ISSN:0003-2654    

DOI:10.1039/AN95782FX046

出版商:RSC    

年代:1957

数据来源: RSC

ISSN:0003-2654    

DOI:10.1039/AN95782BX048

出版商:RSC    

年代:1957

数据来源: RSC

ISSN:0003-2654    

DOI:10.1039/AN95782FP141

出版商:RSC    

年代:1957

数据来源: RSC

ISSN:0003-2654    

DOI:10.1039/AN95782BP149

出版商:RSC    

年代:1957

数据来源: RSC

摘要:

NOVEMBER, 1957 Vol. 82, No. 980 THE ANALYST PROCEEDINGS OF THE SOCIETY FOR ANALYTICAL CHEMISTRY OKDIKARY MEETING AN Ordinary Meeting of the Society was held at 6.30 p.m. on Tuesday, Xoveinber 5th, 1957, in the Lecture Theatre of The Royal Institution, 21 Albemarle Street, London, jV.1. The Chair was taken by the President, Dr. J. H. Hamence, M.Sc., F.R.1 .C. A lecture on “Recent Developments in Chelatometry” was given by Dr. Rudolf Pribil. XE\V MEMBERS ORDIXARY MEMBER\ John Derek Cosgrove, B.Sc. (Dunelm.) ; Gino Dicastro, Dr. Biochcm. (Rome) ; Cyril Jack Keattch, A.R.I.C. JVKIOR A~EAIBLR Rayniond Frederick Hall. SORTH OF ENGLAND SECTIOS AN Ordinary Meeting of the Section was held a t 8.15 p.ni. on Saturday, October 5th, 1957, a t the Engineers’ Club, Albert Square, Manchester. The Chair was taken by the Chairman of the Section, hlr.A. N. Leather, B.Sc., F.R.I.C. A discussion on “The Analysis of Trade Effluents” was opened by J. G. Sherratt, B.Sc., F. 13. I, c. MII>LANL>S SECTIOK AX Ordinary Meeting of the Section was held a t 6.30 p.m. on Tuesday, October 8th, 1957, in the Mason Theatre, The University, Edrnund Street, Birmingham, 3. The Chair was taken by the Tice-Chairman of the Section, Dr. S. H. Jenkins, F.R.I.C., F.1nst.S.P. The following paper was presented and discussed : “Analytical Methods in Clinical Biochemistry”, by H. T’arley, M.Sc., F.K.I.C. BIOLOGICAL METHODS GKOCP AN Ordinary Meeting of the Group was held at 6.30 pm. on Wednesday, October 9th, 1957, in “The Feathers”, Tudor Street, London, E.C.4. The Chair was taken by the Vice-chairman of the Group, Dr. J. I. i19. Jones, F.R.I.C. A discussion on “Biological Standards” was opened by J. \V. Lightbown, M.Sc., Dip. Bact., F.P.S. 721

ISSN:0003-2654    

DOI:10.1039/AN9578200721

出版商:RSC    

年代:1957

数据来源: RSC

摘要:

722 KELLIE : THE DEI'EKMINATIOX OF 1"iOXO STEROIDS (17-KETO STEROIDS) [VOl. 82 The Determination of 17-0x0 Steroids (17-Keto Steroids) and 17-Oxogenic Steroids (17-Ketogenic Steroids) A Review BY A. E. KELLIE* SUMMARY OF CONTENTS Introduction. Precursors of 17-ox0 steroids and 17-oxogenic steroids. Determination of 17-0x0 steroids. Determination of 17-0x0 steroids in urine. Hydrolysis of steroid conjugates : (a) glucuroriides and (b) sulphates. Chromatography of 17-0x0 ster0id.s. Identification of individual 17-oxo steroids. Determination of 17-0x0 steroids in blood. Determination of 1.7-oxogenic steroids. Determination of non-ketonic steroid ~~lcohols. ALTHOUGH the presence of androgens in urine WAS first demonstrated by biological methods,lI2 bioassay has now been largely replaced by colorimetric methods, which measure both bio- logically active and inactive androgen metabolites.Not all urinary androgen metabolites are 17-0x0 steroids nor are all 17-0x0 steroids androgenic, and it was fortuitous that two of the first crystalline steroids isolated from urine by B ~ t e n a n d t , ~ ~ ~ androsterone (3a-hydroxy- 5a-androstan-17-one) and dehydroepiandrosterone (3,L-hydroxyandrost-5-en-l7-one), were 17-0x0 steroids possessing biological activity. Zimmerma~in~?~ showed that these and other pure steroids containing an activated methylene group -CH2-CO- could be quantitatively determined by the colour given with m-dinitrobenzene in alkaline solution, and Wu and Chou' applied this reaction to measure the concentration of chromogenic androgen meta- bolites in urine.The Zimmermann reaction for 17-0x0 steroids distracted attention from the fundamental importance of biological activity, but without this simple test the rapid expansion that has taken place in the past 20 years would not have been possible. This reaction forms the basis of all reliable methods for the determination of 17-0x0 steroids. Recently, chemical methods have been established whereby more complex steroids (the 17-oxogenic steroidss) may be converted to 17-0x0 steroids for indirect determin a t ' ion. PRECURSORS OF 17-0XO STEIIOIIIS AND 17-OXOGENIC STEROIDS The use of carbon-14 labelling has proved beyond reasonable doubt in man that the synthesis of the biologically important steroids, including cholesterol, can be achieved from a ~ e t a t e .~ Whereas cholesterol, which can act as a precursor of steroid hormones, is formed by many body tissues, the formation of the steroid hormones is closely associated with the adrenals, the gonads and, during pregnancy, the placenta. Dorfman has proposed that the term biosynthesis should be reserved for those metabolic processes that result in the production of steroid compounds with maximum biologicd activity and that all subsequent changes should be considered to be strictly catabolic. I f this criterion is accepted, then the primary steroid products of the endocrine system (excluding oestrogens) are as follows- OH CH, OH (1) (11) (111) Testosterone 1 lp-Hydroxyandrost-4-ene-3: 17-dione Aldosterone * British Empire Cancer Campaign Research Fellow.November, 19571 .4ND 17-OXOGENIC STEROIDS (17-KETOGENlC STEROIDS) 723 7 H 3 CHIOH CHiOH &=* *=’ o#-OH L O (IV) (V) (VI) Progesterone Corticosterone Cortisol This is, of course, an over-simplification of the true situation, as many other related steroids of lower biological activity are produced that may have physiological importance; as these compounds are catabolised in a similar way, the concept is both apt and useful.In general, compounds of high biological activity (I to VI) after intravenous injection disappear rapidly from the blood; for example, the estimated half-life of progesterone (IV) in circulation is 5 minutes and that of testosterone (I) is 8 minutes. Some of the injected material is excreted in the urine in modified form as water-soluble conjugates.These active compounds have in common the A4-3-ketone structure in ring A of the steroid nucleus and, irrespective of sub- stitution in rings B, C and D, they undergo reduction to the dihydro (VIII) and tetrahydro (IX) derivatives before excretion, as shown by the following forinulae- 0 a- HO a- Conjug.-0 (VIII) Dihydro reduction product (IX) (XI Tetrahydro Conjugated reduction steroid product Although it is uncertain where all the intermediate metabolic changes take place, these compounds are the precursors of the 17-0x0 steroids and the 17-oxogenic steroids that appear in blood and urine. The reactions leading to the reduction of the A4-3-ketone group and the subsequent conjugation with glucuronic acid are catalysed by phosphonuc1eotideslO >l1 and probably take place in the liver, so that peripheral blood contains, in addition to the secreted hormones, conjugates of the tetrahydro and probably also of the dihydro (VIII) products. Whether or not all of these circulating metabolites are excreted in the urine depends upon their clearance by the kidney and, because of this complex situation, the dilemma arises as to whether to carry out determinations on blood or urine and what precisely to determine.illthough the concentration of steroid metabolites in blood is probably close to that in the intracellular fluid to which cells are exposed, these values refer only to the time of sampling and do not allow for diurnal variations in secretion. On the other hand, determinations carried out on 24-hour urine samples correspond to the average secretion throughout the day, but do not cover the possibility of selective renal clearance producing a urinary steroid composition that does not resemble that of the internal In both blood and urine the amounts of hormone and hormone metabolites in the free form are very low com- pared with the amount present in the form of conjugates, and most methods recommended for routine use are designed to measure the latter only.The determination of trace amounts of specific free hormones in blood and urine has been described for several hormone~,~~J* but such methods are highly specialised and technically difficult and lie outside the scope of this review, which is limited to the determination of such metabolites of adrenal and gonadal origin as can be measured by the Zimmermann reaction for 17-0x0 steroids.This includes, on the one hand, the native 17-0x0 steroids and, on the other, those steroid com- pounds that can be converted by specific reactions into compounds giving a colour in the Zimmermann reaction. DETERMINATION OF 1’7-0x0 STEROIDS In contrast to the uncertain meaning of terms such as “corticosteroids,” “11-oxy steroids” and so on, there is no ambiguity about the name 17-0x0 steroids, which embraces a group of related steroids, of diverse metabolic origin, that appear in blood and urine and724 KELLIE : THE DETERMINATION OF 17-OXO STEROIDS (17-KETO STEROIDS) [VOl. 82 have as a common structural feature a carbonyl group at the 17-position of the steroid nucleus. The determination of this class of compound.is usually carried out by the Zimmermann reaction,15P which, under defined condition!;, is reasonably specific. Alternative colour reactions that have been proposed17J8 have on1,y limited usefulness. A polarographic met hod for the determination of keto~teroidsl~ has also had little application. The modification of the Zimmermann reaction proposed by Callow, Callow and Emmens20 has been widely adopted. The reagents (0.2 m'l of 2 per cent. w/v ethanolic nz-dinitrobenzene and 0.2 ml of 2.5 N ethanolic potassium hydroxide) are incubated at 25" 0.1" C for 1 hour with 0.2 ml of an ethanolic solution containing 0 to 100 pg of 17-0x0 steroid and are then diluted with 10 ml of ethanol for absorptiometric measurement. Under these conditions 17-0x0 steroids give at 520mp strong absorption that obeys Beer's law closely.3-0x0 steroids give a typical colour after 5 minutes' incubation, but they differ from 17-0x0 steroids in that the colour fades rapidly, and a t 1 hour there is very little absorption at 520mp. 20-0x0 steroids and A4-3-oxo steroids also react slowly with the reagents in the Zimmermann test to give low general absorption. Cholestan-2-oneZ1 and 2 : 3 : 6-trimethylbenzylidene- acetone22 are not 17-0x0 steroids, but they do give the typical colour reaction of this group. The presence of red or purple pigments ic. urine extracts is sometimes troublesome and colour-correction formulae based on the measurement of background absorption are fre- quently i n a d e q ~ a t e . ~ ~ This difficulty can be 'overcome by preparing a ketonic fraction by means of Girard's2* reagent T (carbohydra:~idomethyltrimethylammonium chloride) or, if it can be shown that the background absorption is linear, by applying the Allen c o r r e c t i ~ n .~ ~ Alternative conditions for the Zimmermann reaction have been described by Holtorff and Koch.26 This reaction is carried out in aqueous ethanol and the period of incubation is longer. The method is less reliable than that described by Callow et nl., because the determination of 17-0x0 steroids in impure extracts shows a departure from the linear response required by Beer's law. Under Callow's conditions, and more especially under those of Holtorff and Koch, individual 17-0x0 steroids have different molecular extinction coefficient~,~7 e.g., 3cr- hydroxydp-androstane-l l : 17-dione (1 1-oxoaetiocholanolone) gives a higher molecular extinction than 3a-hydroxy-5/3-androstan-17-one (aetiocholanolone) and for this reason the determination of a mixture of 17-0x0 steroids of unknown composition is only approxi- mate.3~-Acetoxy-14~-hydroxyandrost-5-en-17-one28 and 3p: 16a-dihydroxyandrost-5-en- 17-one (16a-hydroxydehydroepiandrosterone), recently isolated from urine,29 are rare examples of 17-0x0 steroids that do not give the Zimmermann reaction. Micro-scale methods have been describedz7 930 in which as little as 1 pg of 17-0x0 steroid may be determined; such methods depend for their precision on a low reagent blank. Important factors influencing this value include the purity of the m-dinitrobenzene and of the ethanol and the stability of the ethanolic potassium hydroxide. The method of Wilson,31 based on that of Hamburger,32 gives a satisfactory potassium hydroxide reagent ; the solution is prepared in the presence of ascorbic acid and is stored under nitrogen at 4" C .DETERMINATIOS OF 17-cixo STEROIDS IN URINE Originally, measurements of urinary 17-0x0 steroids involved the preparation of a neutral steroid extract and the determination of a figure for the sum of all neutral 17-0x0 steroids, Later, the pathological significance of the appearance of large amounts of dehydroepiandro- sterone in urine was recognised and the origina.1 method was extended to permit the ratio of 3a- to 3p-hydroxy-17-0x0 steroids to be determined (by precipitation as d i g i t ~ n i d e ~ ~ ) in an attempt to establish a differential diagnosis between adrenal carcinoma and hyperpbasia.This period was also associated with many reports of the isolation of individual compounds; these efforts were invariably carried out on a large scale and were not quantitative. In contemporary methods the total 17-0x0 steroid fraction is analysed quantitatively in terms of individual compounds, small urine samples being used. Substantially all the 17-0x0 steroids in normal urine are excreted as water-soluble con- jugates, and there is good evidence that the total is largely made up of glucuronides and sulphates. that trace amounts of the C,, 3 : 17- diones that have been isolated cannot be conjugated in this way unless they undergo pre- liminary enolisation; the suggestion that they may be excreted as thiazolidines has, as yet, little experimental support.Although 17-0x0 steroid glucuronides and sulphates can be ex- tracted quantitatively from urine36 and have been separated from each other,36 there is no satis- factory method for the quantitative separation. of individual conjugates. In the absence of Lieberman and Dobriner have pointedh'ovember, 19571 AND 17-OXOGEliIC STEROIDS (17-KETOGENIC STEROIDS) 725 such methods, it has been customary to hydrolyse the water-soluble conjugates and to extract the liberated steroids into organic solvents. Because of the difficulties associated with this preliminary hydrolysis, most of the methods suggested have been unsatisfactory. HYDROLYSIS OF 17-OXO STEROID CONJCGATES- Glucuronides are reasonably stable at room temperatures in mildly acid or alkaline solution; sulphates, on the other hand, although stable under alkaline conditions, undergo progressive hydrolysis at acid pH values at room temperature.Both forms of conjugates are completely hydrolysed on being boiled in acid solution and this method of hydrolysis has been the mainstay of so-called quantitative methods for two decades. This method of hydrolysis, presumably because of its convenience, persists, yet it is for most problems entirely unsatisfactory. I t is, perhaps, possible t o justify this procedure in quick methods for scanning a large number of urine samples in a preliminary but even in this application it can give rise to misleading results. Hydrolysis with hot acid produces artefacts of dehydration and sub~titution,~8 and these complicate subsequent analysis of the liberated 17-0x0 steroids.Thus, 3-hydroxy- and 11-hydroxy-17-0x0 steroids are dehydrated to the corresponding A2 and AQ(ll) compounds ; in addition, the 3p-hydroxy-A6 compound dehydro- epiandrosterone is converted in the presence of hot hydrochloric acid (or sodium chloride and sulphuric acid) into the corresponding 3p-chloro compound. If these artefacts were equivalent in colour value in the Zimmermann reaction to their precursors, this would not affect the determination of total 17-0x0 steroids, but not only is this not so, but there is strong evidence that total destruction of their precursors also occurs.3g The effect is more marked when isolated sulphate fractions are h y d r ~ l y s e d ~ ~ and, as this fraction contains a high pro- portion of dehydroepiandrosterone, it is predominantly this compound that is d e s t r ~ y e d .~ ~ Accurate determination of dehydroepiandrosterone is not possible after hydrolysis with hot acid, and no method of determining 17-0x0 steroids can be considered satisfactory if it fails with this important compound. Hydrolysis of 17-0x0 steyoid glucuronides-Hydrolysis by means of enzymes is by far the most satisfactory method for the hydrolysis of 17-0x0 steroid glucuronides. P-Glucuronidase from calf spleen,4O bacteria41 or molluscs42 has been used and the hydrolysis can be carried out by adding the enzyme directly to the buffered urine or to urine extracts containing the conjugates.As enzyme inhibitors have been reported in urine,43 the latter method may be preferable, for, by proceeding in this way, much non-steroid matter is excluded from the enzyme digest and, further, as the dried conjugate extract can be redissolved in small volumes of buffer, much higher concentrations of enzyme can be used. A convenient and prolific source of P-glucuronidase in Britain is found in the visceral hump of the common limpet (Patella vulgata)" and from this source preparations with an activity of 106 Fishmann units per gram can readily be prepared. This enzyme brings about complete hydrolysis of 17-0x0- steroid glucuronides (16 hours at 40" C) without the formation of artefacts. Hydrolysis of 17-0x0 stevoid sulphates-According to Bitman and C ~ h e n , ~ ~ 3/3-hydroxy- A5-steroid sulphates are hydrolysed when treated with acetate buffer at pH 4.7.This hydrolysis is promoted by limpet preparations because steroid sulphatases are also present. One of these enzymes, studied by Roy,46 has been shown to bring about the complete hydrolysis of steroid sulphates that have the 3/3-hydroxy-A5 or the 3/3-hydroxy-5~ structure, as in dehydroepiandrosterone sulphate and 3/3-hydroxy-5a-androstan-17-one (epiandrosterone) sulphate, respectively. Androsterone sulphate (3a-hydroxyda) and aetiocholanolone sulphate (3a-hydroxy-5/3), both of which are present in urine, are not hydrolysed; nevertheless, for methods exclusively concerned with compounds of the 313-hydroxy-A5 configurati~n,~g this enzyme is an ideal hydrolytic catalyst.Attempts to bring about the simultaneous hydrolysis of the glucuronide and sulphate conjugates of neutral steroids by limpet powder are unsatisfactory, as the pH optima of the two enzymes are different4' and, further, owing to the presence of androsterone sulphate and aetiocholanolone sulphate in urine and plasma extracts, hydrolysis is incomplete. By carrying out the hydrolysis with ,B-glucuronidase at pH 4.0 in the presence of 0.03 M potassium dihydrogen phosphate, the sulphatase enzyme is inhibited and only glucuronides are hydro- lysed. If the sulphates are subsequently hydrolysed independently, the mode of conjugation of individual 17-0x0 steroids can be studied. Fortunately, alternative methods are available for the hydrolysis of steroid sulphates.Cohen and O n e s ~ n ~ ~ demonstrated that dioxan in the presence of trichloroacetic acid will726 KELLIE : THE DETERMIKATION OF 17-.OXO STEROIDS (17-KETO STEROIDS) [VOl. 82 bring about the hydrolysis of steroid sulphates irrespective of configuration or degree of unsaturation. Hydrolysis takes place at room temperatures overnight and produces few artefacts. Water must be removed from the reagents and excluded from the reaction, a condition that is difficult to ensure with urine extracts. An alternative method3$ of hydro- lysing steroid sulphates, which does not split glucuronides, is by continuous extraction of an aqueous solution of conjugates adjusted t o pH 1. This method can be readily applied to urine samples and appears to be a mild and quantitative method of hydrolysing the sulphate conjugates.In normal urines, hydrolysis by P-glucuronidase, followed by extraction with ether at pH 1 does not lead to the formation of artefacts and, further, the amount of 17-0x0 steroid liberated from conjugation by these combined methods is a t least as great as that obtained by hydrolysis with hot acid. Gallagher has reported the analysis of an abnormal urine from a case of adrenal hyperplasia for which this did not appear to be true; this urine, after being hydrolysed with 13-glucuronidase and continuously extracted with ether at pH 1, was progressively acidified in stages to 4 N and re-extracted with ether. At each stage further amounts of 17-0x0 steroid were 0btained.4~ According to Cohen and Oneson, hydrogen chloride in dioxan will hydrolyse sulphate and glucuronide conjugates simultaneously; for this purpose, Buehlerso recommended con- tinuous extraction with ether of an aqueous solution adjusted to 7.2 N.Both of these methods make use of strongly acid conditions, which are best avoided. When hydrolysis is complete, the liberated steroids are extracted from the aqueous phase by means of an organic solvent. Many solvents have been used for this purpose, e.g., ether, benzene and methylene dichloride, but, as few comparisons have been made of their effective- ness and selectivity, there are no rational grounds for selecting any particular one, especially as 17-0x0 steroids are markedly hydrophobic and quantitative extraction is not difficult to accomplish. CHROMATOGRAPHY OF 17-OX0 STEROIDS- Complete hydrolysis and the absence of artefacts are essential for satisfactory chromato- graphic separation.Artefacts, when formed, are difficult to separate from each other and frequently interfere with the separation of genuine urinary 17-0x0 steroids ; thus the dehydra- tion product of 3a : 1 l~-dihydroxy-5a-androstan-17-one (1 1 p-hydroxy-androsterone) formed during hydrolysis with hot acid is 3a-hydroxy5a-androst-9( 1 l)-en-17-one, which cannot be completely separated from androsterone by any chromatographic system and consequently confuses the separation of androsterone and aetiocholanolone. After complete hydrolysis of conjugates has been achieved under mild conditions a very large choice of chromatographic systems is available for separation of the mixed 17-0x0 steroids.For precise analysis, chromatography on a column is recommended, but several paper-chromatographic systems have also been described. The early work of the Callowssl and many of the large-scale separations achieved by Dobriner et aL.62~53 were carried out by adsorption on alumina ; alternative, but less satisfactory, adsorbents include silicaw and magnesium ~ilicate.~z Neutral steroids such as the 17-0x0 steroids can be recovered quanti- tatively from alumina, and the potential resolving power of this adsorbent is very high. Dingemanse, Huis in’t Veld and de L a a P described a method for the quantitative analysis of small samples of 17-0x0 steroids, based upon the elution of the individual compounds from alumina by benzene containing increasing concentrations of ethanol.The ethanol con- centration was increased stepwise. This method in its original form and in the many modi- fications to which it gave rise56~57J8 has been of great value in the study of urinary excretion of 17-0X0 steroids. Yet not one of these methods achieved satisfactory resolution of the urinary steroid fraction; the 11-deoxy-17-0x0 steroids were poorly separated and the 11- hydroxy compounds, had they survived the hydrolysis with hot acid used in all these methods, would have been more intractable. The Pond method is also open to criticism, because suction, used to increase the rate of flow through the column, may introduce solvent-vapour spaces in the adsorbent column. A notable improvement of the separation obtainable by the Dingemanse chromato- graphic procedure was obtained by Lakshmanan and Liebe~man,~~ who made use of the principle of gradient elution.In the system alumina/ethanol - benzene, 17-0x0 steroids have curved adsorption isotherms, and for this reason the compounds are eluted from alumina with marked “tailing,” which prevents the separation of adjacent compounds. The effect of the gradient is to make these compounds behave as though they had linear adsorptionNovember, 19571 AND 17-OXOGENIC STEROIDS (17-KETOGENIC STEROIDS) 727 characteristics, so that they are eluted as symmetrical peaks; the movement of the com- pounds under gradient elution is more rapid than in the absence of a continuous gradient and the development of the chromatogram can be completed with a much smaller volume of eluting solvent, A suitable gradient of ethanol in benzene applied by the simple device shown in Fig.1 gives excellent separation of the 17-0x0 steroids and permits quantitative analysis of the main urinary 17-0x0 steroids on small urine samples.30 3 Donor reservoir ethanol in benzene) (2% v/v of Recipient reservoir ethanol in benzene) (0.2% v/v of Fig. 1. Apparatus for separation of 1'7-0x0 steroids by gradient elu- tion. Reproduced by permission of the Editors from Biochem I., 1957, 66, 196 It has been known for many years that the behaviour of compounds adsorbed on alumina is greatly influenced by the activity of the alumina, and it has been customary to specify Brockmann grades60 for the separation of 17-0x0 steroids.The activity of alumina for this purpose is greatly influenced by the apparent moisture content (loss of weight on heating at 100" C), which has a pronounced effect on the separation attainable under gradient elution. This has practical importance, as the adjustment of the moisture content of alumina provides the easiest way of reproducing batches of alumina with standard properties. Various com- plex methods of adjusting the activity of alumina by exposure to aqueous solutions of known vapour pressure are slow and probably unnecessary, as the required activity (corresponding to between 4 and 5 per cent. of water by weight) can be achieved by adding sufficient water dropwise to the alumina exposed in shallow trays and subsequently mixing the alumina to achieve uniform distribution.Many factors other than the moisture content of the alumina affect the resolution of 17-0x0 steroids under gradient elution; these include the dimensions of the column, the rate of flow and the nature of the gradient applied. The theoretical aspects of some of these factors have been studied by DrakeeB1 IDENTIFICATION OF INDIVIDUAL 17-OX0 STERIODS- failure to identify individual compounds other than by the position of elution. A further serious disadvantage of methods based on the Dingemanse method is the Although728 KELLIE: THE DETERMINATION OF 17-0XO STEROIDS (17-KETO STEROIDS) [Vd. 82 it is true that, with identical alumina and solvents, pure compounds behave in a predictable manner, it is unsound to assume for fractions from normal and abnormal urines in the presence of a great excess of non-steroidal matter that the position of elution will be main- tained.Infra-red absorption spectroscopy, introduced to the steroid field by Dobriner, Lieberman, Rhoads, Jones, Williams and BarnesB2 offers an excellent method of characterisa- tion when amounts in excess of 100 pg are available. Modern analytical methods, however, do not provide more than a few micrograms of the minor components, and they necessitate alternative methods of identification. When the amount available after determination is in excess of 5 pg, paper chromatography against standard reference compounds provides additional evidence of identity. BushB3 has described systems suitable for 11-oxy- and 11-deoxy-17-0x0 steroids. When smaller amoiints are available, the complex formed in the Zimmermann reaction, after being determined absorptiometrically, can be extracted from the solution and used to establish identitySs4 Many systems of paper chromatography have been suggested for the separation of 17-0x0 steroids.Conventional two-phase systems have been described by A ~ e l r o d , ~ ~ Heft- man,66 Neher and WettsteinB7 and Savard,68 and a reverse-phase system developed by Krit- chevsky and Kirkag is run on paper impregnated by dipping in stearatochromic chloride, Shull, Sardinas and Nubel'O deposited alumina on paper and used this as an adsorbent material. There is little doubt that the labour and time involved in running paper chromatograms is much less than is necessary for column work; on the other hand, without preliminary purifica- tion of the material to be chromatographed, the former method is less accurate.The dis- proportionate amounts of 11-deoxy- and 11-oxy-17-0x0 steroids is such that overloading of the paper is necessary to provide sufficient of the minor components for assay, and no single system of paper chromatography satisfactorily separates the wide range of compounds present, When the individual compounds have been located, they are eluted from the paper for determination. Invariably in this procedure unspecific chromogenic material is also eluted from the paper; it may arise from the original extract or from the materials of the paper. This blank value for the paper sets a lower limit on the amount of 17-0x0 steroid that can be determined.A wide range of colour reagents has been suggested for the location of 17-0x0 steroids,66171 but few compare in specificity with the reagents for the Zimmermann reaction. Papers are treated by immersion in 2 per cent. w/v ethanolic m-dinitrobenzene and in 2-5 N ethanolic potassium hydroxide and are then dried and developed in a current of warm air. I80 I60 140- 120- 100- 80- 60 40 20 D ( 0 ) - A - - E X - / 0 1 I A 40 60 80 100 I20 140 160 Volume of eluate, ml 17-0x0 steroid excretion patterns of the urine from a normal male subject: (a) chromatogram of glucuronide fraction: (b) chromatogram of sulphate fraction. D, dehydroefiiandrosterone; A, androsterone; E, aetio- cholanolone; OA, 11-oxoandrosterone; OE, 11-oxoaetiocholanolone; HOA, 1 lg-hydroxyandrosterone ; HOE, 1 lg-hydroxyaetiocholanolone; 0, a mixture containing androstane-3 : 17-dione and aetiocholane-3: 17-dione; X, unknown.Reproduced by permission of the Editors from Biochem. I., 1957, 66. 196 Volume of eluate, ml Fig. 2.November, 19571 AND 17-OXOGENIC STEROIDS (17-KETOGENIC STEROIDS) 729 Fig 2 shows typical analyses of the urinary 17-0x0 steroid glucuronide and sulphate fractions obtained by gradient elution. In the former fraction the two main components androsterone (A) and aetiocholanolone (E) represent terminal metabolites of C,, steroids pro- duced by the testis and adrenal, eg., testosterone and dehydroeifiiandrosterone. In addition, aetiocholanolone may also be formed from C,, steroids of the 17-hydroxy-11-deoxy type, e.g., 17-hydroxyprogesterone. 11-Oxoaetiocholanolone and 11 P-hydroxyaetiocholanolone arise mainly, by the loss of the side-chain, from the C,, steroids cortisone and cortisol formed in the adrenal, whereas the corresponding 11-oxyandrosterone derivatives, although also of adrenal origin, arise from 11 P-hydroxyandrost-4-ene-3 : 17-dione and adrenosterone (androst- 4-ene-3 : 11 : 20-trione).The absence of dehydroepiandrosterone from the glucuronide fraction of normal urine is noteworthy. Dehydroefiiandrosterone is frequently the major component of the sulphate fraction ; it is believed to originate exclusively in the adrenal. DETERMINATION OF 17-0XO STEROIDS I N BLOOD Very early attempts by Zimmermann6 to determine levels of 17-0x0 steroids in blood gave results that are now known to be too high. Although in principle methods for the analysis of 17-0x0 steroids in blood are similar to those outlined for urine, the main practical difference arises because the amounts to be measured are very much lower and the steroids are accompanied by large amounts of protein and fat.The proteins can be removed by hydrolysis with hot a ~ i d , ~ ~ ~ ? ~ a procedure that cannot be recommended, or they can be precipitated with ethanol.74 Methods involving precipitation are open to the criticism that 17-0x0 steroids precipitated with the protein, irrespective of whether or not they are bound to the protein in plasma, may result in the loss of some material; the digestion of the protein by proteolytic enzymes appears to offer an alternative procedure.In any event the evidence for protein- bound 17-0x0 steroid is largely circumstantial and, even if such compounds exist, the link may be broken by the extraction and hydrolytic procedures normally used. The levels of free steroids that have been established do not exceed their water solubility. The presence of a large amount of lipid in plasma makes the determination of free 17-0x0 steroids difficult. Although free 17-0x0 steroids have been identified and determined in adrenal blood,76*76 such samples are not normally available, and the levels in peripheral blood are too low to be measured by present methods. 17-0x0 steroids present in peripheral blood are predominently in the conjugated form and, because of this, excess of lipid can be eliminated by partition between light petroleum and 70 per cent.ethanol77 or between benzene and water.74 The water-soluble conjugates remain in the lower phase and can be hydrolysed and separated by established methods. The very small amounts of 17-0x0 steroid liberated necessitate the use of small columns and some modification of the gradient-elution apparatus. Migeon and Plager's and later Migeon'O showed that, when plasma was subjected to acid hydrolysis, preferably by continuous extraction with ether at pH 1, small amounts of 17-0x0 steroids were liberated. The presence of dehydroepiandrosterone and androsterone in this material was indicated by paper chromatography, and sufficient of the former compound has been isolated to confirm the identification by infra-red spectros~opy.~~ With larger samples of plasma it is possible to detect trace amounts of aetioch~lanolone.~~ As this treat- ment is known to hydrolyse steroid sulphates but not glucuronides, it is probable that these three steroids circulate in peripheral blood as sulphates.Migeon was unable to find free 17-0x0 steroids in plasma before or after hydrolysis with p-glucuronidase, although Clayton, Bongiovanni and Papadatosso found 12 per cent. of the circulating 17-oXo steroids as glu- curonides. With larger samples of plasma than normally available, the presence of small amounts of androsterone, aetiocholanolone, 11-oxoaetiocholanolone and 11/3-hydroxyandro- sterone in the glucuronide fraction has been demon~trated.7~ In sharp contrast to the large amount of glucuronide conjugates in the urine, in peripheral blood plasma 17-0x0-steroid sulphates preponderate and for all practical purposes they are the only form of circulating 17-0x0 steroid that can be determined at the present time. The levels of 17-0x0 steroids in plasma show wide variation in normal subjects, with no obvious relation to age, e.g., males 20 to 40 years, 12 to 136 pg of dehydroepiandrosterone per I00 ml and 1 to 43 pg of androsterone per 100 ml; females, 20 to 30 years, 6 to 46 pg of dehydroepiandrosterone per 100 ml and 1 to 27 pg of androsterone per 100 ml.According to Migeon,'B dehydroe9iandrosterone is detectable in the plasma at 4 years of age and rises steadily to puberty; it is claimed that the concentration of this compound in the blood of730 KELLIE : THE DETERMINATION OF 17-OXO STEROIDS (17-KETO STEROIDS) [VOl. 82 adult females shows some variation during the menstrual cycle.Comparison of the con- centration of 17-0x0 steroid conjugates in plasma and in urine shows that the glucuronides are excreted at a very much faster rate than the corresponding sulphates.77 DETERMINATION OF 17-OXOGENIC STEROIDS' It has already been shown that C,, steroids produced in the adrenal glands can be degraded to 17-0x0 steroids, which appear in the urine. In health, only a small proportion of the endo- genous and exogenous corticosteroids follow this metabolic pathway and the major part of the material that can be traced also appears in the urine as C,, metabolites with side-chain intact, but in various stages of reduction and oxidation.These changes, which are in addition to those in the steroid nucleus, result in the formation and excretion of very many different types of C,, metabolites.81 Table I lists the recognised forms of side-chains that may arise from a 17cr-hydroxy- corticosteroid precursor such as cortisol (VI; see p. 723); the list is incomplete and should also contain the corresponding set of compounds lacking the 17a-hydroxyl group, which may be formed from 17-deoxy-C2,-steroid precursors. TABL:E I SUMMARY OF THE REACTIONS OF THE CORTTCOSTEROID SIDE-CH-kIX STRUCTURES CHiOH CHiOH CHI CH, $gH O - O H CHOH I &!H f i - O H &Hot4 (4 ( b ) (4 (dl Porter 17-Oxogenic reaction NaBiO; Triphenyl- Formalde- Acetalde- and tetrazon- hydogenii: hydogenic Silber Molybdo- ium com- com- chromo- h'aBH, KaBH, Side-chain type phosphate reagents pounds pounds gens NaBiO, NaBiO, NaBiO, - - + -i- + + + + 3 + - - - (a) Dihydroxyacetone + 4- + ( b ) Glycerol .. . . (d) 17:2O-Glycol . . - - - It is pertinent to emphasise that many of these compounds are thermolabile and unstable in the presence of alkali and that much early work was rendered invalid by the use of hot-acid hydrolysis and the failure to appreciate the need for mild procedures.s2 It seems likely that many compounds of this class are conjugated as glucuronides. The conjugates are hydrolysed by most /3-glucuronidase preparations, but there is, as yet, insufficient evidence that the enzymic hydrolysis goes to completioii; experimental conditions have been found that give maximum yields of free steroid, but these may coincide with the equilibrium position.The evidence of Norymberski,s3 which suggests a substantial proportion of 17-oxogenic steroid conjugated as sulphate, is not conclusive. Many novel attempts have been made to determine the several structures of this group of related steroids. The earliest method, that of Heard and Sobel,g4 which depended on the reduction of molybdophosphoric acid, is quite unspecific and has been abandoned. Substi- tuted derivatives of triphenyltetrazonium ~hloride,'~ prepared for use in solution in sodium hydroxide, depend on the reducing properties of the a-ketol structure -CO-CHOH- and are more specific. These reagents have been of considerable value as spray reagents in paper chromatography for the location of compounds with the cr-ketol side-chain structure (Table I, type (a)).The Porter and Silber reagent,80 dinitrophenylhydrazine in sulphuric acid, reacts only with 17~-hydroxy-20 : 21-ketols and is specific for the side-chain (dihydroxyacetone) found in cortisone and cortisol (Table I, type (0)). Lowenstein, Corcoran and Pages7 in 1946 introduced chromotropic acid into the steroid field as a means of determining small amounts of formaldehyde; this fragment is split off 20:21-glycols (Table I, type (b)) and 20:21-ketoIs (Table I, type ( a ) ) by periodic acid or by sodium bismuthate" and can be used to deterrnine compounds possessing these structures. The method, although useful with relatively pure comp~unds,~~ is of less value when applied + - - - - + - - - - - (c) 17: 20-Ketol .. - 3-November, 19571 AND 17-OXOGENIC STEROIDS (17-KETOGENIC STEROIDS) 731 to biological materials that normally contain comparatively large amounts of non-steroidal “formaldehydogenic” matter. An analogous method described by Cox and MarriangO is based upon the liberation of acetaldehyde by oxidation of steroid glycols with periodic acid; inspection of the possible side-chain structures shows that this fragment is only produced by 17 : 20-glycols (Table I, type ( d ) ) , and the method has high potential specificity for 21-methyl- 17 : 20-glycol side-chains. I t is noteworthy that oxidations of this type, in which a fragment containing two carbon atoms is split off from the C,, molecule, leave behind a C,,-~~-OXO steroid that is very much more specific in origin than the acetaldehyde.The new 17-0x0 steroid, which, unlike the acetaldehyde, must arise from a steroid progenitor, can be determined by the Zimmerman reaction, and in this way many 17cc-hydroxy-C,, compounds (the 17-0x0 steroids”) can be indirectly determined. The oxidising agents used for this type of degradation include periodic acid, lead tetra-acetate and sodium bismuthate, all of which are glycol-splitting agents. The use of chromic oxide in acetic acid for this purpose seems unnecessarily violent, although this reagent will also convert 17-hydroxy-C,, steroids to 17-0x0 steroids.92 The claim of Smith and Tompsettg3 that 3 : 17-dihydroxyandrostanes are oxidised by bismuthate in the presence of chloride to a product that is chromogenic in the Zimmermann reaction has been s~bstantiated,~4 but does not take place when urine is oxidised, because urea is present.No satisfactory method is available for the oxidation of C,, steroids that do not have a 17u-hydroxy group. Of the available oxidants, sodium bismuthate, introduced by Norymber~ki,~~ offers many practical advantages. With this reagent the oxidation is a surfaced-catalysed reaction that is sensitive to light. Few, if any, soluble by-products are produced and, because of the insolubility of the reagent, oxidation can be stopped by centrifugation or filtration. Traces of bismuthate are removed by the addition of bisulphite. In the original application Norym- berski made use of this reaction to determine compounds containing the side-chain structures found in types (a), (b) and ( d ) (Table I), which are directly oxidised to 17-0x0 steroids.For these compounds Norymberski suggested the name 17-oxogenic steroids. When the method is applied to urine, these 17-0x0 steroids are formed in addition to those excreted in the urine, and the sum total is referred to as the “total 17-0x0 steroids.” A separate and independent determination of the preformed urinary 17-0x0 steroids is necessary in order to obtain the value of the 17-oxogenic steroids by difference. A further important feature of the oxidation of 17-oxogenic steroids in urine is that, for the purpose of this determination, the hydrolysis of the conjugated forms is no longer necessary. It has been shown that, in addition to the oxidative removal of the side-chain, bismuthate, when present in excess, brings about the oxidative destruction of the glucuronic acid part of the molecule; in this way 17-oxogenic steroid glucuronides are converted to 17-0x0 steroid formates (or to the free alcohol), both of which are soluble in organic s01vents.s~ It is no longer necessary to hydrolyse in order to extract the steroid residues for determination.This reaction, which appears to be a general one for all clases of steroid glucuronides, has somewhat different consequences when applied to 17-oxogenic steroid sulphates. Compounds of this type, if indeed examples exist, are oxidised to 17-0x0 steroid sulphates, which remain, like their precursors, water soluble. In contrast to the straightforward determination of 17-oxogenic steroid glucuronides, some form of mild hydrolysis is necessary before extraction, if these compounds are to be included. In a subsequent development of the 17-oxogenic steroid method, Norymberski introduced the use of sodium borohydride as a reducing agent to be used in conjunction with sodium bism~thate.~~ Preliminary treatment of compounds of type (c) (Table I) with borohydride converts them to type ( d ) , so that subsequent oxidation with bismuthate converts all 17%- hydroxy compounds to 17-0x0 steroids. During treatment with borohydride, all 17-0x0 steroids present in the untreated urine are reduced to the corresponding secondary alcohols, which give no reaction in the Zimmermann test : all 17-0x0 steroids present after the oxidation are derived from 17x-hydroxycorticosteroids.This method, which has been of considerable value in following and controlling adrenocorticotrophic hormone and cortisol therapy, deter- mines “total 17~-hydroxycorticosteroids,” a term used to distinguish this determination from that of the Porter and Silber chromogens, which are also loosely referred to as 17-hydroxy- corticost eroids. A further modification of the borohydride - bismuthate method can be used to deter- mine compounds of type (c) (Table I) excl~sively.~~ When urine or urine extracts are treated with excess of sodium bismuthate, molecules having side-chains of types (a), (b),732 KELLIE THE DETERMINATION OF 17-OX0 STEROIDS (17-KETO STEROIDS) [VOl. 82 and ( d ) (Table I) are destroyed by oxidation to 17-0x0 steroids, whereas molecules of type (c) remain unchanged.Subsequent reduction with borohydride reduces all 17-0x0 steroids irrespective of origin to the corresponding alcohol!s and simultaneously converts the unchanged molecules of type ( c ) (Table I) to type ( d ) . The latter can then be determined indirectly as 17-0x0 steroids by a second bismuthate oxidation. This method has been used to measure the l’i~-hydroxy-20-oxo steroids excreted in health, in rheumatoid arthritis and in adreno- genital syndrome.Rs Table I summarises the reactions that can be used to discriminate between the various types of 17-hydroxycorticoid side-chain. Of th:e corresponding 17-deoxy compounds, only 20 : 21-glycols and 20 : 21-ketols give formaldehyde on oxidation, and all other specific reactions are negative.The use of these methods permits the total amount of any class of cortico- steroid to be determined, and this may have clinical value; unfortunately, in the process of determination the precursors are destroyed and their ideiitities are lost. No excretion pattern of corticosteroids, analogous to that of the procedure for 17-0x0 steroids, can be obtained by these oxidative methods, as several 17-oxogenic steroids may give rise to the same 17-0x0 steroid. In most urines the 17-oxogenic steroids on oxidation give rise to ll-oxy-17- 0x0 steroids. Recently, methods have been described that give an excretion pattern for corticosteroids based upon the separation of compounds on partition columns of Celite or on silicag9 $loo; separation of these compounds on paper chromatogranis is also possible.lo1 No attempt has yet been made to determine 17-oxogenic steroids in blood.DETERMINATION OF NON-KETONIC STEROID ALCOHOLS The method for the determination of non-ketonic steroid alcohols described below, although not strictly for 17-0x0 steroids or 17-oxogenic steroids, is related to these determina- tions, because it is based upon a reversed Zimmermann reaction. In contrast to the variety of methods available for the measurement of ketonic alcoholic steroids, very few general methods have been suggested for the determination of steroids that have no functional group other than the alicyclic alcohol group. The Liebermann - Burchardt reaction for cholesterol and the sulphuric acid chromogenic reactions used in the determination of pregnane-3cr : 2 0 ~ r - d i o l ~ ~ ~ and pregnane-3a : 16cc : 20a-trioPo3 are colour reactions of some specificity, and cannot be applied to a whole class of compounds.It is, however, possible to esterify alcohols quantitatively and in so doing to incorporate a chromogenic group into the ester, which can be used to determi:ne the alcohol indirectly. Engel, Patterson, Wilson and Schinkello4 showed that it was possible to prepare hemi-3 : 5-dinitrophthalates of primary and secondary alcohols and to determine the ester quantitatively by means of the red colour formed on treatment with methanolic potassium hydroxide. A more practical method of carrying out this type of determination is to esterify with 3:5-dinitrobenzoyl chloride in the presence of pyridine.lo5 By ana.logy with the Zimmermann reaction, if a steroid ester containing a m-dinitrophenyl group is treated with alkali in the presence of excess of a compound containing an activated met hylene group, a characteristic Zimmermann- like colour is obtained.Esterifications carried out with 3 : 5-dinitrobenzoyl chloride are rapid and complete, and the resulting esters when dissolved in acetone (which acts as the compound containing the activated methylene group) give an intense colour when treated with weak ethanolic potassium hydroxide solution. The colour reaches the maximum intensity more rapidly than in a conventional Zimmermann reaction and has a maximum absorption at 560 mp. Application of this method to the non-ketonic fraction of urines obtained from patients with adrenal carcinoma and adrenal tumour has shown that in these clinical conditions there is a substantial change in the amount and pattern of excretion.lo6 The changes in the non- ketonic fraction are complementary to changes already observed in the ketonic fraction by using the conventional methods for 17-0x0 steroids.REFERENCES 1. 2. 3. 4. 5. 6. - . Ibid.. I Q R 6 24s A7 Loewe, S., Voss, H. E., Large, F., and Wahner, A,, Klin. Wochschr., 1928, 1 , 1376. Funk, C., and Harrow, B., Pvoc. SOC. Exp. Biol. Med., 1929, 26, 325. Butenandt, A., 2. attgew Chem., 1932,45, 655. Butenandt, A., and Dannenbaum, H., 2. PhYSiOl, Chem., 1934, 229, 192. Zimmermann, W., Ibid., 1935, 233, 257.November, 19571 AND 17-OXOGENIC STEROIDS (17-KETOGESIC STEROIDS) 733 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. 35. 36. 37. 38 39 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. 65. 66. 67. 68. 69. 70. 71. 72. 73. 74. Brooks, C. J . IV., and Norymberski, J . K., Biochem. J . , 1953, 55, 371. Heard, R. D. H., Blight, E. G., Cam, M . C., Jellinck, P. H., O’Donnell, V. J., Rao, B. G., and Tomkins, G., and Isselbacker, K. J., J . Amer. Chem. Soc., 1954, 76, 3100. Storey, I. D. E., and Dutton, G. J., Biochem. J . , 1955, 59, 279. Kellie, A. E., and Smith, E. R., Ibid., 1957, 66, 490. Ayres, P. J., Garrod, O., Simpson, S. A., and Tait, J . F., Ibid., 1957, 65, 639. Morris, C. J. 0. R., and Williams, D. C., Ibid., 1953, 54, 470.Zimmermann, W.. 2. Vilam.-Horm.-u. Fevmentforsch., 1951, 4, 456; Chenz. Absfr., 1952, 46, 4597. Webb, J . L., Recent Progr. Hormone Res., 1956, 12, 45. -, Vitam. u. Horm., 1944, 5, 1. Dirscherl, W., and Zilliken, F., ivaturwissenschaften, 1943, 31, 349. Pincus, G.. Endocrinology, 1943, 32, 176. Barnett, J.. Henly, .%.,and Morris, C. J . 0. R., Biochem. J . , 1946, 40, 445. Callow. Ii. H.. Callow. R. K.. and Emmens. C. W.. Ibid.. 1938. 32. 1312. I , Broadbent, I.’E,, and’Klyne,’ W., Ibid., 1954, 56, xxx. Holtz, -4. H., Nature, 1954, 174, 316. Frazer, R. W., Forbes, A. P., Albright, A,, Sulkowitch, H., and Reifenstein, E. C., jun., J . Clin. ’ Endocrkn., 1941, 1, 234. Girard, A., and Sandulesco, G., Helv. Chew Acta, 1936, 19, 1095 Allen. 11’. 31.. T. Cli+z. Endocvin..1950. 10. 71. Holtorff, A. F.: and Koch, F. C.; J . BioZ. Chem., 1940, 135, 377. Wilson, H., Fed. Proc., 1950, 9, 246. St. Andre, A. F., MacPhillamy, H. B., Nelson, T. d., Shabica, A. Chem. Soc., 1952, 74, 5506.- Fotherby, K., Colas, A,, A4therden, S. M., and Marrian. G. F.. Kellie. A. E.. and Wade. A. P.. Ibid.. 1957. 66. 196. C., and Scholtz, C. R., J . Amer. Biodkeilz. J., 1956, 64, SOP. Wilson, H., -4rch. Biochem. Biophys.,’ 1954,’52,’ 217. Hamburger, C., Acta Endow., Copenhagen, 1952, 9, 129. Langstroth, G. O., Talbot, K. B., and Fineman, J., J . Biol. Chem., 1939, 130, 585. Lieberman, S., and Dobriner, K., Recent Progr. Hormone Res., 1948, 3, 71. Edwards, R. XV. H., Kellie, A. E., and \Vade, A. P., Mem. Soc. Endocvin., 1953, 2, 53. Kellie, 4 . E., and Wade, A.P., Acta Endow., Copenhagen, 1956, 23, 357. M.R.C. Committee on Clinical Endocrinology, Lancet, 1951, ii, 585. Dorfman, R. I., in Pincus, G., and Thimann, K. V., Editors, “The Hormones,” Academic Press Lieberman, S., Mond, B., and Symles, E., Recent Progr. Hormorze Res., 1954, 9, 113. Talbot, K. B., Ryan, J., and Wolfe, J . K., J . Biol. Chevtz., 1943, 151, 607. Katzman, P. X., Kinsella, R. A., jun., and Doyle, M . L., Science, 1952, 116, 524. Stitch, S. R., and Halkerston, I. D. K., Nature, 1953, 172, 398. Abul Fadl, $1. A. &I., Biochem. J . , 1957, 65, 1 6 ~ . Dodgson, K. S., and Spenser, B., Ibid., 1953, 55, 315. Bitman, J , , and Cohen, S. L., J . Biol. Chem., 1951, 191, 351. Roy, A. B., Biochenz. J., 1956, 62, 41. Dodgson, K. S., Lewis, J .I. M., and Spenser, B., Ibid., 1953, 55, 253. Cohen, S. L.. and Oneson, I. B., J . Biol. Chem., 1953, 204, 245. Gallagher, T. F., Bradlow, L., Fukushima, D. K., Kritchevsky, T. H., Stokem, $I., Eidinoff, M. L., Buehler, H. J., Katzman, P. A., and Doisy, E. A., Proc. Soc. Ex$. Biol. Med., 1951, 78, 3. Callow, N. H., and Callow, R. K., Biochem. J . , 1939, 33, 931. Dobriner, K., Lieberman, S., and Rhoads, C. P., J . Biol. Chem., 1948, 173, 241. Lieberman, S., Dobriner, K., Hill, B. R., Fieser, L. F., and Rhoads, C. P., Ibid., 1948, 172, 263. Rubin, B. L., Rosenkrantz, H., Dorfman, R. I., and Pincus, G., J . Clin. Endocrin. G. Metabolism, Dingemanse, E., Huis in’t Veld, L. G., and de Laat, B. M., J . Clin. Endocrin., 1946, 6, 535. Robinson, A. &I., and Goulden, F., Brit.J . Cancer, 1949, 3, 62. Zygmuntomicz, A. S., Wood, XI., Christo, E., and Talbot, N. B., J . Clin. Eizdocvin., 1951, 11, 578. Pond, &I. H., Lancet, 1951, ii, 906. Lakshmanan, T. K., and Lieberman, S., Arch. Biochevz. Biophys., 1954, 53, 258. Brockmann, H., and Schadder, H., Ber., 1941, 74, 73. Drake, B., A r k . Kemi, 1955, 8, 1. Dobriner, K., Lieberman, S., Rhoads, C. P., Jones, S. R., \Villiams, V. Z., and Barnes, R. B., J . Biol. Chenz., 1948, 172, 297. Bush, I. E., Biochem. J . , 1952, 50, 370. Kellie, -4. E., and Smith, E. R., Nature, 1956, 176, 323. Axelrod, L. R., Fed. Proc., 1952, 11, 182. Heftman, E., Chem. Rev., 1955, 55, 679. Iieher. R.. and Wettstein. A.. Helv. Chim. Acta. 1952. 35. 276. Inc., New York, 1948, Volume I, p. 467. Hellman, L., and Dobriner, K., Recent Pvogv.Hormovze Res., 1954, 9, 411. 1953, 13, 568. , ~~ I I , Savard, K., J . Biol. ?he&., 1953, 202, 457. Kritchevsky, D., and Kirk, M. R., J . Amer. Chem. Soc., 1952, 74, 4484. Shiill, G. M . , Sardinas, J . L., and Nubel, R. C., Arch. Biochem., 1952, 37, 186. Kritchevsky, D., and Kirk, hf. R., Arch. Biochem. Biophys., 1952, 35, 346. Dumazert. C.. and Valensi. G.. Comfit. Rend. SOC. Biol., 1952, 146, 471. Gardner, L. I., J . Clin. Endocjiiz. LGetabolism, 1953, 13, 941. Kellie, A. E., and Smith, E. R., Biochem. I., 1957, 66, 491.734 c- 10. 76. 77. 78. 79. 80. 81. 82. 83. 84. 85. 86. 87. 88. 89. 90. 91. 92. 93. 94. 95. 96. 97. 98. 99. 100. 101. 102. 103. 104. 105. 106. ELWELL AND PEAKE: THE DETERMIXATION OF OXYGEK IX TITANIUM [vol. 82 Bush, I. E., Swale, J., and Patterson, J., Biociiem. J . , 1956, 62, 1 6 ~ . Romanoff, E. B., Hudson, P., and Pincus, G., J . Clin. Endocvin. & Metaholisnz, 1953, 13, 1546. Pfiffner, J. J., Wintersteiner, O., and Vars, H. M., J . BioE. Chew., 1935, 111, 585. Migeon, C. J., and Plager, J. E., Recent Prop Hornzone Res., 1954, 9, 235. Migeon, C. J., in “A Symposium on Adrenal Function in Infants and Children,” State University Clayton, G. W., Bongiovanni, A. M., and Papadatos, C., J . Clin. Endocrin. & 1Metaholisn2, 1955, Dorfman, R. I., and Ungar, I?., “Netabolism of Steroid Hormones,” Burgess Publishing Co., Mason, H. L., Recent Progr. Hormone Res., 1954, 9, 267. Norymberski, J. K., Mem. SOC. Endocrin., 1953, 2, 50. Heard. R. D. H.. and Sobel. H.. I . Biol. Chem.. 1946. 165. 687. of Sew York, Syracuse, New York, 1954, p. 96. 15, 693. Minneapolis, 1953, p. 69. -, - - Chen, C., and Tewell, H. E.: Fed.*Proc., 1951, 10, 377. Porter, C. C., and Silber, R. H., 1. Biol. Chem., 1950, 185, 201. Lowenstein, B. E., Corcoran, A. C., and Page, I. H., Endocrinology, 1946, 39, 82. Edwards, R. W. H., and Kellie, A. E.. Biochem. T.. 1954. 56. 207. Simpson; S. A,, Tait, J. F., Wettstein, A., P;eh&,’R., Euw,’ J. yon, and Reichstein, T., Experi- nzentia. 1953. 9. 333. ~I I ~~~ Cox, R. I., and Marrian, G. F., Biochem. J . , 1952, 52, 339. Brooks, C. J. W., and Norymberski, J . K., Chew G. Ind., 1952, 804. Wilson, H., and Fairbanks, R., Arch. Biochenz. Biophys., 1955, 54, 440. Smith, D. C., and Tompsett, S. L., Analyst, 1955, 80, 397. Norymberski, J . K., and Stubbs, R. D., Biochem. J., 1956, 64, 176. Norymberski, J . K., Nature, 1952, 170, 1074. Appleby, J. I., Gibson, G., Norymberski, J . K., and Stubbs, I<. D., Biochem. J., 1955, 60, 453. Appleby, J . I., and Norymberski, J . K., Ibid., 1955, 60, 460. __- , Ann. Rheum. Dis., 1955, 14, 172. Cook’, E. R., Dell, B., and Wareham, D. J., An,zZyst, 1955, 80, 215. Johnson, D. F., Heftmann, E., and Hayden, A L., Acta Endocr., C o ~ e n l ~ o g c n , 1956, 23, 341. Bush, I. E., Biochem. J., 1955, 55, xiv. Sommerville, I. F., Gough, N., and Marrian, G. F., J . Endocrinol., 1948, 5, 247. Bongiovanni, A. M., and Clayton, G. W., Bull. Johns Hopkins Hosp., 1954, 94, 180. Engel, L. L., Patterson, H. R., Wilson, H., and Schinkel, .M., J . Biol. C h e w . , 1950, 183, 47. Kellie, A. E., Smith, E. R., and Wade, A. P., Biochem. J . , 1953, 53, 578. Kellie, A. E., and Wade, A. P., Ibid., 1953, 53, 582. COURTAULD INSTITUTE OF BIOCHEMISTRY ~IIDDLESEX HOSPITAL MEDICAL SCHOOL LONDON, W. 1 Jwae 61h, 1957

ISSN:0003-2654    

DOI:10.1039/AN9578200722

出版商:RSC    

年代:1957

数据来源: RSC

摘要:

734 ELWELL AND PEAKE: THE DETERMIXATION OF OXYGEK IX TITANIUM [VOl. 82 The Determination of Oxygen in Titanium and Titanium Alloys, Based on the Principle of Chlorination BY W. T. ELWELL AND D. M. PEAKE A chemical procedure for the determination of oxygen in titanium and titanium-base alloys is described. The solid sample is admixed with graphite and chlorinated in an atmosphere of argon and the products of the reaction are subsequently isolated, excess of chlorine being removed by reaction with antimony. The liberated carbon monoxide has been shown to bear a stoicheiometric relationship to the amount of oxygen present in the sample, and in the final stages of the procedure the purified carbon monoxide is oxidised to carbon dioxide and weighed. The results were reproducible when the procedure was applied to the analysis of titanium - manganese alloys : this potential application is an outstanding advantage over the vacuum fusion procedure.With a single apparatus, about twelve determinations can be completed in a normal working week of 5 days. This is about the same rate as that a t which oxygen can be determined by the macro vacuum fusion procedure, but, whereas a complete vacuum fusion unit costs about A2500 and requires fairly constant attention, apparatus for the recommended procedure costs about i120 and i t is estimated that three units could be operated simultaneously. THE most reliable procedure for the determination of oxygen in titanium and its alloys is that based on the principle of vacuum fusion.1.'~~3~47j~617 In this procedure, the sample isNovember, 19571 735 heated at 1600” C in a vacuum in the presence of an excess of carbon, and oxygen is quanti- tatively liberated as carbon monoxide.The necessary apparatus is expensive and its opera- tion requires specialised skill. The vacuum fusion procedure can be applied to the determina- tion of oxygen in titanium and most of its common alloys, with the exception of those con- taining manganese, which readily volatilises, condenses on the cooler parts of the apparatus and combines with some of the oxygen. The need for a relatively simple chemical method for the determination of oxygen in titanium-base samples, particularly those containing manganese, has been apparent for a long time. A chemical procedure proposed in 1951 by Corbetts is similar to the chlorination method of Colbeck, Craven and Murrayg for the determination of non-metallic inclusions in steels.Corbett’s procedure is based on chlorination of the heated sample, whereby titanium is converted into the volatile tetrachloride, and oxygen is indirectly determined in the residue of titania by determining its titanium content. This procedure is very restricted in its applications and involves the assumption that any carbon in the sample is converted into carbon monoxide ; a suitable correction for loss of oxygen in this way must be applied, Further, it cannot be used in the presence of metals, such as aluminium, that give residues of uncertain composition, A similar procedure based on bromination has been proposed by Milner, Hegedus and Dyorsky,lo but criticisms are equally applicable to both halogenation procedures, More recently, a procedure has been proposed in which the sample is heated in the presence of excess of carbon in a stream of bromine vapour,ll whereby titanium and the common alloying elements are converted to their corresponding volatile bromides, and oxygen is evolved as carbon monoxide.After a series of purification stages, carbon monoxide is oxidised to carbon dioxide and weighed, and the weight is used in calculating the oxygen content of the sample. This procedure has met with a mixed reception in the USA., and “further development is necessary before it can be completely relied upon.’’12 In addition, our own experience with the method has not been entirely satisfactory.13 Bromination of titanium-base materials is slow and, in order to make the procedure reasonably rapid, it is necessary to limit the weight of sample to about 1 g, which in turn necessitates a rigid control on the operating conditions and a very careful evaluation of the blank.Further, the sample for bromination must be in the form of thin slices, and the preparation of such samples increases the risk of atmospheric oxidation. Other chemical procedures have been either tried or suggested. These include the use of fluorine at elevated temperatures, when oxygen in the sample is liberated as gaseous oxygen14; a similar procedure involves the use of hydrofluoric acid, which converts the oxygen into water.15 Bromine trifluoride has also been suggested as a reagent for this determination,le but with all fluorine-containing compounds there is extreme difficulty in obtaining them in an anhydrous condition and specially constructed apparatus, e g ., made of nickel, must be used. In all these chemical procedures the apparatus must be rendered free from oxygen by the passage of an inert gas, and the need to provide a system free from moisture is of paramount importance. Unfortunately, the relatively simple expedient of determining oxygen by reduction in either an atmosphere of hydrogen or in the presence of carbon, as has been applied to steel and copper-base materials, cannot be applied in the determination of oxygen in titanium and titanium al10ys.l~ Attempts have been made to determine oxygen in titanium alloys by other methods, including selective solution,ls radioactivation,lj by making use of the Ti0 spectral bands emitted when samples are examined spectrographically in an evacuated s y ~ t e m ~ ~ 9 l ~ and by means of the mass spectrometer.15 s20+’1 All these procedures, however, have limitations, and none is entirely satisfactory.ASD TITANIUM ALLOYS, BASED ON THE PRINCIPLE OF CHLORINATION DEVELOPNENT OF THE METHOD Consideration of the various factors underlying the possible development of a chemical procedure suggested that a method based on chlorination was worthy of further consideration. Chlorine reacts more readily with titanium and its alloys than bromine does, and consequently a larger sample can be taken with correspondingly less dependence both on sampling and on a careful evaluation of the blank.For the determination of oxygen, chlorine has other advan- tages over bromine, e g , , it does not condense in narrow-bore tubing or at inaccessible bends736 ELWELL AND PEAKE: THE DETERMINATIOS OF OXYGEN IN TITANIUhI ['C'Ol. 82 and, more important, it requires no carrier gas during halogenation and so permits a higher concentration of halogen to be present during the reaction. Ti + 2C1,-+TiCl4. At the same temperature no reaction occurs between titanium dioxide and chlorine, but, when the oxide is in intimate contact with excess of carbon, the reaction under suitable conditions is thought to proceed according to the followilig equation- TiO, + 2C + 2C1,-+TiC14 + 2CO. When chlorine is passed over titanium a t 800" C, the following reaction takes place- Under the conditions used in the development of this procedure, however, this reaction does not occur. Hence, neither titanium tetrachloride nor carbon monoxide could be detected after prolonged chlorination of a heated mixture of graphite and titania. It was also observed that, when reference samples prepared by fusing titanium refined by the iodide process and weighed amounts of titania in an argon-arc iurnace were similarly examined, there was a tendency for the recoveries to be low if the oxygen content exceeded about 0.2 per cent., and a corresponding amount of titania was always found in the residue.It is known that up to about 15 per cent. of oxygen, by weight, can exist in solution in titanium, hence the above-mentioned reaction can be expressed by the following equation- Ti T- 0 -+ 2C1, + C-TiCl, + CO.The contemplated procedure, therefore, was to chlorinate the heated sample in the presence of excess of carbon, use being made of argon as a purge-gas, and to convert the oxygen quantitatively to carbon monoxide, which would subsequently be purified and then oxidised to carbon dioxide and weighed. The final oxidation and weighing would follow conventional lines by using a converter of heated copper oxide and then a weighing tube filled with Carbosorb soda asbestos. A valid criticism of the use of chlorine under these conditions lies in its ability to react with carbon monoxide to form carbonyl chloride. A tentative procedure was therefore evolved on the assumption that, after chlorination, the ensuing gaseous mixture would contain titanium tetrachloride, the chlorides of' the alloying constituents, excess of chlorine, carbon monoxide and some carbonyl chloride.Titanium tetrachloride has b.p. 136" C and m.p. - 30" C, and therefore condenses as a liquid a t room temperature, which affords a ready and simple means of removing almost the whole of the titanium and other volatile metallic chlorides that have very similar properties. A study of the equilibrium reaction betwleen carbon monoxide, chlorine and carboiiyl chloride indicated that a high temperature favours the dissociation of carbonyl chloride, but recombination of the gases a t a lower temperature had to be considered. It was essential, therefore, that chlorine from the carbonyl chloride should be removed, not by the conventional method of freezing, but from a stream of hot gas at a temperature as near as possible to that a t which the carbonyl chloride was decomposed.Decomposition of carbonyl chloride was achieved by passing the gas leaving the combustion tube through a silica tube maintained a t 800" c. Finely divided tin that was nominally free from carbon was tried at one stage, and this met with some success. The reaction between tin and chlorine proceeds with incandescence, and stannic chloride is the main product of the reaction. Stannic chloride has b.p. 114" C and m.p. - 33" C, and, being a liquid at room temperature, is easily removed from the reaction zone. Unfortunately, the reaction proceeds one stage further and a large amount of stannous chloride is also formed.Stannous chloride is a solid with m.p. 246" C and its presence rapidly chokes the tube. The powdered metal was used in order to obtain a large surface area and, although this appeared to be satisfactory chemically, the column soon became choked ; the sticks of antimony subsequently used were entirely satisfactory. Excess of chlorine reacts readily with the antimony and, contrary to expectation, there was no visible sign of chlorine passing beyond the first few pieces of the very loosely packed metal. Antimony was placed in a vertical column so that liquid antimony pentachloride could be collected in a catch-pot a t the base of the column. Passage of gas vertically through the Many attempts were made to remove excess of chlorine from the gas stream.In the next series of experiments, tin was replaced by antimony.- 2 3 a P Y OI l i g . 1. \-iem of apparatus for determining oxygen i n titaniunl and titanium alloysNovember, 19571 737 column was not entirely satisfactory, as a small amount of mist, presumably antimony tri- chloride, was carried forward by convection currents, and the copper oxide in the converter soon became poisoned. When the gas stream was allowed to enter a t the top of the column, solid antimony trichloride was formed and seriously impeded the flow of gas. This idea was soon abandoned in favour of the original idea of passing the gas leaving the combustion tube up through the column, but the problem of the antimony trichloride mist had still to be solved, and several unsuccessful attempts were made to solve it.Plugs of glass-wool, filter- paper and a water trap were separately tried, but none was entirely satisfactory. A scrubber containing Carbosorb soda asbestos admixed with glass beads was tried, and this was reason- ably successful over periods of a t least 5 days under conditions of daily working, but low erratic results were associated with the incorporation of this scrubber and, finally, a scrubber con- taining an intimate mixture of manganese dioxide and glass balls was found to be entirely satisfactory. In each determination it was necessary to place the sample in the cold combustion tube, sweep out the air with dry argon, raise the temperature of the combustion tube and then cool it before the next sample could be examined. This was time-consuming, and, in order to make the procedure more practicable, the apparatus was redesigned so that heating of the combustion tube was uninterrupted and an almost continuous supply of samples could be examined.The way in which this was achieved will be clear from the description of the apparatus. APPARATUS- AND TITANICM ALLOYS, BASED ON THE PRINCIPLE OF CHLOKIKATION METHOD The apparatus is shown in Figs. 1 and 2. The Arnold bubblers, A,, A, and A,, are each constructed as a single unit and contain sulphuric acid, sp.gr. 1.84; A, and A, have B24 joints, and B19 cones are attached to their outlet tubes. The purifier for argon, B, is a Pyrex-glass tube, 1 inch in diameter and 18 inches long. The tube is heated electrically to 250" C along 9 inches of its length and is lagged with asbestos; a sheet of asbestos paper covers the tube to prevent any local penetration of the glass by the heating wire.The pressure release valve, C, has a B24 joint and a pressure head of about 7 inches of mercury. The scrubber for carbon dioxide, D, has a B19 joint a t each end; the inlet half of the tube is filled with Carbosorb soda asbestos and the other half with anhydrone. E, and E, are Quickfit and Quartz S61/2 glass taps. The drying tube, F, has a B19 joint at each end and is filled with phosphorus pentoxide. G is a removable B24 cone and H is a Quickfit and Quartz SC8/6 Pyrex-glass tap having a bore wide enough to allow the silica sample boat and its contents to pass freely into the combustion tube, I.(Tap H was lubricated with Edwards' silicone grease, but the centre part of the tap was not greased so that none could be picked up by the sample boat.) The combustion tube, I, is a silica tube, 1.25 inches in diameter and 24 inches long, with a B24 joint a t each end. This tube is heated electrically to maintain a working temperature of 825" C and the position of the sample boat during a determination is controlled by a silica rod that is locked in position by a polythene collar and tongue. The optimum position of the rod allows the sample boat to be housed in the hottest part of the tube. The chloride trap, J, has B24 joints and two outlets; one outlet allows the gaseous effluent to pass forward and the other allows the liquid products of the reaction to be trans- ferred, by means of screw clip K, to the chloride reservoir, L, which has a B24 joint, and from which the liquid products are subsequently removed from the system.The silica tube, 31, has a diameter of 0.4 inch and is heated to 800" C over 1.5 inches of its length as close to the trap for antimony pentachloride, N, as is practicable. The trap for antimony pentachloride has a B24 joint and is emptied by means of a tap having a core made of Teflon. The cone of the trap and the glass column, 0, are constructed as a single unit with trap N. Column 0 is made of Pyrex-glass tubing, 1 inch in diameter and 29 inches long. It has a filter-cone made of china at its base and is filled with sticks of antimony of diameter 0.25 inch and about 0.5 inch long. The column has a side arm about 1 inch from the top and a solid stopper in a B24 socket, P, at its upper end.The glass tube, Q, has a B19 cone a t each end and is connected to the side arm of column 0 by a MS5/12 spherical joint. The tube has a diameter of 0.8 inch and is 12 inches long; it is filled with a mixture of manganese dioxide and glass beads with a plug of glass-wool at each end. Tap H is joined to the combustion tube by a B24 cone.LI v ........ .... c ........ I IUI // A,. A, and A, - Arnold bubblers B - Purifier for argon contain- ing copper turnings and copper oxide i n wire form C -- Pressure release valve D - Scrubber for carbon di- oxide containing Carbo- sorb soda asbestos and an h ydrone El and E2 = Taps (Quickfit and Quartz S61/2) F -- Drying tube containing phosphorus pentoxide G - Removable B24 cone H - Pyrex-glass tap (Quickfit and Quartz SC 816) I Combustion tube J Trap for chloride K - Screw clip L - Reservoir for chloride M Silica tube N Trap for antimony penta- 0 : Column containing sticks of P ~ Solid stopper i n 624 socket Q = Scrubber containing man- R : Tube containing copper S ~ Drying tube containing chloride antimony ganese dioxide oxide anhydrone J'ig.2. Apparatns for determining oxyarn in titanium T = Weighing tube for carbon di- oxide containing Carbosorb soda asbestos and anhydrone U - Scrubber for carbon dioxide containing Carbosorb soda asbestos and anhydrone V ~ Flowmeter and titanium alloysNovember, 19571 AND TITANIUM ALLOYS, BASED ON THE PRINCIPLE OF CHLORINATION 739 in diameter and 6 inches long, and is filled with anhydrone with a plug of glass-wool at each end.The weighing tube for carbon dioxide, T, is a U-tube filled with Carbosorb soda asbestos and anhydrone with a small plug of glass-wool between the reagents and a similar plug a t each end. The scrubber for carbon dioxide, U, has a B14 joint at each end; the inlet half of the tube is filled with Carbosorb soda asbestos and the other half with anhydrone. V is a Aowmeter. With four exceptions all conical and spherical joints are sealed with black Apiezon vacuum wax. The wax must not penetrate more than half-way down the joints on A, and A, or more than t inch down the joint between H and I. The unwaxed joints are the joint at C, cone G and the joint between I and J ; the joints must, however, be lubricated with a silicone grease (Edwards' silicone grease was used).When standard joints are not used and when flexibility is not required, connections are made glass-to-glass by using poly(viny1 chloride) tubing. Lengths of poly(viny1 chloride) tubing are also used to transfer argon and chlorine from the cylinders to the apparatus. The furnaces heating B, I, M and R are controlled by Sunvic switches. The apparatus as shown in Fig. 1 is mounted on a five-ply board, 54 inches x 29 inches Y inch, covered with 16 s.w.g. aluminium sheet to prevent unsightly charring when the waxed joints are heated. Terry spring-clips are used extensively for securing components to the board. (The furnace heating I is held in position by wooden supports on the back of the board.) An elongated hole is cut for the larger tap, H, which enables it to be moved from its normal position during assembly and dismantling.The board is pivoted on an axle, which in turn is supported on a wooden frame. The tilt of the board should be sufficient in both directions to enable the sample boat to slide freely into and out of the furnace. In the rest position, there should be a slight clockwise tilt; the apparatus is held in this position by a simple swivel catch mounted on the front support of the wooden frame. The apparatus must be operated under conditions of good ventilation, preferably in a fume-cupboard. REAGENTS-- Chlorine-As it is extremely difficult to obtain chlorine free from oxygen and as it is desirable that chlorine with a very low total oxygen content should be used, this must be specified when the material is ordered from the supplier.Freshly supplied cylinders of liquid chlorine tend to have a high oxygen content, but after about one-third of the liquid has been used the oxygen content is considerably reduced. Apertures are cut in the board to take the furnaces heating B and I. The maximum angle of tilt required is about 30". Argon-Free from oxygen. Graphite-Rods obtained from Johnson, illatthey & Co., 10 mm in diameter The powdered material is prepared by grinding the rods in a In order to minimise pick-up of oxygen, and 30 cm long, were used. pencil sharpener retained exclusively for this work. the graphite should be powdered immediately before use. Carbosorb soda asbestos, 10 to 14-mesh R.S.S.Anhydrone, 10 to 14-mesh B.S.S. Copper oxide, wire form. Manganese dioxide-Coarse grade, 20 to 40-mesh B.S.S. Antimony-The metal in the form of sticks. Sulphuric acid, sp.gr. 1.84. Nitric acid, sp.gr. 1.42. Hydrochloric acid, sp.gr. 1.18. Hydrofluoric acid, 40 per cent. w/w. Ethanol. Ether. PROCEDURE FOR PREPARING THE SAMPLE- Cut the sample into pieces of about 1 g and clean them by pickling for about 3 minutes in a mixture of 25 ml of nitric acid and 25 ml of hydrochloric acid, containing about 5 ml of hydrofluoric acid. Wash the pieces free from acid with water, rinse them with 10ml of ethanol and then 10 ml of ether, and then dry them in a stream of compressed air for about 1 minute; this is usually sufficient to remove the ether completely.740 ELWELL AKD PEAKE: THE DETERMINATION OF OXYGEN IN TITAKIUM [VOl.82 PROCEDURE FOR DETERMINING THE BLANK- Determine the blank on a weighed sample of titanium (about 5 g), refined by the iodide process and of known oxygen content. Transfer the prepared sample to a previously ignited silica sample boat and fill with freshly ground graphite; level with a spatula and ensure that the sample does not protrude above the graphite. When very pure chlorine is used, it is sufficiently accurate to determine the blank on the graphite only. Place the sample boat in an air-oven at 105” C for 30 minutes and then transfer it, without cooling, to the apparatus. Continue as outlined in the procedure for deterrnining oxygen. The value of the blank B in milligrams is given by- 2< x Y B = W - - - - 01.03636 or, when graphite alone is used, by- B = I.1V where W = increase in weight of weighing tube T in milligrams, X = weight of titanium taken in grams, and Y = percentage of oxygen in the titanium.PROCEDURE FOR DETERMINING OXYGEN- Assemble the apparatus and heat the furnaces to the appropriate temperatures. Close taps E, and E,, open tap H and sweep out the apparatus with chlorine for 10 minutes. After assembly or partial dismantling and re-assembly of the apparatus, ensure stabilisation of the system by passing argon alone at the recommended rate of flow over heated graphite. The increase in weight of weighing tube T should not exceed 1.5 mg per hour. The preliminary gassing with chlorine is advised if at any time the chlorine supply has been stopped for more than 5 hours.If the column of antimony, 0, has been completely renewed, heat the base of the column gently to initiate the reaction. Stop the flow of chlorine, open tap El, and then sweep out with argon for 30 minutes. Close tap 13, open tap E, and remove cone G and care- fully remove the silicone grease from the socket. Place the prepared sample, about 5 g, in the glass tube between cone G and tap E,, lightly grease cone G and replace; ensure that the cone is held firmly in position, e g . , by means of a rubber band. Purge the chamber for 10 minutes with argon, close tap E, and stop the flow of argon. Isolate the argon train by closing tap El, then open tap H. Weigh the weighing tube T, insert in the train, then open the weighing tube taps.Tilt the apparatus sufficiently to allow the sample to slide into position in the com- bustion tube, I. Chlorinate the sample for 2 hours at a rate of flow of about 120ml per minute, At this rate of flow the sample should be completely chlorinated and the trap for antimony pentachloride full. It is important to ensure that, both in determining the blank and determining oxygen in the sample, the volume of antimony pentachloride formed is about equal. Stop the flow of chlorine, open tap El and sweep out the apparatus for 30 minutes with argon at a rate of flow of about 1001 ml per minute. During the initial stage of this sweeping-out period, antimony trichloride mist is visible in column 0. Stop the flow of argon at this stage for about 2 minutes; this enables the mist to condense and prevents rapid saturation of tube Q.At the end of this period, re-weigh the weighing tube, T, and record the increase in weight in milligrams. Till: the apparatus to allow the sample boat to slide into the glass tube between G and E, and, for a further determination, repeat the pro- cedure from “Close tap H, open tap E, and remove cone G. . . .’’ The oxygen content of the sample is given by- 0,03636 x (w - B, per cent. X where W = increase in weight of weighing tube T in milligrams, B = value of the blank in milligrams, and X = weight of sample in grams. oxygen content. Note that it is an advantage to examine periodically a suitable reference sample of known R E s u L r s Numerous results obtained during the course of the development work are not reported here.In most of the experiments, samples were used whose oxygen content had been deter- mined by the vacuum fusion procedure and invariably, when discrepant results were obtained,November, 19571 741 some minor modification to the chemical procedure was made before the determination was repeated. Finally, a bar of commercially pure double-melted titanium of 4 inch diameter was obtained and its precise oxygen content was determined by a vacuum fusion procedure.' An independent value for oxygen was supplied by Mr. E. Booth of the United Kingdom Atomic Energy Authority, who used the semi-micro vacuum fusion method of Booth, Bryant and Parker.*Z In this way it was shown that the oxygen content of the bar was uniform throughout at 0.102 Replicate determinations of oxygen by the proposed procedure on this material gave the following results- 0.10, 0.11, 0.12, 0.12, 0.10, 0.11, 0.09, 0.09, 0.10, 0.10, 0.09 and 0.09 per cent.The average value by the vacuum fusion procedure was 0.10 per cent., and it can be seen that the results by the proposed procedure are in good agreement with the established value. In view of the inability to determine oxygen in manganese-containing alloys by the vacuum fusion technique, the proposed procedure was applied to two rods of double-melted titanium containing 4 per cent. of aluminium and 4 per cent. of manganese. The results of replicate determinations of oxygen by the proposed procedure were as follows- AND TITANIUM ALLOYS, BASED ON THE PRIXCIPLE OF CHLORINATION 0.005 per cent.Rod No. 1 Rod No. 2 . . . . 0.13, 0.13, 0.12 and 0.12 per cent. 0.18, 0.17, 0.18, 0.18 and 0.16 per cent. The reproducibility of the results of these tests is satisfactory, and there is no reason to doubt, the established oxygen contents, but unfortunately it is not possible to compare them with any results obtained by an alternative reliable procedure. Two further manganese-bearing materials were examined, but oxygen results ranged from 0.10 to 0.24 per cent. over a total of about twelve determinations. These samples were very difficult to prepare; they were so hard that as many as ten hack-saw blades had to be TABLE I DETERMINATIOK OF OXYGEN IN TITANIUM-BASE MATERIALS Oxygen found by- -_I_ r-- A > proposed Type of material method, % Titanium refined by the iodide Commercially pure titanium 0.34 0.10 0.37 0.09 0.09 0.08 0.14 process .. . . . . 0.018 Titanium- 124 per cent. of tin - 2& per cent. of alu- minium alloy . . . . Titanium - 24 per cent. of tin - 5 per cent .of aluminium Titanium refined by the iodide process + 0.2 per cent. of oxygen . . . . .. alloy * . . . . . .~ 0.11 0.55 0.11 0.09 0.09 0.11 0.12 0.13 0.17 0.20 vacuum fusion method, /O 0.012 0.34 0.10 0.38 0.11 0.08 0.10 0.13 0.13 0.53 0.13 0.08 0.10 0.11 01 0.12 0.12 0.15 0.20* vacuum proposed fusion method, method, 0 , a % 0.09 0.10 0.13 0.13 0.14 0.14 0.12 0.13 0.11 0.13 0.12 0.13 0.15 0.14 0.49 0.53 0.10 0.10 0.07 0.08 0.08 0.10 0.12 0.11 0.09 0.11 0.16 0.17 0.23 0.20* * Theoretical value. vacuum proposed fusion method, method, % 0 I /O 0.20 0.20 0.10 0.10 0.11 0.13 0.20 0.20 0.07 0.08 0.20 0.20 0.21 0.20 0.12 0.13 0.07 0.08 0.12 0.13 0.11 0.11 0.10 0.10 0.1 4 0.13 used in the preparation of a 5-g sample.Although every care was taken to minimise the heat evolved during the sawing operation, it is conceivable that the high results are due to atmo- spheric oxidation. This practical difficulty is associated with the sampling of all titanium- base materials, particularly alloys, and the preliminary pickling treatment incorporated in the recommended method aims at minimising errors introduced in this way.742 ELWELL AND PEAKE [Vol. 82 Table I shows the results of determinations of oxygen in various samples by the reconi- mended method and by the vacuum fusion procedure, together with two theoretical values.It can be seen that the results are all in good agreement. APPLICATION OF THE RECOMMENDED PROCEDURE TO SAMPLES OF ZIRCONIUM Samples of zirconium were examined by the chemical procedure, but there was a practical difficulty to overcome, as zirconium tetrachloride is a solid and chokes the tube leading to the chloride trap. There is little doubt about the successful application of this procedure to the determination of oxygen in zirconium and its alloys, perhaps after some minor modification, but a serious attempt to extend the recommended procedure has not been made, CONCLUSIONS The method is no more rapid than the existing macro vacuum fusion procedure, but the cost of the equipment represents a considerable fmancial saving. The installation of a com- plete vacuum fusion unit costs about A2500, whereas the equipment necessary for the proposed chemical procedure costs only about E120.The time taken to complete a single determination is about 23 hours and, allowing for stabilisation of the train and periodic blank evaluations, it is estimated that, with one apparatus, about twelve determinations can be completed in a normal working week of 5 days. Because of the actual working time needed to manipulate a single apparatus, one person should be able to operate three trains simultaneously with a corresponding increase in the number of samples examined. The outstanding advantage of the chemical procedure is its potential application to the determination of oxygen in alloys containing manganese. 1. 2. 3.4. 5. 6. c 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. REFERENCES Sloman, H. A,, J . Inst. Mefals, 1945, 71, 391. Sloman, H. A,, and Harvey, C. A., Ibid., 1962, 80, 391. Derge, G., J . Met., 1949, 1, 31. Walter, D. I., Anal. Chem., 1950, 22, 297. Stanley, J. K., Von Hoene, J., and Weiner, G., Ibid., 1951, 23, 377. McDonald, R. S., Fagel, J. E., and Balis, E. W., Ibid., 1955, 27, 1632. “The Analysis of Titanium and its Alloys,” Imperial Chemical Industries Limited, London, 1956, Corbett, J. A , , Analyst, 1951, 76, 652. Colbeck, E. W., Craven, S. W., and Murray, IT. J., “The Seventh Report of the Heterogeneity of Steel Ingots Committee,” Special Report h’o. 16, The Iron and Steel Institute, London, 193i, p. 124. Nilner, T., Hegedus, A,, and Dyorsky, M., Kohdszati Lapok, 1955, 12, 554. Codell, M., and Norwitz, G., Anal. Chem., 1955, 27, 1083. “Analytical Chemistry of Titanium,” Titanium Engineering Bulletin No. 3, Titanium Metals Cor- Roberts, R. O., personal communication. Kellogg, H. H., J . Met., 1951, 3, 137. “Proceedings of Symposium of Analysis and hIetallography of Titanium,” Armour Research Hoekstra, H. R., and Katz, J . J., Anal. Chqm., 1953, 25, 1608. “Gmelin Handbuch der anorganischen Chemie, System No. 41, Verlag Chemie, G.m.b.H., Wein- Ehrlich, I?., 2. Elektrochem., 1939, 45, 362. Tudowitch, K. L., “Quantitative SpectrographLC Analysis for Oxygen in Titanium Metal,” Armour Kirshenbaum, A. D., Mossman, R. A., and Grosse, 4 . V., Trans. Amer. Soc. Met., 1964, 46, 525. Kirshenbaum, 4 . D., and Grosse, A. V., Anal‘. Chim. Acta, 1957, 16, 225. Booth, E., Bryant, F. J., and Parker, A,, Analyst, 1957, 82, 50. p. 54. . poration of America. Foundation, Illinois Institute of Technology, Chicago, Illinois, 1951. heim, 1953, pp. 117 and 203. Research Foundation, Illinois Institute of ’Technology, Chicago, Illinois, NP-4486, 1952. RESEARCH DEPARTMENT IMPERIAL CHEMICAL INDUSTRIES LIMITED NETALS DIVISION KYNOCH WORKS, WITTON, BIRMINGHAM June 5th, 1957

ISSN:0003-2654    

DOI:10.1039/AN9578200734

出版商:RSC    

年代:1957

数据来源: RSC

摘要:

November, 19571 FLETCHER AXD WARDLE 743 The Determination of Tellurium in Lead and Lead Alloys BY N. W. FLETCHER AND R. WARDLE A new absorptiometric method for the determination of tellurium in lead and lead alloys is described, in which use is made of the absorption of tellurium bromide in hydrobromic acid. The method is more sensitive than those based on the use of iodides. The alloy is dissolved in a mixture of bromine and hydrobromic acid and tellurium is precipitated with stannous bromide. After filtration, the tellurium is redissolved in a bromine - hydrobromic acid mixture, most of the bromine is removed by boiling the solution and the remaining traces by addition of ascorbic acid. A standard amount of hydrobromic acid is added and the absorption is measured at 442 mp.There is no interference from other alloying elements, as only selenium is precipitated under these conditions and this is volatilised during the removal of free bromine. The range of the method can be altered by selecting a suitable weight of sample, 2 g of sample being required for the range 0.005 to 0.02 per cent. NIELSCH and Boltz have published papers on methods for determining antimony,l bismuth,2 ~ o p p e r , ~ iron4 and tin6 by measuring the absorptions of their bromides in concentrated hydro- bromic acid. In an attempt to apply their work to the analysis of lead alloys, the absorption spectra, in hydrobromic acid, of the bromides of aluminium, antimony, arsenic, bismuth, cadmium, calcium, copper, iron, lead, magnesium, nickel, selenium, tellurium, tin and zinc were measured.Tellurium bromide, like the bromides of antimony, bismuth, copper, iron and tin, was found to have a spectrum of analytical interest. The absorptions of the bromides of the remaining nine elements mentioned above were negligible a t the concentrations normally found in lead alloys. There was a considerable overlap of the absorption spectra of the bromides of antimony, bismuth, copper, iron, tellurium and tin, and, as the absorption of the other elements could not be suppressed, a separation was required before measurements could be made. Tellurium could easily be precipitated in the elementary form and a method based on this precipitation, followed by re-dissolution of the tellurium in hydrobromic acid, has been developed. EXPERIMEKTAL REMOVAL OF FREE BROAIISE FROM SOLUTIOX- The concentrated hydrobromic acid used, 46 to 48 per cent., contained sufficient free bromine to give a high blank reading, which led to errors when only small amounts of tellurium were present.So reference to a suitable method for removing bromine was encountered in the literature, but, during other work in this laboratory, hydrazine hydrate was found to be effective and was used in some of the preliminary experiments. The reaction, however, was slow, and it was necessary to allow the solutions to stand for a t least 1 hour before measurement. At a later stage, ascorbic acid was found to be much superior to hydrazine hydrate for removing bromine, the reaction being instantaneous. The earlier work was repeated with ascorbic acid in place of hydrazine hydrate.MEASUREMENT OF ABSORPTION SPECTRUM- The absorption spectrum of a solution of 0.28 mg of tellurium in a mixture of 40 ml of AnalaR hydrobromic acid, 46 to 48 per cent., 5 ml of a 10 per cent. solution of ascorbic acid and 5 ml of distilled water was measured by using a Unicam SP600 spectrophotometer. A blank solution of the reagents was prepared without tellurium and its absorption spectrum was measured in the same way. The absorption spectra are shown in Fig. 1, and all sub- sequent measurements were made at 442 mp.744 FLETCHER AND WARDLE : ?HE DETERMIBATION OF [Vol. 82 PREPARATION OF STANDARD TELLURIUM SOLUTION- A stock solution was prepared by dissolving 0.500 g of pure tellurium in hydrobromic acid and diluting to 1 litre with more hydrobromic acid.For use, 10 ml of the stock solu- tion were diluted to 100 ml with hydrobromic acid and aliquots were taken as required. EFFECT OF HYDROBROMIC ACID CONCENTRATION- Solutions containing 0-2 mg of tellurium, different amounts of hydrobromic acid and 5 ml of a 10 per cent, solution of ascorbic acid were prepared. The solutions were diluted to 50 ml with distilled water and their optical densities were measured, with the following results- Hydrobromic acid present, ml . . 40 35 30 28 Optical density . . .. . . 0.421 0,419 0.373 0.132 Although the optical density varies with the hydrobromic acid concentration, this variation can be eliminated when a fixed amount of acid is measured out by pipette for each determination. The fixed amount of concentrated hydrobromic acid chosen was 35 ml.Wavelength, rnp Fig. 1. Absorption spectra: curve A, blank solution; curve B, tellurium bromide in hydrobromic acid COMPLIANCE WITH BEER’S LAW- Solutions containing 0 to 0.4 mg of tellurium were measured into 50-ml calibrated flasks arid 35 ml of hydrobromic acid and 5 ml of a 10 per cent. solution of ascorbic acid were added. The solutions were diluted to the mark with distilled water and their optical densities were measured at 442 mp in 4-cm cells. The figures in the final column represent the optical density per mg of tellurium as measured under these conditions ; the almost constant value shows that Beer’s law is obeyed. The results are shown in Table I. TABLE I COMPLIANCE WITH BEER’S L,~W Tellurium taken, Optical density Optical-density Optical-density difference per mg difference mg of tellurium 0.0 0.003 0.1 0.2 11 0.208 2.08 0.2 0.411 0.408 2.04 0.3 0.616 0,613 2.04 0.4 0.818 0.815 2.04 - - CHOICE OF PRECIPITANT- Sodium hypophosphite was initially used as the precipitant under different conditions, but the recoveries were only about 85 per cent.November, 19571 TELLURIUM IN LEAD AND LEAD ALLOYS 745 Different amounts of tellurium were precipitated from solutions of hydrobromic acid with a 25 per cent.solution of stannous bromide and were allowed to stand for 1 hour to coagulate. The precipitates were collected on No. 4 sintered-glass crucibles and dissolved in hydrobromic acid containing 5 per cent. of bromine. The solutions were evaporated t o small bulk and were then diluted to 50 ml by adding 5 ml of a 10 per cent.solution of ascorbic acid, 35 ml of hydrobromic acid and distilled water. Tellurium was determined in each solution by the proposed method, the results being as follows- Tellurium taken, ing . . 0.1 0.25 0.3 0.4 Recovery, "/o . . . . 102 103 97 100 Five millilitres of a 25 per cent. solution of stannous bromide were chosen for the precipitation of tellurium, APPLICATION OF THE METHOD TO SYNTHETIC MIXTURES- Determinations of tellurium, added as a standard solution, were carried out on 1 g of a lead - antimony alloy containing 12 per cent. of antimony, the results being as follows- Tellurium added, mg . . 0.1 0.2 0.3 0.4 liecovery, 7; . . . . 97, 95 96, 92 104,94 103.99 The recoveries show that the method is applicable to lead - antimony alloys in the absence of interfering elements.EFFECT OF OTHER ELEMEXTS- Selenium is the only other element normally present in lead alloys that is precipitated by stannous bromide and so its effect was examined. The effect of arsenic was also investigated in case of possible co-precipitation with tellurium. Solutions of selenium and arsenic were added to solutions containing 35 ml of hydro- bromic acid and 5 ml of a 10 per cent. solution of ascorbic acid and the resulting solutions were diluted to 50ml. The optical densities of the solutions were measured, the results being as follows- Arsenic added, mg. . . . 0 20 0 Selenium added, mg . . 0 0 2 Optical density . . . . 0,011 0,021 selenium precipitated Recoveries of tellurium from mixtures of tellurium and selenium were made and it was found that the selenium was volatilised during the evaporation to small bulk without loss of tellurium, provided that dry spots were not allowed to form on the bottom of the beaker.The optimum volume after evaporation was found to be approximately 3 ml. REPRODUCIBILITY OF RESULTS- To test the reproducibility of the results by the proposed method, an alloy was prepared having an approximate composition of 85.5 per cent. of lead, 12 per cent. of antimony, 2.5 per cent. of tin and 0.01 per cent. of tellurium, and the tellurium content was separately deter- mined by eight analysts. It can be seen from Table I1 that results by the method were reproducible and that the variation between the results obtained by different analysts was negligible.TABLE I1 The results are shown in Table 11. DETERMINATIOX OF TELLURIUM BY DIFFERENT ANALYSTS Analyst No. Tellurium found, yo Mean, O0 1 0.0086, 0.0089, 0.0089, 0*0090 0.0089 - 0.0088, 0.0084, 0.0092, 0,0089 0.0088 3 0.0094, 0.0082, 0.0095, 0.0097 0.0092 4 0.0089, 0.0090, 0.0089, 0.0093 0.0090 5 0.0085, 0.0088, 0.0086, 0,0086 0.0086 6 0.0090, 0,0096, 0.0091 0.0092 I 0.0105, 0.0097, 0.0095, 0.0097 0.0099 8 0.0086, 0.0086, 0.0086, 0.0089 0.0087 1 METHOD REAGENTS- Hydrobromic acid, 46 to 48 per certt. w/v-Analytical-reagent grade. Bf omine - hydrobromic acid mirfwe-Add 5 ml of analytical-reagent grade bromine ta 95 ml of the hydrobromic acid.746 FLETCHER AND WARDLE [Vol. 82 Stannous bromide solution, 25 per cent.-Dissolve 10 g of pure tin in 50 nil of the hydro- Wash soZz4tio.Pz-Dilute a mixture of 40 ml of the hydrobromic acid and 20 ml of the Ascorbic acid solution, 10 per cent.-Dissolve 10 g of ascorbic acid in distilled water and bromic acid, filter if necessary and dilute to 100 ml with distilled water.stannous bromide solution to 200 ml with distilled water. dilute to 100 ml. PROCEDURE- For tellurium contents in the range 0.005 to 0.02 per cent., dissolve 2 g of sample in 20 ml of bromine - hydrobromic acid mixture with gentle heating. (For tellurium contents outside this range, select a suitable weight of sample.) Boil the solution for 5 minutes t o remove excess of free bromine and then cool it to about 80" C, add 5 ml of stannous bromide solution, mix well and allow it to stand for 1 hour.Then filter the solution through a Xo. 4 sintered- glass crucible fitted with a narrow rubber band that does not come into contact with the filtrate. Use the wash solution to transfer and wash the precipitate from the beaker to the sintered-glass crucible and, finally, wash the beaker and precipitate with distilled water. Retain the beaker. Disconnect the source of suction from the 13uchner flask and place a receiver inside it. By means of a pipette, allow 10 ml of bromine - hydrobromic acid mixture to run down the walls of the crucible to dissolve any tellurium not on the sintered disc. After 3 to 4 minutes, re-connect the source of suction and wash the crucible with distilled water, collecting the filtrate and washings in the receiver. Transfer these back to the original beaker and evaporate to about 3 ml.The last part of the evaporation must be done slowly, and on no account must dry spots be allowed to form on the bottom of the beaker. Allow the beaker to cool and add, by means of a pipette, 38 ml of hydrobromic acid and 6 ml of ascorbic acid solution. Transfer the contents to a 50-ml calibrated flask together with the few millilitres of distilled water used to .wash out the beaker. Dilute the solution to the mark with distilled water and measure its optical density at 442 mp in a 4-cm cell. Carry out a blank determination on the same amounts of the reagents, but omitting the sample. After subtracting the optical density of the blank solution, read the amount of tellurium present from a calibration curve. We express our thanks to the Directors of tE,e Chloride Electrical Storage Company Ltd. and to Dr. &!I. Barak for permission to publish this paper. REFERENCES 1. 2. 3. 4. 5. EXIDE WORKS, CLIFTON JWSCTION Wielsch, W., and Boltz, G., Xikrochim. Acta, 1!)54, 313. -,- , Anal. Chim. Acta, 1954, 11, 438. -,- , 2. anal. Chem., 1954, 142, 94. -,- , Ibid., 1954, 142, 102. -,- , Ibid., 1954, 142, 109. THE CHLORIDE ELECTRICAL STORAGE COMPAXY LTU. SWINTON, MANCHESTER First submitted, July Igth, I956 Amended, J24ne 25th' 1957

ISSN:0003-2654    

DOI:10.1039/AN9578200743

出版商:RSC    

年代:1957

数据来源: RSC

摘要:

November, 19571 FLETCHER AND WARDLE The Determination of Bismuth in Lead and Lead Alloys BY N. W. FLETCHER AND R. WARDLE A new spectrophotometric method for the determination of bismuth in lead and lead alloys is described, in which use is made of the absorption of bismuth bromide in hydrobromic acid. The alloy is dissolved in a mixture of bromine and hydrobromic acid, perchloric acid is added and bromine and volatile bromides are removed by heating. A standard amount of hydrobromic acid and then ascorbic acid, which removes the last traces of free bromine and suppresses the interference caused by copper and iron, are added, and the absorption is measured a t 376 mp. Tellurium is the only element that interferes and it can be removed by preliminary precipitation with stannous bromide.The weight of sample is limited to 1 g and the lower limit of the method is 0*001 per cent. 747 As reported previously,l the absorption spectra of a number of elements have been measured in an attempt to apply the work of Nielsch and Bo1tz2~3~4~5~s to the analysis of lead alloys, and a satisfactory method was developed for the determination of tellurium. By applying the principles involved, a method has also been developed for the determination of bismuth in lead alloys. EXPERIMEXTAL MEASUREMENT OF ABSORPTION SPECTRUM- The absorption spectrum of a solution of 0.1 mg of bismuth in a mixture of 40 ml of AnalaR hydrobromic acid, 46 to 48 per cent., 5 ml of a 10 per cent. solution of ascorbic acid and 5 ml of distilled water was measured by using a Unicam SP600 spectrophotometer.A blank solution of the reagents was prepared without bismuth and its absorption spectrum was measured in the same way. The absorption spectra are shown in Fig. 1, and it can be seen that the absorption of the bismuth solution was a t a maximum a t 375 mp. I00 A 350 375 400 425 450 Wavelength, mp Fig. 1. Absorption spectra: curve A , blank solution; curve B, bismuth bromide in hydro- bromic acid EFFECT OF HYDROBROMIC ACID CONCENTRATION- and 5 ml of a 10 per cent. solution of ascorbic acid were prepared. Solutions containing 0.1 mg of bismuth, different amounts of hydrobromic acid, water The final volume of each748 FLETCHER AND WARDLE : THE DETERMINATION OF [Vol. 82 solutioii was 50 ml, and optical densities were measured at 375 mp, the results being as follows- Hydrobromic acid present, ml .. 40 3.5 30 25 20 10 These results show that the optical density was dependent on the hydrobromic acid concentra- tion, but variations could be avoided by using a fixed amount of acid; 25ml was chosen, as the interference caused by the bromides of copper and iron decreased rapidly with reduced acid concentration. REMOVAL OF FREE BROMINE- The concentrated hydrobromic acid used, 4ti to 48 per cent., contained sufficient free bromine to give a high blank reading, which led to errors when only small amounts of bismuth Optical density . . .. . . 0.678 0.673 0.670 0,670 0.661 0.648 were present. the interference caused by the bromides of copper and iron.’ The addition of ascorbic acid to the solution removed free bromine and also suppressed COMPLIANCE WITH BEER’S LAW- Solutions were prepared that contained 0 to 0-16 mg of bismuth in a mixture of 25 ml of hydrobromic acid and 5 ml of a 10 per cent.solution of ascorbic acid. The solutions were diluted to 50 ml with distilled water and their optical densities were measured at 375 mp in 4-cm cells. The figures in the final column represent the optical density per mg of bismuth as measured under these conditions; the almost constant value show that Beer’s law is obeyed. TABLE 1 The results are shown in Table I. COMPLIANCE WITH BEER’S LAW 131smuth taken, Optical density Optical-density Optical-density difference per mg tliilerence mg of bismuth 0.0 0.04 0-08 0.12 0.14 0.16 0.004 0.266 0.525 0.778 0.910 1.050 - 0.262 0.521 04’74 01.906 1.046 - 6.55 6.51 6.45 8-47 6.64 EFFECT OF PERCHLORIC ACID- It was known that the bromides of antimony, arsenic and tin in liydrobromic acid absorbed at 375 mp, and it was proposed to remove these elements by volatilisation as the bromides from perchloric acid.Various amounts of perchloric acid were added to solutions of 0.1 mg of bismuth in 25 ml of hydrobromic acid. After mixing, 5ml of a 10 per cent. solution of ascorbic acid were added, the solutions were diluted to 50ml with water and their optical densities were measured at 375 mp, the results being as follows- Perchloric acid added, ml . . 0.0 1.0 2 .O 4.0 From these results it can be seen that perchloric acid has little effect on the optical density. EFFECT OF LEAD AND LEAD ALLOYS- A large number of determinations of bismuth were made on synthetic solutions and on an alloy prepared from high-purity lead and Specpure antimony.The conditions and amounts of reagents were varied; it was found that the best recoveries were obtained by using the proposed method. Determinations of bismuth were carried out by the proposed method on 1 g of a lead - antimony - tin alloy containing 12 per cent. of antimony and 2 per cent. of tin, to which 0 to 0.12 mg of bismuth had been added. The results are shown in Table I1 and it can be seen that lead, antimony and tin did not interfere. EFFECT OF OTHER ELEMENTS- No interference was caused by 60 mg of tin, 15 mg of silver, 10 mg of aluminium, 10 mg of calcium, 10 mg of nickel, 4 mg of arsenic, 4 mg of cadmium, 1 mg of selenium, 1 mg of mag- nesium and 1 mg of zinc when added to I-g samples of a lead - antimony alloy when the bismuth Optical density .. . . 0.655 0.660 0.655 0.657November, 1957; BISMUTH IN LEAD AND LEAD ALLOYS 749 in the mixtures was determined by the proposed method. The weights of the elements given are not necessarily the maximum tolerable, but are the maximum amounts likely to he encountered in lead alloys of interest. TABLE I1 RECOVERY OF BISMUTH FROM A LEAD - ANTIMONY - TIN ALLOY Bismuth recovered, A- T- Bismuth added, Optical densit!. Optical-density mg difference mg ._ - - 0.0 0.233 0.04 0.492 0,259 0.040 100 0.083 0.752 0.519 0.080 96 0.117 0.085 0.752 0.116 100 Tellurium can be removed by precipitation with stannous bromide before the volatilisation in the presence of perchloric acid.It is essential to use tin free from bismuth and low in copper and iron, e.g., AnalaR or Specpure, for the preparation of stannous bromide. The completeness of the removal of tellurium was shown when added bismuth was determined on 1 g of a lead - antimony alloy containing 12 per cent. of antimony both with and without the addition of 1 mg of tellurium, the results being 0.0034 and 0.0031 per cent., respectively. Much work was done to eliminate interference caused by copper and iron by varying the conditions and by the addition of other reagents, but the work was unsuccessful until ascorbic acid was added to the solutions. Two solutions, each containing 1 mg of selenium, nickel, zinc, copper, iron and magnesium, 4 mg of arsenic and cadmium and 10 mg of calcium and aluminium in an excess of a bromine - hydrobromic acid mixture, were heated to the fuming-point with perchloric acid and then diluted to 50 ml with 25 ml of hydrobromic acid, 5 ml of a 10 per cent.solution of ascorbic acid and water, and their optical densities were measured at 375mp, the results being as follows- Small amounts of copper, iron and tellurium interfere. Solution So. . . . . Blank 1 2 Optical density . . . . 0.037 0.049 0.048 The addition of 5ml of a 10 per cent. solution of ascorbic acid eliminated the inter- ference caused by 5 mg of copper and 5 mg of iron present in solution together. REPRODUCIBILITY OF RESULTS- .4n alloy containing 12 per cent. of antimony, 2 per cent. of tin, 86 per cent. of lead and approximately 0.01 per cent.of bismuth was prepared and the bismuth was determined separately by eight analysts. TABLE I11 DETERMISATIOK OF BISMUTH BY DIFFERENT ANALYSTS hnalyst No. Bismuth found, 7 ; Nean, yo The results are shown in Table 111. 1 0.0129, 0.0132, 0.0131, 0.0130 0.0131 2 0,0125, 0.0132, 0.0129, 0.0128 0.0128 3 0.0123. 0.0124, 0.0122, 0.0122 0.0123 4 0.0125, 0,0124, 0,0125 0.0125 5 0.0130, 0.0138, 0.0127, 0.0130 0.0131 6 0.0122, 0.0122, 0.0122, 0.0123 0.0122 7 0.0133, 0.0128, 0.0128, 0.0124 0,0128 8 0.0127, 0.0127, 0.0126, 0,0127 0.0127 It can be seen from Table I11 that variations between results obtained by different analysts are negligible and that reproducibility is satisfactory. METHOD RE-4GEKTS- Hydrobromic acid, 46 to 48 per cent. w/v-Analytical-reagent grade.Bromine - hydrobromic acid mixture-Add 5 ml of analytical-reagent grade bromine to 95 ml of the hydrobromic acid.750 FLETCHER AND WARDLE LVol. 82 Perchloric acid, 60 per cent. w/v-Analytical-seagent grade. Ascorbic acid solution, 10 per cent.-Dissolve 10 g of ascorbic acid in distilled water and Stannous bromide solution, 25 per cent.-Dissolve 10 g of analytical-reagent grade tin in Wash solution-Dilute a mixture of 40 ml of hydrobromic acid and 10 ml of the starinous dilute to 100 ml. 50 ml of the hydrobromic acid and dilute to 100 nil with distilled water. bromide solution to 200 ml with distilled water. PROCEDURE IN THE ABSENCE OF TELLURIUM- Dissolve 1 g of sample in 20 ml of bromine - hydrobromic acid mixture with gentle heating if necessary.When dissolution is complete, add 10 ml of perchloric acid and boil the solution on a hot-plate until lead bromide begins to crystallise. Remove the beaker from the hot- plate immediately and continue the heating over the flame of a bunsen burner, swirling the beaker vigorously to prevent the contents from bumping and spitting. Continue to heat until the solution is clear and the volume has been reduced to 2 to 3 ml. If a white residue is present in the hot solution, a further 5 ml of hydrobromic acid and 5 ml of perchloric acid should be added and the volatilisation repeated. Then cool and transfer the solution to a 50-ml calibrated flask containing 25 ml of hydro- bromic acid, using distilled water to assist the transfer and to wash the beaker. Add 5 ml of ascorbic acid solution, dilute to the mark with distilled water and adjust the temperature to 20" i.2" c . Measure the optical density with a Unicam SP600 spectrophotometer at 375 mp in a 4-cm cell with the violet filter in position, and the instrument adjusted against water. Carry out a blank determination on the reagents alone in an identical manner. Read the bismuth content from a calibration curve. PROCEDURE IN THE PRESENCE OF TELLURIUM- Dissolve 1 g of sample in 20 ml of bromine - h:ydrobromic acid mixture and boil to remove most of the bromine. Add 1 ml of stannous bromide solution and set aside for 1 hour, filter and collect the precipitated tellurium in a No. 4 sintered-glass crucible and the filtrate in a suitable receiver. Use not more than 20 rrnl of the wash solution in small volumes to transfer the precipitate to the sintered-glass crucible and wash the precipitate and beaker. Then wash the precipitate and beaker with distilled water. Transfer the combined filtrate and washings to the original beaker, add 10ml of perchloric acid and continue from this point as for the procedure in the absence of tellurium. We express our thanks to the Directors of The Chloride Electrical Storage Company Ltd. and to Dr. M. Barak for permission to publish this paper. REFERENCES 1. 2. 3. 4. 6. 7. EXIDE WORKS, CLIFTON JUXCTIOS Fletcher, N. W., and Wardle, R., Analyst, 1957, 82, 743. Nielsch, W., and Boltz, G., Anal. Chim. Acta, 1!354, 11, 438. -, --, Mikrochim. Acta, 1954, 313. -, -, 2. anal. Cham., 1954, 142, 94. - - , Ibid., 1954, 142, 109. Stoliarova, I. A,, Zhur. Anal. Khim., 1953, 8, 210; Aptal. Abstv., 1954, 1, 1239. 5. -, --, Ibid., 1954, 142, 102. THE CHLORIDE ELECTRIC.4L STORAGE COMPANY LTD. First submitted, July 19th, 1956 Amended, Jttne 25th, 1967 SWINTON, ~\IANCHESTER

ISSN:0003-2654    

DOI:10.1039/AN9578200747

出版商:RSC    

年代:1957

数据来源: RSC

摘要:

November, 19573 BUDOWSKI AND BOXDI 781 Determination of Vitamin A by Conversion to Anhydrovitamin A* BY P. BUDOWSKI AXD A. BOND1 Vitamin A may be converted to anhydrovitamin X in benzene solution at room temperature, in the presence of toluene-p-sulphonic acid as a catalyst. The increase in extinction at 399 mp, which results from the dehydration, is proportional to the amount of vitamin A present and can be used for the determination of this vitamin in unsaponifiable extracts. The method exhibits a high degree of specificity and has been applied to a variety of products, including vitamin-A concentrates, fish oils, margarine, butter and poultry mashes, without further purification of the unsapouifiable extracts. THE routine determination of vitamin A by ultra-violet absorption or by the antimony trichloride reaction presents many difficulties with low-potency materials, particularly fortified feeds and foods.Results are often unreliable, in spite of cumbersome chromato- graphic purification procedures. In the work described here, an attempt has been made to arrive a t a simple and accurate assay procedure by making use of a reaction that is more specific for vitamin A and less subject to interference than other non-biological tests. Such a reaction is found in the conversion of vitamin A (I) to anhydrovitamin A (11), which is known to occur in anhydrous solvents in the presence of traces of mineral acids1Jt39415- Vitamin X Anhydrovitamin A The appearance of the retro structure, together with the lengthening of the double-bond system, results in a considerable displacement of the absorption maximum towards the visible region, an increase in extinction and the formation of a fine structure. By measuring the change in extinction a t appropriate wavelengths, it has been found possible, under certain experimental conditions, to determine the vitamin-A content of unsaponifiable extracts from a variety of materials without further purification.Dehydration procedures involving ethanolic hydrogen chloride have been used occasion- ally for the detection of vitamin A in biological materials when doubts existed as to its presen~e.~3~3’ In one of them an attempt a t quantitative measurements has been made.5 Besides ethanolic hydrogen chloride, the use of chloroformic hydrogen chloride has also been describeda8 For quantitative measurements, however, it was found that greater rapidity and better control of the reaction were achieved by adaptation of a procedure described by Shantz9 for the preparation of anhydrovitamin A and which involves the use of benzene as a solvent and of toluene-p-sulphonic acid as a catalyst. EXPERIMENTAL Preliminary tests had shown that the formation of anhydrovitamin A from vitamin A in benzene solution could be carried out at room temperature in the presence of suitable amounts of toluene-p-sulphonic acid, provided that the vitamin was in the alcohol form.Routine saponification and extraction procedureslO were followed for the preparation of the unsaponi- fiable matter, since no special study of these methods was deemed necessary in connection with the present problem.ABSORPTION SPECTRA OF VITAMIS A ASD ANHYDROVITAYIS .A : DIFFERENCE SPECTRUM- Fig. 1 shows the absorption spectra in benzene of the unsaponifiable matter from a United States Pharmacopoeia reference standard, before and after application of the dehydra- tion procedure described below. The U.S.P. reference standard consists of gelatin capsules * Communication from the Agricultural Research Station, Rehovot, 1957 Series, No. 209752 containing 250 mg of a solution of crystalline vitamin-A acetate in cotton-seed oil, at a level of 10,000 i.u. per g. The spectrum of vitamin A has a maximum at 331 mp, and anhydrovitamin A exhibits maxima at 358, 377 and 399mp, minima at 364 and 389mp, and an inflexion from 340 to 345mp.It should be noted that the use af benzene as a solvent results in a batho- chromic shift of absorption by about 7 mp in relation to ethanol. The purity of the anhydro- vitamin A obtained by dehydration can be judged from the relative extinction values at the three peaks. The ratios E(399)/E(377) and E(358)/E(377) are found to be equal to 0.868 and 0.692, respectively, while the corresponding ratios calculated from Shantz's data9 for crvstalline anhvdrovitamin A in ethanol (maxima at 351, 371 and 392 mp) are 0.870 and BUDOWSKI AND BONDI: DETERMI"J4TION OF VITAMIN A BY [Vol. 82 0.690, respectiGely . 10 Wavelength, mp Fig. I. Xbsorption spectra of unsaponifiable matter from U.S.P. reference standard: curve A, before dehydration; curve B, after dehydration Fig. 1 also illustrates the yield from the dehydration reaction, as given by the relative heights of the absorption bands of anhydrovitamin A at 377 mp and vitamin A at 331 mp.This ratio is found to be 1.59, while, for crystalline anhydrovitamin A and vitamin A in ethanol, the corresponding ratio of the molecular extinction coefficients is 1.89. This figure is based on extinction coefficients of 1835 for vitamin All and 3860 for anhydrovitamin A.9 If it is assumed that comparison of such ratios is valid for different solvents, the yield of the dehydration reaction is calculated as 1.59/1*89, or 84 per cent. Robeson and Baxterl2 found a yield of about 60 per cent. for dehydration in ethanolic hydrogen chloride. 300 340 380 4 20 Wavelength, rnp Fig. 2 . Absorption spectra of unsaponifiable matter from a poultry mash: curve A, before dehydra- tion: curve B, after dehydration Fig.2 shows the absorption curves of the unsaponifiable matter from a fortified poultry mash, before and after dehydration. This is a typical example of a highly impure vitamin-X extract, in which the presence of the vitamin is indicated by a slight hump on the absorption spectrum in the region of the maximum absorption of vitamin A. Similarly, the spectrum of the anhydrovitamin A obtained by dehydration is superimposed upon a strong backgroundNovember, 195’71 CONVERSIOS TO ANHYDROVITAMIX A 753 absorption. Absorption curves very similar to those shown in Fig. 2 were obtained for the unsaponifiable matter of other fortified feeds and foods. Although it is clear that, with such extracts, neither the absorption spectrum of vitamin A nor the curve for anhydrovitamin A can be used for any reasonably accurate determination of vitamin A, because of excessive background absorption, it appears that the “difference spectrum” is relatively independent of irrelevant absorption and can be used for such a purpose.The difference spectrum is obtained by subtracting the optical density of the untreated solution, E,, from that of the treated (dehydrated) extract, E,, and plotting the difference, E D = E, - E,, against wavelength. It is necessary to point out that such a difference spectrum differs from the absorption spectra of both vitamin A and anhydro- vitamin -4, and, in fact, corresponds to no specific compound. It should not be confused with the difference spectrum obtained by destructive irradiation13 or selective adsorption of vitamin A,14 which is expected to be identical with the absorption spectrum of pure vitamin A.Fig. 3 shows the difference spectra of the C.S.P. reference standard and the poultry mash referred to on p. 752. 0 Fig. 3. Difference spectra: curve A, U.S.P. refer- ence standard; curve B, poultry mash In order to compare the spectra at different concentrations of vitamin A, the curves have been made to coincide at the 399-mp absorption peak, by using extinction ratios, EDh/399, rather than extinctions, ED, in accordance with the principle put forth by Oser, Melnick and Pader.15 It can be seen that, in spite of the very different natures and purities of the unsaponifiable extracts, the difference spectra are nearly identical. Very similar difference spectra have been obtained for a variety of products, including concentrates, fish-liver oils, premixes, poultry mashes, margarine and butter.TABLE I EXTINCTION RATIO (EDh/399) /(E,4/377) OF UNSAPONIFIABLE FRACTIOKS FKOI\I VARIOUS MATERIALS E~A/399 Number EoA/377 of samples r---------h--7 examined Range Average U.S.1’. reference standard . . 3 0.909 to 0.916 0.913 Fish-liver oils . . . . . . 4 0.903 to 0.912 0.908 Concentrates . . . . . . 2 0.900 to 0.909 0.905 Premixes . . . . . . . . 4 0.890 to 0.920 0.905 Margarine . . . . . . . . 2 0.864 to 0.866 0.865 Butter . . . . . . * . 2 0.815 to 0.826 0.820 Designation Near the visible region, where the absorption due to vitamin A becomes small, the difference spectra resemble the absorption spectrum of anhydrovitamin A : maxima are found at 399 and 377 mp, and a minimum at 389 mp.At lower wavelengths, however, the shape becomes different. An inflexion occurs at 360 to 365 mp, and below 347 and 348 mp ED becomes negative, since in this region the curve for anhydrovitamin ,4 lies below the spectrum754 BUDOWSKI ASD BONDI: DETERMINATION OF VITAMIN A BY [Vol. 82 of vitamin A. Here, the shape of the difference spectrum becomes more variable, especially for extracts from foods and feeds and for oxidised samples. The constancy of shape of the difference spectrum in the region of absorption due to anhydrovitamin A is illustrated by the results given in Table I for the extinction ratios (EDh/399)/(EDh/377) obtained from different products.Except for a somewhat lower ratio found for butter, the values are fairly constant and lie within the range 0.86 to 0.92. The constant shape of the difference spectra obtained for such widely different samples constitutes proof of the high degree of specificity of the dehydration reaction. The increase in extinction at 377 or 399mp brought about by the dehydration may therefore be expected to correlate with the vitamin-A content of the unsaponifiable extracts. OPTIMUM CONDITIONS OF DEHYDRATION- When dehydration is carried out in benzene under the influence of suitable amounts of toluene-$-sulphonic acid, maximum formation of anhydrovitamin A occurs almost instan- taneously. Further changes, however, result in a subsequent decrease in the amount of this compound.Similar observations have been m,ade when dehydration has occurred in ethanolic hydrogen ~hloride.~ It was found convenient to stop the reaction by neutralisation of the catalyst with alkali, whereupon spectrophotometric readings would remain stable for at least 1 hour. The effect of time of catalysis and concentration of catalyst was tested by using, as before, the unsaponifiable matter from the U.S.P. reference standard and the poultry mash. Some 1-ml portions of a benzene solution of the unsaponifiable matter, which contained 80 i.u. in the case of the U.S.P. reference standard, and 17 i.u. in the case of the poultry mash, were mixed with 4-ml portions of benzene containing different amounts of toluene-p- sulphonic acid. The reactions were stopped after various time intervals by thorough shaking with 5 ml of 0.5 N sodium hydroxide, and the optical densities of the benzene solutions were read at 399 mp against the corresponding blanks, which consisted of 1 ml of unsaponifiable extract plus 4 ml of benzene.The results are given in Table 11. TABLE I1 EFFECT OF CONCENTRATION OF CATALYST AND TINE OF CATALYSIS ON THE INCREASE IN EXTINCTION, E D , AT 399 mp Concentration of catalyst in EI, at 399 mp, after- Designation reaction mixture, 7 h > Irg Per ml 0 minute 1 minute 3 niinutes U.S.P. reference stadard . . Poultry mash . . .. 100 150 200 100 150 200 0.565 0.609 04310 0,585 0.611 0.608 0.602 0.585 - 0.298 0.402 0.399 0.396 0,402 0.395 0,390 0.402 0.374 It is seen that, for all concentrations tested, the readings obtained after a reaction time of 1 minute represent the maximum value, or are very close to it, except for the U.S.P.reference standard, with which the highest concentration tested gave values decreasing from the start. For the analytical procedure, a dehydration time of 1 minute is recommended, together with a concentration of catalyst of 120 pg per ml. Such a concentration results if 1 ml of unsaponifiable extract is mixed with 4 ml of benzene containing 150 pg of toluene-$- sulphonic acid per ml. These conditions, however, are not critical, and the dehydration time and concentration of catalyst can be varied by as much as 10 per cent. with negligible effects on the readings. The effect of temperature was examined by immersing the vessel containing the reaction mixture in a water bath at different temperatures during the 1-minute reaction period.With the U.S.P. reference standard, optical-density readings of 0.610, 0-604, 0.599 and 0.571 were obtained at 18", 28", 38" and 48" C, respectively. Small fluctuations in room tempera- ture will therefore have negligible effects on the readings. The recommended catalyst solution is super-saturated, and crystals will separate at room temperature on standing. Furthermore, toluene-$-sulphonic acid tends to form lessNovember, 19571 CONVERSION TO ANHYDROVITAMIN A 755 soluble hydrates on exposure to moisture. This is accompanied by a loss of catalytic activity. It was found that full catalytic activity could be restored by distilling or removing by boiling ti to 10 per cent.of the solvent, thereby dissolving separated crystals and simultaneously expelling traces of moisture. Upon cooling, the solution remains in a supersaturated state for 30 minutes or more, depending on the room temperature. PROPORTIOSALITY BETWEEN VITAMIN-A CONTENT AND INCREASE IN EXTINCTION AT 399 mp- In order to test the relation between level of vitamin A and increase in extinction at 399 mp, the dehydration procedure described below was applied to a U.S.P. reference standard and to a poultry mash (cantaining 6.5 i.u. per g), graded amounts of unsaponifiable matter being used. Concentration of sample, g per 5 ml (curve B) 0 2 4 6 The results are represented graphically in Fig. 4. I 2 - 0 20 40 60 80 I00 Concentration of vitamin A, i.u.per 5 ml (curve A) Relationship between increase in extinction a t 399mp and concentration of vitamin A: curve A, U.S.P. reference standard; curve B, poultry mash Fig. 4. I t is seen that in both cases the readings are proportional to the level of vitamin A. The straight line obtained from the U.S.P. reference standard permits the calculation of the increase in extinction at 399 mp per i.u. of vitamin A per 5 ml of final solution. The value obtained is 0.0122. This figure can be used for the calculation of the vitamin-A content in unknown extracts in much the same way as the extinction coefficient of vitamin A is used in the ultra-violet absorption method. Indeed, it may be expressed as an extinction coefficient by means of the following equation- = 2030, 0.0122 20 x 0.3 x (E:&), at 399 mp = where the denominator represents the concentration of vitamin A (1 i.u.per 5 ml) expressed as a percentage. NETHOD Saponification and extraction of the unsaponifiable matter are carried out by standard procedures.1° The final extract is made up in dry benzene, so that 1 ml contains between 5 and 6Oi.u. of vitamin A. REAGENT- Catalyst solution-Heat under reflux 15 mg of toluene-$-sulphonic acid monohydrate with 100 ml of redistilled benzene until dissolved. Distil or boil off 10 ml of solvent to drive out moisture, allow the solution to cool, protected from moisture, and re-adjust the volume to 100ml with dry benzene. Activate before use by repeating the distillation procedure. Glassware used in preparing the catalyst solution should be dry.PROCEDURE FOR DEHYDRATION- Mix 1 ml of a benzene solution of unsaponifiable matter with 4 ml of catalyst solution. After 1 minute, neutralise the catalyst by shaking the solution with 5 ml of 0.5 N sodium766 BUDOWSKI AND BONDI: DETERMINATION OF VITAMIN A BY [voi. a2 hydroxide for 1 minute. Clarify the solution by centrifugation or settling. Measure the optical density of the clear solution in a Beckman DU spectrophotonietcr at 399 mp, against a mixture of 1 volume of unsaponifiable matter in benzene and 4 volumes of henzene, set at 100 per cent. transmission. Alternatively, 1 g of sodium carbonate may be used. TABLE 111 RECOVERY OF VITAMIN A ADDED TO UNSAPONIFIlBLE EXTRACTS Total Vitamin A Vitamin A vitamin A Vitamin recovered Designation present, added, found, - i.u.per g i.u. per g i.u. per g 1.u. per g Fish oils- A . . . . B . . . . Margarine Butter A . . Butter . . A .. . . B .. .. FOOdS- Poultry mashes- L .. D (unfortified) . . 24,550 22,660 47,500 52,950 . . 9160 6600 15,720 6560 . . 19.0 32.3 .52.i 33.7 . . 22.1 32.3 55.9 33.8 . . 14.2 16.5 31.4 17.2 .. 6.5 10.5 16.8 10.3 .. 11.2 28.1 40.1 28.9 . . 2.9 2.5 5.3 2.4 . . 0.6 8.0 8.7 8.1 TABLE IV COMPARISON OF RESULTS OBTAINED B,Y THE ANTIMONY TRICHLORIDE METHOD AND DEHYDRATION PROCEDURE Designation Cotiwntraks- A* . . BT . . Fish oils- A * . . . B . . . . c . . . . Feed premix Foods- Margarine A . . Margarine B . . Margarine B$ Chocolate spread Butter A . . Butter R . . Butter B: . . Poultry MasJces- A . . . . A: . . . . B ... . B!: . . . . C ' Vitamin A found by the i.u. per g Vitamin A found by the antimony trichloride method, dehydration procedure, 1.u. per g . . 1,135,000 . . 663,000 .. 26,250 . . 9300 . . 8100 . . 2400 . . . . . . * . . . . . . . * . .. . . * . .. 27.2 26.0 22.5 20.6 26.1 19-4 16.0 8.1 4*'7 25.2 17.6 20.2 7.0 3.8 3.2 1,130,000 554,000 24,550 9160 7700 2350 19.0 21.0 18.5 22.1 14.2 - - * Vitamin-A palmitate (obtained from Merck & Co. Inc.) having a potency of 1,000,000 i.u. per g. t Commercial concentrate, showing an abnormal ultra-violet spectrum. $ Chromatographed on a magnesium oxide - Celite mixture (1 $- l ) , with a light petroleum - acetone mixture (9 + 1) as eluting agent. 0,' / O 101 99 104 1045 103 98 103 96 101November, 19571 CONVERSION TO ANHYDROVITAMIN A 757 CALCULATIOK-The reading thus obtained represents the increase in optical density, ED, caused by dehydration, and can be converted to amount of vitamin A in i.u.per aliquot taken for dehydration (1 ml) by the following equation- E D 0.0122' Amount of vitamin A, i.u. = - where 0.0122 is the optical density increase corresponding to 1 i.u. of vitamin A. figure is obtained when the U.S.P. reference standard is treated by the proposed procedure. RECOVERY TESTS- The results of recovery tests carried out with a number of different samples are shown in Table 111. It should be pointed out that these tests refer to the recovery of vitamin A carried through the dehydration procedure only and do not take into account possible losses caused by saponification and extraction.The satisfactory results shown in Table I11 are proof of the absence of significant inter- ference by substances normally present in unsaponifiable extracts. COMPARISON OF RESULTS BY THE AKTIMONY TRICHLORIDE METHOD WITH THOSE BY THE The results obtained by application of the dehydration procedure to a number of samples are compared in Table IV with those given by the Carr - Price method.1° Aliquots of the same saponified extracts were used in each test. It is seen that with high-potency materials, such as concentrates, fish oils and premixes, the agreement is generally good. Dehydration values are 0 to 6.5 per cent. lower than Carr - Price values. A large discrepancy was observed for concentrate B. This sample, upon closer inspection, revealed an abnormal absorption spectrum, with a maximum at 330 mp (in hexane) and marked inflexions at 348 and 368 mp.After chromatography on alumina, a fraction was obtained that exhibited maxima at 333, 348 and 368 mp (in hexane), which indicated the probable presence of retrovitamin A or some derivative possessing the same chromophoric s t r u ~ t u r e . ~ J ~ J ~ J ~ J ~ Retro or rehydrovitamin A, which has only about 8 per cent. of the biological activity of vitamin A,9 is chromogenic toward antimony trichlorides (absorption maximum at 612 mp), but yields only a small amount of anhydrovitamin X under dehydrating conditions.20 Table IV also shows that Carr - Price values obtained from unsaponifiable extracts of foods and feeds are considerably higher than the corresponding dehydration values.Fre- quently, the antimony trichloride colours are abnormal. This is not surprising, in view of the presence in such extracts of irrelevant chromogenic material, such as sterols, carotenoids, vitamin-A oxidation products and so on. Purification by means of chromatography on magnesium oxide - Celite mixture tends to bring the Carr - Price values closer to the observed dehydration figures. The possibility of applying the dehydration procedure to the unsaponifiable fraction of foods and feeds without further purification gives it a distinct advantage. This APPLICATIOKS OF THE METHOD DEHYDRATIOX PROCEDURE- APPLICATIOS OF THE DEHYDRATION PROCEDURE TO AUTOXIDISED T'ITAMIS-A PALMITATE- Difficulties are often experienced and abnormal colours observed when the Carr - Price method is applied to products that have suffered losses of potency through atmospheric oxidation.The dehydration procedure was applied to vitamin-A palmitate (obtained from Merck & Co. Inc., having a potency of 1,000,000 i.u. per g, purified by chromatography on alumina), dissolved in liquid paraffin at the level of 12,000 i.u. per g and allowed to undergo autoxidation at 37" C in the dark. Figs. 5 and 6 show the changes observed in the absorption curve and difference spectrum, respectively, during autoxidation. As in Fig. 3, extinction ratios have been plotted, rather than extinctions. The absorption spectrum of vitamin-A palmitate, measured in benzene after saponifica- tion, is seen to undergo marked changes, which completely obliterate the characteristic peak originally present at 331 mp.On the other hand, the difference spectrum maintains its characteristic shape, even after 98 per cent. destruction has taken place. Only during the7.58 [Vol. 89 advanced stages of autoxidation is any appreciable distortion of the difference spectrum observed at wavelengths below 370 mp. Table V gives a comparison between results obtained by the antimony trichloride method and the dehydration procedure with vitamin-h palmitate during autoxidation. The two procedures were applied to aliquots of the same unsaponifiable extracts. BUDOWSKI AXD BONDI: DETERMINATION OF TIITAYIK 4 BY I50 9 ‘‘9 “.. ..&, Wavelengrh, rnp Fig. 5 . Changes in the absorption spectrum of vitamin A during autoxidation of vitamin-A palmitate, the extent of oxidation being: curve A, 0.0 per cent.: curve B, 69.1 per cent.; curve C, 92.4 per cent.: curve D, 98.0 per cent.Wavelength, mp Fig. 6 . Changes in the difference spectrum vitamin-il palmitate during autoxidation, the extent of oxidation being: curve A, 0.0 per cent.: curve B, 69.1 per cent. : curve C, 92.4 per cent. ; curve D, 98.0 per cent. !O of It is seen that the Carr - Price values are consistently higher than the dehydration results. This would be expected from the presence of colour-producing oxidation products. That such substances were actually contributing to the values obtained was seen from the greyish brown colours produced by antimony trichloride with the more highly oxidised samples. VJith them the readings were quite uncertain, because they increased rapidly from the start.In Table V, the lowest values read, i.e., those read immediately after mixing, have been recorded, and therefore the results obtained with antimony trichloride represent mini- mum values. I t should be noted that saponification does not eliminate the interference caused by oxidation products of vitamin--4 palmitate in the Carr - Price test.November, 195i] CONVERSIOS TO ANHYDROVITAMIS A TABLE V COWARISOS OF RESULTS GIVEN BY CARR - PRICE METHOD AND DEHYDRATION PROCEDUKE FOR VITAMIN-A PALMITATE UNDERGONG AUTOXIDATION IN Vitamin -\ remaining PARAFFIN SOLUTION AT 37” c Vitamin A found by- 7 7 r~~ - --A aiitimonytrichlor~dc dehydration dntimonytiichlorrtle dehydration -rnllr method, method, method, method, l i o i i r ~ I u per g n 12,000 11,500 100 100 ‘3.5 5690 4590 47 40 28.8 4010 3560 33 31 47.5 1250 880 10.4 8.0 51.5 470 226 3.9 2.0 O / I u per g % J U COKCLVSIOK 750 The limited results presented here do not permit a rigorous evaluation of the dehydration method.However, a number of observations may be made about its usefulness compared with that of any other non-biological assay procedure. In the dehydration method, use is made of a chemical reaction, in combination with spectrophotometric absorption measurements. This method may therefore be expected to display features pertaining to both the Carr - Price test and the ultra-violet absorption procedure. Indeed, dehydration is similar to the development of the blue colour with antimony trichloride, not only in the experimental technique followed, but possibly also in the chemical mechanism i n v o l ~ e d .~ l 3 Further, the products of both reactions are unstable. But, whereas the blue colour must be read rapidly, because of its instability, in the dehydration reaction the product can be stabilised by neutralisation of the catalyst. As a result, the dehydration procedure exhibits the good reproducibility and accuracy of readings (as exemplified by the accurate proportionality between ED and level of vitamin A) of the ultra-violet absorption met hod. The sensitivity of the dehydration method, as measured by the extinction coefficient, (E;>& at 399 mp, is slightly superior to that of the ultra-violet absorption method, but inferior to that of the Carr - Price procedure.Results obtained by the dehydration method are, in general, lower than those obtained by the antimony trichloride procedure, even with materials rich in vitamin A, such as fish oils and concentrates. This should be noted in connection with the higher specificity of the dehydration procedure and the observationz1sz2 that other non-biological tests tend to yield higher values than biological assay. This high degree of specificity, which is evident from the relatively constant shape of the difference spectra obtained from widely different materials, no doubt constitutes the main advantage of the dehydration method. It permits the application of the dehydration pro- cedure to the unsaponifiable fraction of such materials as foods and feeds without further purification.Because chromatography is not needed, time is gained and a potential source of errors is eliminated. 1 7 -. 3. 4 . 5 . 6. x. ,. !). 10. 11. 12. 13. REFERENCES Edisbury, J . R., Gillam, -4. E., Heilbron, 1. M., and Morton, R. A , , Biockeni. J . , 1932, 26, 1164. Meunier, P., Dulou, R., and Vinet, A,, Bull. SOC. Chiin. Baol., 1943, 25, 371. Shantz, E. M., C,awley, J . D., and Embree, N. D., J . Amev. Clzeni. Soc., 1943, 65, 901. Embree, S. D., J . Bid. Chem., 1939, 128, 187. Sebrell, W. H., jun., and Harris, R. S., “The Vitamins,” Academic l’rcss Inc., Sew York, 1954, Ames, S. R., and Harris, P. L., Science, 1954, 120, 391. Gillman, J., Norton, K. B., Rivett, D. E. A,, and Sutton, D X., Uiockevr. J , , 1956, 63, 459. Petracek, F. J , , a.nd Zechmeister, L., J . .4nzer. Chem. Soc., 19.58, 78, 3188. Shantz, E., .I. Bid. Ckem., 1950, 182, ,516. The Association of Vitamin Chemists Inc., ‘Xethods of Vitamin Assay,” Second Edition, Inter- Sebrell, TV. H., jun., and Harris, R. S., 09. c i t . , p. 29. Robeson, C. D., and Baxter, J. G., J . Amer. Chem. Soc., 1947, 69, 136. Gyorgy, I?., Editor, “Vitamin Methods,” Academic Press Inc., Sew Tork, 1950, \‘ohme I, Volume I, pp. 48 to 50. science Publishers Inc., New York. 1951, pp. 28 to 31. pp. 39 t o 52.760 CLARK : 0-DITHIOLS IN ANALYSIS. PART V [Vol. 82 14. 15. 16. 17. 18. 19. Oroshnik, W., Science, 1954, 119, 660. 20. 21. 22. Awapara, J . , Mattson, F. H., Mehl, J. W., and IDeuel, H. J., jun., Science, 1946, 104, 602. Oser, B. L., Nelnick, D., and Pader, M., Ind. Eizg. Chenz., Anal. Ed., 1943, 15, 717. Beutel, R. H., Hinkley, D. F., and Pollak, P. I.. J . Amev. Chem. Soc., 1955, 77, 5166. Oroshnik, W., Karmas, G., and Mebane, 1. D., Ibid., 1952, 74, 295. Oroshnik, IT., and Mebane, A. D., Ibid., 1954, 76, 5719. Budowski, P., and Bondi, A . , unpublished observations. Oser, B. L., 34elnick, D., Pader, >I., Roth, R., and Oser, JI., Ind. Eng. C1zem., Anal. Ed., 1945, Chilcote, 31. E., Guerrant, S. B., and Ellenbergc:r, H. -\., -4nnl. C h i n . , 1949, 21, 1180. 17, 559. AGRICULTURAL RESE.4RCH sT.4TIOX REHOVOT, ISRAEL APvil 29fi1, 1957

ISSN:0003-2654    

DOI:10.1039/AN9578200751

出版商:RSC    

年代:1957

数据来源: RSC