to 6-7 to 6.8 with sodium hydroxide and the whole diluted to 50 ml. Nicotinic acid-Becaiise of the higher levels required for the plate assay, the usud extraction ratio of 1 part of yeast to 500 parts of N hydrochloric acid was inadequate. A satisfactory ratio was found to be 1 to 10; for in a series of tube assays with this ratio a mean value of 509 pg per g was found, compared with 491 pg per g for the usual 1 to 5 0 ratio. One gram of dried yeast is added to 10ml of N hydrochloric acid and autoclaved for 20 minutes at a pressure of 15 lb. The pH value of the cold extract is adjusted to 6.7 to 6-8 with 5 N sodium hydroxide.The volume is adjusted to 20 ml, the extract filtered and the filtrate diluted 1 + 1, 1 -C 3, 1 -1- 7 and I + 15. Pantothenic acid-Two grams of dried yeast are added to 25 ml of 0.1 hl sodium acetate buffer of pH 6-8 and the mixture is steamed for 30 minutes. After cooling, the pH is adjusted to 4.5 with acetic acid and 0.5 g each of takadiastase and papain, together with a few drops of toluene, are added. The enzyme digest is steamed for 10 minutes, cooled and the pH value adjusted to 6.7 to 6.8 with sodium hj-droxide. The volume is adjusted to 50m1, the extract filtered and the filtrate diluted 1 - 1, 1 -t- 3, 1 -+ 7 and 1 4 35. Re fewpice standards-For biotin the standard solutions are prepared t o cover the range 5, 10, 20 and 40 mpg of biotin per ml, for nicotinic acid 1, 2, 4, 8 and 16 pg per ml and for pantothenic acid 0.25, 0.5, 1-0, 2.0 and 4.0 pg of calcium u-pantothenate per ml.PROCEDURE- Place 0.1 ml of the standard and test samples at each dilution in the appropriate hole of the Petri plates and incubate the plates overnight. The plates for the assay of biotin and pantothenic acid may be incubated at 30" or 37" C, whereas the nicotinic acid assay plates have always been incubated, in the present study, at 37" C. Measure the diameters of the growth zones in millimetres. For the standard graph, plot the mean diameters against the concentration of the vitamin used. (As is normally found with the B vitamins and a bacterial culture, doubling the vitamin concentration leads to a 2-mm increase in the diameter of the growth zone.) RESULTS Biotin-The only test sample used in the assay of biotin was dried brewers' yeast.The assay values for two yeasts, A and 13, are shown in Table 11, both for plate and tube methods. The mean assay values obtained for yeast A on the Barton-Wright medium (Table 11) were 1.36 pg per g for the tube and 2.26 pg per g for the plate. These results are significantly different (t, found = 2.61; t, 5 per cent. = 2.0). More marked differences were found with the other yeast, B, and with the Skeggs and Wright medium with both yeasts. For both yeasts the plate assay gave results lower than the tube assay to a lesser or greater degree. I t is interesting to note that there is good agreement between the results from the two media both in the plate and tube assays.The range of biotin for the plate assay is 5 to 40 mpg per ml, whereas for the tube assay it is 0.05 to 0.6mpg per ml. The tube assay is about one hundred times more sensitive than the plate assay, which is in general agreement with the relative sensitivities of the two methods. During the course of these experiments, Axelrod and Hofman's publication12 appeared, in which the authors state that on hydrolysing biotin with 2 N hydrochloric acid a loss of 80 per cent. took place, whereas with 2 N sulphuric acid the loss was negligible. These esprimctits were repeatetl with 1 pg of biotin dissolved in 15 In1 of water and by hydrolysing -After filtration, the extract is diluted 1 i- I, 1 -}- 3, and 1 -+ 7 for assay. The mixture is incubated for 3 days at 37" C.18 with 2 AT sulphuric acid and 6 N hydrochloric acid. A loss of 50 per cent.of the biotin was found with sulphuric and complete destruction with hydrochloric acid. In recovery experi- ments with yeast and added biotin subjected to extraction with 5 N hydrochloric or 2 N sulphuric acid, recoveries were fairly good. For example, by using 5 N hydrochloric acid and 1.5 g of yeast with and without 2.0, 1.0 and 0.5 pg of biotin, a mean value of 36.9 mpg per ml of biotin was found for the yeast, and 71.2, 59-8 and 44.2 mpg per ml, respectively, MORRIS AND JONES : A PLATE-ASSAY TECHNIQUE FOR BIOTIN, [Vol. 78 1 ; I OT IN-- A A A x R n B 13 TABLE I1 ASSAY RESULTS OF THE BIOTIN, NICOTINIC ACID AND PANTOTHENIC ACID CONTENT OF YEAST BY THE PLATE AND TUBE TECHNIQUES Medium Barton-Wright Rarton-Wright W’right and Skeggs Wright and Skeggs Rarton-Wright Barton-Wright Wright and Skeggs Wright and Skcggs Method 1 Tube Plate Tube Plate Tube Plate Tube Plate Mean, ug per g 1.36 I -26 1.41 1.19 1-05 0-70 0-97 0.8 I S.E.of mean 0.024 0.027 0.033 0.046 0.023 0.009 0.012 0-021 S.E. of single estimate 0.12 0-16 0.09 0.16 0.09 0.05 0.03 0.09 NICOTISIC ACID- A Connor Johnson Tube 489.0 7.33 28.39 A Connor Johnson Plate 450-0 4.13 23.30 I’ANTOTHENIC ACID- 885 Connor Johnson Tube 73.0 0.67 2-24 710 Connor Johnson Tube 70-1 0.96 3-82 7 1 0 Connor Johnson Plate 63.9 1.01 4-30 S85 Connor Johnson Plate 64.1 0-89 6.25 TABLE I11 NICOTINIC ACID ASSAY VALUES OF A YEAST BY THE TUBE AND PLATE METHODS S.E. of a single Extraction Nicotinic acid, Mean, S.E.of mean estimation PFLg Per g fG Per g Tube ussay- 0.1 g yeast +- 50 nil HC1, 440 470 495 489 7.33 28.39 151b, 20min. 475 460 540 480 500 500 460 510 530 475 500 515 Y l d r ussay- 1.0 g yeast 4- 10 ml -Y HCI, 439 456 431 460 15 lb, 20 min. 436 472 452 434 464 429 432 480 407 479 457 457 480 472 445 466 470 404 488 467 433 440 423 479 468 436 445 481 412 4.13 23.30 for yeast with added biotin, when the samples were diluted to 50 nil after extraction, i.e.,. recoveries of added biotin were 86, 115 and 73 per cent. Similar figures were obtained when 2 N sulphuric acid was used. It would appear that yeast prevents destruction of biotin during acid hydrolysis. Nicotinic acid-The results by both the tube and plate assay techniques for a dried The plate method appears to give values statistically lower yeast are shown in Table 111.than the tube (t, found = 6.42; t, 5 per cent. = 2.02).Jan., 19531 NICOTINIC ACID AND PANTOTHENIC ACID 19 A comparison has been made between the two methods for a variety of materials (Table IV). It is possible that variations in the method of extraction are necessary for different materials. This paper deals essentially with yeast and yeast products; the require- ments for other foods were not investigated. TABLE IV NICOTINIC ACID ASSAY VALUES OF VARIOUS MATERIALS BY TUBE AND PLATE METHODS Nicotinic acid, pg per g r Tube assay Plate assay A \ Oats . . . . . . . . . . . . 8.0 9.7; 6.3 Dried milk . . . . . . . . . . 7.96 4.8; 6.3 Wheat .. . . . . .. . . 48.3; 45.0 43.4; 44.9 Maize .. . . . . . . . . 19.1 8.1; 10.4 Dried blood . . . . . . . . . . 60.0 58.5; 55.8 Pea meal . . . . . . . . . . 32.0 39.0 207.0 Peanut flour . . . . .. . . 292.0 Yeast extract (Difco) . . . . . . 295.0 306.0 Meat extract (Lemco) . . . . . . 1420.0; 1195.0 896.0; 966.0; 980.0 Malt extract .. .. .. .. 63.0; 52.0 Puntothenic ucid-The values recorded in the pantothenic acid assay of two yeasts are shown in Table V. With both yeasts, lower results, which are statistically significant, are recorded with the plate technique (t, found for yeast 885 = 6.32; t, 5 per cent. = 2.02: and for yeast 710, t, found = 4-15; t, 5 per cent. = 2.27). TABLE V PANTOTHENIC ACID ASSAY OF TWO YEASTS BY THE PLATE TUBE METHODS Piate assay 70 56 65 71 59 63 63 64 65 62 56 63 65 62 67 64 67 65 70 68 65 76 58 63 64 65 67 63 63 67 76 59 65 50 59 Yeast 886- I J Yeast 710- 66 57 68 58 72 60 60 68 6 5 ) 70 60 65 Mean 64.1 63.9 S.E.of S.E. of single mean estimate Tube assay 75 7 1 70 0.89 5 2 5 71 I 68 64 67 697 67 73 I 70 73 1-01 4.30 75 64 :; T;J 76 69 Mean 73.0 70.1 AND THE S.E. of mean 047 0.96 S.E. of a single estima- tion 2.24 3.82 The extraction method used was that of Harrison,13 with the omission of treatment by sodium hydroxide. For yeast samples, this treatment was found to be unnecessary, since 110 more pantothenic acid was extracted when sodium hydroxide was used. This extraction method appeared to be the most satisfactory available, although Neilands and Strong1* found that chicken or pigeon-liver enzymes released more pantothenic acid from the bound form.Unfortunately, the difficulty of obtaining a sufficient quantity of fresh chicken or pigeon livers limits the use of these enzymes.SO CORBETT THE DETERMINATION OF MAGNESIUM [Vol. 78 In the pantothenic and biotin assays the growth zones appear more rapidly than in the nicotinic assay and it has been possible to measure the zone diameters after 4+ hours incubation. This has made it possible to report an assay value for a test sample on the same day as that on which the assay has been started. This short-incubation modification has not been fully investigated, but it may be of value, especially for the purpose of works control. The authors wish to thank Miss M. Copus, Miss A. Braddick and Mr. K. J. Stevens for technical assistance, Mr. H. V. Thorne for assistance with the statistical analyses, and the Directors of Beecham Research Laboratories Limited for permission to publish this work. REFERENCES 1. 2. 3. --,- , Ibid., 1950, 75, 608. 4. 5. 6. 7. Williams, T. I., Nature, 1948, 161, 19. 8. 9. 10. 11. 12. 13. 14. Bacharacli, A. L., and Cuthbertson, W. F. J., Analyst, 1948, 73, 334. Jones, A,, and Morris, S., Ibid., 1949, 74, 333. Jones, A., Ibid., 1951, 76, 588. Bacharach, A. L., Nature, 1947, 160, 640. Genghof, 11. S., Partridge, C. W. H., and Carpenter, P. H., Arch. Biochem., 1948, 17, 413, Connor Johnson, B., “Methods of Vitamin Determination,” Burgess Publishing Co., Minnesota, 1848. Wright, L. D., and Skeggs, H. R., Proc. Soc. Exp. Biol. Med., 1944, 56, 95. Barton-Wright, E. C., “Practical Methods for the Microbiological Assay of the Vitamin B Coniplcx and Essential Amino-Acids,” Ashe Labs. Ltd., London, 1946. Snell, E. E., and Wright, J . L., J . Biol. Chem., 1941, 139, 675. Axelrod, A. E., and Hofman, K., Ibid., 1950, 187, 23. Harrison, J. S., Analyst, 1949, 74, 597. Neilands, J . B., and Strong, F. M., Arch. Biochem., 1948, 19, 287. REECHAM RESEARCH LABORATORIES LIMITED RROCKHAM PARK I4rst submitted, January 28th, 1952 Amended, July 31st, 1952 BETCHWORTH, SU RHEY

ISSN:0003-2654    

DOI:10.1039/AN9537800015

出版商:RSC    

年代:1953

数据来源: RSC

摘要:

SO CORBETT THE DETERMINATION OF MAGNESIUM [Vol. 78 The Determination of Magnesium and Aluminium in Titanium Metal BY J. A. CORBETT This paper deals with the separation of titanium from aluminium and magnesium by precipitation with cupferron and extraction with chloroform. The aluminium is determined colorimetrically with “aluminon” and the magnesium gravimetrically with ammonium phosphate after separation from the titanium. TITANIUM metal produced by the Kroll process contains from 0-2 to 0-3 per cent. of magnesium and sometimes small amounts of aluminium. The titanium is received as a powder and is pressed and melted in an arc furnace in an atmosphere of argon. It is of interest to know how much magnesium is retained in the metal after this procedure. It is necessary to remove the titanium completely from solution in order to estimate the magnesium and aluminium.The precipitation of large amounts of titanium as hydroxide or dioxide should be avoided because of adsorption. Wartman, Walker, Fuller, Cook and Anderson1 have suggested precipitation of the titanium, at a pH value of 5 , and saturation of the solution with sulphur dioxide to produce a crystalline precipitate that does not adsorb other metal ions. The separation of elements that can be precipitated with cupferrori and extracted with solvents is finding increasing use. Meunir2 found that titanium could be removed from a hydrochloric acid solution by means of cupferron and subsequent extraction with chloroform, and that small amounts of aluminium could be extracted in the pH range 2 to 5 in the same manner.A review of the use of this technique for separating metallic ions has been published by Furman, Mason and Pek01a.~ Rodden4 has reported the use of cupferron and chloroform in the separation of small amounts of various metallic ions. Short5 separated small amounts of iron in the same way.Jan., 19531 AND ALUMINIUM I N TITANIUM METAL 21 Subsequent to the completion of this work a similar method for the separation of magnesium from titanium has been reported by the Titanium Metals Corporation.6 EXPERIMENTAL Cupferron and chloroform seemed possibly to be applicable in separating titanium, aluminium and magnesium, and a method consisting in the following steps was investigated. (i) Complete separation of the titanium and any iron impurity with cupferron and chloroform.(ii) Adjustment of the pH of the aqueous phase from (i) to a value of 3.5 and extraction of the aluminium with cupferron and chloroform. The aluminium could be recovered from the chloroform and determined colorimetrically. (iii) After destruction of the organic material in the aqueous phase from (ii) the calcium could be precipitated with ammonium oxalate and the magnesium determined in the filtrate. SEPARATION OF TITANIVM- Titanium metal can be dissolved most readily in dilute sulphuric acid; the reaction with dilute hydrochloric acid is slower, but quite satisfactory. Because of the subsequent necessity to adjust the pH to 3.5 in a small bulk, hydrochloric acid was chosen as the acid medium. It was considered that 500mg was the maximum amount of titanium that could be handled without the use of excessive amounts of cupferron and chloroform.Tests showed that 7 g of cupferron was necessary to remove 500 mg of titanium from 100 ml of N hydrochloric acid solution. To ensure complete separation it was necessary to test the aqueous layer with a small amount of cupferron ; a white precipitate of nitrosophenylhydroxylamine indicated that all the titanium was removed. Better separations were attained with pure cupferron. With brown impure cupferron there was an excessive amount of organic material in the aqueous phase and some insoluble material at the interface between the chloroform and aqueous layers. At least four additions of chloroform were necessary to extract the cupferron precipitate.Complete extraction was easily identified as both phases became colourless. The presence of ammonium chloride was found to help the extraction. SEPARATION AND DETERMIXATION OF ALUMINIUM- No difficulty was experienced in separating the aluminium from the magnesium, The aqueous phase from the titanium separation was evaporated until most of the hydrochloric acid was removed, and the pH was adjusted to a value of 3.5 in a bulk of 20ml. The separation was made with small amounts of cupferron and chloroform, as only micro amounts of aluminium were present. The aluminium was recovered from the chloroform layer by evaporation and gentle ignition. The residue was dissolved in hydrochloric acid, and the aluminium estimated colorimetrically with aluminon. The results from known amounts of aluminium are tabulated in Table 11.No aluminium was found in the titanium.* ESTIMATION OF MAGNESIUM- Marsden7 has outlined the conditions for precipitating micro amounts of calcium with ammonium oxalate, and magnesium with ammonium phosphate. Both chemical and spectrographic analyses have shown that there is no calcium in the titanium under investiga- tion. The aqueous phase from the aluminium separation contains the magnesium, chloroform and organic material. The chloroform is removed by evaporation, but the organic material must be destroyed. Attempts were made to destroy the organic material by repeated evaporations with hydrochloric acid and hydrogen peroxide. Many evaporations were necessary and there was uncertainty as to when all the organic material had been removed. Tests were made to find if sulphuric acid, which could be used to destroy the organic material, would affect the final determination of magnesium.Aliquots of a * This has been verified by spectrographic tests conducted by the Defence Research Laboratories, Department of Supply, Melbourne, Australia.22 CORRETT THE DETERMINATION OF MAGSESIUM [Vol. 58 standardised magnesium chloride solution were measured into 100-ml beakers, 2 ml of diluted sulphuric acid (1 + 1) were added and the solutions evaporated until the sulphuric acid fumed. The temperature was raised and the solutions evaporated to dryness. The magnesium sulphate was then dissolved in 5 ml of 2 N hydrochloric acid, and the magnesium determined as under Method (p.23). Table I shows the results compared with those obtained from solutions containing no sulphuric acid, but that were otherwise treated in the $%me manner. TABLE I (hTVIPARISON OF KESL'LTS I?; THE DETERMINATION OF MAGNESIUM BOTH WITH AND WITHOUT SULPHURIC ACID PRESENT JVith sulphuric acid, RQ a jiricsiu 111 a.dtled, magnesium found, magnesium found, m g mg mg 0.55 0.50, 0.46 0.50, 0-48 1-10 1.06, 1.08 1.04, 1.07 2.20 2.18, 2-20 2.18, 2-18 With hydrochloric acid, The results show no significant difference due to the sulphuric acid treatment, and the The high sulphuric - nitric acid treatment was adopted to destroy the organic material. temperature needed to remove the sulphuric acid also decomposed any ammonium salts. METHOI) FOR THE SEPARATION OF TITANIUM REAGENTS- Hydrochloric acid-A 5 N solution.Czlpferron.-Dissolve 9 g of high grade cupferron (white) in 100 ml of water and filter. Ch,broform-Reagent grade. PROCEDURE- Dissolve 0.5 g of sample turnings or powder in 30 ml of 5 N hydrochloric acid solution at boiling-point. When solution is complete, make up the bulk of the solution to 20 ml with concentrated hydrochloric acid and dilute to 100 ml with water. Add 2 g of ammonium chloride and cool to 10°C. Transfer the solution to a 500-ml separating funnel and add 80ml of the 9 per cent. cupferron solution. Shake vigorously for at least 1 minute until the precipitate is well coagulated, and set aside for 5 minutes. Add 50ml of chloroform and shake for 1 minute, then set aside until the layers separate. Remove the chloroform layer, and repeat the separation with 25 ml of chloroform.Add 5 ml of 9 per cent. cupferron solution; a white precipitate, which dissolves, shows that all the titanium is precipitated. Add Z5ml of chloroform, shake for 1 minute and set aside until the layers are separated. Repeat the extraction until both layers are colourless. Filter the aqueous solution through a No. 42 filter-paper, and evaporate on a hot-plate. Evaporate the solution almost to dryness, dilute to 2Oml with water and dissolve any solid salts. Adjust the pH to 3.5 with dilute hydrochloric acid or dilute ammonium hydroxide; measure the pH with a glass electrode. Cool to 10" C and transfer the solution to a 125-ml separating funnel. Add 2 ml of 9 per cent. cupferron solution, shake for 1 minute and set aside for 5 minutes.Add 10ml of chloroform and shake for 1 minute; wait until the layers are separated, and then run the chloroform layer into a platinum basin. Repeat the extraction with 10-ml lots of chloroform, the chloroform being stored in the platinum basin. Three extractions are usually sufficient to remove the aluminium. METHOD FOR THE DETERMINATIOS OF ALI-MIKIUM REAGENTS-- AZuuminon-Dissolve 0-2 g in water and make up to I00 ml. Amvzoniuwz acetate-Dissolve 20 g in water and make up to 100 ml. PROCEDURE- remove organic material at a tempcrature not exceeding 500" C. Evaporate the chloroform extract to dryness in the platinum basin. Ignite gently to Add 5 ml of ConcentratedJan., 19531 AND ALUMINIUM IN TITANIUM METAL 23 hydrochloric acid and heat to dissolve any residue.Transfer to a 50-ml calibrated flask and wash out the basin with 30 ml of water. Add 10 ml of 20 per cent. ammonium acetate solution and 5 ml of 0.2 per cent. aluminon. Adjust the pH to between 4.9 and 5.0 with the aid of a glass electrode, by adding dilute hydrochloric acid and dilute ammonium hydroxide. Make up the solution to 50ml and measure the transmission of the colour with a suitable photo-electric colorimeter in a 1-cm glass cell and a green filter with maximum transmission at 5300 A. Estimate the aluminium concentration from a previously prepared standard graph. METHOD FOR THE DETERMINATION OF MAGNESIUM REAGENTS- Hydrochloric acid-A 2 N solution. .4mmonium Phosphate (dibasic)-Dissolve 5 g of ammonium phosphate in 50 ml of water.PROCEDURE- Transfer the aqueous layer from the cupferron extraction to a 100-ml beaker, add S ml. of diluted sulphuric acid (1 + 1) and 5 mi of concentrated nitric acid. Evaporate and fume to remove the organic material; add more nitric acid if necessary. Remove the cover and raise the temperature. Cool and dissolve in 5 ml of 2 N hydrochloric acid solution, add 16 ml of water and bring to the boil. Remove from the hot-plate and add 5 ml of 10 per cent. ammonium phosphate and 5 ml of diluted ammonium hydroxide (1 + 1) ; stir vigorously and set aside in a cool place overnight. Filter through a Whatman No. 40 filter-paper and wash with 1 per cent. ammonium hydroxide solution. Dissolve the precipitate in the original beaker with the minimum amount of 2 N hydrochloric acid solution.Add 5 ml of 2 N hydro- chloric acid solution and boil. Add 15 ml of water and reprecipitate the magnesium as previously outlined. Set aside overnight, filter through a 9-cm Whatman No. 40 filter-paper and wash with 1 per cent. ammonium hydroxide solution. Transfer the precipitate to a tared platinum crucible, dry and char carefully. Ignite the precipitate and take the iisual precautions to obtain a white residue of magnesium pyrophosphate. Cool and weigh as magnesium pyrophosphate. Evaporate to dryness and remove the ammonium salts. Roil and evaporate to dryness. RESULTS Synthetic samples were prepared from refined “van Arkel” titanium metal and standardised solutions of magnesium chloride and aluminium chloridc. The titanium was reported to contain a “faint trace” of niagnesium and no aluminium.* Half-gram amounts of titanium were taken and dissolved in dilute hydrochloric: acid, and known amounts of magnesium and aluminium were added.The recovery of aluminium and magnesium 13)- the method described above is shown in Table 11. The results by the proposed method are compared in Table I11 with those obtained by precipitating the titanium at a pII of 6 with sulphur dioxide. It will be seen that the results given by precipitation of the titanium are lower than those given by cupferron separation. RECOVERY OF ALUMINIUM AND MAGNESIUM FROM SYNTHETIC MIXTURES r .. T Itanium added, mg 500 5 00 5 00 500 500 ijluminium added, mg 0.025 0-050 0.030 0.015 0.010 Magnesium added, mg 1.5 1.0 0.75 0.50 0.10 Aluminium found, mg 0.02 8 0.050 0.030 0.018 0.013 Magnesium found, m g 1-46 1.05 0.73 0.50 0.08 Spectrographic analysis irom the Defence Research Laboratories, Department of Supply, Melbourne, A u stra ii a.24 MILLER AND CHALMEKS [Vol.78 TABLE I11 COMPARISON OF RESULTS FOR MAGNESIUM BY THE CUPFERKON METHOD WITH THOSE BY PRECIPITATIOK OF THE TITANIUM AT pH 6 WITH SULPHUR DIOXIDE Material magnesium, magnesium, Titanium powder ABP . . . . . . 0.28, 0.27 0-26, 0.24 >? ANZ .. .. . . 0.28, 0.26 0.24, 0.26 Melted bar ARP . . .. .. .. 0.07, 0.09 0.05, 0.04 79 ANZ . . .. .. .. 0.11, 0.12 0.08, 0.06 By cupferron method, By precipitation method, % 0 8 /O REFERENCES 1. 2. 3. 4. 5. 6. 7. Wartman, F. S., Walker, J. P., Fuller, H. C., Cook, M. A, and Anderson, E. L., “I’roduction of Meunir, P., Cornpt. Rend., 1934, 199, 1250. Furman, N. H., Mason, W. B., and Pekola, J . S., Anal. Chppn., 1949, 21, 1325. Rodden, C. J., “Analytical Chemistry of the Manhattan Project,” National Nuclear Energy Series Short, H. G., Analyst, 1950, 75, 420. “Handbook on Titanium Metal,” Titanium Metals Corporation of America, Fourth Edition, J u l ~ - , Marsden, A. W-., J . SOC. Chem. I1.td., 1941, 60, 21. Ductile Titanium at Boulder City, Nev.,” U.S. Bureau of Mines K.I.4519, August, 1949. Division VIII, Vol. I, McGraw-Hill Rook Co., Inc., 1949, p. 471. 1951, p. 44. COMMONWEALTH SCIENTIFIC AND INDUSTRIAL RESEARCH ORGANISATION PHSSICAL METALLURGY SECTION UNIVERSITY OF MELBOURNE CARLTON, S . 3 , VICTORIA AUSTRALI.~

ISSN:0003-2654    

DOI:10.1039/AN9537800020

出版商:RSC    

年代:1953

数据来源: RSC

摘要:

24 MILLER AND CHALMEKS [Vol. 78 Micro-analysis of Silicate Rocks Part I P The Precipitation of Silica as 2 : 4Diniethylquinoline Silicomolybdate and its Gravimetric Determination as Silicomolybdic Anhydride BY CHRISTINA C. MILLER AXD ROBERT A. CHALMERSt Under suitable conditions of acidity silicic acid is converted into silico- molybdic acid and then quantitatively precipitated as 2 : 4-dimethylquinoline silicomolybdate, which, after heating a t 120" to 150" C, may be used as a weighing form, but, being slightly hygroscopic, is preferably converted at 500" to 560" C into silicomolybdic anhydride. A method has been evolved for 1 to 2-mg amounts of silica and applied to 2 to 3-mg samples of rocks, which are initially fused with sodium carbonate, the products being acidified with hydrochloric acid.Phosphoric oxide accompanies silica and must be allowed for. The amounts of iron, aluminium, titanium, manganese, vanadium, calcium and magnesium normally present in silicate rocks are without significant influence on the results. Arsenic and germanium must be absent. THE classical procedure for the determination of silica in a silicate rock is based on fusion of the rock with sodium carbonate, acidification of the cooled melt with hydrochloric acid, and separation, after suitable treatment, of the silicic acid formed. Quantitative collection of the silicic acid is uncertain, and the small amount that escapes separation initially is normally carried down afterwards with the hydroxides of iron, aluminium, and so on, arid determined therein.Guthrie and Miller1 adapted this procedure to the analysis of 100-mg samples of rocks, and a comparable micro-procedure, tentative in nature and relating to PO-mg samples, has been utilised for some time in our laboratory.2 Numerous investigators have determined silica by similar means in samples a few milligrams in weight. * Part I is the paper by Miller, C. C., and Chalmers, K. A., Analyst, 1952, 77, 2. t Present address : The University Science Laboratories, South Road, Durham.Jan., 1953) MICRO-ANALYSIS OF SILICATE ROCKS, PART I1 35 As the scale of operations is reduced the amount of silica escaping in the initial stage tends to increase relatively, but it seems to be adequately held by the hydroxide precipitate. _L\s Guthrie and Miller failed to recover silica quantitatively from the ignited hydroxide precipitate, after fusing the oxides with potassium bisulphate and extracting with acid, they determined silica by expelling it from the oxides as silicon tetrafluoride, by means of hydrofluoric acid in the presence of a little sulphuric acid.In micro-analysis the high-temperature ignitions required for silica and the mixed oxides, and the hygroscopicity of the latter, especially if much alumina is present, are undesirable features. A simpler and more direct method of determining silica is highly desirable. Colloidal substances, such as gelatin, albumin and casein, have been used as coagulants for hydrated silica and to assist in its quantitative deposition. Their efficacy with large percentages of silica has not yet been shown. Many investigators have developed methods based on the production of sparingly soluble salts of silicomolybdic acid with organic bases.When these are used as the weighing form, factors favourable to micro-analysis are obtained. D u d 3 outlined the micro-procedure for the gravimetric determination of 0.1 to 2 mg of silica as hexamethylenetetra-amine silicomolybdate, but the method quoted would most likely be inapplicable to rocks because of the low acidity of the final solution. Brabson, Mattraw, Maxwell, Darrow and Needham4 determined 5 to 25 mg of silica gravi- metrically as 8-hydroxyquinoline silicomolybdate in a solution finally 0.7 N with respect to hydrochloric acid; they analysed various materials, including feldspars. On the micro- scale, the appreciable solubility of the precipitate and the hygroscopicity of the anhydrous complex would be troublesome.Wil~on,~ in determining up to 60 mg of silica in numerous materials, used Brabson’s work as a basis and substituted quinoline for 8-hydroxyquinoline. The separated quinoline silicomolybdate, which was more insoluble than the hydroxyquinoline complex, he dissolved in an excess of standard carbonate-free sodium hydroxide, and the excess he determined acidimetrically. On the micro-scale such titrations present difficulties. We aimed to devise, for the determination of silica in a few milligrams of insoluble silicate rocks, a gravimetric procedure based on the precipitation of a complex formed between silicomolybdic acid and an organic base. Quinoline can be used for this purpose, but we have obtained superior results by using 2 :4-dimethylquinoline as the precipitant and silico- molybdic anhydride as the weighing form.In rock analysis, phosphoric oxide, which is usually present, behaves like silica and must be allowed for. The presence in significant amounts of oxides of arsenic and germanium, which would also be expected to behave like silica, is unlikely. The silica contents of various rocks and a refractory have been successfully determined. EXPERIMENTAL T H E FORMAHON AND STABILITY OF SILICOMOLYBDIC ACID- For the quantitative production of silicomolybdic acid from a silicate one must not allow gelatinous silicic acid to separate. It is generally recommended that formation be effected within the pH range 1.0 to 1.5.Wilson,5 who fused his materials with sodium hydroxide, poured the aqueous extracts of the cooled melts into an excess of hydrochloric acid, thus retaining all the silica in solution, and then, in case polymerisation had occurred, rendered the solutions alkaline again and re-acidified them, so that a pH of about 1.5 was attained .after the addition of an excess of ammonium molybdate in the cold. The solution was heated to 80” to 90” C for a limited time (Brabson did not heat at this stage) and, after the .addition of more acid, because quinoline molybdate is sparing soluble, the precipitant was .immediately added, presumably before any breakdown of silicomolybdic acid could occur. We proposed to fuse silicate rocks with sodium carbonate and to add the aqueous suspensions of the cooled melts to fairly concentrated acidified solutions of ammonium molybdate.Since we desired to have a higher initial acidity than was used by Wilson, Strickland’s6 observations concerning the formation of silicomolybdic acid were of considerable interest to us. In order to account for apparent anomalies in the heteropoly acid formation, especially with reference to the photometric determination of silica, Strickland postulates the existence of alpha and beta silicomolybdic acid, two forms with similar chemical properties but different light absorption characteristics. If molybdenum is present as the simple normal ion, Moo4/’, and the solution is acidified with up to 1.45 to 1.5 equivalents of acid per gram-ion of MOO,”, the alpha complex results, whereas, if 3 to 5 equivalents of acid are used, and at the same time the molarity of uncombined molybdenum after completion of9fj MILLER AND CHALMEIIS : [Vol.7s the reaction is at least 0.05 and the ionic strength of the other ions is less than 0.5, beta silicomolybdic acid is quantitatively formed within 5 to 10 minutes and remains stable as such for about 15 minutes longer. Slow spontaneous conversion into the alpha form occurs over a number of hours. In the presence of an excess of molybdate, both forms are fairly stable towards acid, especially in the cold, but in 1 to 2 N hydrochloric or sulphuric acid they should riot be left more than 30 minutes and, in a hot solution, not more than a few minutes.; THE GRAVIMETRIC DETERMJNATION OF SILICA AS QUIKOLINE SILICOMOLYBDATE- Quantitative results were obtained both in Wilson's esperiments under conditions in which alpha silicomolybdic acid would be formed, and in our experiments, under conditions in which the production of the beta complex from the maximum amount of silica under consideration would be ensured.The composition of the weighed precipitates, subsequent t o drying at 120" to 150" C, was assumed to be H,SiO,.12Mo0,.4C,H,N, and the procedure mas as follows- Four to twelve milligrams of dried powdered quartz were fused with 100mg of sodium carbonate in a platinum crucible, the cooled melt was dissolved in 2 to 3 ml of water and the extract was added quickly with swirling to a mixture of 6ml of 10 per cent. w/v ammonium molybdate and 1.8 ml of 6.0 N hydrochloric acid (this gives about 3 equivalents of acid per gram-ion of MOO,").The crucible was rinsed (at this point the volume was 12 to 15 ml and for 12 mg of silica the molar concentration of uncombined molybdenum exceeded 0.05), the solution diluted with water to 50 ml (pH was about l - O ) , and after 5 minutes heated to 70" to 75" C in a water-bath maintained at 75" C. Then 6.2 ml of 6-0 N hydrochloric acid were added quickly, with stirring, and immediately afterwards 6 ml of quinoline reagent ( 2 g in 100 ml of 0.25 M hydrochloric acid). The mixture was digested at 75" C for 10 minutes, cooled to at least 15" C, and the precipitate was separated on a sintered-glass crucible and washed with cold water, dried at 120" to 150" C for 1 hour, cooled and weighed.All heating must be done in a water-bath as otherwise some rnolybdic acid may separate. The amount of acid added initially and the total acid used can be increased by 25 per cent. without significantly affecting the determination. Less acid is undesirable since, if only 4 mg of silica are present, the large excess of molybdate available for buffering tends to cause deposition of molybdic acid. It makes little difference whether the second addition of acid is made before or after heating, but, since breakdown of silicomolybdic acid occurs more readily at higher acidity, it is safer to add the acid afterwards. The amount of precipitant added may be safely doubled if desired. When the whole experiment is done at room temperature a positive error of 4 per cent.may be incurred, and the precipitate becomes slimy and difficult to filter. The error is geatly reduced if the precipitate formed is digested at 75°C for 10 minutes before cooling. Longer digestion does not improve the precipitate further. The precipitates that have been sucked dry on the filter-funnel are effectively wetted only after churning with wash liquid. Filtrates invariably become turbid during filtration because of re-precipitation of quinoline silicomolybdate that has been dissolved by the wash water. Quinoline silicomolybdate is appreciably hygroscopic ; it absorbs about 0-5 per cent. of water, and should be kept over a desiccant before it is weighed. It is stable up to 150" C at least, but slight decomposition probably occurs a t 180°C. There was a tendency towards a negative error of about 1 per cent.in determining the maximum amount of silica. Since the percentage of silica in the precipitate was found to be essentially correct, the error was probably due to the silica not all being in the form of silico- rnolybdic acid. For this reason, about 8 mg is recommended as the upper limit for silica in the prescribed method. It is better to saturate the wash water with the precipitate. OTHER ORGANIC BASES AS PKECIPIT-4NTS FOR SILICOMOLYBI~IC ACIII- In order to find if even more insoluble silicomolybdates could be prepared, a number of substituted quinolines and analogous compounds were substituted for quinoline in the above rnethod. Their performance w* assessed by determining silica and also by comparing theJan., 10531 MICRO-ANALYSIS OF SILICATE ROCKS.PART I1 27 turbidities obtained in the filtrates from the silicomolybdate precipitates with those obtained from solutions similar in composition to the filtrates and containing known small amounts of silicomolybdic acid. 2- and 8-Methylquinolines, isoquinoline and 3-met hylisoquinoline and 2 : 3-, 2 : 4-, 2 : 6-, 2 : 7- and 2 : S-dimethylquinolines, all without special purification, gave reasonable results for silica. 2 : 3 : 5 : 6-Tetramethylpyridine silicomolybdate was more soluble than the other complexes. 2-Chloroquinoline and the niolybdate of p-naphtho- quinoline were too insoluble in hydrochloric acid to be used. The di-substituted quinolines gave the best results, the 2:4-, 2:6- and 2:8-compounds being slightly superior to the others.As the 2:4-compound was obtainable commercially it was used in all subsequent work. 2 : 4-DimethyZquinoZi~ze---Quantitative results were generally obtained in the determina- tion of 4 to 8 mg of silica under the conditions prescribed for quinoline, and the solubility error was negligible. Dimethylquinoline silicomolybdate was hygroscopic, like the quinoline complex. Since hygroscopicity is undesirable in gravimetric micro-determinations, the effect of converting the complex to silicomolybdic anhydride, which is non-hygroscopic, was examined. Conversion was rapid at temperatures between 500" and 600" C, which represents the upper limit for thermostability, in contradistinction to the value of over 800" C given by Dupuis and Duvals for the anhydride obtained from other silicomolybdates.Free access for air was essential in order to prevent partial reduction of molybdenumv1. As a check on the stoicheiometry of the reactions, two portions of silica were converted into dimethylquinoline silicomolybdate, filtered on paper and ignited directly to silica; the final weight was corrected for a small amount of contaminating molybdic oxide. The weights found were within Similarly, two 0-25-g portions of silicomolybdic anhydride obtained from the dimethylquinoline complexes were converted into silica. The values of the ratio, weight of silica to weight of silicomolybdic anhydride, were 0.03352 and 0.03358, as against the theoretical value of 0-03360. 12 pg of the weights (8 to 10 mg) taken. METHODS FOR THE MICRO-DETERMINATION OF SILICA IS SILICATE ROCKS AS SILICOMOLYBDIC ANHYDRIDE \{'lien the procedure on p. 26 was reduced in scale and applied to a variety of rocks, it was found that the fusion products of some of the rocks failed to dissolve in the acid molybdate solution.Since breakdown of silicomolybdic acid had to be avoided, prolonged heating, especially after the addition of more acid, was not permissible. Wilson's m e t h ~ d , ~ which consisted in adding the fusion product to an excess of acid, heating and making the solution alkaline again before forming silicomolybdic acid, was troublesome on the micro- scale. Fortunately the more complicated procedure proved to be unnecessary. It was found that acidification in the absence of molybdate and solution of the fusion products could be achieved without appreciable polymerisation of silicic acid, which was then converted into silicomolybdic acid and determined as usual.The following procedures are submitted for the determination of silica in silicate rocks. The first is applicable only when the fusion products dissolve readily in cold acid molybdate solutions ; the second is generally applicable. .APPARATUS- Use Pyrex or similar glassware throughout. R E AGENTS- All reagents should be of recognised analytical purity. Hydrochloric acid, 6.0 N-Prepare at intervals by distillation from reagent grade con- Distilled water-Redistil in a Pyrex still at 3-day intervals. Ammonium mol-ybdate solutim-Store in a polythene bottle and renew once a fortnight. 2 :2-Dimet~~~Zqui~ztoEi?.re--~istil before use and collect the fraction that boils between centrated acid.264" and 266" C. PROCEDURE 1- With the aid of a stoppered weighing-stick, weigh into a 1-ml crucible enough (2 to 4 mg) of finely-powdered dried rock to give 1 to 2 mg of silica. Mix it intimately with 25 mg of anhydrous sodium carbonate and fuse cautiously over a micro-bunsen flame during 15 minutes.28 MILLER AND CHALMEKS : p o l . 7s To the cooled crucible, add 0.6 ml of water and then heat it on a water-bath and stir its contents with a platinum wire in order to disintegrate the cake as completely as possible. Next add the suspension quickly to a mixture of 1 6 m l of 10 per cent. w/v ammonium molybdate solution and 0-45 ml of 6.0 N hydrochloric acid contained in a tall beaker of external height 5 cm and diameter 2 cm.In order to accomplish the transfer effectively, hold the tilted crucible in platinum-tipped forceps just within the beaker and, while swirling the contents of the latter, dislodge the contents of the crucible into the beaker with a quick jerking move- ment (the height of the beaker precludes loss by spattering), and immediately afterwards rinse the crucible with a fine jet of water (about 1 ml) from a wash-bottle. After the transfer of most of the suspended solid, place the crucible and lid in the beaker. Dilute the solution with water to 12 to 13 ml, while removing and rinsing the crucible and lid and transferring the solution to a 25-ml beaker. After 5 minutes heat the solution for 3 to 4 minutes in a water-bath at 75" C, add, with constant stirring, during 30 seconds, 1.55 ml of 6.0 N hydro- chloric acid, and then, during 90 seconds, 1.5 ml of 2:4-dimethylquinoline (2 g in 100 ml of 0.25 N hydrochloric acid).Digest the precipitate at 70" to 75" C for 10 minutes, cool in ice-water to about 5" C and filter through a tared 3-ml sintered-porcelain crucible. Rub the walls of the beaker with a polythene rod, transfer all the precipitate to the filter without letting it pack on the disc, using 6 ml of ice-cold water, and wash it four times with 0-75-ml portions of the latter. Wipe the exterior of the crucible and place it in a micro-crucible oven. Heat the precipitate, first at 120" to 150" C for 15 minutes and then, in order to expel dimethylquinoline, a t 250" to 300" C for 5 minutes.Lastly, heat the crucible and a similar one for use as a tare for 20 minutes at 500" to 560" C in an electric furnace, allowing free access of air. Cool both crucibles for 30 minutes in suitable cavities drilled in deep aluminium blocks, transfer to the balance case for 5 minutes, and then weigh. Repeat and re-weigh until the weight is constant. PROCEUIJRE 2- Omit the ammonium molybdate from the beaker to which the suspension is added and, after placing the crucible and lid in the beaker, heat the whole in a water-bath at 75°C until the melt completely dissolves (this usually takes less than 5 minutes). Remove the beaker from the bath, cool, dilute with not more than 6 ml of water, and add 1.5 ml of ammonium molybdate solution with thorough mixing.Filter through a Miller filter-stickg into a 25-ml tall beaker, using about 1.5 ml of 0.1 N hydrochloric acid and then 1.5 ml of water to rinse crucible and beaker. Heat the filtrate for 3 to 4 minutes in a water-bath at 75" C and continue as outlined above. Proceed as above with the fusion and disintegration of the cooled melt. NOTES ON THE PROCEDUKES- A blank experiment was carried out with all the reagents, in accordance with the above procedures, but, as the amount of silicomolybdate was so small, no filtration was done and, instead, the quantity was related to the turbidity of the final solution. Comparison was made with solutions similarly prepared, except that most of the sodium carbonate, which was mainly responsible for the turbidity, was omitted, and to which 1 to 5 pg of silica as silicate had been added.The slight turbidity of the filtrates in all normal experiments was equivalent to 1 pg. With these two factors taken into account, the correction to be applied to the weights of silica obtained was about -2 pg. The sintered porcelain crucibles were freed from silicomolybdic anhydride by treatment, in turn, with cold 3 N ammonium hydroxide, water, dilute acid and then water. The reagent blank amounted to about 3 pg of silica. INFLUENCE OF OTHER CONSTITUENTS, JKCLUDING PHOSPHATE AND VANADATE- No simple means was found of preventing interference from phosphate in the deter- mination of silicate under the conditions given for rocks. Phosphate was precipitated as 2 : 4-dimethylquinoline phosphomolybdate and converted in to phosphomolybdic anhydride (P20,.24M00,), which was markedly hygroscopic. The conversion was essentially stoicheio- metric, the errors found in the determination of 0-432, 1.204 and 1.882 mg of phosphoric oxide being +1, +4 and -13 pg, respectively.Although the hygroscopicity of the pre- cipitate was troublesome when milligram quantities of phosphate were used, the small amountsJan., 19531 MICRO-ANALYSIS OF SILICATE ROCKS. PART I1 29 of phosphate in silicates caused no difficulty. In rocks containing both titania and phosphoric oxide, no combination is likely to occur that will prevent the quantitative deposition of both dimethylquinoline silicomolybdate and dimethylquinoline phosphomolybdate. The same applies to the minor components, zirconia and phosphate. Vanadate, even in large quantities, gave no precipitate with ammonium molybdate and dimethylquinoline.Of 6 mg of vanadium pentoxide present with 1-7 mg of silica, about 7 per cent. was found in the silicomolybdic anhydride, and the result for "silica," calculated as usual, was 1.2 per cent. high, showing that vanadium pentoxide was probably co-precipitated (c: f. Caiii and HostetterlO). THE CORRECTION FOR PHOSPHORIC OXIDE- I t was desirable to have a rapid means of determining phosphoric oxide, in order not to detract from the value of the comparatively rapid method outlined for silica (plus phosphoric oxide). A suitable spectrophotometric method, applicable in the presence of silicate, has been evolved and is described in a separate paper.ll RESULTS SILICA IN SODIUM SILICATE SOLUTIONS- Standard solutions for immediate use were prepared by fusing appropriate portions of 99-94 per cent.pure freshly-ignited powdered quartz with sodium carbonate and dissolving the cooled melts in water. The results in Table I, which show no systematic error, were obtained in accordance with procedure 1 (p. 27) and have been corrected for the small amount of silica in the reagents and filtrates. Weight aliquots of the aqueous extracts were taken. TABLE I DETERMINATIOX OF SILICA IN SODIUM SILICATE SOLUTIONS (PROCEDURE 1) IVeight of silica taken, mg 0.474 0.608 0.782 0.806 0-812 0.994 1.116 Weight of silica Error, taken, Pg mg + 1 1.158 + 2 1.265 + 3 1-415 -2 1.439 + 1 1.47 1 + 2 1.623 - 6 Error, Pg f l -1 0 + I +2 -3 Weight of silica taken, mg 1.710 1.906 1.988 2.005 2.049 2.110 Error, ILg 0 +6 +2 - 2 -3 - 6 The stability of silicomolybdic acid in 0-8 N hydrochloric acid at 75" C-In a few experi- ments based on procedure 1 the solutions were left for various periods at 76" C after the second addition of acid and before the addition of 2 :4-dimethylquinoline reagent.The following set of results relating to 2-mg amounts of silica clearly shows the necessity for proceeding quickly with the precipitation. Period of heating bcforc adding precipitant, minutes 0 10 22 30 Error in dcterrnination of silica, pg . . .. .. -3 - 28 - 44 - 54 Experiments were also made essentially in accordance with procedure 2 (p. 28), except that, after the addition of the sodium silicate solutions to the acid, the period of contact with the acid at 17" and 75" C was varied before the dilution and addition of the molybdate. The results shown in Table I1 have no definite trend and indicate that the silicic acid is not so affected by a short period of contact with hydrochloric acid, even at 75" C, as to prevent its conversion into the heteropoly acid. The addition of molybdate after the dilution with water apparently does not influence the results.SILICA IN THE PRESEXCE OF OTHER ELEMENTS- An average of 1-6 mg of silica was used in each experiment. Appropriate salts (analytical reagent grade) of the various elements were incorporated in the 6 N hydrochloric acid to which standard solutions of sodium silicate were added in accordance with the conditions given in procedure 2 (p.28). Standard phosphate solutions were prepared from dried potassium dihydrogen phosphate. Blank runs were made on all solutions and the requisite small corrections applied. The precipitates of silicomolybdic anhydride were examined by30 MILLER AND CHALMERS : [Vol. 7 8 TABLE I1 DETERMINATION OF SILICA IN SODIUM SILICATE SOLUTIONS (PROCEDURE Time of Weight of standing with silica taken, the acid, mg minutes 1.574 2 1.528 2 1.601 c) 1.648 5 1.496 1 0 1.658 120 Error, P!3 0 - 1 ._ 5 - 3 0 1 2 Time of Weight of standing with silica taken, the acid, mg minutes 1.211 0 1.604 2 1-846 2 1-511 6 1.516 5 1.738 5 1.203 10 1.686 15 1.209 20 1.204 30 1.600 30 2) Error, CLg - 4 - 4 + 3 -3 -4 +:I - 1 -t 2 + :3 4- 5 c - I Temperature 17" C 'l'~ml)c~ature 75" C appropriate micro-tests for the elements originally present with the silicate.In all esperi- ments in which phosphoric oxide was present, the stoicheiometric production of phospho- molybdic anhydride (P20,.24M00,) was assumed, and its weight was deducted from that of the combined anhydrides. All these results are shown in Table 111. TABLE I11 DETERMINATIOK OF SILICA IX THE PRESENCE OF OTHER ELEMENTS Approximate amounts of other elements as oxides, mg 0.09 (P,O,) 0.09 (P205) + 0.15 (TiOJ 0.15 (TiO,) 3.0 (A1,OJ 3.0 (Fe,O,) 1-5 (CaO) + 1.5 (MgO) 0.07 (V205) Error on about 1.6 mg of silica, CLg -4, -5, -3, - 2 -6, +I. 3.3 -1, f l -66, + 4 -3, -6 -7, -6 -4, - 3 Approximate amounts of foreign oxide in precipitate, CLg - 15 1 nil nil As the amounts of aluminium, iron, calcium and magnesium oxides added were far in excess of those that would be associated with the silica in 2 to 3 mg of silicate rocks, and the amounts of titania and phosphoric oxide considerably exceeded the likely maxima, it was concluded from the results above that their influence would, in general, be negligible.The favourable factor for the conversion of silicomolybdic anhydride into silica nullifies the effect of contaminating titania. The vanadium oxide content of rocks is norinally much too small to have a significant effect. SILICA IN INSOLUBLE SILICATES- The results in Table IV refer to micro-analyses on sample weights ranging from 3.0 to 3-3 rng. The feldspars were used as issued by the U.S. Bureau of Standards, but all the other samples, except the olivine-basalt, which was part of the original sample analysed on the semi-micro scale by Guthrie and Miller,l were powders that had passed through a 300-mesh sieve.The state of subdivision of the olivine-basalt specimen was scarcely adequate for micro-analysis, but it was of interest to include this substance. All the samples, save the flint clay, which was dried at 140" C, were dried at 105" to 110" C before use. For all the Inicro- analyses, except those indicated, procedure 2 (p. 28) was used and correction was made for silica in the reagents and filtrates (p. 28). All the results shown have been corrected for the phosphoric oxide content of the materials, which was determined by the rapid spectrophoto- metric method referred to on p. 29 (1.00 per cent. of phosphoric oxide = 0.85 per cent.of silica). Analyses made by the classical procedure were conducted on 77 to 106 mg of material in accordance with Guthrie and Miller's instructions.Jan., 19531 MICRO-ANALYSIS OF SILICATE ROCKS. PART I1 TABLE IV DETERMINATION OF SILICA IN SILICATE ROCKS AND A REFRACTORY Silicate :kFFeldspar So. 70. . . . *Feldspar No. 9 9 . . . . *Flint clay S o . 97 ''Burnt refractory No. 76 . . Rnalcitc syenite. . . . Kinkell tholeiitc . . Phonolite . . . . Olivine-basalt . . .. Approximate amounts of certain components SiO, by f > PzO, inicro- A1,0,, Fe,O,, TiO,, CaO, M,gO, content, method, % % % % /O Yo Yo A 18 19 39 38 16 12 19 12 10 1 4 2 0.33 0.37 16 3 a 4 0.65 0.64 6 1 2 1 0.11 0.13 13 2 10 13 0.46 0.50 66.9 66.8 66.8 (65-6) S (66-3)s 68.4 68.2 42.8 42.8 55.3 55.0 (551) 56.7 56.9 47.5 47.8 [48*5] .56-8 56-8 44.6 43.9 (43.8)s 31 S O , by classical method, 66.66i O / , o (j i;.f i 6 t -i"S7t 54.9 55.3 55.5 5 6 4 56.G 47.5 47-8 56.9 56.9 44.0; (51.88) :F t7.S Bureau of Standards sample. t U S . Bureau of Standards certificate value. We used only the fractions of the day and thc refractory 5 Early results obtained by procedure 1. With tlic olivine-basalt a sinall residue was removed before Surveyed as a whole, the results obtained on the milligram samples show no systematic error and compare favourably with those based on the classical procedure. The standard deviation was calculated from the unbracketed results for all the samples, except the olivine- basalt which was justifiably excluded, by means of the formulal2- t h a t passed a 300-mesh sieve.Guthrie and Miller's figure. eitecting thc precipitation. where A' is the number of observations, k , the number of samples, and Sd12 . . . Zdk2, the sum of the squares of the deviations from the means for samples 1 . . . k . The value is 0.13. Twenty-three representative results shown by Hillebrand and Lundell13 for the analysis by the classical macro-procedure of six silicates and refractories give a corresponding standard deviation of 0-073. In the new method, the necessity for simultaneously determining phosphoric oxide is a minor disadvantage compared with that of determining silica in the mixed oxides, as in the classical procedure. A great advantage of the new method is the speed with which the total silica content of a rock can be determined. \Ve gratefully acknowledge a maintenance grant to one of us (R. A. C.) from the Depart- ment of Scientific and Industrial Research, and a grant from Imperial Chemical Industries Limited. We are indebted to Principal H. B. Nisbet, D.Sc., F.R.I.C., and to Dr. M. Pryde of the Heriot-Watt College for samples of substituted quinolines. REFERENCES 1. 2. Miller, C. C., Chem. G. Ifid., 1946, 26. 3. Duval, C . , Anal. Chhinz. Acta, 1947, I, 33. 4. Guthrie, IT-. C. A., and Miller, C. C., Min. Mag., Land., 1933, 23, 403. Brabson, J. A., Mattraw, H. C., Maxwell, G. E., Darrow, A., and Needliain, M. F., .??znl. Clcem., 1948, 20, 504.32 CHALMERS : MICRO-ANALYSIS [Vol. 78 5. 6. 7. 8. 9. 10. 11. 12. 13. Wilson, H. N., Analyst, 1949, 74, 243. Strickland, J. D. H., Chem. G Ind., 1950, 393. _- , personal communication. Dupuis, T., and Duval, C . , Anal. Chim. A d a , 1950, 4, 50. Miller, C. C., J. Chew. SOC., 1939, 1962. Cain, J. R., and Hostetter, J. C., J . Amer. Chew. SOC., 1911, 43, 2552. Chalmers, R. A., Analyst, 1953, 78, 32. Davies, 0. I,., “Statistical Methods in Research and Production,” Oliver and Boyd, London, 1947, p. 31. Hillebrand, W. F., and Lundell, G. E. F., “Applied Inorganic Analysis,” John U‘iley and Sons Inc., New Sork, 1929, p. 656. CHEMISTRY DEPARTMENT THE UXIVERSITY, EDINBURGH, 9 May 30th, 1952

ISSN:0003-2654    

DOI:10.1039/AN9537800024

出版商:RSC    

年代:1953

数据来源: RSC

摘要:

CHALMERS : MICRO-ANALYSIS [Vol. 78 Micro-analysis of Silicate Rocks Part I11 The Spectrophotometnic Deterrnination of Phosphoric Oxide in the Presence of Silica BY ROBERT A. CHALMERS* Orthophosphate is converted into phosphomdybdate under acid con- ditions, which prevent formation of silicomolybdate, and then reduced to molybdenum blue with ferrous ammonium sulphate. A rapid method has been devised for determining 1 to 150-pg amounts of phosphoric oxide and is applied to 5-mg samples of silicate rocks, following their fusion with sodium carbonate and addition of hydrochloric acid. Interference by iron" and vanadiumV is eliminated by prior reduction in a silver reductor. Arsenic and germanium must be absent, but the amounts of aluminium, titanium, calcium and magnesium normally present do not interfere.THE object of this investigation was to devise a rapid method for the determination of phosphoric oxide in samples of silicate rocks weighing a few milligrams. The need for a method sensitive to about 1 pg arose because phosphoric oxide accompanied the silica in the new method of determining silica described in Part I1 of this series.' Hecht2 determined phosphoric oxide gravimetrically as phosphomolybdic anhydride in 10 to 20-mg samples of rocks after removing silica by means of hydrofluoric and nitric acids. The method is time-consuming and a photometric method would appear to be more suitable. Sumerous studies have been made of the colorimetric determination of phosphorus, but only a few methods have been applicable in the presence of silicate.The applicable methods can be divided into two groups: those in which silicate interference is suppressed by means of certain hydroxycarboxylic acids, and those in which formation of silicomolybdic acid is prevented by a high acid concentration. Zimmermann3 used citric acid as suppressor and determined up to 100 pg of phosphoric oxide in 20 ml of solution in the presence of 5 mg of silica and 1 mg of iron. Fiske and Subbarow4 determined 0.46 to 3-68mg of phosphoric oxide in 100 ml of solution in the presence of 90 to 720 mg of silica by making the solution 0.6X with respect to sulphuric acid. Rockstein and Herron5 recently determined up to $7 pg of phosphoric oxide in 10 ml of solution that was 0.76 IV in sulphuric acid, and con- sidered this acidity adequate to suppress interference by 7-5 mg of silica.EXPERIMENTAL Silicate rocks are readily broken up by fusion with sodium carbonate and subsequent solution of the melts in hydrochloric acid. This process converts the phosphorus compounds into orthophosphates. In these solutions it was proposed to form phosphomolybdic acid at an acidity high enough to prevent the formation of silicomolybdic acid and then to reduce: * Present address : The University Science Laboratories, South Road, Durham.Jan., 19531 OF SILICATE ICOCKS. PART 111 33 the former, by means of a suitable reductant, to molybdenum blue. Rockstein and Herron6 added to a neutral phosphate solution first a solution of ammonium molybdate in 0.83 AT sulphuric acid and then a solution of ferrous sulphate as reducing agent, so that the find acid concentration was 0.76 N .The following modified procedure, suitable for phosphate solutions that were initially acid, was used as a basis for examining the factors likely to affect a colorimetric method for silicate rock analysis. Half a millilitre of 6 N hydrochloric acid was added to 2 ml of the phosphate solution. Then 2 ml of 10 per cent. w/'v ferrous ammonium sulphate that was 0-15 N with respcct to sulphuric acid and 2-5 ml of 2.5 per cent. w/v ammonium sulphate containing 8 per cent. V/Y of sulphuric acid were added in turn, each being thoroughly mixed on addition. The solution rh\III\\ Glass wool-- Fig. 1 . Detail of attachment of the silver reductor was diluted to 10 ml in a calibrated flask, mixed thoroughly and placed in a water-bath at 25p C for 10 minutes.The optical density was determined spectrophotometrically at 750 mp in 1-cm glass cells. Choice of wavelength--The absorption spectrum was examined over the range 500 to 900mp, by means of a. Unicam spectrophotometer Model SP500. The wavelength of 750 mp was selected for measurements since it gave maximum absorption with minimum deviation for small changes in wavelength. Rate of colour development and stability of colour-At 17" C addition of 5 mg of silica as silicate, or the use of an additional 0.1 ml of 6 N hydrochloric acid, slowed the rate of reduction so that as much as 30 minutes was required for complete colour development. -it 50" C, if silicate was present, some silicomolybdic acid was formed and reduced.At 25" C reduction was complete in 10 minutes even in the presence of 5 mg of silica or 0.6 in1 of 6 -Y hydrochloric acid. The colour intensity was stable, within 2 per cent., for at least 2 hours even in the presence of silicate, if kept at room temperature after colour development wits complete. Eflect of dzflerezzt r e a p i t coitccntvations-If the concentration of molybdatc was doubled, silicate interfered ; this interference could be suppressed by increasing the acid concentration, there being no significant effect on the intensity of colour due to phosphate if this was done. The colour intensity was reduced by about 5 per cent. when the con- centration of ferrous ammonium sulphate was halved. The amount of hydrochloric acid A blank containing the reagents only was run for comparison.34 CHALMERS MICRO-ANALYSIS [I-01.78 present could be varied from 0.5 to 0.6 ml of 6 N without affecting the colour intensity. The order of addition of the molybdate and reductant was immaterial. Eflect of various ions-Examination of the effect of various ions, when present at the maximum concentrations likely to be found in silicate rocks, was made under the conditions established. Silicate did not interfere even when the equivalent of 5 mg of silica was added. Titanium, aluminium, calcium and magnesium had no significant effect. Arsenate gave a positive error, but arsenic seldom occurs in significant amounts in silicate rocks and was not further considered; nor, for a similar reason, was germanium. IronI'I gave a negative error, which was attributed to its effect on the potential of the reducing couple.Vanadate gave a negative error if the reductant was added after the molybdate, the error being of the order to be expected from formation of phosphovanadomolybdic acid and reduction of the molybdenumV1 in the heteropoly acid. When the molybdate was added last the vanadate was reduced by the ironJ1, a corresponding amount of ironlIr being formed. The interference of ironlI1 and vanadate was eliminated by reducing them to the bi- and quadrivalent states, respectively, by means of a silver reductor, which was incorporated, as shown in Fig. 1, in one arm of the capillary of an appropriate suction apparatus, so that the rock solution could be transferred to a 10-ml calibrated flask and simultaneously reduced.A length of 10 cm of the reductor limb was loosely packed with silver granules (those passing B.S.S. 36-mesh sieve), which were held in place by glass wool plugs at both ends. The small amount of peroxide produced in the reductor6 did not affect the results. Since molybdate is reduced in the silver reductor, it was necessary to add the acid molybdate after reduction. The small increase in the concentration of ironII arising from iron present in silicate rocks had no effect on the intensity of colour due to phosphate. When the acid molybdate was added to solutions that had passed through the reductor, an opalescence often appeared in the solution; this was xttributed to interaction between the molybdate and silver chloride carried over in solution from the reductor.The turbidity could be removed by centrifugation. METHOD 12 I .- AGE S TS--. &411 reagents should be of recognised analytical quality. Acid moZybdate solution-Dissolve 12.5 g of ammonium molybdate in 250 ml of water :md add with stirring a cold solution of 40 ml of concentrated sulphuric acid in 200 ml of water. Dilute to 500 ml. The solution keeps indefinitely. Ferrous ammonium sulplzate solution-Dissolve 5 g of ferrous ammonium sulphate in 30 ml of water containing 7-5 ml of N sulphuric acid and dilute to 50 ml. The solution must be used on the day of its preparation. Hydrochloric acid-Dilute concentrated acid to 6.0 N. PKOCEIIURE FOIi ROCKS- IVeigh a 5-mg sample into a 6-ml platinum crucible, mix it intimately with 25 to 30 mg of anhydrous sodium carbonate, and fuse over a micro-bunsen flame during 15 minutes.Extract the cooled melt with 0.5 ml of water, heating on a water-bath, cool, add 0.6 to 0.65 ml of 6 N hydrochloric acid and warm to 75" C, if necessary, to ensure complete solution. Transfer the solution to a 10-mi calibrated flask via the silver reductor (Fig. l), using three I-ml portions of water to rinse the crucible and reductor, the latter being allowed to drain between rinses. Add 2 ml of ferrous ammonium sulphate solution and 24ml of acid iiiolybdate solution, mixing between additions, and dilute to the mark. Mix thoroughly, place in a water-bath at 25" C for 10 minutes, centrifuge, and determine the optical density at 750 mp in I-cm glass cells in a spectrophotometer, and compare with a blank containing the reagents and prepared in the same way as the samples.RESULTS CAT.IBRATION CI:RVE- Dried potassium dihydrogen phosphate (0.958 g) was dissolved in water and diluted accurately to 500 ml. Portions of this stock solution were diluted ten and hundred-fold to provide solutions containing 100 pg of phosphoric oxide per 11.11 and 10 pg of phosphoric oxide per ml, respectively. All solutions were made approximately 0.001 N in hydrochloric acid to prevent adsorption of phosphate on the glass. Suitable volumes of standard solutionJan., 19531 OF SILICATE ROCKS. PART I11 35 were transferred by means of a pipette to 10-ml beakers, 0.56 ml of 6 N hydrochloric acid was added to each, and the solutions were passed through the reductor and treated from that stage onwards as in the procedure above.From the results in Table I it is seen that Beer’s law is obeyed for amounts of phosphoric oxide up to about 25 p g ; above that limit there is a slight departure from linearity. TABLE I CALIBRATION DATA FOR PHOSPHORIC OXIDE Phosphoric oxide, pg 0 2.5 5 12.5 25 50 100 175 Optical density . . 0.000 0.015 0.028 0.071 0.144 0.278 0.544 0.93 PHOSPHORIC OXIDE IN THE PRESENCE OF OTHER ELEMEXTS- Appropriate reagent-grade salts of the various elements were incorporated in the 6 N hydrochloric acid added to the standard phosphate solutions. Blank runs were made on these salts and corrections applied if necessary. Silicate formed by fusing precipitated silica with sodium carbonate and dissolving in water was added. An appropriate in- crease was made in the amount of hydrochloric acid added so that the final acid concentration remained the same throughout.From the results in Tables I1 and 111 it is seen that the elements usually found in silicate rocks are without significant effect on the determination. TABLE I1 CALIBRATIOW DATA F O I ~ PHOSPHORIC OXIDE IK THE PRESESCB OF 34mg OF SILICA Phosphoric oxide, pg . . 0 5 12.5 60 100 175 Optical density . . . . - 0.001 0.03 1 0.070 0.28 1 0.547 0.93 TABLE I11 DETERMINATIOK OF PHOSPHORIC OXIDE IN THE PRESESCE OF OTHER ELEMENTS Approximate amounts of other elements as oxides, mg 1.0 (Fe,O,) 2.0 (A1,OJ 0.5 (TiO,) 0.025 (V,O,) 0.5 (CaO) + 0.5 (MgO) 0.025 (AsZOS) Error on 20 pg of phosphoric oxide, PQ 0 -1 -1 -1 -1 +6 Error on 50 pg of phosphoric oxide, K - 1 - 1 -1 0 0 Error on 150 pg of phosphoric oxide, 1Lg 0 0 -2 0 0 - PHOSPHORIC OXIDE IN IKSOLUBLE SILICATES- ,4 number of analyses were made on 5-mg quantities of U.S.Bureau of Standards samples. The samples were not dried before use. As no standard samples were available of rocks with relatively high phosphoric oxide content, synthetic samples were prepared by fusing 4-mg quantities of Bureau of Standards Feldspar No. 70 with 25 mg of sodium carbonate and adding, with the hydrochloric acid, salts of iron and titanium equivalent to 0.5 mg of ferric oxide and 0.25 mg of titanium dioxide. Appropriate volumes of standard phosphate solution were added and the determinations were completed in the usual way. The synthetic rock samples contained the equivalent of 50 per cent, of silica, 14 per cent.of alumina, 10 per cent. of ferric oxide and 5 per cent. of titania, based on a sample weight of h mg. The results are shown in Table IV. The error is of the order of 1 pg of phosphoric oxide for amounts up to 50 pg, and 4 pg for amounts of the order of 150 pg, with a definite tendency towards a negative error when the maximum amount of phosphoric oxide is determined in the presence of large amounts of mixed oxides that are rich in titania. Silicate rocks seldom contain more than 0.5 per cent. of phosphoric oxide, although as much as 3 per cent. has been reported. No claim is made for a high degree of accuracy in this determination.36 CHALMERS [L'ol. 78 TABLE IV DETERMINATION OF PHOSPHORIC OXIDE IN SILICATE ROCKS AND A REFRACTORY p205 content, p205 by Silicatc micro-method, % % *Feldspar No. 7 0 . . .. .. . . . . 0.01 0.012t 0.01 0.16 0.11 0.10 "Feldspar No. 99. . .. . . .. .. 0.14 0.142t "Flint clay Xo. 97 . . .. . . .. 0.09 o-ost *Burnt refractory So. 70 . . .. .. 0.10 0.069f Synthetic rock A . , I . . . .. Synthetic rock 13 . . .. . . .. Synthetic rock C . . .. .. .. 0.5 1 0.51$ 0.5 1 1.00 2.93 3.011 2.93 0.97 1.01$ * U.S. Bureau of Standards sample. t U.S. Bureau of Standards certificate xralue. that passed a 300-mesh sieve were used. $ Calculated from phosphate added. Of the clay and tlic refractory only those portions I gratefully acknowledge a maintenance grant from the Department of Scientific and I wish to thank Dr. Christina C. Miller for helpful discussions and Industrial Research. criticism during this work. REFERENCES 1. 2. 3. 4. 5. 6. Miller, C. C . , and Chalmers, 13. A., Analyst, 1953, 78, 24. Hecht, F., Mikrochinz. Ada, 1937, 2, 188. Zimmermann, M., Angew. Chem., '4, 1960, 62, 291. Fiske, C. H., and Subbarow, Y., J . Riol. Chem., 1925, 66, 376. Rockstein, M., and Herron, P. W., Anal. Chem., 1951, 23, 1500. Miller, C. C., and Chalmers, R. -4., ,4naZyst, 1962, 77, 2. CHEMISTRY DEPARTMENT THE UNIVERSITY, EDINBURGH, 9 June 3vd, 1952

ISSN:0003-2654    

DOI:10.1039/AN9537800032

出版商:RSC    

年代:1953

数据来源: RSC

摘要:

Jan., 1'3631 MOSSEL AN11 TIE BRUIN 37 The Detection of Preservatives in Beverages a Fermentation Test, with Special Reference to Brominated Compounds A fermentation test has been developed for detecting preservatives, other than benzoic and sulphurous acids, in soft drinks. The substrate is brought to a pH value of 5-4 f 0.1 and, if necessary, is enriched with 0.25 per cent. of Difco yeast extract. Thereupon it is inoculated with approximately 104 cells per ml of bakers' yeast (Saccharomyces cerevisiae) and transferred to Einhorn (Smith) fermentation tubes. The gas formed is measured after incubation for 36 and 48 hours at a temperature of 24" f 1" C. Brominated acetic acid derivatives can be detected when present in beverages at concentrations corresponding to bromine concentrations of 0.6 to 2 mg per kilogram.Sulphurous and benzoic acids at concentrations of 75 mg per kilogram do not interfere with the test, owing to the rather high pH value. Eosin, a food-colouring material that is used in The Netherlands, only inhibits the fermentation when at least 50 mg of bromine is present per kilogram of sample. Essential oils, tannins or other compounds present in natural fruit juices do not interfere at the concentrations that occur in fruit drinks. Citrus terpenes, which are present in some beverages, can be inactivated by adding liver particles to the fermentation medium. Under these circum- stances threshold levels of detection of the various brominated preservatives vary from 1 to 10mg o f bromine per kilogram. SINCE the discovery by Dickensl in 1933 of the pronounced SH-enzyme inhibiting action of halogen acetic acid derivatives, Genevois, Cayrol and Nicolaieff2 and Genevois and NicolaiefP have introduced esters of bromoacetic acid as preservatives, e.g., in soft drinks4 Chemical methods for detecting them have been developed almost simultaneously and have been improved markedly in the course of time.5,6,7,8,9,10,11~1z~13,14 The weak point in these methods for analysing foods for brominated preservatives was that, although the presence of organically combined bromine could be shown, it could not be demonstrated explicitly that the bromine atoms were part of an antimicrobial compound.We attempted to solve the problem of detecting brominated preservatives by applying the principle outlined by Kluyver in 1914,15 viz., by using a direct biological method that will reveal the antimicrobial properties of the substrate under investigation.I t is evident that a biological method only indicates the presence of active concentra- tions of preservatives in general. Therefore, a positive test-even in combination with a positive chemical test for an organic bromine compound-is only an indication, but no proof, of the presence of a brominated preservative. A proof of the presence of a brominated preservative can only be given by isolating the bromine compound present and demonstrating its antimicrobial properties with the help of the biological technique. PRINCIPLES OF THE METHOD The biological method for the detection of antimicrobial action is based on the general principle that unpreserved beverages are fermented on inoculation with a suitable yeast, whereas fermentation is inhibited in preserved beverages.SENSITIVITY- A sensitive technique for the detection of preservatives accoiding to the Kluyver principle had been developed already.l6 In the method, which actually is a modification of Heim and Poe's gasometric technique,17 bakers' yeast (Saccharomyces cerevisiae) at an initial densit!. of the order of lo4 cells per ml is used as a test strain. This low inoculum permits the detection of preservatives at rather low concentrations.38 MOSSEL AND 1)17 BRITIN THE 1)ETECTION OF 1'IIF:SEKVATIVES [Vol. 78 SELECTIVITY- -4 certain degree of selectivity in the method is required to ensure no interference from benzoic and sulphurous acids, which in many countries are permitted in soft drinks at concentrations of 75 to 100 mg per kilogram.In order to inactivate these preservatives, use was made of their lack of appreciable antimicrobial action unless a critical concentration of undissociated molecules is present, i.e., unless the pH of the medium falls below a critical value.1~~19~20~21~22~23~24~~s~z6~~7 This can be explained by assuming that the ions are unable to penetrate the cells sufficiently to exert their inhibitory action, owing to the polar character of the protoplasma, whereas undissociated molecules penetrate readily.28 Preliminary investigations showed that sulphurous and benzoic acids, when present at concentrations up to 100 mg per kilogram, were devoid of any significant action towards the inoculum of 104 cells per ml at an initial pH >, 5.4 & 0.1.RELIABILITY- Some soft drinks are prepared from sugar, acid and artificial colour and flavour only (imitation beverages) and so lack the nutrients required for the activities of the yeast cells used as an inoculum.29 The same may hold true for fruit beverages prepared from juices poor in such nutrients. Therefore, in such media, fermentation may not occur, even if preservatives are absent. To avoid these "false positives" it is necessary to enrich such beverages before inocula- tion with some source of nutrients, preferably yeast extract. The addition of 0.25 per cent. w/v of Difco dehydrated yeast extract ensured a sufficient supply of nutrients for optimal development of the inoculum used.ACCURACY- order of 10 per cent. PROCEDURE- Substrate-Expel the carbon dioxide from the soft drink under investigation by pouring 100ml slowly into a sterile 500-ml Erlenmeyer flask and, after plugging, shake gently €or about 10 minutes. Adjust the pH of the substrate to 5.4 & 0.1 by adding a few drops (about 3 drops per 100ml) of a fresh 30 per cent. w/w aqueous potassium hydroxide solution. Estimate for this purpose the number of drops required in a trial experiment, in which the drink is titrated with the alkali under electrometric control. Finally, dissolve 0.25 per cent. w/v of yeast extract in the substrate, preferably by adding 2 per cent. v/v of a freshly sterilised (15 minutes at 121" C) 12Q per cent. w/v aqueous solution of Difco yeast extract.If the sample is not contained in the original bottle or if, for other reasons, the sterility of the sample is doubted, sterilise the substrate. This should be done by filtration over sintered-glass filters rather than by heat treatment, as the latter might convert part of the bromoacetic acid preservative to inactive bromide ion. Inocwlwm-Starting with a fresh preparation of commercial bakers' yeast, i.e., one stored in a refrigerator for not more than 3 days, prepare under aseptic conditions a 10 per cent. W/I- dispersion in sterile saline. Dilute this dispersion aseptically three successive times 1 + 9, to give an inoculum with a density of the order of lo6 cells per ml. Inocwlation and iitczzbation-Inoculate the substrate with 1 per cent.v/v of the yeast suspension containing 106 cells per ml and transfer it to at least two sterilised Einhorn (Smith) ferment ation tubes. Incubate at 24" 5 I" C and measure the volumes of gas formed after 36 and 48 hours. The reproducibility of the volumes of gas formed in replicate determinations is of the METHOD EXPERIMENTAL RESULTS GESERAL- representative model experiments are reported here. broth at an initial pH of 5.4 Out of the experimental material gathered in the course of time the results of a few In these experiments a 10 per cent. w/v sucrose-0.25 per cent. w/v yeast extract 0-1 was used as base medium.Jan., 19531 IX BEVERAGES H Y A FEKMESTATION TEST 39, The preservatives incorporated were: (i) ethyl bromoacetate, b.p. 154" to 156" C ; (ii) commercial preparation 1, bromine content (in organic combination) 0.28 per cent.w/w*, pH = 2.2; (iii) commercial preparation 2, bromine content (in organic combination) 0.51 per cent. w/w*, pH = 44; (iv) sodium benzoate, more than 99 per cent. pure; and (v) sodium sulphite, containing 233.3 per cent. of sulphur dioxide. The results obtained with these preservatives are recorded in Table I ; the data recorded represent the averages of two series of two duplicates. TABLE I F E K M E ~ T X T I O S OF SUC'ROSE - YEAST EXTRACT MEDIUM OF AN INITIAL pH O F 5.4 5 0.1 BY THE STAXDARD ISOCYLC'M AT 24" 1 ° C "Total" fermentation corresponds to formation of 178 ml of gas T'ncorrected volume of gas formed at 24" C Concentration of mg per kq Presen-ative added added preservative, - None ... . . . . . . . Sulphurous acid. . . . . . . . 75 100 Renzoic acid . . . . . . . . 7.5 100 Ethyl bromoacetate . . . . .. 2 (Brt) 1 -) 0.5 73 5 .> 2 77 I 97 1 7' 0.5 >> Commercial preparation 1 . . .. 10 )f Commercial preparation 2 . . .. 2 7 % I I after 36 hours, after 48 hours, in 1 ml 174 179 12 174 174 0 0 0 0 0 15 0 0 0 29 t Br: bromine present in organic linkage, mg per kilogram. I t is clear from tlicse data that neither sulphurous acid nor benzoic acid interferes with the test when present at normal commercial concentrations, whereas brominated preservatives in general can be detected when present at bromine concentrations from 2 mg per kilogram upwards. The more effective preparations even inhibit fermentation at concentrations of 0.5 mg of bromine per kilogram.Since the concentrations of brominated preservatives used in practice are of the order of 1 to 10 mg of bromine per kilogram,12~13~30~31 the test developed was suitable for the purpose in hand. INTERFERENCE FRON EOSIN- 'The use of eosin (potassium tetrabromofluorescein) is not forbidden in some beverages, such as lemonades, in The nether land^.^^ Since this dye possesses some antimicrobial activity,% it was essential to study any influence that this colouring might have on the fermentation under the test conditions. For this purpose, quantities of eosin corresponding with bromine levels of 0.5 to 100 mg per kilogram were added to the standard sucrose - yeast extract medium. The amount of gas formed in these media upon inoculation and incubation is reported in Table 11.These experiments reveal that eosin starts to interfere with the fermentation tests at levels corresponding to bromine concentrations of the order of 50 mg per kilogram, i.e., 95 times the threshold value for the least active brominated preservative. INTERFERENCE FROM ESSENTIAL OILS, TANNINS OR GLYCOSIDES I'RESEKT IN FRUIT JUICES- The experiments reported in the previous sections were all carried out with semi- synthetic media. Before the fermentation test could be accepted as a method of detecting * The authors are greatly indebted to Professor J . A. C. van Pinxteren for carrying out these deter- minations.40 MOSSBL AND DE BRUIN: THE DETECTION OF PRESEKVATJVES ~ 1 7 0 1 . 7s preservatives in fruit drinks, it was necessary to study whether the essential oils and tannins, or other compounds present in natural fruit juices, tend to inhibit the fermentation.For this purpose common Dutch fruits and some citrus fruits were hand-pressed, filtered through cheese cloth, pasteurised for 1 minute at 80" C and then inoculated with approxi- mately 1oP cells per ml of Saccharomyces cerevisiae. Lemon and grapefruit juice are known to be high in acid content; this may inhibit the fermentation and so suggest slight antimicrobial activity. Fermentation experiments with these juices were therefore also carried out after adjusting the pH value to 3.5. In addition, lemon juice was enriched with 6 per cent. of dextrose because of its low sugar content. The results obtained are summarised in Table 111.TABLE I1 INFLTJENCE OF EOSIN ox THE STANDARD FERMENTATION TEST Uncorrected volume of gas fornicd a t 34" C f 1 1':osin level, after 36 hours, after 48 hours, mg o f bromine per kg m 1 ml 0 174 - 0.5 1 2 6 10 20 50 100 17& 176 17+ 1 7 i 16 15 9 3 k TABLE 111 FERMENTATION OF FRUIT JUICES BY THE STANDARD IXOCIILUM Fruit . . . . . . . . . . . . .. Apple Blueberry . . . . . . . . Cherry . . . . . . . . Grapefruit . . . . . . . . Grapefruit (pH = 3-6) . . .. Lemon (pH = 3.6) . . . . . . Orange . . . . . . . . Pineapple* . . . . . . . . Raspberry . . . . . . . . Blackcurrant Lemon . . . . . . . . Lemon (pH = 3.5) 60;, dextrose Redcurrant . . . . . . . . * Natural undiluted juice of pineapples, Uncorrected volume of Juice gas formed at 24°C ----A- 7 r-.-----A-----7 D after 36 hours, after 48 hours, rn 1 ml PH " L O 1.351 3.1 17g - 1-34!) 3- 1 17* 1.355 3.4 17b - 1.351 2.8 17; __ - 3.5 17h -- - 10 1-346 3.0 0 1.342 9.7 0 0 - 3.6 t 5 17$ __ 3.5 13 17i 1.348 3.0 l i b -.- 1.353 3.5 17 ii 1 -340 3.6 17* - - 174 1.349 3.0 5 processed in KO.2 cans by a ('alifornian packer. One concludes from these data that the essential oils, tannins and other compounds present in the juices, other than blueberry and redcurrant, do not interfere with the fermenta- tion test. Neither of the inhibitory berry juices contained free benzoic acid, which is reported to have been detected in cranberries.= Hence, some other constituents of these juices must have slight antimicrobial properties. This fact is of little importance in the detection of brominated preservatives, since a dilution of 10 per cent.of these juices (the normal level. of fruit juices in fruit drinks) with the standard 10 per cent. sucrose - 0-25 per cent. yeast extract medium was fermented at the same rate as the sugar - yeast extract medium itself. It is therefore not to be feared that essential oils, tannins or other compounds, present in fruit juices-with the possible exception of cranberries, which are not used in soft drinks in The Netherlands-will inhibit the fermentation of fruit drinks derived from them.Jan., 19531 IN BEVERAGES BY A FERMENTATION TEST 41 Recently it has been shown, however, that pure orange oil possesses fairly strong anti- microbial properties% and that this is due to autoxidation products of d-lim~nene.~ ~36 Traces.of terpenes may be present in cloudy citrus soft drinks and so may cause false positives for brominated preservatives. It was therefore desirable to look for a method of neutralising the antimicrobial action of these terpenes. Since the antimicrobial properties of the terpenes depend on the presence of peroxides, it was worth trying to neutralise their action by reducing the redox potential of the medium. Additions of up to 1 per cent. of ferrous sulphate, ascorbic acid and sodium thioglycollate proved unsuccessful in this respect, but the addition of a few pieces of boiled liver, as suggested by Tar~zzi,~' gave the desired effect. It was necessary, however, to study how far such an addition of SH-compounds could interfere with the limiting values for the detection of the brominated preservatives.' For this purpose, the experiments reported in Table I were repeated in the presence of about 150 2-mm cubes of liver (weighing roughly 14g) per tube.For comparison, parallel tests were run with: (i) a highly inhibitory concentration, i e . , 0.05 per cent. w/v, of Guinea orange oil terpenes (peroxide value = 8.5 milli-equivalents of oxygen per kilogram) dissolved in 96 per cent. ethyl alcohol, the alcohol concentration in the culture medium being 1 per cent. v/v; (ii) the same, but with 150 pieces of liver per tube; and (iii) a "semi-blank," vix., a culture medium with 1 per cent. v/v of ethyl alcohol. TABLE IV IKFLUENCE OF CITRUS TERPENES AND LIVER 01; THE STANDARD FERMENTATIOW TEST Uncorrected volumc of gas formed at 24" C Medium f 1 after 36 hours, after 48 hours, ml ml Standard: 10% sucrose, 0.2Syo yeast extract, initial pH = 6.4 Standard + 0.050/6 of terpenes in alcohol ( l y " v/v) 0.1 174 .. * * l 3 Standard + 1% v/v of 96% ethyl alcohol . . .. .. Standard -+ terpenes + liver . . . . .. . . .. . . 5 Standard + ethyl bromoacetate -10 mg Br/kg + liver . . . . 1 93 >9 N 5 mg Br/kg + liver . . .. 1 Y9 99 N 2 mg Br/kg + liver . . . . 2 77 99 - 1 mg Br/kg +liver . . .. 5 Y Y 97 -0.5 mg Rr/kg + liver . . . . 5 33 99 -15 mg Br/kg $- liver . . 2 93 99 -10 mg Br/kg 4- liver . . 4 99 3 3 - 6 mg Br/kg + liver . . 11 99 99 - 2mg Br/kg + liver . . 0 99 3 9 - 1 mg Br/kg + liver . . 2 9, > 9 -0.5 mg Br/kg + liver . . 1 4 . . Standard + comrncrcial preparation I -20 mg Br/kg $- liver .. 1 >> 93 N 2 mg Br/kg -1 liver . . 1;i Standard + conimcrcial preparation 2 - 5 mg Br/kg + liver . . - - 0 2 6 16 5 7 16 0 1 9 - 174 When comparing the limits for the detection of the various brominated preservatives in the absence and in LIMITS Ethvl bromoacetate the presence of liver, the results in Table V are obtaiied. TABLE V FOR THE DETECTION OF BROMINATED PRESERVATIVES Levels Compound 7- 1 KO liver, mg of bromine pcr kg Ln the presence of liver, mg of bromine per kg .. . . . . .. .. < 0.6 2 Commercial preparation 1 . . .. .. . . 2 Commercial preparation 2 . . .. . . . . d 0.5 1 0 1 As might be expected,l the addition of liver increases the limiting values, but this increase is not too serious with ethyl bromoacetate and commercial preparation 2.Even for the weak preservatives (commercial preparation 1) the limiting value, although considerably42 MOSSEL -4ND DE BRUIN [Vol. 7s higher, is just of the order of the levels used in pra~tice.l~J39309~~ Moreover, addition of liver is only desirable in citrus soft drinks and even then only if terpenes are shown to be present. A combination of eosin and terpenes is not to be feared since eosin is only used in the red beverages, such as the raspberry and cherry varieties. COXCLUSIONS The selective fermentation test for the detection of preservatives developed in this study serves the purpose of indicating the presence of antimicrobial agents other than sulphurous and benzoic acids at the legally accepted concentrations. The three brominated preservatives investigated inhibit the fermentation a t levels of 0.5 to 2 mg of bromine per kilogram upwards, whereas neither the various compounds derived from fruit juices present in fruit drinks, nor eosin at concentrations below 50 mg of bromine per kilogram, interferes with the test.If, exceptionally, citrus terpenes are present in a lemonade, the test can still be satis- factorily modified so as to indicate the presence of the preservatives sought. 1 . 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 36. 36. 37. REFERBKCES Dickens, F., Hiocherri. .I., 1933, 27, 1141. Genevois, L., Cayrol, P., and Xicolaieff, T., Compt. Rend. SOC. J j i o l . , .P(li,is, 1933, 112, 1382. Genevois, L., and Nicolaieff, T., Ibid., 1933, 112, 1385.Clarenburg, A., van Bsveld, L. W., and Reith, J. F., Phavni. Weekhl., 1941, 78, 57. Hahn, F. L., Mikrochem., 1935, 17, 822. Chelle, L., and Vitte, G., Ann. Falsif., 1936, 29, 98. Florentin, I)., and Munsch, M., Ibid., 1936, 29, 104. Reith, J. F., Chem. Weekbl., 1940, 37, 319 and 535. Deshusses, J., Mitt. Lebensun. Hyg., Bevn, 1942, 33, 138. von Fellenberg, Th., Ibid., 1944, 35, 367. Eisenberg, W. V., and Wilson, J. B., .J. Ars. Ofl. Agric. Ckein., 1947, 30, 363. Eeckhout, G., Fennentatio, 1948, 17. von Fellenberg, Th., Mitt. Lebensin. Hyg., Bertz, 1961, 42, 72. van Pinxteren, J . A. C., Analyst, 1952, 77, 367. Kluyver, A. J., Thesis, Delft, 1914, p. 129. Mossel, D. A. A., Nature, 1950, 166, 188. Heim, H. C., and Poe, C. l?., Food Tech., Champuzgn, 1948, 2, 23. Vermast, P. G. F., Biochern. Z . , 1921, 125, 106. Waterman, H. I., and Kuiper, Y., Rec. Trnu. Chin?. I’uys-Bas, 1924, 43, 323. Cruess, W. V., and Richert, P. H., .I. Bact., 1929, 17, 363. Goshorn, R. H., Degering, E. F., and Tetrault, 1’. h., Ind. I:ng. Cliein., 1938, 30, 646. Hoffman, C., Schweitzer, T. R., and Dalby, G., Ibid., 1941, 33, 719. Rahn, O., and Conn, J. E., Ibid., 1944, 36, 185. Sabalitschka, Th., and Marx, H., Pharm. Z., 1947, 83, 187. von Schelhorn, M., Dtsch. Lebet.ismitt.-Rclscli., 1949, 45, 255. -, Ibid., 1950, 46, 151. -, Ibid., 1981, 47, 16 and 128. Jacobs, M. H., Cold Spr. H u b . Synip. Quloit. J j i o l . , 1940, 8, 30. Rogosa, M., J . Hact., 1944, 47, 159. Liithi, H., Mitt. Lebensm. Hyg., Bern, 1946, 37, 378. Florentin, D., Ann. Falsif., 1960, 43, 328. Reith, J. F., Voeding, 1942, 3, 220. Schweizer, Ch., Milt. Lebensm. Hvg., Uerti, 1928, 19, 1. TVinton, A. L., and Winton, I<. B., “Structure and Composition of I;oo(ls,” John Wiley 8: Sons, Inc., Yew York, 1935, Vol. 11, p. 827. Mossel, .I). X. A , , .Vatwe, 1951, 168, 999. Zukerman, I., Ibid., 19.51, 168, 617. Tarozzi, G., Zbl. H a k t . , Aht. I, 1905, 38, 61!k. CENTRAL ISSTITUTC FOR r\TUTKITION l<ESEP.KCH, T.Y.0. ITTRECHT, THE SETHERL4SDS

ISSN:0003-2654    

DOI:10.1039/AN9537800037

出版商:RSC    

年代:1953

数据来源: RSC

摘要:

Jan., 19531 FELISTEIN ANI) KLENDSHOJ 43 The Analysis of “ General Unknowns ” in Toxicology BY M. FELDSTEIN AXD N. C. KLENDSHOJ A modification of the Stas - Otto procedure for the analysis of poisons is presented. The modification is designed as a rapid sorting method in cases for which the analyst has no guidance from the autopsy findings as to the nature of a suspected poison. Toxic agents are divided into six groups, the division being based on the physical and chemical properties of members of each group. The departure from the classical Stas - Otto procedure involves the use of a large continuous liquid - liquid extractor and chloroform as an extraction medium for the separation of non-volatile organic poisons. These substances are separated in succession from an aqueous extract of tissue and fall naturally into the following groups : acid - chloroform soluble, sodium hydroxide - chloroforni soluble and ammonium hydroxide - chloroform - isopropyl alcohol soluble.IN the operation of a toxicological laboratory concerned with the investigation of sudden and violent death, approximately 30 per cent. of the samples submitted for analysis are classified as “general unknowns.” This means, simply, that pathological examination at autopsy has revealed no organic cause of death and that poisoning of some type may be indicated. In addition, many such samples are submitted for analysis in order that the possibility of death from toxic agents may be ruled out. An analysis of this kind for general unknowns presents an extremely difficult problem, owing to the great range of possible toxic agents.Many of the procedures listed in the standard textbooks on toxicology and in the recent literature are designed for the analysis of specific poisons, when these are known. For “general unknowns” the toxic agent must be sought through some procedure that classifies poisons as a group. The usual approach to such a problem follows the Stas - Otto procedure. This procedure has been modified to render it adaptable to the search for toxic materials in the general unknown. The modified procedure is not intended to replace the reliable Stas - Otto method, but has been designed to afford the analyst a rapid sorting system for those poisoning cases in which there is no guidance from the autopsy as to the nature of the suspected poison.The procedure has been tested over a period of 4 years and has been found satisfactory. Substances to be looked for have been divided into the following six groups- Acid - chloroform soluble non-volatiles. Sodium hydroxide - chloroform soluble non-volatiles. Group 1. Volatiles. Group 2. Group 3. Group 4. Ammonium hydroxide - chloroform - isopropyl alcohol soluble non- Group 5. Heavy metals. Group 6. Anions. volatiles. Each of these groups is separated and tested for toxic agents. The procedures that are used for the isolation of each of these groups will be described in detail under the heading of each group. The assignment of a particular substance to one of these groups generally follows from a consideration of its physico-chemical properties.If a particular poison is found, a new portion of tissue is taken and subjected to specific quantitative analysis for that substance. METHOD PKOCEDUKE- transfer it to a 2-litre flask. poisons. usual manner. flask. I’olatiles (Group 1)-Grind a portion of tissue, about 600 g, in a IVaring blendor and Steam distil from acid solution in order to remove the volatile Collect 500ml of distillate and test for volatile agents (including alcohol) in the L4 cid - chloroform soluble non-volatiles (Group 2)-Filter the residue in the distilling Suspend the insoluble precipitate in 500 ml of boiling water to extract any further44 FELDSTEIN A N ) KLENDSHOJ THE ANALYSIS OF [Vol. 78 toxic materials and filter. Place a litre aliquot in a continuous liquid - liquid extractor similar to the one designed by Hershberger and Wolfel (Fig.1). Extract with chloroform at laboratory temperature for a minimum of 2 hours. Filter the chloroform extract, which contains the members of this group, and dry it over anhydrous sodium sulphate for several hours. Distil off the chloroform until approxi- mately 50 ml are left in the flask. Transfer the contents of the flask with a few rinsings of chloroform into a small beaker and evaporate to dryness on a steam-bath. The members of this group, consisting of barbiturates, salicylates, phenacetin, acetanilide and so forth, Combine the filtrates and measure their volume. Standard-taper toint-- about 26 inches Capacity 1200-1400 m' i - 4 about I 2 inches about 18 inches Fig. 1. Continuous liquid - liquid extractor remain in the beaker. The absence of the members of this group is usually indicated by the absence of any residue.In order to determine the completeness of extraction of members of this group, known amounts of substances were added to brain tissue, which was then subjected to the procedure described. Table I shows the results of these experiments. Apply the usual qualitative tests to this residue. TABLE I RECOVERY OF MEMBERS OF THE ACID - CHLOROFORM SOLUBLE GROUP ADDED TO 6oog OF BRAIh' Substance Amount added, Amount recowred, mg mg Phenobarbital . . ., .. * . 15 14 Nembutal . . . . .. .. .. 12 11 Salicylic acid . . . . .. .. .. 17 17 Phenacetin . . . . .. . . .. 30 27 Acetanilidc . . . . .. .. . . 15 13 Sodium hydroxide - chloroform soluble mn-volatiles (Group +-Make the aqueous liquid remaining in the extraction chamber after removal of the acid - chloroform soluble groupJan., 19531 “ GENERAL UNKNOWNS ” I N TOXICOLOGY 45 alkaline to litmus-paper with sodium hydroxide and extract continuously.Treat the chloroform extract exactly as for group 2. Table I1 shows the results of recovery experiments with members of this group, including codeine, atropine, strychnine, thonzylamine and procaine. TABLE 11 RECOVERY OF MEMBERS OF THE SODIUM HYDROXIDE - CHLOHOFORM SOLUBLE GROUP ADDED TO 500g OF BRAIN Substance Amount added, Aniount recovered, mg mg Stryclininc . . .. .. . . .. 10 9 Thonzylamine (neohetramine) . . . . 25 25 Codeine . . . . . . . . .. 10 9 Atropine . . . . . . .. . . 5 4 10 9 Procaine (novocaine) ... . .. 10 10 Ammo.tzium hydroxide - chloroform - isopyopyl alcohol soluble 7tou-volatiles (Group 4)- Make the aqueous liquid remaining in the extraction chamber after the removal of group 3 acid to litmus with dilute hydrochloric acid, and then alkaline with ammonium hydroxide, and once again extract continuously with a mixture of 5 parts of chloroform and 1 part of isopropyl alcohol. Table I11 shows recovery cspei-i- ments with morphine. which is the most important member of this group. Treat the extract exactly as indicated above. TABLE I11 RECOVERY OF MORPHIXE (GROUP 4) ADDED TO 5OOg OF BRAIN Morphine added, Morphine recovered, mg I l l g 20 1s 1 Tt I 5 1 0 s ) 4 Heavy metals (Grou9 5)-Place a 100-g portion of ground tissue in a 3-litre Erlenincyer flask and add about 200ml of concentrated nitric acid.Place the flask on a steam-bath for a minimum of 12 hours. After cooling, filter the material through glass wool to remove the solidified fat. Transfer the solution to an 800-ml flask with ground-glass joints (Fig. 2) and add 5 ml of concentrated sulphuric acid. Distil the contents until charring occurs. Cautiously add a mixture of 3 parts of nitric acid and 1 part of perchloric acid until the charred material is completely oxidised. Distil until fumes of sulphur dioxide appear. Allow the flask to cool, add 10 ml of distilled water and distil again until fumes of sulphur dioxide appear. Transfer the contents of the flask with several Ei-ml portions of distilled water to a 125-ml Erlenmeyer flask. This solution contains any heavy metals that were present in the tissue.Test the solution qualitatively and quantitatively in the usual manner to determine the members of this group. Table IV shows the results of experiments in which known amounts of metals have been added to tissue and subjected to the procedure described. ‘TABLE IV RECOVERY OF METALS ADDED TO 100g OF BRAIN Substance Amount added, Amount recovered, mg mg Lead . . . . . . .. . . . . 0.10 0.10 0.20 0.19 Arsenic . . . . . . . . 0.50 0.75 0.49 0.76 Zinc . . . . .. . * . . . . 10.00 9.70 20-00 19.50 Mercury . . . . . . . . . . 0-50 0.42 1.00 0.95 Anions (Group 6)-The members of this group are separated into two classes: those The first that cannot be destroyed by ashing and those that are destroyed by ashing.46 FELDSTEIN AND KLENUSHOJ THE ANALI'SIS OF [Vol.78 sub-group, which includes borate, iodide, bromide and nitrate, is isolated by ashing a portion of the tissue in the presence of sodium carbonate. The material is usually ashed in a porcelain evaporating dish in a muffle furnace regulated a t 600" C. Substances that cannot be ashed, e.g., nitrite and chlorate, are separated from tissue by dialysis against distilled water. The dialysate is concentrated by evaporation in a vacuum oven at 50' C. Fig. 2. Distillation apparatus No quantitative recovery data are presented for these groups, since most of the substances have been determined by procedures described in the literature. DISCUSSION OF RESULTS The first four groups of toxic agents require only one sample of tissue and the next three require one sample each, making in all a total of four samples of tissue.When large amounts of tissue are submitted, 500g are taken for the first four groups and 1oOg for the others. The procedure provides for the separation of each of these groups into distinct and separate fractions. The usual procedure with general unknowns is to subject the four portions to analysis at the same time. Whilst the volatiles are being distilled, or the non-volatiles are being extracted, the metals and anions can be prepared so that at the end of a working day the six groups are separated and ready for analysis at the same time. The tests for volatile materials are too numerous to warrant repeating here. The literature is also replete with tests for the non-volatiles, metals and anions.The advantage of the liquid - liquid extractor lies in producing relatively pure residues, and it is often possible by mere inspection to rule out the presence of a complete group. The sulphuric acid solution of the metal group can be examined according to various schemes. If the presence of large amounts of metal is suspected, it can be analysed according to standard qualitative pro- cedures. It has been our practice to separate each of the groups mentioned and then perform qualitative tests to establish the identity of any particular agent within each group. For For smaller amounts, specific micro-methods can' be applied.Jan., 19531 “GENERAL UNKNOWNS” IN TOXICOLOGY 47 example, should our general unknown analysis indicate the presence of a barbiturate, a new sample of tissue would be analysed for barbiturates by ultra-violet absorption techniques.It has been found distinctly advantageous to reserve portions of each tissue submitted for analysis in a container kept under deep-freeze refrigeration. This provides a convenient method for storing tissue samples. The reserve tissue is usually ground in a Waring blendor and weighed in 10 to 15 or 100-g portions into waxed containers, which are stored under deep-freeze refrigeration. When a particular toxic agent is found by qualitative analysis, there is already at hand a weighed and frozen portion of the tissue for quantitative analysis. Experience has shown that tissues stored in this fashion show no change in concentration of any of the toxic agents, including alcohol and the volatiles. For example, analyses on portions of tissue stored over a period of 6 months show identical results for alcohol content. Extraction of the aqueous samples is done in the cold, since the chloroform vapours are condensed, and the cooled liquid is permitted to drop through the sample. Emulsions are rarely formed; even when one is, it is broken as the extraction continues. No details for individual analyses have been presented for the reasons given. REFEKESCE 1. Hershberger, E. B., and Wolfe. S. K., J . Biol. Chenz., 1940, 133, 669. DIVISION OF TOXICOLOGY UNIVERSITY OF BUFFALO, SCHOOL OF MEDICINE THE BIOCHEMISTRY LABORATORY BUFFALO GENERAL HOSPITAL BUFFALO, KEW YORK First submitted, March 13th, 1952 Amended, August l l t k , 1952

ISSN:0003-2654    

DOI:10.1039/AN9537800043

出版商:RSC    

年代:1953

数据来源: RSC