摘要:

NOVEMBER, 1958 THE ANALYST Vol. 93, No. I I12 Determination of Trace Amounts of Cobalt in Alumina by Atomic-absorption Spectroscopy RT B. FLEET, I<. V. LIBERTY AND T.S. IXIEST (Chemistry Department, Impevial College, Loqzdon, S. W.7) After the dissolution of alumina by hydrochloric acid in a sealed tube at 270" C traces of cobalt are determined by measurement of atomic absorption a t 240.7 nm. In one of the two procedures described cobalt, in the range 60 to 250 p.p.ni. in alumina, is determined by aspiration directly into an air- acetylene or nitrous oxide - acetylene flame, and in the other procedure cobalt, in the range 10 t o 100 p.p.m. in alumina, is determined by co-precipitation on hydrated manganese dioxide, followed by extraction into isobutyl methyl ketone as its 8-hydroxyquinoline complex.The extract is sprayed into an air - propane flame and absorbance measurements are made as before. hTo interferences wcre found from the other elements likely to be present in alumina. The nitrous oxide - acetylene flame is lzss favourable than air - acetylene for the determination of cobalt because of the unfavourable effect of the strongly reducing cyanogen zone of the flame and because of considerable loss of atomic cobalt caused by ionisation. SEVERAL resonance lines have been used in the determination of cobalt by atoniic-absorption ~ p e c t r o s c o p y , ~ ~ ~ ~ ~ ~ ~ ~ ~ but that at 240.73 nni appears to be the most sensitive, provided that a sufficiently narrow slit is used to isolate the line from others nearby,G e.g., those at 240.63 and 240.88 nm.The sensitivity of the 210-73 nm line was confirmed in the present study, as shown in Table I. Generally the measurement is best made in an oxidising air - acetylene name. steels and nicl~el,~ but there is otherwise a sparsity of information on the interference of other metal ions, particularly those which form refractory oxides, such as aluminium. TABLE I (about 4 F.p.ni. aqueous Co2+/air - propanc flame) Procedures have been described for cobalt in copper CHOICE OF ATOMIC LINE FOR MEASUREMENT Spectral-line wavelength, nm Remarks on absorption 345.3 No absorption observed 353.3 No absorption observed 240.7 0.22 absorbacce unit 242-5 0.14 absorbance unit 252.1 0.08 absorbance unit 352.9 No absorption observed 340.5 Insufficient intensity to make measurements possible The determination of cobalt in nearly pure alumina by atomic-absorption spectroscopy presents problems, such as sample dissolution to produce a medium suitable for nebulisation into a conventional burrier and the matrix effect of large amounts of aluminium on the atomisation of small amounts of cobalt.In this communication we propose two analytical methods. The first of these permits the determination to be effected on a simple atomic- absorption apparatus with a relatively cool, air - propane flame. I t involves collection of the traces of cobalt by co-precipitation on manganese dioxide and solvent extraction with a hexone solution of 8-hydroxyquinoline. This method is sensitive and precise, but is more time consuming than the alternative procedure, which relies on the use of a relatively hot, air - acetylene flame and more sophisticated apparatus.The latter procedure does not 701 0 SAC and the authors.702 FLEET, LIBERTY AND WEST: DETERMINATION OF COBALT IN [,4nalyst, 1701. 93 require chemical pre-manipulation other than sample dissolution, but it is less sensitive. The sensitivity of the latter procedure could be increased to equal that of the former by combina- tion with the collection - extraction process used in the cool-flame method, but only by sacrificing its advantages of rapidity and simplicity. EXPERIMENTAL AFPARATUS- Techtron AA4 atomic-absorption syectrophotometer, with a lamp and detector modulated at 285 C.P.S. and fitted with an R213 photomultiplier; 10-cm slot burner heads for air - propane (AB42) and air - acetylene (AB41), and 5-cm slot burner head for nitrous oxide - acetylene (AB40); and a Hilger & Watts (FL157) hollow-cathode lamp for cobalt.INSTRUMENTAL SETTINGS- Wavelength peaked on 240.7 nin; slit width 50 p (minimum compatible with signal strength available from lamp); lamp current 29 mA; air pressure 15 lb per sq. inch; nitrous oxide pressure 16 lb per sq. inch; propane pressure sufficient to give a non-luminous flame; acetylene pressure sufficient to give a “rich” slightly luminous flame with air or to give a 15-mm high red feather with nitrous oxide; height of measurement in air - propane flame, 15 mm; in air - acetylene, 13 mm; in nitrous oxide - acetylene, 15 mm; solution uptake rates in air - propane and air - acetylene were 1 ml per 14 seconds for water, 1 ml per 17 seconds for AlCl, solution, 1 ml per 13-5 seconds for hexone and in nitrous oxide - acetylene, 1 ml per 11.6 seconds for water and 1 ml for 15.2 seconds for AlC1, solution.(AlCl, solution equivalent to 0.1 g of Al,O, per 2.5 ml.) PRESSURE VESSEL FOR DISSOLUTION OF ALUMINA- A stainless-steel pressure vessel operating at 3000 lb per sq. inch pressure at 2‘70” C, similar to that described elsewhere,* and supplied by A.W.R.E., Aldermaston, was used. The vessel is loaded with the requisite amount of solid carbon dioxide to equalise pressure on both sides of a sealed inner silica tube containing the alumina sample and concentrated hydro- chloric acid at 270” C. SILICA EIGESTION TUBES- 23 cm long, sealed at one end.REAGENTS- These were made from 2 to 26mm wall silica tubing with an inner bore of 5 mm, about The tubing was obtained from Messrs. Hanovia Limited. All reagents were of analytical-reagent grade, except where stated. Stock cobalt solutiom, 1000 and 100 fi.fi.irt.--Prepare by dissolving 4.0372 g and 04037 g Lower concentrations were Sodium hydroxide solution, 4 M. Hydrochloric acid, concentrated (about 35 per cent.). 8-Hydroxyquinoline, 5 per cent. (0.35 M) in 2 M acetic acid. Potassium alumiaium sulplzate, saturated soli,dion-About 1 1.4 g of KAl(SO,),. 12E1,O Alztminizlm chloride solutiofc-Aluminium foil, 50 g, was dissolved in excess of hydro- Sodium acetate trihydrate, solid. Hydroxylammonium chloride, solid. of CoCl,.6H20, respectively, in water and diluting to 1 litre.prepared by dilution immediately before use. were dissolved in 100 g of water. chloric acid (1 + 1) and diluted to 1 litre with distilled water. ANALYTICAL PROCEDURES (1) DISSOLUTION OF ALUMINA SAMPLE- Weigh accurately about 0-1 g of alumina into the bottom of a silica tube of wdl thickness 2.0 to 2.5 mm and bore 5 mm, add sufficient concentrated hydrochloric acid to fill a 9 to 10-cm length. Draw out the upper part of the tube, cool and seal off quickly, keeping an even wall thickness and avoiding the formation of a vacuum in the tube (a vacuum could cause loss of contents on opening the tube). The final length should be about 19 cm.r\TOiWnber, 19681 ALUMINA BY ATOMIC-ABSORPTION SPECTROSCOPY 703 The tubes should be identified by lightly marking with a diamond and placed in the inner container of the pressure vessel.Add sufficient solid carbon dioxide to equalise the pressure at 270" C (weights calculated from a formula given elsewheres). Screw down the pressure block and tighten the release valve, and then the compression bolts. Heat the vessel in a hot-air oven at 270" C and leave overnight (about 15 hours). After this period remove from the oven and cool the vessel to room temperature; allow the carbon dioxide to escape by cautiously opening the release valve. Loosen the com- pression bolts and unscrew the head. Remove the tubes, ensure that they are all at room temperature and brcak open at a scratch made neaw the top. Transfer the solution into a 25-ml beaker and dissolve any crystals in a little water.The solutions obtained tail be treated by the analytical methods described below. (2) C:1L18RL4T10X CURVE FOR COBALT IN ALUMINIUM SALTS AND DETERMINATION OF COBALT I N SOLID ALUMINA I N THE AIR-PROPANE FLAME- Calibration curve-Pipette 8 ml of the potassium aluminium sulphate solution into each of a series of 100-ml beakers and add 0.5 to 9-0-ml aliquots of standard 2 p g per ml cobalt solutioil. Add 5 nil of 4 M sodium hydroxide solution and 1.0 ml of 0.01 M potassium per- mangaiiate solution to each. Bubble sulphur dioxide through the solutions until the purple colour has been discharged and all the permanganate has been converted into manganese dioxide. Filter immediately on pulp pads and wash the pads with about 20ml of distilled water and allow to drain.Place a clean 100-ml beaker under each funnel and dissolve the manganese dioxide on each pad with about 2 ml of concentrated hydrochloric acid. Wash the pads with about 20 rnl of water and add a few crystals of hydroxylammonium chloride (about 10mg) to each beaker in order to ensure complete reduction of manganese to the bivalent state. Render the solutions nearly neutral (about pH 6) by the addition of solid sodium acetate trihydrate, by using test-papers or a pH meter. Transfer the solutions to 100-rnl separating funnels via a 25-ml measuring cylinder. Make each solution (with washings) up to 25 ml. Pipette 2.0 ml of 5 per cent. 8-hydroxyquinoline reagent into each funnel and add 10 ml of isobutyl methyl ketone. Shake the funnels vigorously for 1 minute in order to effect the extraction, discard the aqueous layer in each funnel and spray the organic extracts into the air - propane flame.Measure the absorbance values uewus the pure solvent with the instrumental settings given above. The calibration curve is linear from 1 to 10 pg of cobalt. Alumina in the form of 0.1-g samples can be treated in the same way after dissolution in the pressure vessel, cf. (1). (3) CALIBRATION CCRVES FOR COBALT IN ALUMINIUM SALTS AND DETERMINATION OF COBALT Transfer by pipette 11.0 ml of the aluminiuni chloride solution into each of a series of 25-ml graduated flasks and add 0.5 to 10.0-ml aliquots of standard 100 pg per ml cobalt solution. Adjust the voluine to 25ml. These solutions contain the equivalent of a 0.1-g samplc of alumina per 2-5 ml of solution.The solutions can be sprayed directly into the air - acetylene and nitrous oxide - acetylene flames. The absorbance is measured against a blank that contains no added cobalt. Alumina samples can be determined in a similar manner, following dissolution in the pressure vessel, by adjusting the volume of the hydrochloric acid solution to 2.5ml with distilled water followed by direct spraying. ANALYSIS WITH AN AIR - PROPANE FLAME- In the exploratory experiments cobalt was determined by measurement of its atomic absorption in an air - propane flame at 240.73 nm in the presence of increasing amounts of acldecl potassium aluminium sulphate. It soon became apparent that, at the temperature in the mantle of the air - propane flame, aluminium oxide clotlets, formed by degradation of the aluminium salt in the strongly oxidising primary cone, were not sufiiciently dissociated to release the cobalt atoms into the flame.The addition of releasing agents such as strontium and lanthanum salts was of little avail, and for small amounts of cobalt in a medium corre- sponding tQ that which would be obtained on dissolution of 0.1 g of alumina no reproducible I N SOLID ALL'MINA IK AIR - ACETYLENE AND NITROUS OXIDE - ACETYLENE FLAMES-704 FLEET, LIBERTY AND WEST: DETERMINATION OF COBALT IN [Analyst, Vol. 93 signals could be obtained. It was, therefore, found necessary to separate the cobalt from the matrix element by using a procedure similar to that devised by Marshall and Wests €or traces of nickel and iron in alumina. As atomic-absorption measurements provide a chenlically specific analytical procedure selectivity of extraction is unnecessa-y, except that the traces of cobalt must be separated from the bulk of the matrix element, aluminium.For this reason, 8-hydroxyquinoline was selected as the ligand and isobutyl methyl ketone(hexone) as the solvent. Cobalt(II), having an electrovalence of two and a co-ordination number of six, forms a co-ordination unsaturated Co.Ox, complex and it is, therefore, necessary to select a solvent that will displace the two co-ordinated water molecules remaining in the complex by two oxygen-donating solvent molecules. A systematic examination of various solvents showed that hexone offered optima conditions, both for ease of quantitative extraction ar,d efficiency of nebulisation and at omisat ion.Quantitative extraction of the cobalt could only be obtained at pH 6 or greater than 6 in which region aluminium was also sufficiently well extracted to cause serious consumption of reagent. It was also found impossible to keep such large amounts of aluminium 8-hydroxy quinolinate in solution. Furthermore, the amounts of aluminium remaining in the extract interfered with the atomic-absorption measurements. The masking of such large amounts of aluminium to permit extractive concentration of the cobalt(I1) did not prove feasible and the alternative of extracting aluminium to leave the cobalt behind in solution did not appear to be attractive or practicable. Consequently a collection technique was used to pre concentrate and separate the cobalt from the aluminium.A survey of the over-all experi- mental design showed that hydrated manganese dioxide would be an excellent collector in view of the subsequent use of 8-hydroxyquinoline as an extractant. Although manganese dioxide has been used previously for iron and nickels we could find no previous record of the collection of cobalt(I1) hydroxide in this way.l0 The manganese dioxide was generated iic sitz by the action of sulphur dioxide on permanganate in a strongly alkaline solution. At this pH the aluminium is held in solution quantitatively as sodium aluminate. I t was subse quently proved that from 0.5 to 50 pg of cobalt could be collected quantitatively on 1 mg of manganese dioxide in this way from 80 ml of solution containing the equivalent of 0.1 g of alumina.The hydrated manganese dioxide is easily filtered on a pulp pad and, after washing briefly, can be dissolved in a small amount of concentrated hydrochloric acid washed through the pad. The addition of a small amount of hydroxylammonium chloride to reduce man ganese( IV) is beneficial. Direct aspiration of the dissolved manganese dioxide sample into the air - propane flame yielded cobalt signals, but at a much lower sensitivity than that which could be obtained by subsequent extraction with 8-hydroxyquinoline in hexone at yH 6 Manganese (11) is not extracted at this pH and consequently does not interfere in any way with the determination of cobalt. Aluminium does not pass through the collection extraction procedures in detectable amounts.Interference studies were carried out on range of other metals likely to be present in nearly pure alumina or aluminium salts. This was done by adding a 100-fold excess of Ca2+, Cr3+, Fe3+, Mg2+ or Ni2+ to 5 pg of Co2t in solution containing the equivalent of 0.1 g of Al,O,. All these measurements were made against a blank carried out on the aluminium solution without addition of cobalt or the other ions. In all cases these blanks themselves measured against pure solvent revealed only negligible amounts of cobalt present in the aluminium salt. In the air - propane flame the calibration curve for pure aqueous solutions was a neat straight line through the origin giving absorbance readings of 0.05 and 0.425 for 1 and 10 p.p.m of cobalt, respectively, in a 10-cm path length. Following extraction into hexone - 8-hydroxy quinoline, a near linear calibration curve was obtained with absorbances of 0.012 and 0.12 corresponding to 0.1 p.p.m.and 1.2 p.p.m. of Co2+ in the organic phase. The limit of detection (0-005 absorbance unit) for the entire analytical procedure (collection in the presence of alumina, extraction and measurement) was 0.04pg of cobalt per ml of hexone and the analytical range was 0.1 to 1.Opg of cobalt per ml of hexone (equivalent to 10 to 1OOpg of cobalt per g of Al,O,). At the level of 0-5 pg of cobalt per ml of hexone the relative standard deviation of the entire analytical procedure was found to be 2-13 per cent. The time required for a complete analysis following dissolution of the sample was about 1 hour.No interference was found.November, 19681 ALUMINA BY ATOMIC-ABSORPTION SPECTROSCOPY 705 As the air - propane flame was insufficiently hot to dissociate thealuminaclotlets efficiently, attention was turned to the air - acetylene flame as an atom reservoir for cobalt in the presence of large amounts of aluminium. The oxy-acetylene flame might be still more favourable, but its high burning velocity makes it unsuitable for support at a long-slot burner of the atomic-absorption type. However, the air - acetylene flame has been used in its oxidising mode for the determination of cobalt and in its fuel-rich (luminous) form for aluminium. Experiments were, therefore, made on a solution containing aluminium chloride equivalent to 0.1 g of A1,0, per 2.5 ml of solution and varying amounts of cobalt, on a 10-cm long, air - acetylene slot burner.Satisfactory results were obtained in this way, but at a sensitivity approximately only one half of that obtained in the air - acetylene flame in the absence of aluminium. There is little doubt that this is partly caused by the increased viscosity of the solutions containing high concentrations of dissolved salts, because there was a 20 per cent. decrease in the rate of uptake of the solution. A further factor may arise from the formation within the nebulising chamber of larger droplets with a subsequent smaller specific surface in the flame. However, the fact that the best cobalt signals could only be obtained for aluminium-coiit aining solutions in a luminous reducing flame is also significant.Experi- ments showed that these are the conditions that most favour the atomisation of aluminium rather than cobalt. The best signals for pure cobalt solutions were obtained in a leax air - acetylene flame. Obviously minimisation of the partial pressure of atomic oxygen in the flame is necessary to induce increased thermal dissociation of the A1,0, clotlets. These conditions are, however, not too favourable for efficient atomisation of cobalt. This balance of flame conhtions is revealed in the criticality of air - acetylene ratios and of the position of measurement in the flame (about 1.3 cm above the burner head), as shown in Fig. 1 . ANALYSIS WITH AN AIR - ACETYLENE FLA4ME- 0 L A 5 10 I5 20 25 30 Height, mm Fig.1. Graph of absorbance against height of measurement above burner head for cobalt in aluminium chloride solutions, with an air - acetyl- ene flame Best results were obtained at maximum lamp current and minimum slit width (see Fig. 2). No interference was observed from 50-fold excesses of the same ions used in the air - propane flame work. In this instance, in the presence of 0.1 g of Al,O, the limit of detection 0.005 absorbance unit) was 0.22 pg of cobalt per ml, with an analytical range of 2 to 10 pg per ml (equivalent to 50 to 250 pg of cobalt per g of A1,0,). At the level of 4 pg of cobalt per ml the relative standard deviation of the method was 1.4 per cent. The time required for a complete analysis following dissolution of the sample was only a few minutes. ANALYSIS WITH A NITROUS OXIDE - ACETYLENE FLAME- Recent workll has shown that the temperature of the type of nitrous oxide - acetylene 50" C.This compares with a value flame used for atomic-absorption measurements is 2780"706 0.4 (3[ i 4 0 2 /- 0.1 ! I I I I I I (4 - - - I I I L 0 50 100 150 200 of about 2300" C for the air - acetylene flame. This flame was, therefore, investigated in a similar manner. In this instance aqueous solutions of cobalt gave only one quarter of the signal obtained in the air - acetylene flame. This is only partly accounted for by the shorter (5 cm) path length of the flame. of the flame is not a decisive factor in dissociating the A1,0,. This is borne out by the lower atomic-absorption signals that are obtained in a stoicheiometric (and therefore hotter) nitrous oxide - acetylene flame, as opposed to a fuel-rich flame.It would appear, therefore, that the higher temperatur Cobalt, jig Fig. 3. Calibration curve a t 240.7 nm for cobalt in aluminium solutions in the nitrous oxide - acetylene flame. Net absorbance against pg of cobalt per 0.1 g of alumina It was felt that the higher temperature of the nitrous oxide flame might cause loss of atomic cobalt by ionisation, and this was borne out to a certain extent by the increase in signal (about 30 per cent.) obtained by addition of an excess of potassium ions. Because of the high ionisation potential of cobalt, 7.8 eV, it is not clear, however, if this effect is cause entirely by suppression of ionisation. In the absence of aluminium, cobalt yields a highe signal in a stoicheiometric nitrous oxide - acetylene flame than in a fuel-rich flame.It is apparent, therefore, that the strongly reducing cyanogen zone of the atomic-absorption type of nitrous oxide - acetylene flame has an adverse effect on the atomisation of cobalt, but in beneficial to a limited extent in dissociating alumina clotlets. Because of both of theseNovember, 19681 ALUMINA BY ATOMIC-ABSORPTION SPECTROSCOPY 707 effects, however, this flame is inferior to the air - acetylene flame as an atom reservoir for cobalt. In the slightly fuel-rich (red feather) nitrous oxide - acetylene flame, in the presence of 0.1 g of Al,O,, a detection limit (0.005 absorbance unit) of 1 pg of cobalt per ml was obtained, with an analytical range of 2 to 20 pg of cobalt per ml (equivalent to 50 to 500 p g of cobalt per g of Al,O,).A relative standard deviation of i 3.0 per cent. was obtained at the Fig. 3 shows the nature of the analytical curve obtained for cobalt in the presence of 0.1 g of alumina. DISCUSSION AND CONCLUSIONS pg of cobalt per ml level. Two analytical procedures have been developed for the determination of trace amounts of cobalt in alumina or aluminium salts. The collection - extraction procedure offers a ensitive but more time-consuming procedure that can be applied with simple equipment , while the direct procedure, in which an air - acetylene or nitrous oxide - acetylene flame is used, offers manipulative simplicity and rapidity. In order to determine cobalt in the presence of aluminium the use of hot flames is necessary to produce dissociation of alumina clotlets.Despite its unfavourable effect on the determination of cobalt (alone) it is best to use a fuel-rich flame in order to establish conditions where a compromise is struck between reasonably effective liberation of cobalt from the matrix and the unfavourable effect of a reducing medium on maintaining a population of cobalt atoms. The nitrous oxide - acetylene flame produces considerable ionisation of cobalt, and its strongly reducing cyanogen zone, which appears in the hottest part of the mantle above the primary zone, has a strongly adverse effect on atomic-absorption measurements of cobalt atoms. TABLE I1 DETERMINATION OF KNOWN AMOUNTS OF COBALT I N THE PRESENCE OF A1,UMINIUM (= 0.1 g Of Al,O,) BY ATOMIC-A4BSORPTION PROCEDURES Flame medium All amounts are stated in pg 7- A I Air - propane* Air - acctylene Nitrous oxide - acetylcne * & Taken Found Taken Found Taken Found 3.0 3.2 9.0 9.2 7-5 7.8 3.0 3-0 9.0 8-9 7.5 7.5 3.0 3.1 9.0 9.0 7.5 8.0 4.9 5.0 20.0 20.0 30.0 30.5 4.9 5.1 20.0 20.1 30.0 28.5 4.9 4.9 20.0 20.5 30.0 28.5 4.9 4.8 20.0 19.8 30.0 29.0 4.9 4.7 47.5 46-5 47.5 47.5 4.9 4.9 47-5 47-0 47.5 48.5 7.5 7.6 47.5 47.0 47.5 48.7 7-5 7.8 - - 47.5 48.0 7.5 7.5 7.5 7-4 - - - - - - - - * The following extractive separation was as given under Experimental. TABLE I11 DETERMINATION OF APPROXIMATELY KNOWN AMOUNTS OF COBALT, pg, IN SAMPLES BY A SEMI-QUANTITATIVE METHOD WITH A RELATIVE PRECISION OF k50 PER CENT. OF A1,O3, AND COMPARISON WITH SPECTROGRAPHIC RESULTS THAT WERE OBTAINED Comparative Alumina Analytical Cobalt found in spectrographic sample No.method 0.1 g of Al,O,, pg result 1 Air - acetylene 12.7 10.0 1 Nitrous oxide - acetylene 13.5 10.0 2 Air - propane 1.0 0.5 3 Air - propane None detected < 0.5708 FLEET, LIBERTY AND WEST The analytical range for both the acetylene flames was found to be closely similar. The slope of the calibration curve in the air-supported flame was considerably steeper. Although the method described here is devised for trace amounts of cobalt in alumina it is equally applicable to aluminium salts, as shown in Tables XI and 111. We thank A.W.R.E., Aldermaston, for financial support in aid of this work, and for the loan of the pressure vessel used to digest the samples of alumina. We also thank the Courtauld Research Foundation for the award of a grant for the purchase of the Techtron AA. flame spectrophot ometer. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. REFERENCES Allan, J. E., Natwe, 1960, 187, 110. Gatehouse, B. M., and Willis, J. B., Spectrochim. Acta., 1961, 17, 710. Robinson, J. W., Analyt. Chtem., 1961, 33, 1067. Allan, J. E., Spectrochim. Acta, 1962, 18, 259. hEcPherson, G. L., Price, J. W., and Scaife, P. H., Nature, 1963, 199, 371. Elwell, W. T., and Gidley, J. A. F., “Atomic-Absorption Spectrophotornetry,” Second Edition Pergamon Press, Oxford, London, Edinburgh, New York, Toronto, Sydney, Paris and Braun schweig, 1966, p. 88. Elwell, U‘. T., and Scholes, I. R., “Analysis of Copper and its Alloys,” Pergamon Press, Oxford Lolidon, Edinburgh, New York, Toronto, Sydney, Paris and Braunschweig, 1966. Gordon, C. L., Schlecht, TV. G., and Wichers, E., J . Res. Nutn. Bur. Stand., 1944, 33, 457. Marshall, G., and West, T. S., Talanta, 1967, 14, 823. West, T. S., Analytica Chim,. Acta, 1961, 25, 405. Kirkbright, G. F., Peters, M. K., Sargent, M., and West, T. S., Tabaizta, 1968, 15, 663. Received November 2 lst, 1066

ISSN:0003-2654    

DOI:10.1039/AN9689300701

出版商:RSC    

年代:1968

数据来源: RSC

摘要:

Aizalyst, November, 1968, Vol. 93, pp. 709-714 709 The Determination of Silicon by Atomic-absorption Spectrophotometr y, with Particular Reference to Steel, Cast Iron, Aluminium Alloys and Cement B Y W. J. PRICE AND J. T. H. ROOS (Pye Unicam Ltd., York Street, Cambridge) The determination of silicon by atomic-absorption spectrophotometry in silicon-containing and siliceous materials has been investigated. Means are outlined by which many types of sample can be completely dissolved without loss of silicon, thus avoiding the need for fusion. Full operating details are given, and the results, accuracy and sensitivity of the method are discussed. INTRODUCTION of the nitrous oxide - acetylene flame1 s 2 made possible the determination , by atomic-absorption spectrophotometry, of about twenty-five more elements than had previously been possible when only air - propane and air - acetylene gas mixtures were available.These elements include aluminium, barium, silicon, titanium and vanadium. Bowman and Willis3 investigated the application of the nitrous oxide - acetylene flame, and have used it for the determination of, amongst others, vanadium in steel and fuel oil, and aluminium, silicon and titanium in bauxite. These authors also discussed problems of ionisation at the high temperature reached by this flame, investigated certain interferences and give results for the analysis of standard samples of the materials mentioned. Two other papers discuss the determination of silicon by atomic-absorption spectro- hotometry. In one method4 the authors precipitate silica (from cement samples) and collect the precipitate, which is then fused with sodium carbonate.The melt is dissolved in water, and silicon determined in the resulting solution. However, the introduction of a fusion tech- nique clearly detrzcts from the convenience of the atomic-absorption method. In the other method5 silicon was determined in steels by using chemically analysed (N.B.S.) samples as standards. It was also noted that with the preparation procedure given, the samples and standards were stable for only 1 or 2 hours. The present paper describes the direct determination of silicon, by atomic absorption, a steel, cast iron, aluminium alloys and cement, without the need either for precipitation of the silica or for an alkaline fusion technique. An independent calibration is prepared from pure chemicals.EXPERIMENTAL APPARATUS AND REAGENTS- The work was performed with a Unicam SP90 atomic absorption spectrophotometer tted with an SP91 multiple lamp turret and SP94 nitrous oxide system, and used in con- function with an SP93 air compressor unit. A high spectral output silicon lamp was used with a lamp holder specially designed to facilitate lining up the smaller cathode characteristic of this type of lamp. Scale expansion up to three times was used when necessary for the determination of low concentrations of silicon in steels. Silicon stock solution, about 2000 p.$.vz. of silicon-This was prepared from pure sodium silicate and standardised gravimetrically. Alternatively, it may be prepared from pure silica dried by heating to 500" C, the correct weight taken, fused in the minimum amount of sodium carbonate, and made up to volume in water.0 SAC and the authors.710 PRICE AND ROOS: THE DETERMINATION OF SILICON BY [Analyst, Vol. 93 Vanadium stock SoZution, 2-5 per cent.-A 12-5-g sample of pure vanadium trichloride was dissolved in about 200ml of water and 10ml of hydrochloric acid (spgr. 1.16); the solution was then filtered and diluted to 500 ml. Iron stock solution, 5 per cent. Fe3+-A 50-g sample of high purity iron (B.C.S. 260/2) was weighed into a 1-litre beaker, 200 ml of water, together with 500 ml of hydrochloric acid were added, and the beaker heated on a hot-plate. About 100ml of hydrogen peroxide (50 vol.) were added carefully in small portions.After the reaction had stopped, the solution was boiled for a few minutes to decompose remaining traces of hydrogen peroxide and after cooling, the solution diluted to 1 litre in a calibrated flask. AZumi&m stock solution, 2 per cent.-High purity silicon-free aluminium foil (e.g Specpure or similar grade), 20 g, was dissolved in 200 ml of hydrochloric acid and diluted to 1 litre in a calibrated flask. Stock solutions of calcium, phosphorus and sodium (5000 p.p.m.) were also prepared All solutions except aluminium (4.v.) were made with analytical-reagent grade chemicals and de-ionised water, and stored in polythene bottles. All acids used were of analytical-reagent grade, stock concentrated. INVESTIGATION O F OPTIMUM INSTRUMENTAL CONDITIONS- As silicon is one of the less sensitive elements that can be determined by atomic-absorption spectrophotometry, it was considered necessary to establish the instrumental condition giving the greatest sensitivity for silicon.Consequently, “optimum conditions” refer here to those instrument settings that result in the greatest silicon sensitivity. Two solutions were prepared containing 200 and 400 p.p.m. of silicon. With a mono chromator slit of 0.1 mm the absorption of each solution at 251-6 nm was measured a different acetylene flow-rates between 3.5 and 5.0 litres per minute and different height between 0.5 and 2.0cm of the light path above the burner top (“observation height’’ The nitrous oxide flow-rate was fixed at 5.0 litres per minute. The best sensitivity was obtained with an acetylene flow-rate of 4-2 litres per minute and observation height of 1.0 on The effect of slit widths between 0.05 mm and 0.20 mm was also investigated when using the conditions of best sensitivity.As expected, best sensitivity was obtained with slit width between 0.05 mm and 0.1 mm, and sensitivity fell off slightly with wider slits. Under the chosen conditions, summarised in Table I, a sensitivity (i.e., that concentration of the element, in p.p.m., causing an absorption of 1 per cent.) of 8 to 10 p.p.m. was obtained for silicon, and a reproducibility at the 400 p.p.m. level of +3 p.p.m. proved also to give the highest signal-to-noise ratio. These condition TABLE I RECOMMENDED CONDITIONS FOR THE DETERMINATION OF SILICON Wavelength .. .. .... 251.6nm Slit width . . . . . . . . 0.1 mm (equivalent to a spectral band width of 0-6 nm) Observation height . . .. . . 1.0 to 1.3 CII~* Nitrous oxide flow-rate . . . . 5.0 litres per minute Acetylene flow-rate . . .. . . 4.0 to 4.5 litres per minute* * These were found to vary slightly from instrument to instrument. EFFECT OF FOREIGN IONS- To investigate the effect of other ions on the silicon absorption, several series of solution were prepared containing 200 p.p.m. of silicon and different concentrations (0 to 3000 p.p.mn of aluminium, calcium, iron, phosphorus and sodium. Of these elements, aluminium, calcium iron and sodium enhanced the silicon absorption ; phosphate ion exhibited a suppressive effect From the results (Fig. 1) it can be seen that the enhancement effect of the interfering ions do not increase linearly as the concentration of ion increases, but levels off if a sufficient amount of interfering ion is present.If, therefore, a sufficient amount of interfering ion is added both to the sample solutions and to the standards, the interference effect will be compensated for. similarly to give the same over-all enhancement. aluminium and iron; in the presence of at least 3000 p.p.m. of aluminium, vanadium or iron Furthermore, the combined enhancing effect of two or more interfering ions flattens on Vanadium added as vanadium trichloride was found to have an effect similar to thatNovember, 3 9681 ATOMIC-ABSORPTION SPECTROPHOTOMETRY 711 0.06, I I I 0 lo00 2000 Concentration of added elements, pg per ml Fig.1. absorption. nil Effect of interfering ions on silicon Silicon concentration is 200 pg per The addition of calcium, iron, magnesium, aluminium, manganese, zinc cn ;opper up to 000 p.p.m. was found to have no further effect on the silicon absorption (kigs. 2 and 3). Moderate amounts of sodium (less than 500 p.p,m.) were found to cause slight enhancement of silicon, even in the presence of aluminium or vanadium. 8 rd n Silicon concentration, p g per mI Silicon concentration, Ng per ml Fig. 2. Effect of added aluminium (1000 Fig. 3. Effect of added vanadium (5000 p.p.m. of vanadium trichloride) on silicon absorp- tion: a, silicon plus vanadium with and without added ions (as stated in text) ; b, silicon alone g per ml) on silicon absorption: a, silicon plus aluminium with and without added interfering ons; b, silicon alone AMPLE PREPARATION- The separation of silicon as silica and subsequent fusion of the precipitate with, e.g., sodium hydroxide, are superfluous steps if the silicon can be brought directly into solution.larke6 has described a rapid method for the dissolution of steels and cast irons that does not involve precipitation of silica; hydrochloric acid attack, followed by treatment with hydrogen peroxide, results in the complete dissolution (except for carbon) of samples containing less than about 1 per cent. of silicon. The possibility of using hydrofluoric acid to dissolve samples containing more than 1 per cent. of silicon was investigated. Langmyhr and Graff7 found that hydrofluoric acid was suitable reagent for the dissolution of siliceous material, and showed that no silicon was lost by volatilisation as silicon tetrafluoride when such material was dissolved in an excess of hydrofluoric acid, provided that the solutions were not heated.This was confirmed during712 PRICE AND ROOS: THE DETERMINATION OF SILICON BY [Analyst, Vol. 93 the present work, and was also found to hold true for silicon dissolved in mixtures of hydro- fluoric and hydrochloric acids. However, as expected, the hydrofluoric acid was found t o attack glassware with the dissolution of significant amounts of silicon (Fig. 4) ; polythene apparatus was , therefore, used whenever hydrofluoric acid was present in the soh tion. Time, minutes Fig. 4. Dissolution of silicon glassware by 1 ml of hydrofluoric acid plus 6ml of hydrochloric acid diluted to 100 ml Because of the slight enhancement by sodium of the silicon absorption, and becaus'e silicon calibration graphs were prepared from sodium silicate solutions, it was necessary to add sodium to sample solutions containing more than 100 p.p.m.of silicon (corresponding to about 200 p.p.m. of sodium in the sodium silicate solutions). The following dissolution procedures were eventually adopted. CEMENT- Weigh out 0.4 g of the sample into a polythene beaker and add 10 ml of water. Stir the mixture with a polythene rod and add 6 ml of hydrochloric acid (spgr. 1-16> and 0.5 m of hydrofluoric acid (40 per cent. w/w) (addition of more hydrofluoric acid may result in the precipitation of calcium fluoride).When dissolution is complete, add 20 ml of vanadium solution and 10 ml of sodium stock solution, and dilute to 100 ml. Prepare a silicon calibration graph (0 to 600 p.p.m.) by diluting suitable volumes of silicon stock solution, 20 ml of vanadium stock solution and 6 ml of hydrochloric acid to 100 ml. CAST IRON- Weigh out 1 g of sample into a polythene beaker and add 8 ml and 2 ml of hydrochlorid acid and nitric acid (sp.gr. 1.42) , respectively. Cover the beaker with a clock-glass and head on an asbestos mat on a hot-plate until the reaction has stopped. Cool, wash the clock-glass and add 2 ml of hydrofluoric acid. Filter the solution if necessary, add 5 ml of sodium stock solution and dilute to 100 ml. Prepare the calibration graph for silicon (0 to 400 ?.p.m.of silicon) by diluting suitable volumes of silicon stock solution and 20 ml of iron solution to 100 ml. It is unnecessa-y to add vanadium as the large excess of iron present in both standard solutions and sample will overcome any possible interference from other constituents of th sample. STEEL- Weigh out 1 g of the material into a glass beaker and add 5 to 10 ml of water followed by 10 ml of hydrochloric acid. Heat the solution and add, carefully and in small portions 15 ml of hydrogen peroxide (5.0 vol). Boil the solution, allow to cool and filter into a 100-m calibrated flask. The calibration graph (0 to 100 p.p.m. o silicon) is prepared as for cast iron. Alternatively, steel samples can be dissolved in the same way as samples of cast iron but omitting the addition of sodium solution. Dilute to the mark with water.November, 19681 ATOMIC-ABSOKPTION SPECTROPHOTOhIETRY 71 3 ALUZ~~IKIUAZ ALLOYS- To 0.5 g of the sample in a polythene beaker add 10 ml of water and, in small portions, a total of 8 ml of hydrochloric acid, Cover the beaker with a clock-glass, and allow the reaction to subside. Heat on an asbestos mat on a hot-plate until all reaction ceases.Cool and add 2 ml of hydrofluoric acid. (For alloys containing more than 2 per cent. of silicon, add 5 ml of sodium stock solution before finally making the solution up to volume.) Prepare a silicon calibration graph (0 to 250 p.p.m. of silicon) as for cast iron, but add 25ml of aluminium stock solution in place of the iron. Addition of vanadium is rendered unnecessary by the presence of aluminium in both standard solutions and samples.In every experiment the standard containing no added silicon was run as the reagent blank. Add 15 ml of hydrogen peroxide (50 1701.) in small portions. Filter the solution into a polythene flask and dilute to 100 ml with water. RESULTS The results obtained for the determination of silicon in several standard samples are summarised in Table 11. When analyses were performed in duplicate, the results are bracketed together in the table. The coefficient of variation at the 0.5 per cent. silicon level (in steel) was found to be 0.012 per cent., or 2.4 per cent. of the mean. TABLE I1 COMPARISON OF RESULTS FOR STANDARDISED MATERIALS Silicon, per cent. Sample No. N.B.S. 1013 N.B.S. 1015 I3.C.S. 206/2 B.C.S.170/3 Private samplc H.C.S. 241/1 B.C.S. 251/1 U.C.S. 253/1 B.C.S. 336 B.C.S. 339 B.C.S. 33.5 K.C.S. 151/1 K.C.S. ,016,/1 H.C.S. 268 Type Nominal composition Cement CaO 64% ; Fe,03 3%;A1,03 3% Cement CaO 61%; Al,O, 5%;MgO 4%, Cast iron P 1.37% ; Mn 0.32% Cast iron Mn 0.76% ; P 0.71 ; CU 0.600,; Cast iron High-speed steel Low alloy steel Low alloy steel Stainless steel Chrome - vanadium Cr 12,40,/, Stainless steel Cr 18.6% ; 9.476 Aluminium alloy Aluminium alloy =\luminiuIn alloy Fe203 3% - W 19.6% ; Co 5.7% ; Cr 50,L Mn 1.5% ; Mo 1.6% Ni 1%; Cr 17; Cr 17-60,/, ; Ni 9-57; ; Mo 2.4% steel Cu 4% ; Mg 1.4% ; Ni 2% Cu 474 ; Mg 0.7476 ; Mn 0.73% Cu 1.3476 ; Mg 0.56% r Standard analysis SiO, 24.1 7; SiO, 20.67; 3.42 2.47 2.14 0.33 0.41 0-65 0.5 1 0.36 0.82 0.38 0-74 2.43 1 Atomic absorption SiO, 23.5% SiO, 23.7% SiO, 20.07b SiO, 19.9% 3.43 2.53 0.30 0.39 0.65 0.5 1 0.53 0.36 0.83 0.72 2.46 DISCUSSION Enhancement of silicon absorption by elements such as aluminium, calcium, sodium and vanadium can be explained in terms of suppression of the ionisation of silicon in the hot nitrous oxide flame (about 3000" C) by the added element.Evidence for ionisation in this flame exists for magnesium8 and copper, both of which have ionisation energies just slightly lower than that of siiicon. If this assumption is correct, it would appear that silicon is ionised t o the extent of 10 to 15 per cent. No satisfactory explanation for the suppression of silicon absorption by phosphate is as yet forthcoming. The presence of phosphorus in steel and in cast iron has a negligible influence on the silicon response, presumably because of the complexing effect of iron itself.The magnitude of the effect of phosphorus on silicon in the ahsence of iron can be interpolated from Fig. 1, and an example of the determination of714 PRICE AND ROOS silicon in cast iron in the presence of a comparatively high amount of phosphorus is included in Table 11. The majority of replicate results agree to within 3-2 per cent. of the certificate value. When silicon is present as a minor constituent, as for instance in steel where it is usually present to an extent less than 1 per cent., a coefficient of variation of 2.4 per cent. of content was found; at higher silicon levels the precision improves significantly to give acoefficient of variation better than 1 per cent.The lowest concentration level at which silicon c m be determined with acceptable accuracy depends mainly on sample dilution and instrumental conditions. As interfering effects can be largely overcome, the limits both of detection and determination are not influenced by the matrix elements. With a silicoii sensitivity of 8 p.p.m. and a detection limit in aqueous solution of 3 p.p.m., concentrations of about 30 p.p.m. can be measured with a 10 per cent. coefficient of variation. This represents 0.03 per cent. of silicon if a 1-g sample is made up to 100 ml of solution. This is probably the practical limit of determination for silicon in steel or aluminium by the methods described, though clearly it can be improved by replication; lower concentrations can be determined with a greater proportional error.The presence of silicon has hitherto caused difficulty in analytical methods, both because it has been thought to be difficult to bring completely into solution and also because of the interference effects it has caused. With the knowledge that silicon can be retained completely in solution, and not lost either by precipitation as silica or volatilisation as silicon tetrafluoride, a new approach can now be made to the complete analysis of many different materials by atomic-absorption spectrophotometry. It should first be decided whether or not silicon is to be determined by atomic absorption, and if the silicon content is less than 20 per cent. it is reasonable to expect that it can.For silicon levels between 5 and 20 per cent. a small sample weight can be taken to e- lisure dissolution o f silicon and to enable all or most major elements, including silicon, to be determined in this one solution. A second, more concentrated, solution, with silicon removed as silicon dioxide, may have to be prepared for the determination of minor elements. If silicon is present at a concentration of less than 5 per cent., larger amounts of sample (up to 1 g per 100 ml) should be used. Minor elements are determined in the same solution, and major elements after suitable dilutions have been made. The presence of hydrofluoric acid necessitates the use of beakers and flasks constructed from material other than glass. Polythene beakers are quite satisfactory, provided strong heating is not required. Hard polythene measuring cylinders, which have been calibrated prior to use, enable volumes of 100ml to be measured with an accuracy of better than 0-5 per cent. REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. Willis, J. B., Nature, 1965,207, 715. Amos, M. D., and Willis, J. B., Spectrochim. Acta, 1966, 22, 1325. Bowman, J. A., and Willis, J. E., Analytica Chim. Acta, 1067, 39, 1210. Capacho-Delgado, L., and Manning, D. C., Analyst, 1967, 92, 553. McAulifYe, J. J., Atomic Absorfition Nfysletter, 1967, 6, 69. “Chemical Analysis for Iron Foundries, B.C.I.R.A. Methods of Analysis Sub-committee (Co-ord., W. E. Clarke), Allen St Unwin, London, 1967, p. 121. Langmyhr, F. J., and Graff, P. R., Analytica Chim. Acta, 1959, 21, 334. Elwell, W. T., and Gidley, J. A. F., “Atomic-Absorption Spectrophotometry,” Second Edition, Pergamon Press, Oxford, London, Edinburgh, New York, Toronto, Sydney, Pans and Braun- schweig, 1966. Received April 8th, 1968

ISSN:0003-2654    

DOI:10.1039/AN9689300709

出版商:RSC    

年代:1968

数据来源: RSC

摘要:

Analysf, November, 1968, Vol. 93, $9. 715-719 716 Determination of Sodium-22 in Rain Water by Using a Low Level p-Counter (Ministry of Technology, BY B. A. BURDEN London, S.E. 1) Laboratory of the Government Chemist, Cornwall House, Stamford Street, A method has been devcloped for the separation froin rain water of sodium-22, sufficiently radiochemically pure for it to be measured by low level p-counting. A 50-litre sample of rain water is evaporated to about 1 litre and the concentrate passed through a cation-exchange column. The sodium is eluted with 0.7 N hydrochloric acid solution. The sodium fraction is then purified, precipitated with a-methoxyphenylacetic acid, converted into sodium chloride and counted in a low background anti-coincidence f-counter. SCHEMES of analysis have been worked out and used for the determination of most of the radionuclides present in rain and tap waters.They include fission nuclides, natural activities resultiiig from the uranium and thorium series and manganese-54 which, although not a fission nuclide, has been present in fall-out since the Russian tests of 1961. Sodium-22 and sodium-24 are formed in small amounts in the upper atmosphere by cosmic-ray spallation of argon.1 Sodium-22 is also formed by the fast-neutron reaction 23Na(n,2n)22Na in thermo- nuclear explosions. Sodium-22 was first detected in rain in 1957 (0-007 pCi per litre),2 but a sharp increase was reported in 1963 following the first thermonuclear tests in the atm~sphere.~ During examinations of fall-out contamination in aircraft engine washings in this laboratory, in 1963, sodium-22 was suspected, although its presence was not later substantiated.Sub- stantial amounts of sodium-22 had, however, been reported in the urine of eskimos eating caribou meat4 and in milk.596 The half-life of sodium-22 is 2.6 years, and it decays by positron emission (90 per cent.) and electron capture (10 per cent.) to an excited state of neon-22. De-excitation to the ground state takes place by emission of a 1-28-MeV y-photon. y-Spectrometry can be used to determine sodium-22, but the levels are much too low to be detected by an ordinary y-spectrometer. Perkins and Nielson5 have obtained results by using a 4096 channel multi-parameter memory spectrometer. Positron counting3 is a much more specific way of determining sodium-22, but it was found too expensive to set up for the work that was to be undertaken in this laboratory.EXPERIMENTAL A 50-litre sample of rain water was collected on the roof of Cornwall Xouse, Stamford Street, London, S.E.1, by using the sampling method of Wood, Metson and Richards.7 The sample is reduced in volume by evaporation and the concentrate filtered to remove any insoluble matter. The insoluble matter is treated with hydrofluoric and perchloric acids to remove silica and the residue treated with dilute hydrochloric acid, the solution being combined with the concentrate and made up to 1 litre with distilled water. A small aliquot is removed for the determination of sodium by flame photometry, and the remainder passed through a cation- exchange column, which absorbs all of the cations.The sodium and some of the potassium 0 SAC; Crown Copyright Reserved.716 BURDEN : DETERMINATION OF SODIUM-22 IN RAIN WATER [Analyst, 1‘01. 93 and caesium are eluted from the column in the first fraction, with 0.7 N hydrochloric acid as eluant. The eluate is contaminated with ruthenium, which is removed by two ruthenium oxide scavenges, after which the filtrate is evaporated to dryness with perchloric acid. The perchlorate residue is extracted with ar,hydrous ethyl acetate, which dissolves sodium per- chlorate in preference to caesium and potassium perchlorates. The residue is filtered off and dry hydrogen chloride gas passed through the filtrate to precipitate sodium chloride. The sodium chloride is then dissolved in water and sodium precipitated with a-methoxyphenylacetic acid reagent, which separates the sodium from the remaining traces of potassium and caesium.The acid salt of the organic acid is converted into sodium chloride with butanol- dry hydrogen chloride gas mixture, filtered, mounted and counted in a low background anti-coincidence p-counter. The weight of sodium chloride gives the chemical recovery of the counting source, by using the sodium content of the rain water determined by flame photometry and making allowance for the sodium present in all of the reagents added. Although the method takes about 7 days to complete, the actual practical work after the ion-exchange separation takes 1 day. Because of the time elapsed between the start aid finish, any sodium-24 (half-life 15.4 hours) present initially would have decayed away before the source was counted.METHOD APPARATUS- Low background (about 1 count per minute) anti-coincidence 13-particle counters were used. Ion-exchange colunzns-A column of length 20 cm and diameter 3 cm was prepared with Amberlite resin CE-120 (200 mesh) and converted into the hydrogen iorm with concentrated hydrochloric acid. De-mountable Pevs9ex Jilter-stick, 16-mz internal dianzeteif--Filter-papers, Whatman No. 42, 21-mm diameter, for me with filter-stick. The flow-rate used was 0.8 ml per minute. I?z*frared lamps. Heating ring, 1200 watts. Flask, 10-Zitre capacity. Beakevs aiad centrifuge tubes. Platirtum crzicible. Pi$ et f es . REAGENTS- of radiochemical contamination. acid and dilute with water to give abmt 5 mg of ruthenium per ml of solution.Use analytical-reagent grade reagents whenever possible and check them for the pre$ecce Rzfthe?zium carriev solution-Dissolve ruthenium trichloride in concentrated hydrochloric Hydrojzroric acid sdi.ttion*, 40 per cent., aqueous. Hydrochloric acid, 10, 0.7 and 0.2 N. PerchZorz’c acid sol::tion, 60 Per C Z ~ . aqueous. Bronzifze water--Saturate water with bromine. EtlzaizoE, 95 irjer ceut. aqueous solztion. Eutnnol - 2iydroge:z chloride-Saturate but anol with dry hydrogen chloride gas. E thjd acetate. Tetmmeth_vlarrsmoizi.Ytm hydroxide solution, 25 $er cepJt. aqtwozks. Dry hydrogen chloride gas-Prepare by dropping concentrated hydrochloric acid on 10 concentrated sulphuric acid. cc-n4et~~oxyphe~z-~iZacetic acidg--Dissolve 5-0 g of oc-methoxyphenylacetic acid in 26.4 in1 of absolute ethanol, add 4.1 ml of tetramethylammonium hydroxide solution and make up to 37.7 ml with distilled water.The reagent is 0.8 N with respect to the acid and 0.3 N with respect to the base. The sodium sa!t precipitated is the acid salt [C,H,(CH,O)CHCOONa.C,H,(CH,O)- CHCOOH] containing 6.49 per cent. of sodium. “Gelvn solzdion”-A 5 per cent. w/v methanolic solution of poly(viny1 acetate).November, 19681 BY USING A LOW LEVEL P-COUNTER 7 17 PROCEDURE- Step 1-Evaporate the 50-litre sample of rain water to about 900 ml on a heating ring. Filter the concentrate through a filter-paper (Whatman No. 30). Step 2-Ignite the filter-paper containing the insoluble material in a platinum crucible, allow to cool, and then treat the residue with hydrofluoric and perchloric acid solutions to remove the silica; evaporate to dryness.Step 3-Treat the residue with 10 ml of 10 N hydrochloric acid and 10 ml of water (remove and discard any remaining insoluble matter by filtration), and combine it with the main filtrate from step 1 and make up to 1 litre with water. Stee 4-Dilute 10 ml of this solution to 1 litre for the determination of sodium by flame photometry. Step 5-Pass the remaining 990 ml through the ion-exchange column. Wash the column with 300 ml of 0.1 N hydrochloric acid. Elute the sodium from the column with 300 ml of 0.7 N hydrochloric acid, evaporate the eluate to about 20 ml and add 5 mg of ruthenium carrier, followed by 1 ml of bromine water.Step 6-Add tetramethylammonium hydroxide solution (this is used because a strong base is required for complete precipitation) until the solution is alkaline, then add 2 ml of ethanol and boil to coagulate the ruthenium oxide precipitate. Remove the ruthenium oxide by filtration and repeat the ruthenium scavenge on the filtrate, and again filter off the ruthen- ium oxide. Step 7-Acidify the filtrate from step 6 with 5 ml of perchloric acid solution and evaporate to dryness under an infrared lamp, and finally heat on a hot-plate to remove the last traces of perchloric acid. (All traces of perchloric acid must be removed because of risk of explosion with organic material.) Step 8-C00l, add ethyl acetate to the residue and warm to dissolve the sodium per- chlorate, filter and reject any insoluble matter.Step 9-Pass dry hydrogen chloride gas into the filtrate to precipitate sodium chloride, centrifuge and discard the supernatant liquid. Step 10-Dissolve the sodium chloride in water, add 10 ml of a-methoxyphenylacetic acid reagent and allow to cool in an ice-bath for 40 minutes, with occasional stirring. Transfer the tube to a water-bath at 20" C, stir for 5 minutes, centrifuge and discard the supernatant liquid. Step 11-Add 5 ml of butanol - hydrogen chloride mixture to the precipitate to convert the sodium salt of a-methoxyphenylacetic acid into sodium chloride, centrifuge and discard the supernatant liquid. Wash the sodium chloride with 5ml of butanol, centrifuge and discard the washing. Step 12-Transfer the precipitate from step 11 with butanol to the filter-stick containing two tared Whatman No.42 (21-mm diameter) filter-papers and wash once with 2 ml of butanol. Step 13-Separate the filter-papers, dry in an oven at 80" C, re-weigh, mount on a 1-inch aluminium planchet by using Gelva solution and count in a low background @-counter. Check for the absence of caesium and potassium by counting the source covered with a 51 mg per cm2 aluminium absorber. Average transmission for the counter used was 15 per cent. for sodium-22, 26 per cent. for caesium-137 and 50 per cent. for potassium-40. Step 14-Calculate the chemical recovery from the weight of sodium chloride on the filter, the sodium content of the rain water and the sodium content of the reagents used after the ion-exchange separation.STANDARDISATION OF THE /3-COUNTER- Use a standard sodium-22 solution (obtainable from the Radiochemical Centre, Amer- sham) to prepare a series of sources varying in thickness from 6 to 60 mg per cm2. Count each source on a low level @-counter, both normally and covered with a 51 mg per cm2 aluminium absorber, and plot an efficiency against source weight graph. LIMITS OF DETECTION- By using a counter with a background of 1-2 counts per minute and a counting time of 1440 minutes, the minimum detectable activity is 0.4 pCi of sodium-22, based on three standard deviations of the background count. This gives 0.008 pCi per litre for a 50-litre sample.718 BURDEN: DETERMINATION OF SODIUM-22 I N RAIN WATER [ArtajySt, VOl. 93 RESULTS Table I shows the amounts of total sodium found in rain water and the reagents used.TABLE I CONTENT OF INACTIVE SODIUM FOUND IN RAIN AND REAGENTS Sodium, Sample p.p.m. Rain water (1) . . .. .. .. .. .. 3.8 Rain water (2) . . .. .. .. .. .. 2.8 Tetramethylammonium hydroxide solution . . 800 Ruthenium carrier solution . . .. .. .. 0.6 Rain water (3) . . .. .. .. .. .. 0.9 a-Methoxyphenylacetic acid solution . . .. 100 Table I1 shows the radiochemical recovery of sodium-22 added to different types of water. The activity added was sufficient to make it possible to neglect in the calculation any sodium- 22 present in the sample originally. TABLE I1 RADIOCHEMICAL RECOVERY OF SODIUM-22 Sodium-22 added, Sodium-22 found, Radiochemical Type of water pCi pCi recovery, per cent. Distilled .. . . .. .. 928 890 96 Rain . . .. .. .. .. 4640 4509 97 Soft . . .. .. .. .. 4640 4500 97 Hard . . .. .. .. .. 4640 4310 93 Intermediate . . .. .. .. 4640 4456 96 Table I11 shows the decontamination factors for the principal radionuclides, with a half-life greater than 30 days, likely to be present in rain water. Manganese-54 was not included in these experiments because it is neither positron active nor 13-active and hence would not be counted by a Geiger counter. Results given in Table I11 were obtained by adding active isotopes to various samples of rain water and then taking the latter through the above procedure. The amount of con- tamination caused by each isotope was measured in the final sodium chloride source. With ruthenium, the chemistry of which is complex, the active species was added in the form of ruthenium trichloride, which may not match the chemical species in fall-out.The recommended procedure, however, affords adequate decontamination from ruthenium species so far encountered in rain water. TABLE I11 DECONTAMINATION FACTORS Activity added, Activity found, disintegrations disintegrations Decontamination Nuclide per minute per minute factor Ruthenium-lO6/rhodium-106. . .. 8 x lo8 50 1.6 x 105 Caesium-137 . . . . . . . . 2.6 x 104 2 1.3 x 104 Strontium-90/yttrium-90 . . .. 2 x 105 1 2 x 105 Cerium-l44/praseodymium-l44 . . 1.25 x 105 3 4.2 x 104 Promethium-147 . . .. . . 4.5 x 105 10 4.5 x 104 Zirconium-95 . . .. .. . . 8.0 x 104 1 8.0 x 104 Antimony-l25/tellurium-125m . . 1.25 x 10' 15 8.6 x 104 *RaD.E.F... .. .. . . 2.22 x 104 0.8 2.7 x 104 Potassium-40 . . .. .. .. 1416 0.6 2.3 x 10' * RaD.E.F. is an equilibrium mixture of lead-210 (RaD), bismufh-210 (RaE) and polonium-210 (RaF).November, 19681 BY USING A LOW LEVEL P-COUNTER 719 TABLE IV RESULTS OBTAINED FROM RAIN-WATER SAMPLES Activity of sodium-22, pCi per litre Activity of caesium-137, pCi per litre Date of collection of rain water December, 1964, to February, 1965 . . 0.18 f 0-03* 12.3 May to June, 1965 . . .. .. . . 0.13 f 0.02 18.8 October to November, 1965 . . .. 0.04 f O - O l t 2.7 * 90 per cent. confidence limits on counting. t Very high rainfall. Table IV shows the results obtained from rain-water samples, the results for caesium-137 being included for comparison. These values are in close agreement with those obtained in June, 1964, by Perkins,lo who used a multi-parameter y-spectrometer.They are about half the peak levels observed by RodeP in 1963. La1 and Peters11 calculated the production rate for sodium-22 produced by cosmic rays at latitude 50” N. If it is assumed that all of this is carried down uniformly in rain, the calculated activity in London rain water is 0.01 pCi per litre, which is considerably less than the observed values. The author thanks Dr. D. I. Coomber for his helpful advice, and the Government Chemist, Ministry of Technology, for permission to publish this paper. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. REFERENCES Perkins, R. W., Nielson, J. M., and Thomas, C. W., Science, N.Y., 1964, 146, 762. Marquey, L., Costa, N. L., and Almeida, I. G., Nuovo Cim., 1957, 6, 1292. Rodel, W., J . Geophys. Res., 1965, 70, 4447. Palmer, H. E., 4th Symposium in Radioactivity in Scandinavia, Riso, Denmark, 1964; Contract Perkins, R. W., and Nielson, J. M., Nature, 1965, 205, 866. De Bortali, M., Gaglioni, P., and Malvaciani, A., Health and Safety Laboratories Report No. 155, Wood, R., Metson, P., and Richards, L. A., Nature, 1964, 203, 617. Norman, N., Beck, J. C., and Browne, J. S. L., J. Lab. Clin. Med., 1957,50, 308. Reeve, W., and Cristoffel, J., Analyt. Chem., 1957, 29, 102. Perkins, R. W., “Radiochemical Methods of Analysis,” International Atomic Energy Agency, Lal, D., and Peters, B., Prog. Elem. Part. Cosm. Ray Phys., 1962, 6, 1. Received April lst, 1968 AT(45-1) CONF-747-1; Nuc~. S C ~ . Abstr., 1964, 43799. 1965. Vienna, 1965, Volume 11, p. 65.

ISSN:0003-2654    

DOI:10.1039/AN9689300715

出版商:RSC    

年代:1968

数据来源: RSC

摘要:

720 Analyst, November, 1968, Vol. 93, $9. 720-721 An Investigation of the Degraded Neutron Flux in a 14-MeV Neu tron-activation Cell BY P. E. FRANCOIS (Department of Physics, University of Aston in Birmingham, Gosta Green, Biriningham 4) hleasurements have been made of the fluxes of thermal and indium resonance neutrons near the target of a 14-MeV neutron generator in a small irradiation cell. These measurements can be used to assess the importance of interfering reactions in fast-neutron activation analysis. For a fast- neutron output of loe neutrons per second the measured sub-cadmium flux was 3.8 x lo4 neutrons per cm2 per second, and the flux at the 1.4-eV indium resonance was 1-3 x lo3 neutrons per cm2 per second per eV. FAST-NEUTRON activation analysis is often performed with the source in the middle of a relatively small cell.In these conditions there may be, at the sample position, a considerable flux of neutrons scattered from the walls with degraded energies, and it is useful to know the extent of this contamination to assess the importance of any interfering reactions. The most important neutrons are likely to be those with energies in the thermal and resonance regions. The fluxes of neutrons with energies below the cadmium cut-off and in the indium resonance region in a typical assembly have been measured in this Department. The results depend on the exact arrangement of moderating and absorbing material, but they will serve as a guide for workers with similar assemblies who wish to evaluate the necessity to take precautions against interfering reactions.EXPERIMENTAL PROCEDURE Neutrons (14 MeV) were produced by the D-T reaction at the centre of a cell about 1.5 X 1.5 x 2 m. The walls consisted of concrete and earth on two sides and about 1 m of concrete on the others. The neutron output was measured by the associated particle tech- nique, and the results are normalised to an output of lo9 neutrons per second. The thermal- neutron flux was measured by irradiating two indium foils, one with a cadmium cover, in symmetrical positions 5 cm from the target. The resonance flux was measured by subse- quently irradiating two indium foils, both in cadmium covers, but one also covered by 0.4 mm of indium, which served to shield that foil from neutrons with energies equal to the indium resonance. All of the indium foils were 1-27 cm in diameter and 0.28 mm thick.The activities of the foils were measured with an end-window Geiger counter. This counter was calibrated by using an indium foil irradiated in a thermal-neutron facility. The activity of this foil was measured on the Geiger counter, and on a scintillation counter whose efficiency for indium-116m y-rays was obtained from the calculations of Vegars, Marsden and Heath.l RESULTS THERMAL NEUTRONS- The activities of the bare and the cadmium-covered foils were followed for 1 hour after the end of the irradiation. Both decay curves showed the presence of a mixture of activities, but the difference fell with a half-life that agrees with the expected 54 minutes of indium-116m. 0 SAC and the author.FRANCOIS 721 The saturation activity caused by neutrons below the cadmium cut-off was estimated to be 3.03 x 103 disintegrations per second for a fast-neutron output of 4.88 x los neutrons per second.A small correction needs to be applied for the depression of thennal-neutron flux within the foil; this was estimated from data compiled at the Argonne laboratory.2 Assuming an activation cross-section of 155 barns, the calculated value of the therrnal-neutron flux was 3.8 x lo4 neutrons per cm2 per second for a fast-neutron output of lo9 neutrons per second. RESONANCE NEUTRONS- The decay curves of the two foils again indicated a mixture of activities, but the dif- ference between the counts fell with the expected half-life of indium-ll6m. The saturation disintegration rate caused by the resonance neutrons at the end of bombardment was estimated to be 0-745 x lo3 disintegrations per second for a fast-neutron output of 6.7 x lo8 neutrons per second.A correction for self-shielding must again be applied, and a correction is also necessary to allow for the difference in distribution within the foil of the activity produced by the resonance neutrons and that by thermal neutrons in the calibration foil. This has been estimated by assuming an exponential absorption of @-particles within the foil with a linear absorption coefficient, calculated by the method of Gleason, Taylor and TabernJ3 of 1-25 x lo2 cm-l. To calculate the neutron flux, it is necessary to make assumptions about the shape of the neutron spectrum and the variation of the activation cross-section. If we assume that the activation is mainly caused by the large resonance at 1.4 eV, and the spectrum is constant over the width of this resonance, we can integrate on the assumption that the cross-section varies in accordance with the Breit - Wigner expression. If the peak cross-section is taken to be 30,000 barns and the width of the resonance 0.07 eV, then the observed activa- tion corresponds to a neutron flux at the 1.4-eV indium resonance of 1.3 x 103 neutrons per cm2 per second per eV for a fast-neutron output of lo9 neutrons per second. REFERENCES 1. 2. 3. Vegars, S. H., Marsden, L. L., and Heath, R. L., U S . Atomic Energy Commission Report, 1DO 16370, “Reactor Physics Constants,” U.S. Atoinic Energy Commission Report, ANL 5800, 1963. Gleason, G. I., Taylor, J. D., and Tabern, D. L., Nucleonics, 1951, 8, No. 5, 12. 1958. Received March 27th, 1968

ISSN:0003-2654    

DOI:10.1039/AN9689300720

出版商:RSC    

年代:1968

数据来源: RSC

摘要:

722 AstaZyst, November, 1968, Vol. 93, PP. 722-728 The Micro Determination of Calcium in Mammalian Hard Tissues BY C. ROBINSON AND J. A. WEATHERELL (Biologicat Research Unit, Dental School and Hospital, University of Leeds) A spectrophotometric method is described for the determination of microgram amounts of calcium. The procedure is extremely rapid and simple, enabling amounts of calcium in the range 2 to 3.5p.g per ml to be determined with an accuracy of about &-1 per cent. (standard deviation). The method, based on the ability of calcein (fluorescein-3,3’-bismethylimino- discetic acid) to complex with calcium ion, measures the reduction in light absorption of a calcein solution at 506 nm on the addition of calcium. THE analysis of many small enamel particles, 20 to 50pg in weight, required a method of calcium determination that was both rapid and simple to carry out.An accuracy of about 2 1 per cent. (standard deviation) was necessary in order to examine small variations in calcium concentration and calcium-to-phosphorus ratios. Existing techniques were either insufficiently sensitive or too laborious for this purpose. Calcein (fluorescein-3,3’-bismethyliminodiacetic acid) , a dye that forrns a fluorescent complex with calcium selectively in the presence of magnesium, has proved valuable in previous investigations of mineralised biological tissues. The calcium - calcein complex was used as an indicator in the complexometric titration of calcium with EDTA.ls2 To increase the sensitivity of this method, the authors attempted to determine calcium by measuring the amount of fluorescence with a spectrophotofluorimeter. Over a narrow range of concen- trations of about 0.5 to 2 pg of calcium per ml, this technique proved moderately successful.Two difficulties were encountered, however, the first being the relationship between calcium concentration and fluorescence, which was not linear (Fig. l), and the second, the instability of the photofluorimetric instruments available, which prevented accurate reproducibility. 0.5 I .o I .5 2.0 2.5 Calcium concentration, pg per rnl Fig. 1. Relationship between calcium concentration and fluorescence w Wavelength, nm Fig. 2. Absorption spectrum of calcein solution read against a solution containing calcein and calcium When calcium is added to an aqueous solution of calcein, the resulting fluorescence is accompanied by an apparent decrease in the brown colour of the solution.This prompted 0 SAC and the authors.ROBINSON AND WEATHERELL 723 an investigation of the absorption profiles of calcein and calcein - calcium. The profile results from a combination of fluorescence and light absorption. By reading the calcein blank against the calcium - calcein test solution, a maximum E value is obtained at 506 nm and a minimum at 480 nm (Fig. 2). If the cells are reversed the peaks are inverted, 480 nm becoming the maximum and 506nm the minimum. It has proved possible to measure calcium concentration accurately by reading at 506 nm. The calcein - calcium complex and the calcein solutions are not stable and the optical densities begin to decrease significantly about half an hour after making up the dye.If the dye solution is placed in an ice-bath, however, satisfactory results can be obtained for at least 30 minutes. This limitation is not very restrictive, as the procedure is so simple that several internal standards, together with twelve to sixteen test samples, can be analysed in 30 minutes. In the range 2 to 3.5 pg per ml, the determination has an accuracy of +_1 per cent. (standard deviation). Outside this range, there is some deviation from linearity, as can be seen in Fig. 3. 0.8 - 0-7 - 0.6 - .n 0 3 0.5 - 0.4 - 0.3 - 0.2 - 0.1 - 6 0.01 I I 1 I 1 I 2 3 4 5 Concentration of calcium, p g per ml Fig. 3. Effect of increasing concentrations of calcein on calibration curves.The figure a t the end of each curve refers to the concentration of calcein, mg per cent. EXPERIMENTAL WAVELENGTH OF MAXIMUM ABSORPTION- Fig. 2 shows the absorption profile of a solution of calcium-free calcein, read against a calcium - calcein solution. The maximum at 506 nm and the minimum at 480 nm can be reversed by reading the solution of calcium - calcein against that of the calcium-free solution of dye.724 ROBINSON AND WEATHERELL: THE MICRO DETERMINATION OF [AnUlySt, VOl. 93 Wavelength, nm Fig. 4. Absorption spectrum of a solution containing calcein read against distilled water Wavelength, nm Fig. 5 . Absorption spectrum of a solution containing calcein and calcium read against distilled water The 506 nm maximum reading does not seem to correspond precisely with Amax.of calcein (Fig..4), Amax. of calcium - calcein (Fig. 5) or to the excitation maximum of the calcium - calcein complex (Fig. 6). The readings obtained at this wavelength appear to result mainly Wavelength, nm Fig. 6. Excitation spectrum of a solution containing calcium and calcein from the fluorescence emitted by the calcium - calcein complex. This is suggested by the effects that increasing dye concentrations have on the readings, which can be explained b i quenching, and, as the excitation maximum is so far from 506 nm, this suggests that there is also some reduction in absorption. There is perhaps also some reduction in the absorptiorNovember, 19681 CALCIUM I N MAMMALIAN HARD TISSUES 725 of the dye caused by the addition of calcium.Fig. 7 shows the fluorescence-emission spectrum of a calcium - calcein solution, and gives some indication of the extent of fluorescence at the wavelength of 506 nm. 25 t Wavelength, nm Fig. 7. Fluorescence-emission spectrum of a solution containing calcein and calcium EFFECT OF DYE CONCENTRATION- Fig. 3 shows how increasing concentrations of dye affect the calibration graphs. Above a certain concentration the readings fall, which could be caused by quenching of fluorescence by free calcein. This is strongly supported by the fact that increasing concentrations of dye affect fluorescence in a similar manner. The optimum concentration of the dye seemed to be about 5 mg per cent. In the present work a Koch-Light preparation was used. EFFECT OF PERCHLORIC ACID CONCENTRATION- Large variations in pH affect the optical density of the calcium - calcein complex and attempts were made to keep the concentration of perchloric acid fairly constant.The particles of enamel, bone or dentine were weighed and dissolved in 2 M analytical- reagent grade perchloric acid. For a sample of enamel, x p g of the tissue was dissolved in x pl of acid. This procedure arose from the fact that enamel contains about 36 per cent. of calcium, so that the dilution produced a solution containing an optimum concentration of calcium, i e . , about 3 pg of calcium per ml. The amount of acid present was 8-33 pl of 2 M perchloric acid. For other tissues, e.g., bone or dentine, a similar procedure was followed, the aim being to prepare solutions keeping the calcium and acid concentrations as close as possible at these levels.The calibration solutions (e.g., 2-0, 2.5, 3.0 and 3.5 pg of calcium per ml) also contained similar amounts of acid. As, however, these were made up by progressively diluting a stock solution, which contained both calcium and acid, the more dilute calcium calibration solutions contained also slightly less acid. Only the 3.0 pg per ml calcium calibration solution con- tained exactly 8.33 p1 of 2 M perchloric acid per ml. The effect of this small variation in acid concentration on the determined values for calcium was below 4 1 per cent. The deter- mination can tolerate a two-fold increase in the concentration of acid, i.e., 2 x 8-33pl of 2 M perchloric acid per ml, with no significant reduction in accuracy.Higher concentrations of acid than this tend to produce lower readings. The dissolved enamel was diluted to 0.12 x ml with de-ionised water. EFFECT OF MAGNESIUM AND PHOSPHATE CONCENTRATIONS- Magnesium can have an effect on the results. The readings of a 3.0 pg per ml calcium solution rose as the molar ratio of magnesium to calcium increased from 0 to 0.1 (Fig. 8). Subsequently, the increase took the form of a series of plateaux. When phosphate is present726 ROBINSON AND WEATHERELL: THE MICRO DETERMINATION OF [AnUiySt, VOl. 93 - E n a L 0 M .- 5 - a -0 c .- E 0 Y f3 Magnesium - calcium molar ratio Fig. 8. Effect of increasing concentrations of magnesium on a solution containing 3.0 pg of calcium per ml the effect of magnesium is reduced (Fig.9). At a magnesium-to-phosphorus ratio of 0.08, the apparent increase in calcium concentration is about 1.3 per cent., at a magnesium-to- phosphorus ratio of 0.15, the increase in the apparent calcium concentration was 2.0 per cent., where a plateau was reached extending to a magnesium-to-phosphorus ratio of 1.5. Here a further increase in the determined calcium content occurred to about 4 per cent., where a Magnesium -phosphate molar ratio Fig. 9. Effect of increasing magnesium - phosphorus molar ratios in a solution containing 3-0 p g of calcium per ml second plateau was reached. In the present context, i e . , for bone, dentine or dental ename, (expected magnesium-to-phosphorus ratio about 0.04) ,3 there is thus no significant interference by magnesium. The effect of greater concentrations of magnesium could be blanketed b3 adding magnesium to bring the magnesium-to-calcium molar ratio to between 0.15 and 1.0 This level corresponds to the plateaux shown in Figs.8 and 9. Equimolar amounts of phos- phate do not affect the determination of calcium. REAGENTS- Make up to 1OOml with de-ionised water. minations should be completed within 30 minutes of making up this dye solution. METHOD Calcein soldion-Dissolve 5 mg of calcein in 25 ml of ice-cold 3 M potassium hydroxide Place this solution in an ice-bath. All deter- CALIBRATION SOLUTIONS- Calcium carbonate, 12-5 mg, dried for 1 hour at 105" C is dissolved in 13.9 ml of 2 & perchloric acid and diluted to 1 litre with de-ionised water. Aliquots of this solution, 10 12.5, 15 and 17.5 ml, are diluted to 25 ml to give solutions containing 2.0, 2.5, 3.0 and 3-5 p~ of calcium per ml, respectively.November, 19681 CALCIUM I N MAMMALIAN HARD TISSUES 727 PROCEDURE- Pipette 1 ml of water and 1 ml of calcein solution into a l-cm path-length glass cuvette; mix thoroughly with a polythene rod.Retain this water - calcein blank throughout the procedure. In the second cuvette mix 1 ml of sample and 1 ml of calcein solution. Read the water - calcein blank against this test solution at 506 nm. Make up the calcein solution immediately before the determination. RESULTS Fig. 10 shows a typical calibration graph. Concentration of calcium, pg per ml Fig. 10. Typical calibration curve for the range 2.0 to 3.5 pg of calcium per ml Table I compares the results obtained by analysing solutions of bone, dentine and enamel by complexometric titration,2 with those obtained by diluting the solutions 100 times and determining calcium by the technique described above.TABLE I COMPARISON OF TITRATION WITH SPECTROPHOTOMETRIC MICRO DETERMINATION Expressed as pg of calcium per ml of solution r \ Titration results, Spectrophotometric results, L Material analysed duplicate determinations triplicate determinations Enamel .. .. . . 2-84 2.87 Dentine . . .. .. 3-28 3.33 Bone . . .. .. 2.99 2.96 DISCUSSION The method described provides an extremely simple and rapid technique for the deter- mination of micro amounts of calcium. The basis of the method, i.e., the reduction in the apparent absorption of calcium - calcein solutions at 506 nm is empirical, as the peak does not seem to correspond to any single one of the spectra shown in Figs.4, 5, 6 or 7. It is not clear to what extent it is caused by fluorescence and to what extent by absorption, but. the accuracy and repeatability of the method, together with its great simplicity, justify the empirical approach. Some restriction is imposed by the instability of the dye, but the ease with which the determinations are performed permits the analysis of a fairly large number of samples in the 30 minutes available.728 ROBINSON AND WEATHERELL The calcium concentration of the solutions analysed should lie between 2.0 and 3.5 pg of calcium per ml. For mammalian hard tissues this presents no difficulty, as the approximate percentage of calcium is known. The results presented in Table I agree with those obtained by an established complexo- metric procedure. Interference by magnesium has been described. Its effect is appreciable in the absence of phosphate, but can be reduced by adding phosphate, or adjusting the magnesium con- centration to within the range corresponding to the plateau shown in Fig. 8. For mammalian hard tissues the effect of magnesium is much reduced by the indigenous phosphate. The mechanism of this reduction is not clearly understood, but is probably caused by phosphate reducing the tendency of magnesium ions to aggregate with calcein molecules in solution. The sensitivity of the procedure can be increased considerably by using micro cells. We thank the Medical Research Council for a grant in support of this work, and Mr. G. Naylor for technical assistance. REFERENCES 1. 2. 3. Mori, K., Archs Biochem. Biophys., 1969, 83, 552. Weatherell, J. A., Clinica Chim. Ada, 1960, 5, 61. Brudevold, F., in Kugelmass, I. N., and Sognnaes, R. F., Editors, “Chemistry and Prevention of Received April 22nd, 1968 Dental Caries,” Charles C. Thomas, Springfield, Illinois, 1962, p. 34.

ISSN:0003-2654    

DOI:10.1039/AN9689300722

出版商:RSC    

年代:1968

数据来源: RSC

摘要:

Analyst, November, 1968, Vol. 93, p p . 729-731 729 Determination of Hydrof luoric Acid in Inhibited Red Fuming Nitric Acid BY E. I;. CROOMES AND R. C. McNUTT* (Army Propulsion Laboratory and Center, Research and Development Directorate, U.S. Army Missile Command, Redstone Arsenal, Alabama 36809) A simple method has been devised for directly measuring fluoride-ion concentrations in inhibited red fuming nitric acid, which is commonly used in propellents. The technique involves the use of a newly developed fluoride- sensitive electrode, and the concentrations are determined by measuring the potential developed between this electrode and a reference electrode. A special procedure, involving a buffer solution, was devised for weighing the inhibited nitric acid before neutralisation, dilution and measurement.The results are more accurate than those obtainable by the commonly used, and more complex, indirect methods. For the particular instance of inhibited red fuming nitric acid, the technique also proved superior to another proposed direct method in that it is effective with samples as small as 1 g. RED fuming nitric acid is a common and effective oxidant for liquid-propellent missiles, but it is a very corrosive acid that will attack most metal containers. The corrosive action can be inhibited to a large extent by the addition of hydrofluoric acid in concentrations of about 0.5 M. For both assay and storage monitoring purposes, there is a need for a reliable technique of determining fluoride-ion concentrations in inhibited red fuming nitric acid.Until recently the determination of fluorides has been difficult; the only analytical tech- niques available were based entirely on indirect methods, such as the effect of fluoride on redox potential of metal-ion complexes and on bleaching techniques such as that used by Johnson and Jones,l which involved the removal of iron from ferron by fluoride. However, the recent development of the Orion fluoride-ion electrodeR makes it possible to use direct measurement techniques2 With this electrode, the direct measurement of fluoride-ion concentration is as simple as pH measurement. One such technique, which provides a simple, sensitive and accurate means of deter- inining the hydrofluoric acid in inhibited nitric acid, is described below. Another direct method, likewise based on the Orion electrode, was also considered and is discussed briefly; however, this second technique was found unsuitable for this analysis.EXPERIMENTAL GENERAL TECHNIQUE- This method basically consists in weighing a solution of the inhibited nitric acid, neutralis- ing the acid and measuring the potential developed between the fluoride-sensitive electrode and a convenient reference electrode, in this instance a saturated calomel electrode. MATERIALS- A standard fluoride stock solution was prepared from a weighed amount of analytical- reagent grade sodium fluoride. Aliquots of the stock solution were used to prepare known concentrations over the desired concentration range in a buffer solution of 0.1 M acetic acid and 0.1 M sodium acetate.* Present address : Department of Chemistry, Athens College, Athens, Alabama 3561 1. 0 S.4C and the authors.730 CROOMES AND MCNUTT: DETERMINATION OF HYDROFLUORIC ACID [APzdySt, VOl. 93 PROCEDURE- The procedure described below was designed specifically for the hydrofluoric acid concentration range usually used in inhibited nitric acid, ie., 0.4 to 0.8 per cent. w/w. About 1 g of the inhibited nitric acid was weighed and diluted to 100 ml with 0.1 M acetic acid - 0.1 M sodium acetate buffer. A 5-ml volume of this solution was removed and diluted to 100ml with the buffer solution. The amount of acid taken was determined accurately by weighing a screw-capped plastic vial containing 25 to 30 ml of buffer solution, and then adding about 1 ml of the sample with the aid of a disposable pipette and a 1-ml rubber suction cup.After the sample had been drawn into the pipette, the tip was quickly inserted into the weighed buffer solution and the acid forced out. In quick succession the vial was closed and re-weighed, and the solution diluted to 100ml in a calibrated flask to provide the dilutions for measurement. In the preparation of the solutions, it is important that the pH be kept within the range of 4 to 6. The pH is slightly low, but a well poised buffer free from interfering anions is provided. The final solution was then transferred to a beaker of suitable size, and the potential (in millivolts) measured with the Orion fluoride-ion electrode vemm a saturated calomel reference electrode. As the resistance of the electrode is high, a high-impedance potentiometer is required.A Beckman expanded-scale pH meter was used for this particular experiment. At 25" C, E = E + 0.05915 pF. The potentiometer was calibrated before running each group of samples. A standard fluoride solution was used to adjust the electrode potential to a pre- determined value, which was obtained in this instance from a calibration graph prepared by plotting the log of the fluoride-ion concentration zleisus the potential in millivolts (Fig. 1); the reproducibility of these specific-ion electrodes is fairly well established as being about that of a good glass electrode for pH. 5 15 30 45 60 -mV versus S.C.E. Fig. 1. Calibration graph with Orion fluoride electrode versus S.C.E. in 0.1 M acetic acid buffer. Measurements were made with a Beckman expanded-scale pH meter S RESULTS AND DISCUSSION This method is rapid, simple and direct, and quite accurate, as indicated by the results given in Table I for known samples in similar ionic conditions.These samples were prepared by adding known amounts of sodium fluoride to fuming nitric acid to simulate the charac- teristics of inhibited red fuming nitric acid, so that the buffer solution would carry the same load as with an inhibited nitric acid sample. If the latter has a lower concentration of hydrofluoric acid than the prepared samples studied, it may be necessary to use a larger sample and to provide a greater buffer capacity for the additional acid. Instead of the 0.1 M acetic acid and sodium acetate, a 1.0 M system could be used.November , 19681 IN INHIBITED RED FUMING NITRIC ACID 731 In Table 11, the results are compared with those obtained by a colorimetric procedure developed by Merwin3 and Steiger,4 which is based on the bleaching action of fluorine on the yellow colour produced by oxidising a solution of titanium sulphate with hydrogen peroxide.This method is at present required by U.S. Army Military Specification. TABLE I DETERMINATION OF HYDROFLUORIC ACID IN KNOWN SAMPLES PREPARED FROM SODIUM FLUORIDE AND FUMING NITRIC ACID Hydrofluoric acid added, Sample No. per cent. 1 0.40 2 0.53 3 0-64 4 0.79 5 0.88 6 1-19 Hydrofluoric acid found, per cent. 0.39 0.52 0.65 0.79 0.90 1.18 TABLE I1 COMPARISON OF METHODS FOR MEASURING HYDROFLUORIC ACID IN INHIBITED RED FUMING NITRIC ACID Sample No.1 2 3 4 5 6 7 8 Merwin - Steiger method 0-62 0-67 0-60 0-58 0.69 0.66 0-65 0.64 Fluoride electrode 0.63 0-61 0-65 0.66 0.59 0.59 0.65 0.67 The unfavourable results from the Menvin - Steiger method were to be expected, as it requires about a 6-g sample to give sufficient fluoride ion to determine its concentration. The nature of inhibited nitric acid portends difficulty in any method that requires the neutralisation of more than a l-g sample. In addition, the technique is complex and time consuming. The standard graph used in the Merwin - Steiger method must be checked frequently to be certain of the results obtained. This necessitates preparing several known samples and a blank for each set of test runs. Analysis of the sample and verification of the standard graph requires several hours.Another direct method for determining fluoride-ion concentration investigated was a titration technique suggested by Li~~gane,~ in which the Orion specific-ion electrode was likewise used. Thorium, calcium, aluminium and zirconium were tried as titrants for the fluoride ion, but this method was not amenable to the analysis of inhibited nitric acid. Thorium gave a nitration curve, but was of little value at the fluoride concentration that would permit the use of a l-g sample. A solution of 0.1 M thorium was prepared and standardised with standard sodium fluoride. With such a high concentration of thorium, it was necessary to neutralise 6 to 7 g of sample to obtain a satisfactory curve. The problem of neutralisation was further complicated by the need to keep the volume of solution small so that the fluoride- ion concentration would not be too low. The other titrants did not give curves at all under the required conditions. We thank Mr. Albert W. Saddler of Technical Micronics Control Inc. for supplying the results obtained by the Menvin - Steiger procedure, which are presented in Table 11. REFERENCES 1. 2. 3. 4. 5. Johnson, W. L., and Jones, M. M., Inorg. Chem., 1966, 5, 1345. Rechnitz, G. A., Chem. Engng News, 1967, June, 146. Merwin, H. E., Amer. J . SC~., 1909, 28, 119. Steiger, G., J. Amer. Chem. SOC., 1908, 30, 219. Lingane, J., J . AnaZyt. Chem., 1967, 39, 881. Received May 2nd, 1968

ISSN:0003-2654    

DOI:10.1039/AN9689300729

出版商:RSC    

年代:1968

数据来源: RSC

摘要:

732 Analyst, November, 1968, Vol. 93, pp. 732-736 Determination of the Total Silicon Content of Water BY P. M. BAKER AND B. R. FARRANT (Analytical Laboratory, The Pernzutit Co. Ltd., Chiswick, London, W.4) A method is described for the determination of total silicon content of water based on evaporation, fusion and subsequent absorptiometric measure- ment of solutions of reduced /i?-molybdosilicic acid. The standard deviation for this procedure, by using samples of between 20 and 60m1, was found to be 0.1 pg of silicon as silica. To meet the specification requirements of new high-pressure boilers a method was required for the determination of lower levels of total silicon in water than had been previously neces- sary. The following method describes an adaptation of Morrison and Wilson’s method,l in which the random errors at low concentration, contributed by atmospheric pollution of the evaporation and fusion stages, are reduced.The method has been tested to determine the standard deviation of analytical results by carrying out silicon determinations on specially purified water. LABORATORY- The method was carried out in a laboratory, shown in Fig. 1, that complied with the following requirements : maximum protection against solid particulate contamination at bench level; the minimum restriction on movement of personnel and on working facilities; and maintenance of the filtration and air control equipment causing the minimum interruption of work. To achieve the above conditions, the maximum possible filtration efficiency of the air must be effected immediately before it passes over the work-bench area.The air, which is first passed through a large surface area 5-pm pre-filter followed by four absolute filters, enters the laboratory through eleven ceiling-mounted absolute filters directly above the work- bench area. Each filter operates at about 60 per cent. of its air-carrying capacity, and has an initial input velocity of 150 feet per minute. The pre-conditioned air is then guided down the working area of the bench by a 4-foot deep, clear Perspex screen installed over the leading edge of the entire bench area. The screen affords protection from operator contaminatidn and from contamination generated within the floor area. The whole plant works on a once- through air system. The optical density measurements were made in 10-cm cuvettes at 805 nm on a Unicam SP500 spectrophotometer, with de-ionised water in the reference cuvette.It is advisable to check the wavelength, as this may vary slightly from one instrument to another. APPARATUS- Heating block-This is a cast aluminium block (9 x 9 x 4 inches) fitted with four 250-watt heaters and thermostat, drilled and turned to take four 30-ml platinum crucibles and a thermometer. To ensure a good heat transfer between the block and the crucibles, a few millilitres of glycerol are introduced into each compartment, the temperature of the block being maintained at between 105” and 110” C. A series of fusion tests is carried out to determine whether any particular crucible gives results consistently different from the others. When not in use the crucibles should be stored in electronic-grade hydrofluoric acid.Polythene bottles-Narrow-necked, 4-oz polythene bottles with caps are used in the colorimetric stage. Each bottle is cleaned initially with 0.75 ml of 5 N hydrochloric acid and 2 ml of 10 per cent. w/v ammonium molybdate in 100 ml of water, heated in a boiling water bath for 1 hour and then emptied. The solutions are then treated as described in the colorimetric determination under “Procedure.” Bottles that give high optical densities must not be used. EXPERIMENTAL Platinum crucibles-Platinum crucibles of 30-ml capacity are used. This procedure is repeated without the heating. 0 SAC and the Permutit Company, Ltd.Fig. 1. The lay-out of the laboratory [To face page 732BAKER AND FARUNT 733 REAGENTS- All reagents are of analytical-reagent grade unless otherwise stated.Polythene beakers, measuring cylinders and funnels should be used for making up reagents, which are then stored in polythene bottles. Treated water-Mixed-bed de-mineralised water further treated with De-Acidite FF regenerated with 201b of sodium hydroxide per cu. foot. This water, which contains less than 0.001 p.p.m. of reactive silicon dioxide, is hereafter referred to as water. Sodium carbonate-A 20 per cent. w/v aqueous solution of micro analytical-reagent grade sodium carbonate was made up and tested as follows. Take 20 g from each batch, make up to 100 ml with water and fdter through a Whatman No. 42 filter-paper into a polythene bottle.Carry out fusion blanks in duplicate on 1 ml of this solution from each batch and select further batches containing less than 5 pg of silicon dioxide per 1 ml of solution. Ammonium moZybdnte-Make up a 10 per cent, w/v solution of ammonium molybdate tetrahydrate in water. This reagent is stable for at least 1 month. Hydrochloric acid-Prepare 5 N hydrochloric acid by dilution of concentrated hydro- cliloi-ic acid with water and standardisation against N sodium carbonate solution. It is found that 5 N hydrochloric acid, purchased as a standard solution, gives high reagent blanks. Oxalic acid-Use a 10 per cent. w/v solution of oxalic acid dihydrate. Reducing agent solution-Make up the first solution by dissolving 0-5 g of l-amho-2- naphthol-4-sulphonic acid and 1.0 g of sodium sulphite, N%S0,.7H20, in 50 ml of water, and filtering this solution through a Whatman No.41 filter-paper into a polythene bottle. Make up a second solution by dissolving 27.4 g of sodium metabisulphite (anhydrous) in 150 ml of water, filtering and mixing with the first solution. This should be freshly made every 3 days. Standard solzction of silica-Fuse 1 g of pure silicon dioxide (Johnson Matthey Specpure grade) with 5 g of micro analytical-reagent grade sodium carbonate until a clear melt is obtained. This solution contains 1000 p.p.m. of silica. CALIBRATION- Take 10 ml of the 1000 p.p.m. silica solution and dilute to 500 ml to obtain a 20 p.p.m. silica solution. Take 20 ml of this solution and dilute to 1 litre to obtain a solution containing 400 pg per litre of silica.For calibration samples, pipette the required volumes of standard silica into clean 4-oz polythene bottles and make each up to 100 ml with water. To each bottle add 1 ml of the 20 per cent. w/v sodium carbonate solution, then proceed as described under “Colorimetric st age.’ ’ Prepare two calibration graphs, one from 20 pg to 1 pg and the other from 5 pg to 0.1 pg. The second graph can be obtained by further dilution of the 400 pg per litre silica solution. Both graphs are linear. FUSION TESTS- Test each crucible by carrying out at least two fusions in it. Transfer by pipette 1 nil of the 20 per cent. w/v sodium carbonate solution into each crucible and evaporate to dryness in the heating block. Fuse at bright red heat for 5 minutes with a Meker burner by using a nickel - clirome support and handling the crucible only with platinum-tipped tongs.Allow the melt to cool and dissolve it in water, the top of the heating block being used as a hot-plate. Make up to 100 ml and proceed as in “Colorimetric stage.” If any crucible gives a high result, fuse in it 1 g of anhydrous sodium carbonate and swirl the melt so that it comes in contact with the whole of the inside of the crucible, dissolve this melt and discard. Repeat the fusion tests with 1 ml of the 20 per cent. w/v sodium carbonate solution. REAGENT BLANK- Into one of the prepared 4-02 polythene bottles place 100 ml of the water and 1 nil of the 20 per cent. w/v sodium carbonate solution (bottle A). Into a second bottle, B, place 5 ml of water and 1 ml of the 20 per cent.w/v sodium carbonate solution. Following the procedure described in the “Colorimetric stage” add the same volumes of the reagents to the contents of both bottles, except that after the addition of oxalic acid to bottle B add a further 95 ml Cool, dissolve the melt in water and make up to 1 litre. ANALYTICAL PROCEDURE734 BAKER AND FARRANT: DETERMINATION OF THE [Analyst, VOl. 93 of water, then add the 1 ml of reducing agent. When the solutions are ready, measure their optical densities at 805 nm against de-ionised water. The bottle A gives the optical density of the reagent $us water blank; bottle B gives reagent blank. The difference in optical density between bottles A and B is equivalent to 95 per cent. of the reactive silicon content of the water as silica, used throughout the method.EVAPORATION- Two determinations can be completed in duplicate in 1 day with four crucibles, by carrying out 40 and 60-ml evaporations on each sample. By using the platinum-tipped tongs remove the four crucibles from the hydrofluoric acid, rinse well with water and, to ensure complete removal of the hydrofluoric acid, heat in the Meker burner to bright red heat, then place in position in the heating block. Transfer by pipette into each crucible 1 ml of 20 per cent. w/v sodium carbonate solution. Allow the contents of the two crucibles selected for the 40-ml evaporations to evaporate to dryness and subject them to the “Fusion test” procedure. To the two remaining crucibles add the first 10-m1 aliquot of the 60-ml evaporation samples.Return the first two crucibles to the heating block, add a further 1 ml of the 20 per cent. w/v sodium carbonate solution and the first 10-ml aliquot of the 40-ml evaporation samples. Introduce subsequent 10-ml aliquots of both 40 and 60-ml evaporation samples into their respective crucibles when the volume has been reduced to about 2 to 3ml. When tlie evaporations are completed, fuse, dissolve the melt, make up to 100 nil and transfer the solutions to 4-02 polythene bottles. Carry out “Fusion test” procedure on the crucibles with 1 ml of the 20 per cent. w/v sodium carbonate solution. COLORIMETRIC STAGE- To each of the bottles containing samples add 1-5 ml of 5 N hydrochloric acid. Swirl the contents of the bottles and introduce into each 2ml of the 10 per cent.w/v ammoniuiii inolybdate solution. Add 2 ml of the 10 per cent. w/v oxalic acid solution to each, swirl, and add 1 ml of the reducing agent. Set aside for at least 15 minutes, after which the colour is stable for several hours. Measure the optical densities in a 10-cm cuvette at 805 nm against de-ionised water. REPORTING- The optical density measurements for the samples less the optical density of the fusion blanks gives the concentration of total silicon as silica in the samples. The optical density of the fusion blanks less the optical density of the bottle A in the “Reagent blank” gives the concentration of non-reactive silicon as silica in the fusion blanks. DEFINITIONS- Reactive silicon content of water-Defined as those forms of silicon, mainly monomeric and dimeric silicic acids, that react with ammonium molybdate in 10 minutes under the conditions of the “Procedure” given above.With the same calibration graph as that used for total silicon content take 100ml of original sample, add 1 ml of the 20 per cent. w/v sodium carbonate solution and proceed as in “Colorimetric stage.” Total silicon content of the water-Defined as the silicon concentration determined after tlie evaporation and fusion procedure. Non-reactive silicon-Defined as the difference between total and reactive silicon. EFFECTS OF OTHER SUBSTANCES- It was found that the increasing concentrations of neutral salts present in the samples caused a decrease in absorption. The effect of 0.2 per cent. w/v neutralised sodium carbonate was found to decrease the optical density figures by about 10 per cent.for concentrations of silica between 0 to 20 pg per 100 ml. Solutions, 100 ml 5 pg 10 tLg 15 Pg 2 0 r g . . 0.167 0-334 0.603 0.872 Sodium silicate solutions .. .. .. .. Sodium silicate solutions plus 0.2 g of sodium carbonate . . 0-150 0.303 0.456 0.609 Close bottles and set aside for 10 minutes. Silica A f \November, 19681 TOTAL SILICON CONTEXT OF WATER 735 The effects of variation in concentration of reagents and interferences from other sub- The method of Morrison and Wilson, stances present were not investigated thoroughly. although not identical, indicates that similar interferences would occur.2 s 3 PRECISION The results are shown in Table I. The figures in the right-hand columns for each batch are the absorbance minus the corresponding blank.In each batch the first blank to be tested was assigned to the first observation on each solution tested, the second blank to the second observation on each solution. Each of the following batches was carried out on consecutive days over a period of 1 week on one homogeneous batch of water. TABLE I RESULTS Solution 1st Batch 2nd Batch 1 Fusion blanks.. . . 0.042 0.042 0.037 0.042 0.0395 Mean 0.042 Mean 2 20-ml evaporation . . 0.048 0.006 0.052 0.010 0.048 0.011 0-038 0.004 Mean 0-0085 Mean 0.007 3 40-1111 evaporation . . 0.050 0.008 0.050 0.008 0-040 0.003 0.044 0.002 Mean 0-0055 Mean 0.005 4 60-m levaporation . . 0-053 0.011 0.051 0.009 0.047 0.010 0.053 0.011 Mean 0.0105 Mean 0.010 3rd Batch 0-038 0,034 0.036 Mean 0.039 0.001 0.045 0.011 Mean 0.006 0-043 0.006 0.049 0.015 Mean 0.010 0.042 0,004 0.041 0.007 Mean 0.0055 Calculations of within batch (Mo) and between (MI) mean squayes- 4th 0-044 0.041 0.043 0.045 0.042 Mean 0-053 0-050 Mean 0.053 0.044 Mean Batch Mean 0-001 0.001 0.001 0.009 0.009 0.009 0.009 0.003 0-006 5th Batch 0.040 0.043 0.0415 Mean 0.045 0.005 0-048 0.005 Mean 0.005 0.050 0.010 0.054 0.011 0.053 0-013 0.048 0.005 Mean 0.0105 Mean 0.009 Solution .. 1 2 3 4 Coding X . . 1000 (X-0.034) 1000 (X-0.001) 1000 (X-O.OO2) 1000 (X-O*003) EX2 .. . . 483 347 494 (CB)2/mn . . 307 203 3 60 CB2/n . . 454 267 413 m .. . . 5 5 5 n . . .. 2 M1 .. .. 14.3 Mo . . .. 6.0 2 16-0 16.0 2 13.3 16.2 Calculations for standard deviations- Solution 2 3 S W ~ = MO ..* . .. 16 16.2 Sb2 = (M, - Mo)/n .. NS NS Stz = MI 4- (n - l)Mo/n.. 16 14.8 (NS = non-significant) Decoding- Data deviation 1 2 Coded units .. .. .. .. sw 2-5 4.0 NS Sb - St - 4.0 Original units . . . . .. . . Sw 0.0025 0*0040 Sb - NS St - 0.0040 Standard Solution Solution 370 270 313 5 2 10.8 11.4 4 11.4 NS 11.1 Solution Soh tion 3 4 4.0 3.4 NS NS 3.9 3.3 0.0040 0.0034 NS NS 0.0039 0.0033 where Sw is the within-batch standard deviation, Sb the between-batch standard deviation and St the total standard deviation of single observation. Conveysion into silicon as silica- Solution St in optical density units St in pg of SiO? per litre 2 20-ml evaporation .. .. 0.004 5 3 40-ml evaporation .. .. 0.004 3 4 60-ml evaporation .. .. 0,003 1-3736 BAKER AND FARRANT DISCUSSION This method has been developed to determine the standard deviation of analytical results, with a practical sample size, in order to meet high pressure boiler specifications which require a maximum of 20 pg per litre of total silicon as silica in feed water.All statistical tests given in this paper are based on the 95 per cent. confidence limit. The limit of detection for each of the evaporations is defined as 4.6520 where o is expressed in pg of silica per litre. For 20-ml evaporations the standard deviation is 5 pg, and the limit of detection is 4.652 x 6 = 23.3 pg. For 40-ml evaporations the standard deviation is 3 pug, and the limit of detection is 4-652 x 3 = 14.0 pg. For 60-ml evaporations the standard deviation is 1.3 pg, and the limit of detection is 4.652 x 1-3 = 6.0 pg. When it is only necessary to detect 20 pug per litre or more of silicon as silica it would appear from the above figures that, under the conditions given in this paper, it would be sufficient to use 40-ml evaporations. However, should smaller concentrations of silica need to be detected, then 60-ml evaporations should be used. It can be assumed that as in each evaporation the standard deviation was 0.1 pg, it should be possible to achieve lower limits of detection by using larger volumes of samples. This paper is published by permission of The Permutit Co. Ltd. REFERENCES 1. Morrison, I. R., and Wilson, A. L., Analyst, 1963,88, 446. 2. - , - , Ibid., 1963, 88, 88. 3. -,- , Ibid., 1963, 88, 100. Received February 6th, 1968

ISSN:0003-2654    

DOI:10.1039/AN9689300732

出版商:RSC    

年代:1968

数据来源: RSC

摘要:

Analyst, November, 1968, Vol. 93, $$. 737-739 737 Screening for Alkaloids in Toxicology by Using Thin-layer Chromatography: A Rapid System Simulating Paper Chromatography BY P. E. HAYWOOD AND M. S. MOSS (Forensic Laboratory, Equine Research Station, Newmarkct, SufloZk) Although numerous thin-layer chromatographic systems for the separation of alkaloids have been reported, none of them compares in abundance of recorded Rp values with the citrate paper system of Curry and Powell. In order to overcome the slowness (which can be a serious disadvantage with this paper system in diagnosis of acute poisoning) without recourse to a new thin-layer chromatographic system and consequent sacrifice of results, a system has been developed that gives lip values which, for practical purposes, are substantially the same as those in the paper system, for the limited range of drugs examined and for their extracts from horse urine.It is unlikely that the correlation is due to chance, and it is considered, therefore, that the more rapid thin-layer chromatographic system can be substituted for the paper system. ALTHOUGH originally “alkaloid” was the name given to basic organic nitrogen compounds of plant origin, the definition has gained wider scope, largely because of the development by the pharmaceutical industry of synthetic drugs related chemically and pharmacologically to these “classic” alkaloids, and to the development of other basic nitrogenous drugs with no evo- lutionary origins at all in classic alkaloid chemistry. An analytical chemist confronted with the need to identify a compound in this class must now search a large and complex group.Chromatographic methods present the most useful approach initially, and the paper-chromato- graphic system described by Curry and Powel1,l in which a citrate buffered paper and butanol- citric acid solvent are used, has many advantages over the other paper systems discussed by Clarke,2 who has recorded RF values on over 450 different alkaloidal compounds by using a slightly modified form of the system. Because of the extensive results on RF values now available (see also Jackson and Moss3) this is undoubtedly the best system for the screening of biological samples for the presence of this type of compound. We have used a system of this kind for several years for the screening of race-horse urine and saliva samples for the presence of “dope,” and for screening human body fluids when acute poisoning has occurred.However, when a rapid result is required, as often happens in human poisoning, the time required for the paper system is a serious disadvantage (about 8 to 10 hours are required for a useful separation). To shorten development time of the chromatogram, without recourse to a new system, which would have meant sacrificing a large amount of data, we have investigated the use of a thin-layer system, with cellulose powder: analogous to the paper system of Curry and Powe1l.l The thin-layer chromatographic system was evaluated by comparing RF values with those obtained with the paper system when a limited number of drugs was used.A larger number of drugs have been compared on these systems by Sunshine, Fike and Laudesman5 but without a statistical analysis. Pure drugs, and extracts from horse urine to which pure 0 SAC and the authors.738 Residue Dissolve in a few ml of 0 . 1 ~ acid. Extract with eaual hydrochloric volumes of Adjust pH to 9.5 in the presence of phos- phate buffer. Extract with chloroform (about 3 times the volume of water). Evap- orate chloroform in the presence of 1 drop of concentrated hydrqchloric acid. Ether (discard) Residue Dissolve in a few drops of methanol for chromatography. Fig. 1 . Extraction scheme used for allrsloids in horse urine drugs had been added, were spotted in methanol solution on the papers and thin-layer plates. Fig. 1 summarises the extraction procedure used.RF values and standard deviations are listed in Table I. Most of these are derived from six separate chromatographic developments for each drug or urine extract investigated. The urine extracts used were from six separate extractions for each of the six drugs, each urine sample used being different (k, 36 separate urine samples were used). It was hoped that this would allow for variations in RF values caused by normal variations in composition of urine, and thus to simulate practical working conditions as far as possible. Although we always ran one thin-layer chromatogram and one paper system in adjacent tanks, with solvent that had been prepared in bulk for the day’s experiment, we did not obtain any better agreement between thin-layer chromatographic and paper RF values than by random comparison of R, values obtained on different days.Development time for the thin-layer chromatographic system was 1 i to 2 hours, and for the paper, 8 to 10 hours, with a 15-cm rise of solvent for each. TABLE I R, VALUES AND STANDARD DEVIATIONS OF DRUGS RUN ON PAPER AND THIN-LAYER CHROMATOGRAMS Clarke’s figures are inserted for comparison RF (paper) r Urine Standard Pure Standard Drug extract deviation drug deviation Strychnine . . . . 27 3.1 23 5-0 Amphetamine . . . . 49 2.5 45 3.0 Nikethamide . . . . 85 1.5 88 2-6 Procaine .. .. 29 2.7 28 3.1 Methylamphetamine . . 54 5.2 53 4.8 Atropine .. .. - 33 4.1 Cocaine . . .. .. - 39 3.1 Xylocaine . . .. - 65 3.7 Methylphenidate , . - - 61 4.3 Phenothiazine . . ..- - 96 1.0 Me thap yrilene .. - 36 3.4 Ephedrine . . .. 48 1.2 45 3.4 - - - - Clarke 24 45 83 27 52 42 33 42 59 59 32 - RF (thin-layer chromatogram) Pure Standard Urinc Standard drug deviation extract deviation h -7 26 4-3 28 4.3 49 2.9 49 2.1 85 2.7 82 1.9 31 3.5 27 4.5 55 4-8 53 6.4 49 3-2 49 6.2 40 3 4 - - 42 4.1 67 4.2 64 5.3 98 1.7 35 2.4 - - - - - - - - - -November, 19681 TOXICOLOGY BY USING THIN-LAYER CHROMATOGRAPHY 739 EXPERIMENTAL Thin-layer chromatographic plates-A 50-g sample of CC41 thin-layer chromatographic cellulose powder (Chromedia-Whatman) was agitated with 100 ml of 5 per cent. w/v sodium dihydrogen citrate. Method of agitation is critical, and the object should be to ensure maximum agitation without aeration, otherwise crazing of adsorbent occurs during drying.We achieved this by using a laboratory stirrer with paddle blades. Plates (5 x 20 cm) were spread 0.25 mm thick with a Shandon “Unoplan” spreader, and left in the air for 15 minutes, followed by drying at 80” C for 1 hour in an oven with an internal circulating fan. Plates were developed in tanks that had been equilibrated previously for at least 1 hour by the use, on the inside walls, of blotting-paper impregnated with solvent. Bufered fiafw-whatman No. 1 paper was dipped in a 5 per cent. w/v solution of sodium dihydrogen citrate, blotted, and dried in air at room temperature, essentially as described by C1arke.l No equilibration was used before development. Solved-For paper and thin-layer chromatograms, 9 parts of butanol were shaken with 1 part of 5 per cent.citric acid solution v/v and the solvent used the same day. Preparation of this solvent differs considerably from Clarke’s method, but RRF values do not appear to be affected appreciably thereby. The advantage over Clarke’s method is that this preparation is homogeneous. Clarke has, however, reported the preparation of a similar homogeneous solvent elsewhere6 but does not correlate the results obtained by using this solvent with those reported elsewhere.2 Location-All spots were located by spraying with a mixture of 1 g of chloroplatinic acid, 18 g of potassium iodide and 600 ml of water. We wish to thank Mr. John Giltrow for technical advice in preparing the films, and H. Reeve Angel and Co. Ltd. for a gift of CC41 cellulose powder. This work was undertaken as part of a research programme under the Scheme for the Suppression of Doping, financed by the Horserace Betting Levy Board, to whom we are grateful. REFERENCES 1. 2. 3. 4. 5. 6 . Curry, A. S., and Powell, H., Nature, 1954, 173, 1143. Clarke, E. G. C., in Lundquist, F., Editor, “Methods of Forensic Science,” Interscience, London and Jackson, J. V., and Moss, M. S., in Smith, I., Editor, “Chromatographic and Electrophoretic Moss, M. S., “Symposium on Identification of Drugs and Poisons,” Pharmaceutical Society of Sunshine, J., Fike, W. W., and Laudesman, H., J. Foreizs. Sci., 1966, 11, 428. Clarke, E. G. C., and Hawkins, A. E., J. Pharm. Pharmac., 1963, 15, 390. New York, 1962, Volume 1, p. 1. Techniques,” Heinemann, London, 1960, p. 394. Great Britain, London, 1965. Received March 15th, 1968

ISSN:0003-2654    

DOI:10.1039/AN9689300737

出版商:RSC    

年代:1968

数据来源: RSC

摘要:

740 Analyst, November, 1968, Vol. 93,pp. 740-748 Ex traction Procedures in Chemical Toxicology BY S. L. TOMPSETT (Department of Clinical Chemistry, University of Edinburgh, Royal Injirmnry, Edinburgh 3) Investigations into the separation by distillation, solvent extraction and the use of ion-exchange resins of drugs from biological materials, with particular reference to urine, are described and discussed. THE number of drugs that can be encountered in chemical toxicology may be described as countless. Excellent techniques involving the use of ultraviolet spectrophotometry, thin-layer chromatography and spectrofluorimetry are available for the detection and determination of many drugs, and there is an extensive and useful literature on these subjects. Before any of these techniques can be applied, it is essential to effect separation of the substances, suspected or unknown, from the biological material under investigation. This material can include blood, urine, gastric contents, tissues (liver and brain) and foodstuffs, all varying greatly in chemical composition.Because of differences in chemical composition, especially the presence or absence of protein and fat, each type of material presents an individual prob- lem. Urine, because it is a solution and usually contains neither fat nor protein, presents the simplest problem. During the last 3 years, a chemical toxicology unit has been working in this Department, and its work is closely associated with that of the Poisoning Treatment Centre at the Royal Infirmary, Edinburgh.The latter deals with the diagnosis and treatment of about 1000 suspected cases of drug overdosage each year.l On rarer occasions, other substances, e.g., cleaning fluids and weed killers may be involved. Few of the cases have proved fatal. The laboratory receives blood (withdrawn from patients on admission), gastric aspirate and lavage, and urine (collected for a t least 8 hours after admission) for chemical analysis. Such tissues as liver and brain are generally not available. Patients admitted to hospital with symptoms of so-called drug overdosage constitute it complex problem. In some cases poisoning is more or less obvious, whereas in others it is a question of deciding whether the symptoms are actually caused by drug overdosage or by organic disease. Drug overdosage may be caused by more than one drug.Frequently the evidence obtained from patient sources is inadequate or even misleading. In such cases, which are becoming more frequent, a system of screening is essential. Of the materials provided, urine has proved the most useful in this respect. An examination of the literature has shown the inadequacy of the data relating to the separation of drugs from biological material, and the present paper is concerned with investi- gations into some aspects of this problem. The biological material that has received the greatest attention in these investigations is urine. Most of the drugs examined possess high extinction ultraviolet spectra, and use has been made of this property, in conjunction with a recording ultraviolet spectrophotometer (Unicam SPSOO), to assess recoveries by the distillation and solvent-extraction procedures.In assessments with ultraviolet spectrophotometry, recordings were made against appropriate blanks. SEPARATION BY DISTILLATION Goldbaum and Domanski2 have shown that many basic drugs can be recovered by distillation in the presence of sodium hydrogen carbonate. The technique is a modification of the usual steam-distillation procedure, but has the advantage that final volumes are much lower. The author3 has applied this procedure to the examination of urine for the presence of some volatile basic drugs (phenothiazines) that could be detected by the use of ultraviolet spectrophotornetry. The following is a description of the procedure as applied to urine.Twenty-five millilitres of urine are diluted to 50 ml with water and 1 g of sodium hydrogen carbonate added. The mixture is distilled in an all-glass distillation apparatus, 30 ml of distillate being collected; 3 ml of 10 N hydrochloric acid are added and the ultraviolet spectrum is recorded over the range 200 to 350 nm (distillate A). 0 SAC and the author.TOMPSETT 741 The distillate, which is now acidic, is diluted to 50ml with water and the complete The results obtained are shown in Table I. procedure repeated (distillate B). TABLE I DISTILLATION PROCEDURE Basic substances- 1. Recoverable in distillate A but not in distillate B. mepyramine (recovery 25 per cent.) and ephedrine (recovery 15 per cent.). original substances, the ultraviolet spectra are unchanged as the result of distillation. Recovery 50 to 80 per cent., depending on the nature of the substance, with the exception of As compared with the Amphetamine Propox yphene Methamphetamine Tigloidine jl-Phenyleth ylamine Phenmetrazine Tranylc ypromine Benzphetamine N-Meth ylephedrine C hlorc yclizine Chlorphedianol Ephedrine Pipradol Me peridine Chlorphentermine C yclizine Methadone Benzhexol Amitryptyline nor-Tryptyline P yridine Nicotine Imipramine Phenindamine Triplolidine Phenyltoloxamine Brompheniramine Dextromethorphan Chlorpheniramine Methoxyphenamine Diethylpropion Chlorprothixene Pheniramine Trifluoperazine Promethazine Pipamazine Trifluopromazine Promazine Methdilazine Chlorpromazine Trimeprazine Methotrimeprazine Rlethap yriline Mep yramine Tripelennamine Triazine substances used as herbicides, Desmetryne, Atrazine, Simazine, Prometone and Pro- As compared with the original metryne .The following substances are recovered in both distillates A and B. substances, the ultraviolet spectra are unchanged as the result of distillation. 2. Diphenhydramine Diphen ylp yraline Orphenadrine Hydrazine devivatives- there are marked changes in the ultraviolet spectra. R, but there are no further changes in the ultraviolet spectra. Neutral and acidic substances- distillation procedures. The following substances are recovered in distillate A but, as compared with the original substances, The “new” substances can be recovered in distillate Phenelzine Isocarboxazid The following are recovered in distillates A and B. The ultraviolet spectra are unchanged bythe Phenacetin Coumarin Paramethadione Troxidone Methsuximide Monohydroxyphenols (phenol, 0- and p-cresol, thymol, guaiacol and chloroxylenol) Benzoic acid Salicylic acid Neutral substances- They are recovered in distillate B, but there is a considerable change in the ultraviolet spectra.The following substances are recovered in distillate A, there being no change in the ultraviolet spectra. Ethchlorvynol Ethinamate Methylpentynol carbamate Methylpentynol Many basic drugs can be recovered by distillation. None, with the exceptions of diphenhydramine, orphenadrine and diphenylpyraline, is recoverable in distillate B. Some neutral drugs and, in particular, some acidic substances, e.g., salicylic acid, benzoic acid and the monohydroxyphenols, are recovered in distillate B.As evidenced by ultraviolet spectra, volatile substances were recovered unchanged, with the exception of the few recorded. No attempt has been made to achieve full recovery. Recoveries over the range 0.5 to 2.0 mg were found to have a linear relationship, and those from urine and water were identical. RECOVERY OF VOLATILE DRUGS FROM DISTILLATES- With the exception of diphenhydramine, orphenadrine and diphenylpyraline, the basic drugs can be recovered by the addition of a slight excess of hydrochloric acid to the distillate and evaporation to dryness, preferably in an all-glass vacuum still.742 TOMPSETT : EXTRACTION PROCEDURES IN [AutaZyst, VOl. 93 Following the addition of sodium hydroxide to the distillate or a concentrate, the free bases can be extracted with diethyl ether, which can then be removed by evaporation at room temperature.There is no loss of base, except with methamphetamine, amphetamine, nicotine and /3-phenylethylamine, when the losses are total if 1-mg amounts are involved. SEPARATION BY EXTRACTION WITH ORGANIC SOLVENTS PROCEDURE WITH CHLOROFORM- Chloroform, which is used universally as a solvent, was used in examinations carried out under the following conditions. To 10 ml of aqueous solution containing 1 mg of the drug under investigation were added for acidic reaction, 1 ml of 10 N hydrochloric acid; for neutral reaction, 1 ml of 10 N hydrochloric acid, followed by solid sodium hydrogen carbonate until evolution of carbon dioxide gas ceased (pH about 7); and for alkaline reaction, 1 ml of 10 N sodium hydroxide.Forty millilitres of chloroform were added and the mixture shaken vigorously for 2 minutes in an 8 x 1-inch glass-stoppered tube, the aqueous phase then being removed and rejected. Anhydrous sodium sulphate was added to effect dehydration. The mixture was filtered and 30 ml of the filtrate evaporated to dryness. Residues were dissolved in 10 ml of N hydrochloric acid or N sodium hydroxide and the ultraviolet spectra recorded against an either N sodium hydroxide or N hydrochloric (or sulphuric) acid. The aqueous phases were appropriate blank over the range 200 to 350 nm. In a further series of experiments, 30 ml of chloroform extract were shaken with 10 ml of separated and the ultraviolet spectra recorded against an appropriate blank over the range 200 to 350 nm.The procedures have been applied to 10-ml volumes of urine, and the results obtained are shown in Table 11. TABLE I1 INFLUENCE OF REACTION (ACID, NEUTRAL AND ALKALINE) ON EXTRACTION WITH CHLOROFORM A. Extraction with chloroform + Substance extracted - Substance not extracted Sulphonamides- Sulphaguanidine . . Phthalylsulphathiazole Succinylsulphathiazole Sulphacetamide . . Sulphamethizole . . Sulphaphenazole . . Sulphame thoxydiazine Sulthiamine . . .. Sulphasomidine . . Sulphamethoxyp yridazine Sulphamezathine . . Sulphamerazine . . Sulphadimethoxine . . Sulphafurazole .. Sulphasomizole .. Sulphanilamide .. Sulphapyridine . . .. Sulphathiazole . . .. Sulphadiazine . . .. Hydrazine deyivatzves- Nialamide .. .. Isocarboxazide .. Iproniazid . . .. Coumayins- Dicoumarin . . .. Warfarin .. .. 4-Hydroxycoumarin . . 7-Hydroxycoumarin . . .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. . . .. .. .. .. .. .. . . .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. Neutral - - - - - - - + + + + + -t- + + + + + + + + + + + - -November, 19681 Indanedione- Phenindione . . Xanthine deyivatives- Caffeine .. Theobromine . . Theophylline . . Monohydroxyphenols- Phenol . . .. 0- and p-Cresol Chloroxylenol . . Guaiacol .. Barbiturates . . Thymol . . .. Miscellaneous- CHEMICAL TOXICOLOGY .. .. .. . . . . .. .. .. .. .. TABLE Acetyl-p-aminophenol . . Phenylbutazone . .Oxyphenbutazone . . Chlordiazepoxide . . Chlordiazepoxide lactam Phenytoin . . .. Morphine . . .. Nitrazepam . , .. Oxazepam . . .. p - Aminophenol .. Phenylacetylurea . . Bibenzonium (bromide) Captodiamine . . .. Vanillic acid diethylamide Glutethimide . . .. Methaqualone . . .. .. .. .. .. .. .. . . .. . . .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. 2-Ami~o-5-chlorobenzophenone* 2-Amino-5-nitrobenzophenone t I I (continued) .. .. .. .. .. .. .. . . .. .. .. .. .. . . .. .. .. .. .. .. .. .. .. .. .. .. f . .. * Produced from chlordiazepoxide. its .. .. .. .. . . .. .. .. .. .. . . .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. Acid + + + + + + + + + + + + + + + + + + + + + - + + - - - - - lactam and Neutral + + + + + + + + + + + + + + + + + + + + + + + + + + + + 743 oxazepam when these substances are heated with 5 N hf&ochloric acid in a boiling wate'r bath for 1 hour.t Produced from nitrazepam when this substance is heated with 5 N hydrochloric acid in a boiling water bath for 1 hour. Extraction of chloroform solutions with aqueous N hydrochloric (or sulphuric) acid or N sodium hydroxide -+ Substance extracted - Substance not extracted B. Barbiturates . . .. Warfarin .. .. 7-Hydroxycoumarin . . 4-Hydroxycoumarin . . Sulphamethoxydiazine Sulthiamine .. .. Dicoumarin . . .. Phenylbutazone . . Oxyphenbutazone . . Acetyl-p-aminophenol . . Morphine . . .. Theobromine . . .. Theophylline . . .. Caffeine .. .. Methaqualone . . .. Glutethimide . . .. Iproniazid . . .. Phenylacetylurea . . Oxazepam . . .. Bibenzonium .. .. p-Aminophenol .. Phenytoin . . . . Sulphapyridine . . .. Sulphadiazine . . .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. . . .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. . . . . N Hydrochloric acid - + + + + + + (Trace) - N Sodium hydroxide + + + + + + + + + + + + + + + + - - - + + + (Turns brown) - - Extracted with 5 N hydrochloric acid.744 TOMPSETT : EXTRACTION PROCEDURES IN [Analyst, Vol. 93 Reference has not been made to well defined acidic substances, e.g., salicylic acid, as their properties are well defined. A substance such as salicylic acid is extractable with chloroform from aqueous solutions with an acidic, but not a neutral, reaction. Salicylic acid can be recovered from chloroform by extraction with aqueous solutions of sodium hydroxide.Reference has also not been made to well defined bases, e.g., quinine, for the same reasons. Quinine can be extracted with chloroform from aqueous solutions with either neutral or alkaline reactions and from chloroform with aqueous solutions of mineral acids. It has been recorded4 that sulphonamides may interfere in the ultraviolet detection and determination of barbiturates; this can occur only with sulphonamides that are soluble in chloroform. Theobromine and theophylline can interfere in the ultraviolet determination of bar- biturates, and both are constituents of numerous pharmaceutical preparations. Caffeine presents unusual extraction properties, as it can be extracted by chloroform from aqueous solutions with an acidic, neutral or alkaline reaction.Extraction of caffeine from chloroform solution cannot be effected with aqueous N sodium hydroxide or N hydrochloric (or sulphuric) acid, but it can with aqueous 5 N hydrochloric acid. The anomalous behaviour of bibenzonium (bromide) merits attention; its extraction as a halide complex probably occurs with acidic and neutral solutions. Chlordiazepoxide, the lactam, oxazepam, and nitrazepam5 are converted into chloro- or nitroaminobenzophenones when heated with 5 N hydrochloric acid for 1 hour in a boiling water bath. Such amines are extracted with chloroform, and heptane, not only from neutral and alkaline aqueous solutions, but also from those with acidic reactions; it is possible that halide complexes are extracted from acidic aqueous solutions.Extraction from a neutral solution will include two groups of substances, vix., weak acids and bases. Some distinction can be made by extraction of the chloroform solution with either aqueous solution of alkali or mineral acid. This is not an entirely clear-cut procedure because ampholytes, e.g., acetyl-9-aminophenol, are extracted from chloroform solution under both conditions. It has been found3 that some of the phenothiazine sulphoxides, e.g., chlorpromazine sulphoxide, could be recovered by chloroform extraction from aqueous solutions with a neutral or alkaline reaction, by using the technique described above, whereas others, eg., perphenazine, could not. To recover the latter group, it was found necessary to use the following modified procedure.To 10 ml of an aqueous solution are added 1 ml of 10 N sodium hydroxide and 40 ml of chloroform. The mixture is shaken vigorously for 2 minutes. Without separation of the phases, dehydration is effected by the addition of anhydrous sodium sulphate. The mixture is filtered and 30 ml of the filtrate are evaporated to dryness. Many of the substances listed in Table I1 can be extracted from chloroform solution with aqueous solutions of sodium hydroxide or mineral acid, or both. Methaqualone and glutethimide are exceptions in that they can only be recovered by the removal of the solvent by evaporation. Bases can be recovered from solution in chloroform by extraction with an aqueous solu- tion of a mineral acid. For this purpose one has a choice of acids, but in some instances correct selection is important. Sulphuric acid must be used in preference to hydrochloric acid if, as with quinine and quinidine, use is to be made of fluorimetry.Sulphuric acid must be used for the extraction of some phenothiazine drugs from chloroform, as in some instances hydrochloric acid is inadequate because of the formation of chloroform-soluble halide com- p l e ~ e s . ~ s6 Under these conditions, emulsion formation is extremely rare. Only one extraction with solvent was carried out and although complete recovery is not attained, recoveries were found to be linearly related over a range of 0.25 to 2.0 mg, hence quantitative results could be obtained. I t would appear from the literature that this concept is being generally accepted.The use of a dehydrating agent, such as anhydrous sodium sulphate, ensures clean, almost water-free extracts, the compositions of which are more readily reproducible. I t would appear that this concept also is being more readily accepted. For extraction purposes, 4 parts of chloroform to 1 part of urine have been used.November, 19681 CHEMICAL TOXICOLOGY 745 Reference should be made to a factor that could influence the effect of the washing of chloroform extracts. Substances that can be recovered by extraction of chloroform solutions with aqueous solutions of alkali or mineral acid may be divided into the following groups. (i) Those extracted by aqueous solutions of alkali but not by those of mineral acid, e.g., acids such as the barbiturates. (ii) Those extracted by aqueous solutions of mineral acid but not by those of alkali, e.g., bases such as quinine.(iii) Those extracted by aqueous solutions of both alkali and mineral acid. Ampholytes, e.g., the sulphonamides, behave in this manner. In addition, other substances, such as theophylline and 7- and 4-hydroxycoumarin, behave in this manner but cannot be described as ampholytes. The “washing” of organic solvent extracts has been general practice, but this has been avoided by the author as such a procedure could result in losses of “unknown” constituents. The separations are preferably carried out in glass-stoppered measuring cylinders or tubes that can be centrifuged, and a teat-pipette is useful for the separation of top layers. These extraction procedures have been applied to normal urine, and the ultraviolet spectra obtained have been re~orded.~ U S E OF OTHER SOLVENTS- Many other solvents have been used for extraction purposes.Diethyl ether is a good general solvent, as it has a low boiling-point and hence can be readily removed by evaporation. Unfortunately, it is highly inflammable, contains anti- oxidants and is not particularly selective. When used as an extractant for urine, many normal constituents are extracted. Ethyl acetate is another good general solvent but is not particularly selective in its action. Because of their insolubility in other organic solvents, it is used for the extraction of chlorothiazide and hydrochlorothiazide from urine.’ t8 No details have been published concerning the limitations imposed by the pH of the aqueous solution on the extraction of these substances by ethyl acetate.Benzene is a selective solvent but, unfortunately, it possesses an intense ultraviolet spectrum, so that traces can cause complications if ultraviolet spectrophotometry is used. Heptane is a selective solvent, and has been successfully used in determinations involving phenothiazine drugs, imipramine etc. ; it is particularly useful in spectrofluorimetry. Morphine can be extracted with organic solvents from neutral aqueous solutions or aqueous solutions containing ammonia; chloroform is not an efficient solvent for this purpose. TABLE I11 SOLVENTS, OTHER THAN CHLOROFORM, USED FOR THE EXTRACTION OF DRUGS FROM AQUEOUS SOLUTIONS Solvent Substance Benzene.. .. - . ... . Quinine, Quinidine*S Heptane . . .. .. . . Phenothiazine drugs8J13t1* Caffeinelo Imipraminee J1 Phen ylbu tazone14 Desipraminet Trimipramine The following are not extractable by heptane : sulphonamides, barbiturates and oxyphenbutazone (a drug and a metabolite of phenylbutazone) Ethyl acetate . . .. .. . . Ethchlorvynol~5 Chlorothiazide’ f H ydrochlorothiazidesz * Selectively separated from metabolites. t A drug and also a metabolite of imipramine. f Not soluble in chloroform.746 TOMPSETT : EXTRACTION PROCEDURES IN [Analyst, Vol. 93 Chloroform - isopropyl alcohol (3 + 1) or butanol - benzene (1 + 1) are more efficient but are not particularly selective and, as a result, many constituents of normal urine are also extracted. Results are summarised in Tables I11 and IV.TABLE IV INFLUENCE OF PH ON THE EXTRACTABILITY OF SOME DIURETIC DRUGS (ALL INSOLUBLE IN CHLOROFORM) WITH ETHYL ACETATE FROM AQUEOUS SOLUTIONS Substance Acid Neutral Alkaline .. - . + + Hydrochlorothiazide . . .. - . + + .. * . + + Hydroflumethiazide .. .. * - + + - - Chlorothiazide . . .. Bendrafluazide . . .. - - SEPARATION WITH ION-EXCHANGE RESINS Ion-exchange resins have not received much application in chemical toxicology, possibly because the procedures involved are slow in comparison with other extraction procedures. The authorle t o 22 has examined the use of ion-exchange resins, not only in connection with problems associated with chemical toxicology, but also with problems associated with clinical biochemistry. To achieve some degree of uniformity, these studies have been con- fined mainly to the use of one resin, Dowex 50W x 12 (200 to 400 mesh). It has been TABLE V CONTAINING DOWEX 50W x 12 BEHAVIOUR OF SOME BASIC AND OTHER SUBSTANCES WHEN PASSED THROUGH A COLUMN Present in the 0.1 N hydrochloric acid eluate and, therefore, not retained Retained by the resin and eluted with 0.5 N Retained b j the resin and eluted with N hydro- hydrochloric acid chloric acid Retained by the resin and eluted with 2.5 N hydrochloric acid Retained by the resin and eluted with 5 N Retained by the resin and eluted with 8 N hydrochloric acid hydrochloric acid A Barbiturates, phenytoin, phenylbutazone, oxyphen- butazone and diazotisable amines obtained by the action of hot solution of alkali or mineral acid on chlorothiazide and hydrochlorothiazide Ammonium ion and many amino-acids Many amino-acids, sulphonddes, adrenalin, theo- bromine, theophylline, P-aminophenol and iso- prenaline Cystine, histidine, caffeine, isoniazid, morphine and the diazotisable amine obtained from oxyphen- butazone by the action of hot mineral acids Nicotine, codeine, brucine, strychnine, amphetamine, paraquat and diquat Quinine, paludrine, mepacrine, the aminobenzo- phenones, produced from chlordiazepoxide, ox- azepam and nitrazepam by the action of hot mineral acid,6 and the diazotisable amine produced from phenylbutazone by the action of hot mineral acid2S B Substance Applied in hydrochloric acid Eluted with Methyldopa2I .. .. .. .. 0.1 N N Hydrochloric acid Meperidinic acid17 .. .. .. 0.1 N 6 N Ammonia solution .... N 5 N Hydrochloric acid Rretylium Hexamethonium } . * Guanethidin Morphine17*18 . . .. .. .. N (a) 2-5 N Hydrochloric acid Codeine 17J8 Brucine Strychnine 21 These substances are recovered from eluates by evaporation to dryness in an all-glass vacuum still. (b) 6 N Ammonia solution (a) 5 N Hydrochloric acid (b) 4 N Ammonia solution - ethanol (20 + 80) .. .. .. N These substances are recovered from eluates by solvent extraction. } - -November, 19681 CHEMICAL TOXICOLOGY 747 found that with some drugs, there is no alternative to the use of cation-exchange resins for extraction purposes. The column characteristics were: resin (cation exchange), Dowex 50W x 12 (H+form), 200 to 400 mesh, height 70 mm and diameter 10 mm. Substances (about 1 mg) in 100 ml of 0.1 N hydrochloric acid were applied to the column and the column was washed with 100 ml of 0.1 N hydrochloric acid. Elution was then carried out with 100-ml volumes of 0.5, 1, 2-5, 5 and 8 N hydrochloric acid in that order.The procedure has been used extensively in urine analysis. Urine can be applied to the column either as 10 ml of urine plus 1 ml of 10 N hydrochloric acid and diluted to 100 ml with water or as 100 ml of urine plus 10 ml of 10 N hydrochloric acid. The particular technique used depended on the nature of the substances to be recovered. Some results are summarised in Table V. The procedures used for the recovery of certain drugs are also referred to in Table V. In several instances, e.g., methyldopa, bretylium, meperidinic acid, guanethidine and hexamethonium, because of their insolubility in organic solvents, this is the only pro- cedure available to achieve concentration. Eluates obtained from cation-exchange resins are not entirely suitable for direct examina- tion by ultraviolet spectrophotometry because of absorbance caused by background material derived from the resin material.With isoprenaline, it has been found satisfactory, provided that an adequate blank has been carried out. During the earlier work on this subject, the author used colorimetric methods to obtain evaluation of the technique, especially as many of the substances examined, e.g., many amino-acids, did not show absorbance within the ultraviolet range. DISCUSSION There were at least two objectives concerned in the present investigation, and the attainment of these cannot be described as complete.Because of the lack of results given in the literature, information was required concerning separation techniques, which are a necessary preliminary to any chemical toxicological pro- cedure. It was hoped to formulate a scheme of approach that could be applied to the examina- tion of the new drugs that are constantly appearing. For routine purposes, only the distilla- tion and solvent-extraction procedures are used. Chloroform is the solvent of the first choice, as low blank values are obtained when applied to normal urine. Ion-exchange resins are used only as a last expedient. The second objective was to obtain sufficient information that could be applied to the rapid screening of urine for the presence of drugs, a procedure that is becoming more and more essential in hospital practice.The following screening procedures have been developed and are now used, followed by ultraviolet spectrophotometry over the range 200 to 350 nm. The following procedure was used. (1) Distillation of urine in the presence of sodium hydrogen carbonate. (2) Extraction with chloroform of urine with a neutral reaction, then (a) evaporation of the chloroform extract to dryness and dissolution of the residue in N hydrochloric acid; (b) extraction of the chloroform solution with N sodium hydroxide; and (c) extraction of the chloroform solution with N sulphuric acid. (3) Extraction with chloroform of urine with an alkaline reaction, then (a) evaporation of the chloroform extract to dryness and dissolution of the residue in N hydrochloric acid; and (b) extraction of the chloroform solution with N hydrochloric or sulphuric acids, followed by 5 N hydrochloric acid.The subsequent procedures can be completed within 1 hour. Extraction with chloroform of urine with an acidic reaction is not normally carried out as many normal ultraviolet-absorbing substances are also extracted and their presence can confuse interpretations. It should be emphasised that although the coverage for screening purposes is large, it is not complete. The analyst should be presented with the specimens of his choice in order that a satis- factory assessment can be made. This is not always the practice as clinicians tend to place much emphasis on blood levels and, as a result, blood may be the only material supplied for analysis.It should be noted, however, that in certain cases, e.g., amphetamine intake and paraquat poisoning,24 Some clinicians are loath to obtain urine from comatose patients.748 TOMPSETT blood levels are usually so low that detection and determination, respectively, are not really possible. Because of the much higher concentrations encountered in urine, the latter is the most suitable material for chemical analysis, at least under these circumstances. Several important books and reviews on these subjects have been ~ritten,2~,~6,27 but our knowledge of them is still incomplete. With some drugs, such factors must be taken into consideration if satisfactory results are to be obtained. Morphine is excreted almost entirely in conjugated form, thus necessitating the use of hot, acid hydrolysis before applying any analytical pro- cedure.Following the ingestion of paracetamol or phenacetin, the recovery of products as P-aminophenol, as the result of hot, acid hydrolysis, appears to produce the most satisfactory resuk28 The barbiturates pose another interesting problem. Following the ingestion of certain barbiturates, e g . , amylobarbitone and pentobarbitone, unchanged barbiturates may not be detectable in the urine. By using special solvent techniques (not described in this paper) it is possible to detect and determine large amounts of metabolites, particularly of the hydroxylated type.29 ,30,31 932 Ultraviolet spectrophotometry, with a recording instrument, has been used extensively in the present investigations.Reference must be made to the work of Bradford and Brackett ,= who have described an extensive scheme involving the use of ultraviolet spectro- photometry and designed mainly for the examination of materials such as suspected pills, capsules and narcotics. The author expresses his thanks to the many pharmaceutical firms who supplied samples of drugs without cost. The conjugation and metabolism of drugs are important factors to be considered. Their investigations were not concerned with urine examination. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. REFERENCES Aitken, R. C. B., Daly, R. J., Kreitman, N., Matthew, H., and Proudfoot, A.T., Brit. M r d . J., 1968, 1. 706. Goldbaum, L. R., and Domanski, T. J., in Stolman, A., Editor, “Progress in Chemical Toxicology,” Volume 2, Academic Press, New York and London, 1965, p. 221. Tompsett, S. L., Acta Pharmac. Tox., in the press. Curry, A. S., in Stewart, C. P., and Stolman, A., Editors, “Toxicology,” Volume 2, Academic Press, New York and London, 1961, p. 153. Tompsett, S. L., J . Clin. Path., 1968, 21, 366. McBay, A. J., and Algeri, E. J., in Stolman, A., Editor, “Progress in Chemical Toxicology,” Volume 1, Academic Press, New York and London, 1963, pp. 168 and 187. Baer, J. E., Leidy, H. L., Brooks, A. V., and Beyer, K. H., J . Pharmac. Exp. Ther., 1959,125,295. Sheppard, H., Mowles, T. F., and Plummer, A. J., J . Pharm. Sci., 1960,49, 722. Brodie, B. B., Udenfriend, S., Dill, W., and Downing, G., J . Biol. Chem., 1947, 168, 311. Farmilo, C. G., and Genest, K., in Stewart, C . P., and Stolman, A., Editors, op. cit., p. 268. Tompsett, S. L., Proc. Ass. Clin. Biochem., 1967, 4, 164. Ragland, J. B., and Kinross-Wright, V. J., Analyt. Chem., 1964, 36, 1356. Ragland, J. B., IGnross-Wright, V. J., and Ragland, R. S., AnaZyt. Biochem., 1965, 12, 60. Burns, J . J., Rose, R. K., Chenkin, J., Goldman, A., Sculert, A., and Brodie, B. B., J . Pharmac. Algeri, E. J., Katsas, G. G., and Luonga, M. A., Amer. J . Clin. Path., 1962, 38, 125. Tompsett, S. L., Clinica Chim. Ada, 1960, 5, 415. -, Acta Pharmac. Tox., 1960, 17, 295. -, Ibid., 1961, 18, 414. -, Ibid., 1962, 19, 368. -, 3rd International Meeting in Forensic Medicine, Immunology, Pathology and Toxicology, April 16-24, 1963, School of Pharmacy, London. -, Bull. Int. Ass. Forens. Tox., 1967, 4, No. 3, 2. Tompsett, S. L., Forshall, W., and Smith, D. C., Acta Pharmac. Tox., 1961, 18, 75. Tompsett, S. L., Bull. Int. Ass. Forens. Tox., 1967, 4, NO. 2, 1. Ken-, F., Patel, A. R., Scott, P. D. R., and Tompsett, S. L., Brit. Med. J . , in the press. Williams, R. T., “Detoxication Mechanisms,” Chapman and Hall, London, 1959. Turner, L. K., in Curry, A. S., Editor, “Methods of Forensic Science,” Volume 4, Interscience Publishers, London, New York and Sydney, 1965, p. 177. Stolman, A., and Stewart, C. P., in Stolman, A., Editor, op. cit., p. 1. Tompsett, S. L., Bull. Int. Ass. Forens. Tox., 1966, 3, No. 5, 1. Moss, M. S., and Jackson, J. V., 3rd International Meeting in Forensic Medicine, Immunology, Moss, M. S., Proc. Ass. Clin. Biochem., 1965, 3, 218. Tompsett, S. L., Ibid., 1965, 3, 286. -, Bull. Int. Ass. Forens. Tox., 1966, 3, No. 2, 3. Bradford, L. W., and Brackett, J. W., Mikrochim. Ada, 1958, 3, 353. Exp. Ther., 1953, 109, 346. Pathology, and Toxicology, April 16-24, 1963, London. Received May 2114 1968

ISSN:0003-2654    

DOI:10.1039/AN9689300740

出版商:RSC    

年代:1968

数据来源: RSC

摘要:

Analyst, November, 1968, Vol. 93, $9. 749-755 749 Recommended Methods for the Evaluation of Drugs PREPARED BY THE JOINT COMMITTEE OF THE PHARMACEUTICAL SOCIETY AND THE SOCIETY FOR ANALYTICAL CHEMISTRY ON METHODS FOR THE EVALUATION OF DRUGS The Chemical Assay of Cascara Dry Extract, Cascara Tablets and Cascara Bark THE Panel, which was set up to recommend standard methods of assay of drugs and their preparations containing anthraquinones, has published Reports on the “Chemical Assay of Senna Fruit and Senna Leaf”l and on the “Chemical Assay of Aloes.”2 The work has now been extended to cascara bark and to certain of its preparations. The constitution of the Panel was: Professor J. M. Rowson (Chairman), Dr. C. Daglish, Professor J. W. Fairbairn, Miss B. Gartside (resigned October, 1967), Miss H.M. Perry and Mr. H. A. Ryan, with Mr. P. W. Shallis as Secretary. REPORT Cascara, the dried bark of Rhamnus purshiana DC., is an anthraquinone purgative which is administered as a dry extract usually in the form of sugar-coated tablets; a liquid extract and an elixir are also used. The most important constituents of the bark are (a) four primary glycosides, cascarosides A and B (glucosides of barbaloin), cascarosides C and D (glucosides of chrysaloin); ( b ) aloins, barbaloin, chrysaloin, which are C-glycosides ; smaller amounts of (c) 0-glycosides and of ( d ) free anthraquinones are also present. The cascarosides are almost tasteless, but the aloins are extremely bitter. Detailed published information on the relative biological activities of the cascarosides and of the aloins is lacking; such preliminary information as was available to the Panel indicated that both groups of compounds possessed similar activities but that they differed in the speed of their action.The work of Fairbairn and Simic3 has shown that in the preparation of cascara dry extract from the bark a significant loss of total glycosides occurred and that the proportion of cascarosides to aloins was lower in the extract than in the bark. Because of the thermolability of the cascarosides, their different speed of action and their tastelessness, the Panel agreed with the recommendation of Fairbairn and Simic3 that a chemical method of evaluating cascara should estimate the cascarosides separately from the aloins. Such a pair of figures can give some measure of the c u e exercised in the pharmaceutical manufacture of extracts of cascara and can also be used to detect adulteration of such preparations by the addition of aloes.Several methods have been published for the chemical assay of cascara, and that recom- mended by Fairbairn and Simic3 was selected by the Panel as the basis for their work. The dry extract was examined first, because of its frequent use in the form of tablets. The method was then applied to the assay of dried cascara bark. CHEMICAL ASSAY The method recommended by the Panel for the assay of cascara preparations is set out in the Appendixes. The use of 70 per cent. ethanol as an extracting solvent in stage (i) was found to be satisfactory. In stage (ii) the free anthraquinones are removed by extraction with carbon tetrachloride; relatively large volumes of this solvent are used to obviate the formation of troublesome emulsions.The presence of 1 per cent. of sodium chloride in the aqueous phase was ineffective for this purpose. The cascarosides are then separated from the aloins by partition in the system ethyl acetate - water. Fairbairn and Simic3 have shown that this partition is 96 to 99 per cent. effective, but the use of recently water-saturated ethyl acetate is essential. They have also shown that the 0-glycosides only represent some 10 per 0 SAC.750 THE CHEMICAL ASSAY OF CASCARA DRY EXTRACT, [A%@bSt, VOl. 93 cent. of total anthracene gIycosides in the extract and that these are fairly evenly distributed between the two phases when partitioned in stage (ii).The Panel concluded that the separate estimation of this small amount of O-glycosides would be an unnecessary elaboration of the method. TABLE I RESULTS OF COLLABORATIVE ASSAYS OF SEVEN SAMPLES OF CASCARA DRY EXTRACT By the method described in Appendix I Sample Laboratory CDEl A .. .. B .. .. C .. .. D .. .. E .. .. CDE2 CDE3 CDE4 CDE5 CDE7 Mean .. .. s.d. . . .. s.d. as yo of mean A .. .. B .. .. C .. .. D .. .. E .. .. Mean .. . . s.d. . . .. s.d. as % of mean A .. .. B .. .. C .. .. D * . .. E .. .. Mean .. .. s.d. . . .. s.d. as % of mean A .. .. B * . .. C .. .. D .. .. E .. .. Mean .. .. s.d. . . .. s.d. as % of mean A .. .. B .. .. C .. .. D .. .. E .. .. Mean . . . I s.d. . . .. s.d. as % of mean CDE6 A .. ..B .. .. C .. .. D .. .. E .. .. Mean . . .. s.d. . . . . s.d. as yo of mean A .. .. B .. .. C .. .. D .. .. E .. .. Mean . . . . s.d. . . .. s.d. as yo of mean .. .. .. .. . . .. .. . . .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. . . .. .. .. .. .. .. .. . . . . .. .. .. .. .. .. .. .. .. .. .. . . .. .. .. Aloins as anhydrous barbaloin, per cent. 4.32 4-84 3-93 3-93 4.21 - 4.00 4.26 4.01 4.04 4.17 0.29 7.0 8-63 8-53 8.30 8-03 8.12 6-71 7-60 7.70 7.88 8.04 7.94 0.53 6-7 8.07 8.03 7.15 7.38 6-96 7-46 7.60 7-40 7-30 7.35 7.47 0-35 4.7 5.68 5.48 5.23 5.23 5-71 4.29 5.00 5.20 5.31 4.70 5-18 0.43 8-3 3.62 3.54 3.43 3.49 3.35 3.65 3.60 3-50 3-39 3.78 3.54 0-13 3.7 2.48 2.49 2.46 2-46 2.83 2.77 2.40 2-50 2-51 2-59 2-55 0.14 5.5 3-38 3-44 3-41 3.42 3-36 3-07 3-60 3-50 3.29 3.02 3- 35 0.18 5.4 Cascarosides as cascaroside A, per cent.6.10 6.86 6-76 6.77 6.34 6-97 6-18 6-27 6-80 6-53 6-56 0.3 1 4-7 7.73 7-38 7.87 7-96 6.72 7.90 7.10 6.80 7.20 6.57 7-32 0-52 7.1 6.98 6.88 7.11 6-81 6-82 7.60 6.60 6.70 6.68 6-47 6.87 0.32 4.7 4.70 4.76 4-45 4.28 4-72 4.79 4.30 4.50 4.27 4.29 4.5 1 0.22 4-9 6-89 7-08 7.65 7-36 6-61 6.99 7.20 6-90 6.52 6.87 7.00 0.32 4.6 6.15 6-34 6-85 6.66 6.03 5-90 6-50 6.30 6.17 6.16 6-31 0.29 4.6 7-01 6.82 7-75 7.58 8-73 6-05 7-20 6.90 6.94 6.62 6-96 0.48 6-9 Total glycosides cascaroside A, per cent. 13-32 14.95 13-33 13.34 12.87 13.39 13.92 13.28 13.53 0.59 4.4 22.07 21.70- 21.84 21-47 20.36 19-23 19.90 19-90 20.45 20.08 20.70 0.99 4.8 20.53 20.36 19.14 19.28 18.53 20.15 19.40 19-23 18.94 18-83 19.44 0-68 3.5 14.25 13.96 13-24 13.08 14-31 12.00 12.71 13.21 13-19 12-19 13.21 0.79 6.0 13.01 13-06 13-56 13.28 12.27 13.16 13.21 12.87 12-26 13.26 12-99 0.43 3.3 10.35 10.57 11.04 10.82 10.82 10-87 10-53 10.53 10-43 10.55 10.65 0.22 2.1 12.71 12.62 13.53 13-38 12.42 11.24 13.21 12-87 12.51 11-74 12.62 0.71 5-6 as 13.38 -November, 19681 CASCARA TABLETS AND CASCARA BARK 751 The aloins present in the ethyl acetate fraction are freed from solvent and are converted into the free anthraquinones by the drastic condition of stage (iii), in which iron(II1) chloride and hydrochloric acid are used.The liberated anthraquinones are then extracted with carbon tetrachloride and are estimated spectrophotometrically in N sodium hydroxide solution. Oxidation and estimation procedures of stage (iii) are the same as those of our Report on the Chemical Assay of Aloes2 The cascarosides present in the aqueous phase of stage (ii) are estimated in stage (iv) by the sarne procedure as that specified for aloins.The Ei:m value for cascaroside A is derived from the work of Fairbairn and S i m i ~ . ~ Results for total glycosides are obtained by summation of the values for cascarosides and for aloins, re-calculated as equivalent cascarosides. Investigations by members of the Panel have shown that results for total glycosides by this process of summation are the same as those obtained by a separate, direct estimation of total glycosides and omitting the ethyl acetate - water partition in the second part of stage (iz). Such a direct estimation of total glycosides would, however, be an acceptable procedure.CASCARA DRY EXTRACT Members of the Panel have applied the method of Appendix I to the assay of seven different samples of cascara dry extract available on the English market. All spectrophoto- meters were initially checked by using a common standard sample of ammonium cobalt sulphate solution. The results of these assays are given in Table I, from which it will be noted that, in general, satisfactory concordance of results was obtained within and between laboratories, but that greater variation existed in the results for aloins than in those for cascarosides or for total glycosides. The extinction ratio value R for aloin assays was within the range 1.8 to 2.1; for cascaroside assays it was between 1-9 and 2.3.CASCARA TABLETS Cascara tablets are a convenient and widely used dosage form of cascara dry extract; they are generally sugar-coated. This preparation may be assayed by the method already described. To allow for individual tablet variation, a number of tablets, equivalent to about 2 6 g of dry extract, should be taken; they are disintegrated by soaking and by trituration in water before the addition of ethanol to about 70 per cent. content as an extracting solvent. An aliquot is then assayed as set out in Appendix 11. Results are expressed in terms of milligrams per tablet. TABLE I1 RESULTS OF COLLABORATIVE ASSAYS OF TWO SAMPLES OF CASCARA TABLETS By the method described in Appendix I1 Sample Laboratory CT 1 A . . .. B . . . . C . . .. D . .. . E . . .. Mean . . . . s.d. . . .. s.d. as yo of mean CT2 A .. .. B .. .. C .. .. D . . . . E .. . . Mean . . .. s.d. . . .. s.d. as % of mean .. .. .. .. .. .. .. .. .. .. . . .. .. Aloins as anhydrous barbaloin, mg per tablet 4-43 4.78 4.56 4-60 5.18 5.10 4.73 4.93 5.16 5.12 4.86 0-28 4.1 4-86 4.63 4-44 4-67 4.28 4.15 4.59 4-91 4.58 0-25 5.5 4.45 4-83 Cascarosides as cascaroside A, mg per tablet 8-14 8-31 9-15 8.94 9.38 9.05 9-16 9.13 8-08 8.15 8.75 0.5 1 5.8 9.98 10.05 10.90 11.03 11-10 10.10 10-57 10.31 9.41 9.32 10.28 0.63 6.1 Total glycosides as cascaroside A, mg per tablet 15-63 16.39 16.85 16-55 18.12 17.67 17-07 17.37 16.79 16.79 16.92 0.69 4.1 17.51 18-22 19.12 18.89 18-56 17-89 17-72 17-26 17-17 17.62 17-99 0.68 3.8 Two commercial samples of sugar-coated cascara tablets were assayed by members of the The figures obtained are recorded in Table I1 and Panel using the method of Appendix 11.these show satisfactory concordance of results within and between laboratories.752 THE CHEMICAL ASSAY OF CASCARA DRY EXTRACT, [Analyst, VOl. 93 CASCARA BARK The Panel has shown that the method of Appendix I can be applied, with minor modi- fications, to the assay of samples of cascara bark. The details are given in Appendix 111. Maceration of the powdered bark in 70 per cent. ethanol overnight gave satisfactory extrac- tion. TABLE I11 RESULTS OF COLLABORATIVE ASSAYS OF FIVE SAMPLES OF CASCARA BARK By the method described in Appendix I11 Sample Laboratory CB 1 A . . .. B . . .. D .. .. E .. .. Mean .. .. s.d. . . ,. s.d.as yo of mean CB2 CB3 CB4 CB5 A .. .. B .. . . D .. .. E .. .. Mean .. .. s.d. . . .. s.d. as yo of mean A .. .. B .. .. D .. .. E .. .. Mean .. .. s.d. . . .. s.d. as yo of mean A .. .. B .. .. D .. .. E .. .. Mean .. .. s.d. . . .. s.d. as % of mean A . . .. B .. .. D .. .. E . . .. Mean . . .. s.d. . . .. s.d. as % of mean .. .. . . .. .. .. .. .. .. .. . . .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. . . .. .. .. Aloins as anhydrous barbaloin, per cent. 2.02 2.06 2.60 2-50 2.23 2.24 2.31 2.12 2.26 0.20 8.8 1.10 16-20 1.27 1.37 1-24 1.21 1.39 1.50 1.29 0.13 10.0 1.56 1-63 1-91 1.95 1-73 1-82 1.93 1.99 1-82 0.16 8-8 1.58 1-55 1.83 1.84 1-74 1-76 1.79 1.81 1.74 0.11 6.3 1.52 1.55 1-79 1.86 1-72 1.72 1.83 1.64 1.70 0.13 7.6 Cascarosides Total glycosides as as cascaroside A, cascaroside A, per cent.4.45 4.79 4-70 4-68 4.77 4-73 4.27 4-60 4-62 0.18 3.9 6-25 6.15 6.31 6.12 8-15 6.13 5-72 5.66 6-06 0.24 4.0 5.31 5.57 5.45 5.25 5-25 5-24 4.82 4-77 5.21 0.28 5.4 7.42 7.65 7.37 7.29 7.10 7.37 6.32 6.45 7.12 0.48 6- 7 5.94 6-01 5.78 5-63 6-94 6-02 4-97 5-15 5.68 0.41 7.2 per cent. 7-86 8.26 9-10 8.90 8.51 8.48 8.18 8-19 8-44 0-41 4-9 8.14 8.21 8.49 8.48 8-23 8-16 8-09 8.23 8-25 0.15 1.8 7.98 8-34 8.69 8.56 8.14 8.28 8.09 8.14 8.28 0.24 2.9 10.13 10.32 10.50 10.43 10.02 10.30 9.31 9.48 10-06 0-44 4.4 8-54 8.66 8-83 8.79 8.83 8.90 8.04 7.89 8.56 0.39 4.6 Five commercial samples of cascara bark were assayed by the method of Appendix I11 and the results obtained by members of the Panel are set out in Table 111. These show good concordance within and between laboratories, although variation does exist between results of assays for aloins.The figures for the ratio value R for cascaroside assays were between 1.9 and 2.2; for aloin assays they were 1.7 to 1.9 and these lower values may suggest slight contamination in the aloin assay. INVESTIGATION OF LIQUID PREPARATIONS OF CASCARA Cascara liquid extract is a concentrated aqueous extract of the bark, to which ethanol is added to give a final content of 21 to 24 per cent. v/v. The Panel has devoted considerable effort to its assay by application of the method of Appendix I. A sample of about 3g, accurately weighed, of liquid extract was diluted to 100ml with 70 per cent. ethanol and a 10-ml aliquot of the dilution was assayed by the method of Appendix I.Three commercialXovember, 19681 CASCARA TABLETS AND CASCARA BARK 753 samples of cascara liquid extract were examined and within each laboratory good concordance of results was obtained. But for each constituent in each of the three samples, involving 45 duplicate assays, there was very wide deviation between results from the different labora- tories. Typical results are those for sample CLE2 in Table IV. TABLE IV RESULTS OF COLLABORATIVE ASSAYS OF SAMPLES OF CASCARA LIQUID EXTRACT CLE2 assayed by the method of Appendix I ; CLE4 assayed within the period of 1 week Sample CLE2 CLE4 Aloins as Cascarosides Total glycosides anhydrous as as barbaloin, cascaroside A, cascaroside A, Laboratory per cent. w/w per cent. w/w per cent. w/w A .... B .. . . C . . .. D . . . . E . . .. Mean . . .. s.d. . . . . s.d. as % of mean A . . .. B . . . . C . . . . D . . . . E .. . . Mean .. .. s.d. . . . . s.d. as yo of mean .. .. . . . . .. . . .. .. .. .. . . . . . . .. . . .. 1.09 1.07 1.56 1-55 1-06 1-06 1.59 1.58 0.98 0-95 1.14 0.48 1.54 1.50 1.63 1.58 1-33 1-57 1-79 1.71 1.33 1-31 1.53 0.16 42.1 10.5 1-12 1.10 1.26 1-32 1-28 1-15 1.41 1.32 1.13 1.07 1.22 0.12 9.8 1-15 1-15 1-29 1.37 1.11 1.06 1-30 1-34 1.12 1-08 1.20 0.12 10.0 2-96 2.89 3.88 3.93 3.04 2-93 4.06 3.96 2-78 2.68 3.31 0.57 3.75 3.66 4-03 4.03 3-33 3-68 4.30 4.20 3-34 3.28 3.76 0.37 9.8 17-2 A detailed study of sample CLEl in one laboratory over a 3-month period showed a fall in aloin content from 0.98 to 0.35 per cent., but the cascarosides remained constant in amount.An examination of the two other samples CLES and CLES in the same laboratory detected no similar degradative changes over a period of 3 to 4 months. Examination of the three samples in the four other laboratories after this period of 3 to 4 months failed to produce any evidence of disappearance of aloins in CLE1; and all results within each of these four labora- tories were in general agreement at the beginning and end of the period of time, but dis- agreement between laboratories was still marked. Attempts were next made to eliminate possible errors due to ageing of the preparation, also to the occurrence of precipitation in the liquid extract itself and in the initial dilution with 70 per cent. ethanol. A freshly filtered sample CLE4 was sent out from a central stock to each laboratory, where it was assayed under closely detailed conditions and within the period of 1 week.The results obtained are given in Table IV, from which it will be noted that deviations of about 10 per cent. were obtained in the assays of each of the compounds. The storage of four separate portions of this sample for a 3-month period under different conditions in one laboratory resulted in one portion showing a fall in aloins content to 0-48 per cent. with only a slight loss of cascarosides. No significant changes occurred in the three other portions. Preliminary work on cascara elixir also gave divergent results, and the Panel has con- cluded that the method of Appendix I cannot be applied with any confidence to liquid pre- parations of cascara.It stresses that concordant results on such a preparation within a laboratory are unlikely to agree with those obtained in a second laboratory. CONCLUSION The Panel recommends the methods of Appendixes I, I1 and I11 for the chemical assay of cascara dry extract, cascara tablets and cascara bark. It cannot recommend any process for the assay of liquid preparations of cascara. The Panel does not have within its terms of reference the establishing of standards for the materials assayed, but it recommends that cascara and its preparations be assessed on the content of total glycosides and on the proportion of cascarosides present therein.754 [Analyst, Vol. 93 It also draws attention to the greater variability of both figures for cascara dry extract than for the bark itself.Thus the total glycoside content of five samples of cascara bark were within the range 8-2 to 10-1 per cent. (Table 111), and the amount of cascarosides varied between 55 and 73 per cent. of the total glycosides present, whereas for the seven samples of cascara dry extract the range for total glycoside content was 10-65 to 20.7 per cent. (Table I), and the amount of cascarosides varied between 34 and 59 per cent. of the total glycosides present. THE CHEMICAL ASSAY OF CASCARA DRY EXTRACT, Appendix I RECOMMENDED METHOD FOR THE CHEMICAL ASSAY OF CASCARA DRY EXTRACT REAGENTS- Carbon tetrachloride. Iron(I1I) chloride solution, 60 per cent. w/v-Analytical-reagent grade. Hydrochloric acid, 36 per cent. wlw. Methanol. Sodium hydroxide, N.Distilled water. Ethanol, 70 per cent. v / v aqueous. Ethyl acetate. PROCEDURE- (i) Weigh accurately about 0.5 g of powdered extract and place it in a 100-ml calibrated flask with 80ml of 70 per cent. ethanol. Shake the mixture occasionally, allow to stand overnight, make the volume up to 100ml with 70 per cent. ethanol, shake well and filter through a Whatman No. 4 filter-paper. (ii) Transfer 10 ml of this solution into a separating funnel, add 10 ml of water and extract with two or three 40-ml portions of carbon tetrachloride. Wash the combined carbon tetrachloride extracts with one 10-ml portion of water, reject the carbon tetrachloride layer and return the washings to the aqueous layer. Extract the combined aqueous layers with five 60-ml portions of freshly prepared water- saturated ethyl acetate and reserve both layers for further work.(Prepare the water- saturated ethyl acetate by shaking 300 ml of ethyl acetate with 30 ml of water for 3 minutes and then allowing the layers to separate.) (iii) Aloins. Transfer the combined ethyl acetate extracts to a suitable flask, distil off the solvent and evaporate just to dryness. Dissolve the residue in 0.3 to 0.5ml of methanol, rinse out with water into a 50-ml calibrated flask and make up to 50 ml with water. Place 20 ml of this solution in a round-bottomed 100-ml flask containing 2 ml of 60 per cent. iron(II1) chloride solution and 12 ml of hydrochloric acid. Attach a water-cooled, double-surface condenser to the flask, place the flask in a bath of continuously boiling water (so that the water level is above that of the liquid level in the flask) and heat for 4 hours.Cool the solution, transfer it into a separating funnel, and rinse out the round-bottomed flask successively with 3 to 4 ml of water, 3 to 4 ml of N sodium hydroxide and 3 to 4 ml of water, adding these rinsings to the contents of the separating funnel. Extract the contents of the separating funnel with three 20-ml portions of carbon tetra- chloride. Wash the combined carbon tetrachloride layers with two 10-rnl portions of water. Reject the washings. Extract the carbon tetrachloride layer with one 15-ml portion and with successive 5-ml portions of N sodium hydroxide until the final sodium hydroxide extract is colourless. Combine the alkaline extracts, heat them in a shallow dish on a bath of boiling water for 5 minutes, with constant stirring to remove carbon tetrachloride, cool and adjust to 50 ml with N sodium hydroxide. Determine the extinction of this solution in a 1-cm cell at 440 nm and at the maximum, at 500 nm, against N sodium hydroxide in a similar cell.The readings should be taken within 1 hour of beginning the alkaline extraction with one 16-ml portion of N sodium hydroxide (step (iii), line 13) and particular care should be taken to carry out these stages of the estimation in subdued light.November, 19681 CASCARA TABLETS AND CASCARA BARK 755 Calculate the percentage of aloins present as anhydrous barbaloin on the assumption that the EiZm value at 500 nm of the red solution obtained from anhydrous barbaloin is 209.Also calculate the ratio value E500JE440nm. If the ratio value is less than 1.7, reject the result. Re-calculate the percentage of aloins present as cascaroside A on the assumption that the E::& value at 500 nm of the red solution obtained from cascaroside A is 125. ( i v ) Cascarosides. Transfer the aqueous layer, reserved in stage (ii), line 6 above, to a 50-ml calibrated flask and make up to 50 ml with water. With 20 ml of this solution, carry out the iron(II1) chloride oxidation, N sodium hydroxide extraction and colorimetric determination as described above for Aloins, stage (iii), lines 4 to 21 (beginning at “Place 20 ml of this solution . . .” to “. . . subdued light.”). Calculate the percentage of cascarosides present as cascaroside A on the assumption that the E:2m value at 500 nm of the red solution obtained from cascaroside A is 125. Also calculate the ratio value If the ratio value is less than 1.8, reject the result.(v) The percentage of total glycosides present as cascaroside A is obtained by adding together the values for aloins in stage (iii) re-calculated as cascaroside A and for cascarosides in stage (iv). nm/E,,, nm. Appendix I1 RECOMMENDED METHOD FOR THE CHEMICAL ASSAY OF CASCARA TABLETS PROCEDURE- (i) Place a number of tablets, equivalent to about 2.5 g of dry extract, in a glass mortar, add 5 to 8 ml of water, and allow to soak for 15 minutes. Then triturate to a smooth paste, transfer to a 500-ml calibrated flask with the remainder of a total of 150 ml of water, and make up to 500 ml with absolute ethanol. Filter if necessary. Calcu- late results as the weight in milligrams present in each tablet of aloins, cascarosides and total glycosides as cascaroside A. (ii) to (v). Take 10 ml of this solution and proceed as in Appendix I (ii) to (v). Appendix 111 RECOMMENDED METHOD FOR THE CHEMICAL ASSAY OF CASCARA BARK Weigh accurately about 1 g of powdered cascara bark . . . . Extract the combined aqueous layers with three 60-ml portions of freshly prepared water-saturated ethyl acetate . . . . Determine the extinction of this solution in either a l-cm or 2-cm cell at . . . . Proceed as in Appendix I (i) to (v) but use the following modifications: Stage (i) line 1 Stage (ii) line 5 Stage (iii) lines 17 and 18 REFERENCES 1. 2. 3. Report of the Joint Committee of the Pharmaceutical Society and Society for Analytical Chemistry, Report of the Joint Committee of the Pharmaceutical Society and Society for Analytical Chemistry, Fairbairn, J. W., and Simic, S., J . Pharm. Phavmac., 1964, 16, 450. Analyst, 1965, 90, 582. Ibid., 1967, 92, 593.

ISSN:0003-2654    

DOI:10.1039/AN9689300749

出版商:RSC    

年代:1968

数据来源: RSC