摘要:

JULY, 1963 Vol. 88, No. 1048 PROCEEDINGS OF THE SOCIETY FOR ANALYTICAL CHEMISTRY SCOTTISH SECTION AND PHYSICAL METHODS GROUP ,2 JOINT Meeting of the Scottish Section with the Physical Methods Group was held on Friday, Ma\- loth, 1963, at the Education Centre of Imperial Chemical Industries Ltd., Nobel Division, St evenst on, A yrshirc. The subject of the meeting was “Modern Aspects of Electro-analytical Chemistry.” The Chair at the morning session was taken by Dr. W. Anderson Caldwell, F.R.I.C., and the following papers were presented and discussed : “Differential Electrolytic Potentiometry ; a Survey and Appraisal of Analytical Applications,” by E. Bishop, B.Sc., A.R.C.S.T., F.R.I.C., R. G. Dhaneshwar, BII.Sc., Ph.D., J . M. Ottaway, R.Sc., and G. D. Short, B.Sc.; “Controlled- potential Coulometric Analysis,” by G.W. C. Milncr, D.Sc., F.R.I.C., A.Inst. P. ; “Electro- chemical ,4pproaches in Automatic Analysis,” by G. Mattock, R.Sc., Ph.D., F.K.I.C. The Chair at the evening session was taken by the Chairman of the Scottish Section, Dr. R. A. Chalmers, and the following papers were presented and discussed : “Polarographic Determina- tion of Oxygen by ( a ) the Wide-bore Dropping-mercury Electrode and (b) Solid Electrodes,” by K. Briggs, Grad.I.E.E., and G. Knowles, B.A., B.Sc. ; “Chromatopolarography and its ,2pplication to Chemical Analysis,” by G. F. Reynolds, M.Sc., Ph.D., M.R.S.H., F.R.I.C. (see summaries below). The session concluded with a discussion on the use of dead-stop titrimetry in industrial analysis, which was opened by A.F. Williams, B.Sc., F.R.I.C. During the afternoon a visit was paid to the Laboratories of Imperial Chemical Industries Ltd., Nobel Division. DIFFERENTIAL ELECTROLYTIC POTENTIOMETRY; A SURVEY AND ,APPRAISAL 01. ANALYTICAL APPLICATIONS MR. E. BISHOP defined differential electrolytic potentiometry as being a method for the study of electrode processes and titrimetric reactions. I t consisted in observing the potential produced across a pair of indicator electrodes immersed in the stirred titration solution by the passage of a minute heavily ballasted current. The response for reversible reactions was in the form of a first differential peak. When zero-current potentiometry gave favourable results, differential electrolytic potentiometry offered the advantages of simplicity, elimination of standard cell and salt bridge, high-speed response and enhanced precision and discrimination, but it would continue to give excellent results when zero-current potentiometrv failed because of unfavourable reaction constants or low titrant concentrations.Differential electrolytic potentiometry was further uniquely advantageous on the microscale and with samples of extreme dilution ; determinations of 10-l1 mole had been satisfactorily conducted both volumetrically and coulomctrically in volumes of 0.5 to 5 ml. He described the method, discussed the operational parameters and assessed its special virtues. After presenting illustrative results for all types of electron-transfer and ion-combination reactions, he gave an appraisal of the method under normal and adverse conditions and on the microscale (see, e.g., Analyst, 1962, 87, 845 and 860, and other parts of the same series).49 1492 PROCEEDINGS [Analyst, Vol. 88 CONTROLLED-POTENTISL COULOMETRIC ANALYSIS DR. G. W. C. MILNER discussed the principles and potentialities of this technique and described apparatus suitable for this kind of analysis. He described the advantages of this technique in unusual analytical problems in metallurgical analysis and gave a general consideration of applications in metallurgical and inorganic analysis. Finally he attempted a critical assessment of the value of controlled-potential coulometry in electro-analysis. ELECTROCHEMICAL APPROACHES IN ,\UTOMATIC ANALYSIS DR. G. MATTOCK said that automatic: analysis was fundament ally more appropriate to plant process control than to laboratory control, since it bypassed the laboratory altogether in logical application.He therefore placed emphasis on automatic equipment suitable for the plant rather than the laboratory, although he recognised that there were many circumstances in which laboratory analyses were unavoidable. In the application o€ electrochemical methods, instruments such as automatic titrators were well known both for laboratory and plant purposes, but there were several inherently simpler measurement principles that lent themselves more readily to continuous operation, and therefore to automatic control, that had not been fully exploited. He described some recent work on electrochemical instruments, with chief reference to the sample- handling techniques necessary to render the operations automatic.The topics discussed by the author were: direct measurements by glass electrodes, e.g., for sodium, potassium and ammonium, continuous potentiometric monitoring, e.g., of cyanides, simple continuous titration techniques, continuous plant coulometry and plant polarography. POLAROGRAPHIC DETERMINATION OF OXYGEN BY (a) THE WIDE-BORE DROPPING-MERCURY ELECTRODE AND ( b ) SOLID ELECTRODES MR. R. BRIGGS said that the concentration of dissolved oxygen was of paramount importance as an index of the condition of a natural water, and measurements of this constituent were necessary in almost every phase of the work of the Water Pollution Research Laboratory. The chemical methods formerly used were being superseded in many applications by physical methods.He described the recent developments in two such methods, one in which the wide-bore dropping-mercury electrode was used, and the other in which membrane-covered solid electrodes were used. The wide-bore dropping-iizercwy electrode-This was based on the classical dropping- mercury electrode, but had the advantages of greater reliability, stability of calibration and greater current-carrying capacity. It consisted of an 0.8-mm capillary, waterproofed by means of a hard silicone coat, set at 45" to the horizontal and supplied with mercury through a hydraulic resistance. A silver - silver chloride reference electrode, a 16-volt source and a meter, complete the circuit. The electrode was kept at -1.5 volts with respect to the solution, and temperature compensation was provided by means of ther- mist ors.A reproducibility of t-2 per cent. over a 30" C temperature range was achieved regardless of the nature of the sample, provided a mixed reagent was added; this increased the sample conductivity, suppressed maxima and increased the pH of the sample to a value greater than 7.0. Instruments based on this system were used extensively to indicate and record dissolved oxygen. A further application was to the measurement of the rate of respira- tion of the contents of "mixed liquor" tanks in which sewage in intimate contact with a biologically active floc was purified by vigorous aeration. The electrode had also been used to study liquid movements associated with polarographic maxima (G.Knowles and M. G. Keen, J . Electroanal. Chem., in the press). Solid electrodes-The dropping-mercury electrode was sensitive to turbulence in the sample, and solid electrodes were more convenient for direct measurement of dissolved oxygen in streams. Many workers had devised systems applicable to their particular problems. The most successful of these were based on the diffusion of oxygen to a polarised electrode through a permeable membrane and a film of suitable electrolyte. The theory of diffusion current for these systems had been discussed, and a solid electrodeJuly, 19631 PROCEEDINGS 493 with a response time of the order of seconds described by K. H. Mancy, D. A. Okun and C. N. lieilley ( J . Electroanal. Chenz., 1962, 4, 65).Recently a galvanic-cell oxygen analyser, comprising a perforated silver foil cathode (27 sq. cm in area) surrounding a lead anode formed by compressing 150 g of lead shot into the form of a cylinder, had been described by F. J. H. Mackereth (Brit. Pat. Applic., 1962, 19294). The membrane consisted of polythene tubing 0.003 inch thick and the electrolyte was saturated potassium hydrogen carbonate solution. The cell had been found to have a reproducibility of 50.5 per cent. over periods of several months, and the current output was at least 300 pA when in equilibrium with water saturated with air at 20” C. This comparatively large current made temperature compensation practicable, so overcoming the main disadvantage of membrane-covered electrodes, the temperature coefficient of at least 5 per cent.per O C. By using a thermistor bridge circuit without an amplifier, compensation to within -+I per cent. was possible over any 10” C range, and portable indicating and recording instruments with three ranges (0” to 12” C, 10” to 22” C,, 22” to 40” C) had been produced. CHROMATOPOLAROGRAPHY AND ITS APPLICATION TO CHEMICAL ANALYSIS DR. G. F. REYNOLDS briefly discussed the principles of chromatopolarography to illustrate that it was a technique in which both chromatography and polarography were applied to the simultaneous determination of components in a sample solution. He gave details of the apparatus and mentioned its important aspects, and also outlined the special requirements of the eluting solutions, which must act as both eluent and base electrolyte, and discussed the problems of de-oxygenation. Dr.Reynolds described several applications to illustrate the value of the technique. He dealt in particular with its use in conjunction with reversed-phase chromatography, where the advantages of the method became apparent in the separation of mixtures of similar organic compounds whose half-wave potentials were almost coincident (see \;Ir. Kemula, in “Proceedings of the International Symposium on Micro- chemistry, Birmingham, 1958,” Pergamon Press, Oxford, 1960, p. 248). He ended by discussing some fundamental studies. WESTERN SECTION AND ATOMIC ABSORPTION SPECTROSCOPY DISCUSSION PANEL A JOINT Meeting of the Western Section and the Atomic Absorption Spectroscopy Discussion Panel of the Physical Methods Group was held at 2.15 p.m. on Wednesday, April 24th, 1963, in the Department of Physical and Inorganic Chemistry, The University, Bristol. The Chair was taken by the Chairman of the Atomic Absorption Spectroscopy Discussion Panel, Rilr. W. T. Elwell, F.R.I.C. A discussion was initiated by J. A. F. Gidley, B.Sc., A.1nst.P. MICROCHEMISTRY GROUP THE fortieth London Discussion Meeting of the Group was held at 6.30 p.m. on Wednesday, May 29th, 1963, at “The Feathers,” Tudor Street, London, E.C.3. The Chair was taken by the Chairman of the Group, Mr. D. W. Wilson, M.Sc., F.R.I.C. A discussion on “Kjeldahl Nitrogen-the Digestion Process” was opened by P. R. W. Raker, M.Sc., A.R.I.C., and S. Jacobs, M.Sc., Ph.D., F.R.I.C.

ISSN:0003-2654    

DOI:10.1039/AN9638800491

出版商:RSC    

年代:1963

数据来源: RSC

摘要:

494 ANDERSON : FORMATION OF NITROL~S OXIDE FROM HYPOSITRITE [Analyst, Vol. 88 The Formation of Nitrous Oxide from Hyponitrite BY J. H. ANDERSOK (Long Ashtoit Reseavch Station, Bvistol) Solutions of sodium hyponitrite buffered at pH values from 2 to 12 pro- duced nitrous oxide spontaneously only in the pH range 6 to 12. The similarity in range and in symmetry about the optimum (pH 9) between the curves relating pH (2) to the initial rate of formation of nitrous oxide from hyponitrite and (ii) to the calculatecl percentage of hydrogen hyponitrite in a hyponitrite solution suggests that nitrous oxide is formed spontaneously from hydrogen hyponitrite, but not Erom hyponitrous acid or the hyponitrite ion. The amount of nitrous oxide formed accounted for the amount of sodium hyponitrite decomposed.The spontaneous formation of nitrous oxide from solutions of sodium hyponitrite buffered between pH 7 and 11 is a pseudo first-order reaction. The amount of nitrate formed from small amounts of hyponitrite by a two-stage oxidation by permanganate was approximately equivalent to hyponitrite-nitrogen. A possible mechanism for the decom- position of hydrogen hyponitrite into nitrous oxide is suggested. THE inorganic nitrogen compound hyponitrite is reputedly unstable in aqueous neutral and acid solutions, giving rise mainly to nitrous oxide, although some workers have claimed that other inorganic nitrogen compounds are produced. I t is, however, stable in alkaline solution. Frear and Burrel1,l for instance, found that a hypmitrite solution did not de- compose at pH 11.3, but at pH 7.3 it produced nitrous oxide spontaneously and rapidljr without forming other nitrogen compounds.Chaudhary, Wilson and Roberts2 suggested that hyponitrous acid decomposed to nitrous oxide as almost the sole nitrogenous product. The work described here was undertaken to investigate the rate of formation of nitrous oxide as a function of pH, and to determine the products of reaction. Hyponitrite can be determined by oxidation to nitrite with a known excess of alkaline permanganate and subsequent oxidation of nitrite to nitrate after the mixture has been a~idified.~ This procedure was used here to study the relation between hyponitrite added and nitrate formed. EXPERIMENTAL PREPARATIOK OF HYPONITRITE- Solid hyponitrite was prepared as the sodium salt by Addison, Gamlen and Thompson’s method4 and as the silver salt by Medina and Xicholas’s method.5 These preparations were not contaminated by nitrite, nitrate or ammonia ; they were free from hydroxylamine, giving no reaction when tested as described by €rear and Burrel1,l and only small amounts of azo-dye after iodine oxidation by a modification of Czsaky’s method.6 The sodium hyponitrite preparation contained carbonate.A solution of sodium hyponitrite was prepared from the silver salt by adding the solid to a small excess of 0-5 N hydrochloric acid. After the precipitate of silver chloride had been removed by centrifugation, a sample of the clear supernatant liquid was rapidly mixed with an excess of sodium hydroxide solution to rnake the sodium hydroxide concentration 0.2 x.This solution, and a solution prepared by dissolving solid sodium hyponitrite in 0-2 N sodium hydroxide, had a light absorption peak at 248 mp, which is in agreement with previous findings for hyponitrite4 Solutions of solid sodium hyponitrite were used for the experiments described below. Reagents were of analytical grade and de-ioiiisecl water was used. FORMATIOX OF NITROUS OXIDE- The formation of nitrous oxide by decomposition of hyponitrite was followed mano- metrically at 30” C in conventional Warburg constant-volume respirorneters7 and 16-ml flasks. In the experiment represented in curie A, Fig. 1 (see p. 498), flasks contained 2.3 ml of buffer solution of known pH in the range 2 to 12 (prepared with use of a wide-range glass electrode) and 0.2 ml of 2 N potassium hydroxide $lus filter-paper in the centre well.Buffer solutions of pH 2 to 7.9 contained 0.1 RI pyrophospliate and 0.1 M citrate; buffer solutions of pH 8 to 12 contained 0.1 M pyrophosphate, 0.1 M borate and 0-1 hi carbonate.July, 19631 XSDERSOS FORMATION O F NITROUS OXI1)E FROM HYI'ONITRITE 495 After tempcrature equilibration, 0.2 ml of a solution of sodium hyponitrite (43-? mg of solid sodium hyponitrite dissolved in 8ml of 0-1 N sodium hydroxide) was tipped in, and flasks were shaken at 80 cycles per minute. The pH values of the reactions were measured after the experiments; these are the values shown on curve A, Fig. 1. The increase in pH caused by the formation of hydroxyl ions from hyponitrite was not greater than 0.3 pH unit.The initial rate of a reaction was calculated from its reaction velocity curve at the point where the formation of nitrous oxide was 15 per cent. of the total evolved. similar technique was used in tlie nitrogen balance experiments recorded in Table I, except that the reaction was begun by adding 8.1 mg of solid sodium hyponitrite through the side-arm of a flask to 3 ml of 31 phosphate buffer solution or 0.1 r\; sodium hydroxide in the main compartment. TABLE I RETATIOS BETWEE?; I-IYPONITKITE ADI)ED, ?; ITKOUS OXIDE FORMATIO?; AT I'AKIOVS pH \7ALLTES AXD IIESIDC'AL HYPONITRITE s,o p1-I of buftcir formed, pmolcs 1 32.1 3 8.3 a -- 0.,.4 7 39.2 9 38.1 11 40- 1 (0.1 1~ NaOH) 18.3 Iiate o f ?;,O formation, pmolcs per 20 minutes x 1.6 3.2 IS 23 7 3 IVeight o f solid Sa,?;,O, added, mg Residual nitrate, pmoles 8.1 0.75 8.0 0.6 8-7 0 8.1 0 8.1 0 8.1 0 8.1 0 Residual hyponitrite determined Total inorganic in samples, nitrogen, pmolcs pn1oles 40.2 31.3 34.4 39.2 38.1 40.1 39.9 Carbonate in solid sodium hyponitrite was measured manomet rically by differential absorption (bj- potassium hydroxide) of caibon dioxide formed when a solution of the hyponitrite was added to 0.5 ?; sulphuric acid.D ETE RM I ?I' AT I o N o F I r\' o R G A N I c s IT EZO G E ?: c o 31 PO u N D s- (a) Ammonia in samples was determined as described by Conway.8 ( b ) Xitrite was determined colorirnetrically by the modification of Nicholas and ?jason's9 method described below. Samples containing 0.005 to 0.200 pmole of nitrite were diluted to 7 nil with water, and 1 ml of sulphanilamide reagent (I per cent.of sulphanilamide in 2 N sulphuric acid) was added. After 2 minutes, 2 ml of a 0.02 per cent. aqueous solution of N-l-naphthylethylenediamine dihydrochloride were added. The light absorption of the resulting azo-dye was determined after 10 minutes in an E.E.L. (Evans Electroselenium Ltd.) colorimeter with a green filter (OGRI ; maximum transmission at 530 mp), against a reagent blank solution. The amount of nitrite in the sample was found from a calibration curve for solutions of sodium nitrite treated in the same way. (c) Sitrate in the presence of nitrite was determined by a modification of Middleton's methodlo for quantitatively reducing nitrate to nitrite. A solution containing 0-25 to 5 pnioles of nitrate plus nitrite was diluted with water to 7.5ml in a round-bottomed glass tube (diameter 2.3 cm; length 10.5 cm). The addition of 2-5 ml of 4 N ammonium hydroxide brought the pH of the solution t o 11.6 and the volume to 10 ml.Nitrite present was deter- mined in a 0.25- to 1.0-ml sample diluted to 7 ml with water. The reduction of nitrate to nitrite was initiated by the addition of 108 mg of zinc powder, maintained in suspension by shaking the reaction mixture at 300 cycles per minute on a reciprocal shaker. The reaction was stopped after 5 minutes by spinning in a centrifuge at 500 g for 1 minute. Nitrite was determined in 0-25 to 1 ml of the clear supernatant liquid. Nitrite was also measured in the supernatant liquid of a reagent blank solution, in which nitrite and nitrate solutions had been replaced by water.This value, which accounted for traces of nitrite or nitrate present in the zinc powder, was always deducted from the value of nitrite determined. Before use, the method was examined: 1. The recovery of 0.25 to 5 pmoles of sodium nitrite (without added nitrate) was 90 to 96 per cent.; mean recovery 92 per cent. 2. The recovery of nitrite formed from 0-25 to 5.0pmoles of potassium nitrate (without added nitrite) was 89 to 96 per cent. of theoretical; mean recovery 91 per cent.496 ANDERSON: FORMATION OF NITROUS OXIDE FROM HYPONITRITE [APzdySt, VOl. 88 The similarity between the mean recoveries shows that the reduction of nitrate to nitrite is 99 per cent. complete. The mean recoveries were not affected by the presence of 25 pmoles of manganous sulphate. Manganous ions, although used by Middleton,lo therefore appear to be unnecessary. The recovery of added nitrite was not affected by phosphate or pyro- phosphate, but, because these compounds diminished the rate of reduction of nitrate to nitrite, it was necessary to re-determine the recovery of nitrite formed from added nitrate.Experi- mentally determined values for nitrate were found by multiplying the recovery of nitrite by the appropriate factors derived from these recovery experiments. Before the nitrate in samples of reaction mixtures was determined it was necessary to adjust them to pH 8 by adding hydrochloric acid or sodium hydroxide, 0.05 ml of 0.005 M phenolphthalein in 70 per cent. ethanol being used as indicator.The subsequent addition of ammonium hydroxide brought the solution within the pH range (11 to 11.6) required for the reduction of nitrate to nitrite. (d) Hyponitrite was determined by a two-stage oxidation with excess of permanganate3 and subsequent determination of the nitrate formed. One millilitre of hyponitrite was mixed with 0.5 ml of 2 N sodium hydroxide in a round-bottomed tube, as used for the nitrate deter- mination; 0.25 ml of 0.1 M potassium permanganate was then added, and the oxidation of hyponitrite to nitrite was allowed to proceed for 7 minutes (stage 1). The addition of 0.5 ml of a solution of 0.005 M manganous sulphate in G N sulphuric acid acidified the permanganate, which was then allowed to oxidise nitrite to nitrate for 7 minutes (stage 2 ) .A pre-determined volume (approximately 0-5 ml) of 0.2 M sodium oxalate, sufficient to reduce all the per- manganate to colourless manganous ions in 5 minutes, was added. After a further 2 minutes, 0.05 ml of phenolphthalein indicator was added down the side of the tube, and the mixture was adjusted to pH 8 by carefully adding 2 N sodium hydroxide solution with swirling. There was no nitrite present in this mixture. The mixture was diluted to 7.5ml with water, and then 2.5 ml of 4 N ammonium hydroxide were added, which brought the pH to 11.0 to 11.6. Zinc powder (100 mg) was then added, and the mixture was shaken for 5 minutes, as described in the nitrate determination. After the zinc powder and the precipitate of manganese hydroxide had been removed by centrifugation, nitrite was determined in samples of the clear supernatant liquid.Values were always corrected by subtracting nitrite found in a reagent blank solution, in which water replaced the hyponitrite solution. When solutions containing 0.5 to 5 pmoles of nitrite or nitrate were treated in this way, the recovery of added nitrite was 78 to 87 per cent. of theoretical (mean recovery 85 per cent .) and the recovery of nitrite formed from added nitrate was 75 to 86 per cent. of theoretical (mean recovery 83 per cent.). The similarity of the two mean recoveries suggests that nitrite is quantitatively oxidised to nitrate by acid permanganate. The observation that no nitrite was present in determinations before reduction with zinc and ammonia is consistent with this suggestion.The oxidation of hyponitrite by alkaline permanganate produces nitrite,3 but, since it is possible that nitrate may be simultaneously produced, the determination must be regarded essentially as a method of determining nitrate. When nitrate and nitrite were present with hyponitrite, their concentrations were subtracted from the amount of nitrate formed from hyponitrite (q., see Table I). A fixed amount (2pmoles) of nitrite was used in replicate determinations, and the recoveries were within the range 78 to 87 per cent. when the periods of reaction were varied independently within the time limits : oxidation with alkaline permanganate (stage l), 5 to 15 minutes; oxidation with acid permanganate (stage 2), 2 to 10 minutes; reduction by oxalate (total period), 7 to 10 minutes.These results show that in the determination there is no significant conversion of nitrate or nitriite to nitrogenous gases, which could disappear from solution. When the permanganate was reduced by acidified solutions of ferrous sulphate or sodium arsenite the recoveries of nitrite were low. Probably these reagents in acid conditions reduced nitrate to nitrogenous gases (nitrogen or nitrous oxide). It was important to minimise the amount of oxalate added, because a large excess diminished the rate of reduction of nitrate to nitrite. STOICHEIOMETKY OF HYPONITRITE OXIDATION- manganate, because nitrous oxide may also be produced. In the determination of hyponitrite, the crucial stage is the oxidation by alkaline per- Stage 1 was studied mano-July, 19631 ANDERSON: FORMATION OF NITROUS OXIDE FROM HYPONITRITE 497 K20,2- = 2N0,- metrically to measure nitrous oxide formed.from the equation- after the hyponitrite determination had been completed. The amounts of nitrate and nitrous oxide formed were then compared with the amounts of hyponitrite added, which was measured as nitrous oxide produced when a similar sample completely decomposed a t approximately pH 8. When 3-2 pmoles of hyponitrite in 0.2 ml of 0-2 N sodium hydroxide were added to a solution of 1 ml of 0.1 M potassium permanganate PLUS 1.3 ml of 0.16 N sodium hydroxide in a Warburg flask, 0.1 pmole of nitrous oxide was immediately formed, and 5.7 pmoles of nitrate were formed from a sample removed after 10 minutes.Thus, of the 3-22 pmoles of hyponitrite added, 89 per cent. was oxidised, 3 per cent. was converted to nitrous oxide and 8 per cent. could not be accounted for. Sitrous oxide formation is therefore not significant during alkaline oxidation. When, however, the hyponitrite solution was added to a solution of 1.3 ml of 0-16 N sulphuric acid plus 1 ml of 0.1 M permanganate, 78 per cent. appeared as nitrous oxide, 16 per cent. was oxidised to nitrate and 6 per cent. could not be accounted for. I t seems that permanganate catalyses the formation of nitrous oxide from hyponitrous acid in sulphuric acid. Similar results were obtained when sodium hyponitrite solutions prepared from silver hyponitrite were mixed with alkaline or acid permanganate. When hydroxylamine sulphate was added to alkaline or acid permanganate, nitrate and large amounts of nitrogenous gases were formed ; hydrazine sulphate formed nitrogenous gases, but no nitrate.RE su LTS Table I shows results of nitrogen-balance experiments in which the amount of hyponitrite added was related to the amount of nitrous oxide formed at various pH values and the residual amount of hyponitrite. The column, “total inorganic nitrogen,” gives the sum of nitrous oxide formed, the residual hyponitrite and the amount of hyponitrite converted to nitrate in the manometric reaction. Since the purpose of the experiments was to make a nitrogen balance, the duration of a reaction, in the range 2 to 6 hours, was not critical. During this period, e.g., after 2 hours in the experiment at pH 9, a sample was removed, mixed with sodium hydroxide, and the determination of hyponitrite was begun immediately.S o nitrite, nitrate, hydroxylamine or ammonia was found in further samples. The small amount of nitrate formed in reactions at pH 1 and 3 was typical of that occasionally produced at low pH values. The amount of nitrous oxide formed (mean, 39.1 pmoles) thus accounts for the hyponitrite disappearing. The sum of nitrous oxide formed and residual hyponitrite in 0.1 N sodium hydroxide is 39.9 pmoles. 3a2N20, + H20 = 2NaOH + N 2 0 39.1 pmoles correspond to 4-15 mg of sodium hyponitrite or 51 per cent. of the weight used. Carbonate (Xa,CO,) accounts for 36 per cent.; water might account for the remaining 13 per cent. The relation between the pH of the reaction and initial rate of formation of nitrous oxide is shown in Fig.1 (curve A). Formation of nitrous oxide occurs in the pH range 6 to 12 with a symmetrical optimum at pH 9. In similar experiments lasting 90 minutes, in which hyponitrite solutions were added to 0-5 N sulphuric acid or 0.2 s sodium hydroxide, there was negligible formation of nitrous oxide. The amounts of nitrous oxide formed when reactions were allowed to proceed to com- pletion (approximately 90 minutes), were the same throughout the pH range 7 to 11 (mean value 9.1 pmoles i 4 per cent.). No hyponitrite, nitrate, hydroxylamine or ammonia and only about 0.005 pmole of nitrite was found in the residual reaction mixtures. In reactions at pH values outside this range, similar amounts of nitrite, but no nitrate, hydroxylamine or ammonia, were found.The mean value of nitrous oxide was obtained from 0-2 in1 of hyponitrite solution (43-2 mg per 8 ml). This solid preparation thus contained about 90 per cent. of sodium hyponitrite. Results similar to those in Fig. 1 (curve A) were obtained with a solution of sodium hyponitrite prepared from silver hyponitrite. The amount of hyponitrite oxidised was found At pH 7, 9 and 11 the hyponitrite rapidly and completely decomposed. According to the equation-498 ANDERSON : FORMATION O F NITROUS OXIDE FROM HYPONITRITE [Analyst, 1701. 88 The relation between the rate of formation of nitrous oxide and pH in Table I bears a general resemblance to Fig. 1 (curve A). The finding (see Table I) that nitrous oxide was formed rapidly at pH 1 contrasted with the general observation throughout this work (see Fig.1, curve A) that hyponitrite was stable in acid solutions. Possibly some impurity in the reagents catalysed this reaction and the formation of small amounts of nitrate (pH 1 and pH 3 in Table I). At pH 3 and pH 5 the sum of the nitrous oxide appearing and the residual hyponitrite is less than the expected amount of 39.1 pinoles. I t is likely that at these pH values some formation of nitrous oxide occurred as the acid reaction mixtures added to sodium hydroxide (stage 1) passed transiently through a pH range of 6 to 12. Fig. 2 shows the relation between the logarithm of the hyponitrite concentration remain- ing at time t (minutes) after the initiation of a manometric experiment at pH 8-6 and t .The amount of hyponitrite present at time t was found by subtracting the manometric reading at time t from the stationary reading observed at the end of the experiment. The linear relationship shows that the spontaneous decomposition of hyponitrite is a pseudo first-order reaction. DISCUSSION OF THE WORK Similar straight lines were obtained at pH values from 7 to 11. C haudhary, Wilson and Roberts2 found that, at pH 7.3, hyponitrite disappears according to a first-order reaction. Frear and Burrelll found that hyponitrite is stable at pH 11.3, but at pH 7.3 it decomposes with a half-life of 11 minutes, forming only nitrous oxide. The finding (Fig. 1, curve A) that formation of nitrous oxide occurs spontaneously in the pH range 6 to 12 is in agreement with these results.Since the reactions in the pH range from 7 to 11 are complete with the formation of no nitrate, hydroxylamine or ammonia and onljr traces of nitrite (Fig. 1, curve A), the amount of nitrous oxide spontaneously formed may be used as a measure of the decomposition of hyponitrite. In residual reaction mixtures outside this pH range no formation of nitrogen compounds other than nitrous oxide was found. Formations of nitrous oxide from hyponitrite in acid conditions were occasionally observed (e.g., pH 1 of Table I). These reactions and formation of nitrate may be catalysed by some impurity in the reagents. PH Fig. 1. The relation between pH and: curve A, the initial rate of formation of nitrous oxide from hyponitrite; curve B, the calculated percentage of hydrogen hyponitrite 2.4 2.0 c 0 .- U m Y c 1.6 s 5 1.2 2 0 U L Y .- .- a x r -1 0.8 0.4 I I I I I 10 20 Time, minutes 0 Fig.2. The relation between log,, hyponitrite concentration remain- ing at time t and t . Reaction mixture: volume 2.5 ml; pH 8.6; 7.8 pmoles of hyponitrite. The rate constant was 0.152 minutes-l for a first-order reactionJuly, 19831 ANDERSON : FORMATION OF NITROUS OXIDE FROM HYPONITRITE 499 Latimerll gives the values below for the acid dissociation constants of hyponitrous acid- H,X20, = HN,O,- + H+; K, = 9 x lo-’ Hh’,O,- = N,O;- + H+; K, = 1 x 10-l’ The percentage of hydrogen hyponitrite ion in a hyponitrite solution at a given hydrogen ion concentration can be calculated from the equation- where [A: = concentration of K2O,2-, [HA] = concentration of HN,O,-, [H,A] = concentration of H2T\j20z, r,4,] = [A] + [HA] -1 [H,A] and ;HI = concentration of hydrogen ions.The relation between the pH of a solution of hyponitrite and the percentage of hydrogen hjrponitrite ion calculated in this way (see Fig. 1, curve B) is a symmetrical curve (pH 5 to 13) with a peak at pH 9. The similarity between the observed results (Fig. 1, curve A) and the theoretical findings (Fig. 1, curve B) suggests that the initial rate of nitrous oxide formation from hyponitrite is a function of the hj-drogen hyponitrite ion concentration and that hypo- nitrous acid and the hyponitrite ion are stable. The equation below would thus describe the spontaneous decomposition of hyponitrite- HY202- = N,O + O W . The spontaneous formation of nitrous oxide from hyponitrite at pH values in the range 7 to 11 (e.g., pH 8.6, Fig.2) is a pseudo first-order reaction. This is in agreement with previous findings.l y 2 The molecule of hydrogen hyponitrite, HS,O,-, contains one hydroxyl group with the hydrogen dissociated and one with the hydrogen not dissociated. Possibly the spontaneous decomposition of hydrogen hj-ponitrite proceeds by an intra- or inter-molecular mechanism involving bonding between 0- and OH. The oxidation of hyponitrite to nitrite by alkaline permanganate produces so little nitrous oxide that the nitrate subsequently formed from nitrite bjr acidifying the permanganate is approximately equivalent to the hyponitrite-nitrogen added (e.g., see Table I). In using alkaline permanganate for determining hyponitrite, it is obviously necessary to take the precaution of measuring any nitrogenous gases produced. I thank Dr. C. F. Timberlake for assistance with the calculations. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. REFERENCES Frear, D. S., and Burrell, R. C., Plant Physiol., 1958, 33, 105; Anal. Chenz., 1955, 27, 1664. Chaudhary, M. T., Wilson, T. G. G., and Roberts, E. R., Biochirn. Biophvs. Acta, 1954, 14, 207. Partington, J. R., “General and lnorganic Chemistry.” Second Edition, Macmillan & Co. Ltd., Addison, C. C., Gamlen, G. A . , and Thompson, R., J. Chenz. SOC., 1952, 338. Ncdina, A., and Nicholas, D. J . D., Natuve, 1957, 179, 533. Czsaky, T. Z., Acta Chem. Scand., 1948, 2, 450. t;mbriet, W. W., Burris, R. H., and Stauffcr, J . F., “Manometric Techniques,” Burgess Publishing Conway, E. J . , “Microdiff usion Analysis and Volumetric Error, ” Third Edition, Crosby Lockwood Nicholas, D. J . D., and Nason, A , , in “Methods of Enzymology,” Academic Press Inc., Xeu York Middleton, I<. R., Claem. & I n d . , 1957, 1147. I,atimer, W. M., “The Oxidation States of the Elements and their Potentials,” Second Edition, Received November 23vd, 1962 London, 1951. Co., Minneapolis, 1957. and Son I,td., London, 1950. and London, 1957, Volume 3, p. 981. Prcntice Hall, New York, 1952, p. 95.

ISSN:0003-2654    

DOI:10.1039/AN9638800494

出版商:RSC    

年代:1963

数据来源: RSC

摘要:

500 NEWMAN : PRECIPITATION OF CUPROUS [Analyst, Vol. 88 Precipitation of Cuprous Thiocyanate from Homogeneous Solution BY E. J. NEWMAN (Department of Chemistry, Sir Johu! Cass College, London, E.C.3 and Hopkin & Williams Ltd., Freshwuter Road, Chadwell Heath, Essex) A method for precipitating cup rous thiocyanate from homogeneous solution is described, in which the character of the precipitate is governed by the rate of the reduction of cupric iclns to the cuprous state. The method is simple and accurate; its main advantages over conventional methods are an over-all reduction in the time required for the determination, and the production of a precipitate that is easy to manipulate and to collect by filtration. THE insolubility of cuprous thiocyanate and the relatively high solubilities of the thiocyanates of many other metals provides a selective method for quantitatively separating copper from a solution containing other metals, as well as providing a gravimetric method for its deter- mination. Consequently, since its introduction by Rivot in 1854,l the quantitative precipita- tion of cuprous thiocyanate has been widely applied.Rivotl precipitated cuprous thiocyanate from hydrochloric acid solution with potassium thiocyanate, after reduction of cupric copper tcl the cuprous state with sulphite. He applied the method to the determination of copper in bronzes, but gave few practical details. This method was used by van r\;ame2,3,4 for determining copper in coinage metals. He reported that only a small amount of free acid should be present during precipitation and that the precipitate should be set aside for 20 hours before filtration.Precipitation of tin, arsenic, antimony and bismuth was prevented by adding tartaric acid. Kolthoff and van der Meene5 studied in detail the conditions for precipitating cuprous thiocyanate. They concluded that the concentration of free acid should not exceed 0-5 N and that the concentration of the excess of thiocyanate precipitant should be less than 0.05 N. Reduction of cupric ions with sulphite was best performed in boiling solution. Similar conclusions on the maximum permissible concentrations of acid and excess of precipitant were reported by Belcher and West,6 who used ammonium ferrous sulphate as the reductant in the cuprous thiocyanate method.Most workers consider it necessary to age the precipitated cuprous thiocyanate before filtration, although there are disagreements about the recommended time periods. Thus 2,’ 6 , 6 9 8 20,293y4 245 and “~everal”~ hours have been specified. Stathislo used ascorbic acid to reduce cupric ions in the thiocyanate method. Whereas previous workers2 9 5 had found that oxidising agents must be excluded, Stathis obtained quantitative precipitation of cuprous thiocyanate from nitric acid solutions of copper alloys. Moreover, the time period between precipitation and filtration was reduced to 30 minutes. EXPERIMENTAL The main disadvantage of the cuprous thiocyanate method is that the rather fine precipitate generally produced may clog or even pass through the filter. I t has a marked tendency to “creep” with consequent risk of loss in transferring it to the filter and difficulty during washing.Heating the solution, even to boiling, with constant stirring, improves the granularity of the wetted precipitate, but is only partly effective in dealing with the precipitate that has crept up the walls of the container. The work described here was undertaken to produce cuprous thiocyanate with good filtration characteristics and that does not require ageing before filtration. The technique of precipitation from homogeneous solutionll depends on the production of one of the reactants in the presence of the others at a rate such that a controlled growth of the crystals of the insoluble product results. To achieve homogeneous precipitation of cuprous thiocyanate, it was decided to control the rate of production of cuprous ions.Three methods of approach can be envisaged for achieving this, namely, (1) the slow liberation of cupric ions from a complex compoundJuly, 19631 THIOCYANATE FROM HOMOGENEOUS SOLUTION 501 into a reducing medium, (2) the generation of a reducing agent in a solution containing cupric ions and (3) the production of conditions under which the reaction of cupric ions with a reducing agent is delayed. Successful precipitation of cuprous thiocyanate from homogeneous solution was achieved by using each of these three approaches. Thus, it was possible to precipitate cuprous thio- cyanate from homogeneous solutions containing amminocupric ions. Ammonia solution was added to cupric sulphate solution until the cupric hydroxide precipitate redissolved.The solution remained clear when ammonium thiocyanate and sodium sulphite were added, but when the solution was heated ammonia was expelled and cuprous thiocyanate was precipi- tated. This procedure would, however, have no practical value for determining copper in the presence of metals having insoluble hydroxides. The controlled generation of a reducing agent was obtained by the hydrolysis of sucrose. Sucrose was added to slightly acidic solutions containing cupric and thiocyanate ions. The solutions remained clear at room temperature, but when they were heated precipitation of cuprous thiocyanate began after a few minutes. The cupric ions were reduced by the glucose and fructose produced by hydrolysis of the sucrose.In this way highly crystalline cuprous thiocyanate with excellent filtration characteristics was produced. Unfortunately, it was impossible to precipitate the copper quantitatively in less than 4 hours, and strict control of pH was necessary. Solutions having acidities in excess of 0-03 N in hydrochloric or sul- phuric acid gave darkened precipitates, and the supernatant liquid developed an opalescence and a sulphurous odour. This was attributed to further reaction of cuprous thiocyanate with fructose to produce cuprous sulphide. Nevertheless, with proper control of the condi- tions, successful gravimetric determinations of copper were made, but this method was abandoned because of practical difficulties. Delayed reduction of the cupric ion was reported by Davis12 in the homogeneous precipi- tation of cuprous tetraphenylborate.Ascorbic acid was added to acid solutions containing cupric and tetraphenylborate ions, and precipitation did not occur until about half a minute had elapsed. This method was not applicable to cuprous thiocyanate ; immediate precipita- tion occurred under a wide range of temperature and acidity conditions when ascorbic acid was added to solutions containing cupric and thiocyanate ions. With hydroxyammonium chloride, which is a milder reducing agent than ascorbic acid, delayed reduction of cupric ions was achieved by temperature control. A solution of hydroxy- ammonium chloride was added to a solution of cupric and thiocyanate ions, with an acidity of about 0.2 N, without precipitation occurring.When this solution was heated on a steam- bath, precipitation of cuprous thiocyanate began within a few minutes and appeared to be complete within about half an hour. The cuprous thiocyanate produced in this way was granular and had excellent filtration properties. A study of the conditions for precipitation resulted in the development of an accurate method for the gravimetric determination of copper. INVESTIGATION OF PRECIPITATION CONDITIONS- An investigation of the reaction variables was carried out to find conditions under which precipitation of copper was complete within the limits of good gravimetric accuracy. These conditions were then applied to gravimetric determinations on solutions containing known amounts of copper, to check the procedure.Preliminary experiments were performed in which completeness of precipitation was judged visually. When ammonium or potassium thiocyanate is added to a solution of a cupric salt, a green colour is produced, presumably due to the formation of a complex between cupric and thiocyanate ions. The disappearance of this colour when the complex is treated with hydroxyammonium chloride and heated is a good indication that precipitation is com- plete. These experiments led to the adoption of the reagents used in the subsequent tests. It was found expedient to use a colorimetric micro-titration procedure for determining the amount of copper remaining in the filtrate after precipitation of the cuprous thiocyanate. Portions (50 ml) of a 1 per cent. solution of cupric sulphate were transferred to separate 150-ml Phillips beakers. To each solution were added in the stated order the amounts of hydrochloric acid, sp.gr.1.18, 3.5 per cent. ammonium thiocyanate solution and 3 per cent, hydroxyammonium chloride solution specified in Table I, the solutions being stirred after each addition. The solutions were heated on a steam-bath for 30 minutes, cooled to room temperature and set aside for 10 minutes.502 KEWhlAN PRECIPITATION OF CUPROUS [Analyst, Vol. 88 After filtration through No. 4 porosity sintered-glass crucibles, 5 ml of each filtrate were transferred to a separating funnel containing 8 ml of water and 2 ml of 5 TU’ sulphuric acid, and treated as described below. Five millilitres of a 0.1 per cent. solution of diethylammonium diethyldithiocarbamate were added, and the mixture was shaken vigorously for 2 minutes and allowed to separate.Any copper remaining in the 5-ml portion of filtrate was extracted into the organic layer as the intense yellow copper dithiocarbarnate. A blank solution, containing 50 ml of water in place of the cupric sulphate solution, was prepared, and, with each test, 5 ml were taken through the procedure described above. The blank solution was titrated from a 2-ml burette with dilute standard cupric sulphate solution (1 ml G 10 pg of copper), the mixture being shaken vigorously after each increment and allowed to separate. This was continued until the colour of the organic layer of the blank matched that of the test. The volume o€ titrant required was therefore a measure of the amount of copper in 5 ml of filtrate.RE s r; LTS- negative error due to the amount of copper remaining in the whole filtrate. Results obtained in this way are shown in Table 1. The last column shows the calculated TABLE I EFFECTS O F DIFFEREXT CONCEPiTKATIONS O F THE REAGENTS Hydrochloric added, ml 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 0.5 1.0 1.5 1.8 2.0 acid Ammonium thiocyanate solution added, ml 7.5 10 12.5 15 20 15 15 15 15 15 15 15 15 15 Hydroxyammonium chloride solution added, ml 15 16 15 15 16 5 10 15 2 0 15 15 15 15 16 Copper in filtrate, 0.17 0.06 0.05 0.03 0-05 0.13 0.04 0.03 0.03 0.03 0-03 0.03 0.09 0.25 Yo DISCUSSION OF RESULTS- 15ml each and about The oDtimum conditions derived from the results shown in Table I reuuire I I of ammonium thiocyanate solution and hydroxyarnmonium chloride solution, 1 ml of hydrochloric acid.The concentration of acid should not exceed 0.3 N, and the concentration of the excess of ammonium thiocyanate should not exceed 0.07 M. Similar results were obtained by using a 4.5 per cent. solution of potassium thiocyanate (= 3-5 per cent. ammonium thiocyanate) and with sulphuric or perchloric acid instead of hydrochloric acid. It was also found that nitric acid in concentrations less than 0-1 N did not affect the results. When the optimum reagent concentrations were used it was found that 20 minutes’ heating was insufficient (error 0-33 per cent.), but that heating for longer than 30 minutes was unnecessary. METHOD FOR THE GRAVIMETRIC DETERMINATION OF COPPER The precipitation conditions derived were applied to the gravimetric determination of copper in solutions prepared from pure electrolytic copper foil and copper-bearing alloys. After the cuprous thiocyanate had been precipitated and collected, it was washed first with a dilute solution of ammonium thiocyanate containing a little ascorbic acid (to maintain reducing conditions) and finally with dilute ethanol to remove any excess of ammonium thiocyanate.JQ, 19631 THIOCY=1NATE FROM HOMOGEXEOUS SOLUTION 503 PROCEDURE- Transfer 50 nil of acidified copper solution, containing about 125mg of copper and having an acid concentration of less than 0-3 N, to a clean unscratched 150-ml Phillips beaker.Add, with mixing, 15 ml of ammonium thiocyanate solution and 15 ml of hydroxyammonium chloride solution.Heat the solution on a steam-bath for 30 minutes, cool it to room tem- perature, and set aside for 10 minutes. Collect the precipitate on a previously dried and weighed So. 4 porosity sintered-glass crucible, and use the wash solution to complete the transfer of the precipitate. byash tlie precipitate a few times with wash solution and finally with dilute ethanol. Dry the crucible and its contents for 2 hours at 110" C, set aside to cool, and weigh. 1 g of cuprous thiocyanate T 0-6226 g of copper. APPLICATIOK OF THE METHOD TO COPPER METAL Pure electrolytic copper foil (24395 g), freshly de-greased and washed with dilute nitric acid, was dissolved in 3 ml of nitric acid, sp.gr. 1-42, and 3 ml of water. When dissolution was complete, 5 ml of sulpliuric acid, sp.gr. 1-84, were added, and the mixture was evaporated to small volume to expel nitric acid.When it was cool, the mixture was diluted with 20 ml of water and again evaporated to small volume; this treatment was repeated twice. The solution was then transferred with water to a 1-litre calibrated flask and diluted to the mark with water at 20" C. This solution was about 0.18 N in sulphuric acid, and 50 ml contained 0.1220 g of copper. Twelve 604-ml portions were tested by the method described above, and the weights of copper calculated from the weights of cuprous thiocyanate obtained lay within the range 0.1219 to 0.1221 g with a mean value of 0.1220 g. By assuming the copper to be 100.0 per cent. pure, the recoveries were thus 100.0 per cent.to within k0.08 per cent. The standard deviation was 0.067 per cent. APPLICATION OF THE METHOD TO COPPER ALLOYS Four British Chemical Standard alloys of copper were analysed by this method. Two solutions, one in sulphuric acid and the other in perchloric acid, were prepared from two of the alloys and one solution was prepared from each of the other alloys. CUPKO-NICKEL 180/1- I n s.uZplzuric acid--A 0-9299-g sample of the alloy was placed in a 150-ml Kjeldahl flask, and 3 ml each of water and nitric acid, sp.gr. 1.42, were added. After the metal had dissolved, 1.5 ml of sulpliuric acid, sp.gr. 1.84, were added, and the solution was evaporated t o small volume to expel the nitric acid. The solution was then cooled, diluted with 50 ml of water, and neutralised to pH 4 with ammonia solution, sp.gr. 0.89.It was then transferred with water to a 250-ml calibrated flask, 15 ml of 5 N sulpliuric acid were added, and the solution was made up to the mark at 20" C with water. I n psrclzZoric mid--A 0.9301-g sample of the alloy was weighed into a tall 150-ml beaker and heated with 7-5 ml of perchloric acid, sp.gr. 1.54, until it dissolved. When cool, tlie solution was transferred with water to a 250-ml calibrated flask, and the solution was diluted to the mark with water. MANGANESE-BRASS "B" 179- 192 suZPJzuric acid-A 1-0601-g sample of the alloy was weighed into a 150-ml Kjeldahl flask, 4 ml each of water and nitric acid were added, and occasional heating was applied until dissolution was complete. Two millilitres of sulpliuric acid were added, and the solution was lieated to fumes for 5 minutes.When cool, the solution was diluted with about 20ml of water, and again lieated to fumes for 5 minutes. I t was then cooled and diluted with 50 ml of water. The turbid solution produced was filtered through a Whatman Xo. 42 filter-paper into a 250-ml flask, and the filter was washed with water. The filtrate and washings were diluted to the mark with water. In perclzloric acid-A 1.0727-g sample was transferred to a tall 150-ml beaker and dissolved, with warming, in 9.5 ml of perchloric acid. When cool, the solution was diluted with 50 ml of water and filtered through paper into a 250-ml calibrated flask. The filter was washed with water, and the filtrate and washings were diluted to the mark with water.Preparation of Alloy Solutions504 NEWMAN PRECIPITATION OF CUPROUS [Analyst, Vol. 88 BRONZE 183/1- A 0.7301-g sample was dissolved in a mixture of 8 ml of water and 12 ml of nitric acid contained in a 150-ml beaker. The solution was evaporated by heating on a steam-bath to reduce the volume to between 5 and 10 ml. It was then diluted with 50 ml of hot water, and digested on the steam-bath for a further 15 minutes. The precipitate of metastannic acid was removed by filtration through a No. 42 filter-paper, and the filter was washed 10 times with a few millilitres of dilute nitric acid (1 + 20). The filtrate and washings were evaporated to small volume (but not to dryness) ; the concentrated solution was then diluted with 50 ml of water, and neutralised to pH 4 with ammonia solution.The solution was transferred to a 250-ml calibrated flask, acidilied with 5 ml of perchloric acid, and diluted to the mark with water. BRONZE “C” 207- A 0.7302-g sample was heated in a tall 150-ml beaker with 9 ml of perchloric acid and 1.5ml of nitric acid. The solution was then heated to fumes until colourless, diluted with 50 ml of hot water, and digested for 15 minutes on a steam-bath. It was then filtered through No. 42 filter-paper (to remove the metastannic acid precipitate) into a 250-ml calibrated flask. The filter was washed with 100 ml of water containing 1 ml of perchloric acid, and the filtrate and washings were diluted to the rnark with water. Procedure Copper was determined in three 50-ml portions of each solution by the method described, except that 0.5 g of sodium hydrogen tartrate was added to each portion of the solution prepared from Bronze 183/1 before the other reagents were added.This alloy contained 0-24 per cent. of antimony and 0.14 per cent. of arsenic, both of which produce precipitates under the conditions of the test unless tartrate is added to complex them. The results obtained are shown in Table I1 and are compared with the certified copper contents. Alloy Cupro-nickel Manganese brass “B” Bronze Bronze “C” TABL:E I1 RESULTS OBTAINED WITH B.C.S. Method of reference preparing Copper number solution found, 180/1 In H,SO, 67-35 67.35 67.32 In HC10, 67.33 67.30 67-36 179 In H,S04 58.73 58.7 1 58.75 In HClO, 58.70 58.73 58.67 183/1 - 84.68 84-68 84-76 207 - 86.76 88-76 86.84 ?& RESULTS COPPER ALLOYS All the results obtained are in good agreement Mean Certified copper copper found, value, % % 67.34 67.36 67.33 58-73 58.8 58.70 86.79 86.84 Range of copper values certified, % 67.32 to 67-40 58.6 to 59.06 84.75 to 84.95 ~~ 86.65 to 87.00 with the certified copper contents, and the replicate determinations on each alloy agree well.alloys do not, therefore, interfere in the determination. metals are shown in Table 111. The other cons&ents of the The certified amounts of the otherJuly, 19631 THIOCYANATE FROM HOMOGENEOUS SOLUTION 505 TABLE I11 OTHER COKSTITUENTS OF THE ALLOYS Alloy Constituents other than copper Cupro-nickel 180/1 . . 30.85 per cent. of Ni; 0.82 per cent. of Fe; 0.81 per cent. of Mn 33-9 per cent. of Zn; 1.03 per cent.of Mn; 0.91 per cent. of Pe; Manganese brass “€3” 179 . . [ 1.62 per cent. of Al; 1.75 per cent. of Sn; 0.78 per cent. of Pb; . . I 1.01 per cent. of Ni 5.17 per cent. of Zn; 5.01 per cent. of Sn; 3.51 per cent. of Pb; 0.51 per cent. of P; 0.51 per cent. of Ni; 0.24 per cent. of Sb; 0.14 per cent. of As Bronze 183/1 . . Bronze “C” 207 . . . . 1 . . 9.80 per cent. of Sn; 2.53 per cent. of Zn; 0.41 per cent. of Pb CONCLUSIONS Precipitation of cuprous thiocyanate from homogeneous solution by controlled reduction of cupric ions to cuprous with hydroxyammonium chloride provides a quick, simple, accurate and specific method for the determination of copper. The precipitates produced by this method have good filtration characteristics and in this respect are far superior to those produced by the classical heterogeneous precipitation procedures, notably in that they do not produce the fine suspension of cuprous thiocyanate that displays the phenomenon of creeping up the sides of apparatus in which it is handled.Precipitation of copper in this way also provides a practicable alternative to electrolysis for removing copper from solutions before certain other metals are determined. It has been used for separating large amounts of copper from small amounts of iron, for the determination of traces of iron in analytical-reagent grade copper salts. It is hoped that this will form the basis of a further communication. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. REFERENCES Rivot, L. E., Compt. Rend., 1854, 38, 868. van Name, R. G., Amev. J . Sci., 1900, 10, 451. -, Ibid., 1902, 13, 20. -, Ibid., 1902, 13, 138. Kolthoff, I. M., and van der Meene, G. H. P., Z . anal. Chem., 1927, 72, 337. Belcher, R., and West, T. S., Anal. Chim. Acta, 1952, 6, 337. Kolthoff, I. M., and Sandell, E. B., “Textbook of Quantitative Inorganic Analysis,” Third Edition, Wilson, C. L., and Wilson, D. W., Editors, “Comprehensive Analytical Chemistry,” Elsevier Vogel, -4. I., “A Textbook of Quantitative Inorganic Analysis,” Second Edition, Longmans, Stathis, E. C., Anal. Chim. Acta, 1957, 16, 21. Gordon, L., Salutsky, M. L., and Willard, H. H., “Precipitation from Homogeneous Solution,” Davis, D. G., Anal. Chem., 1960, 32, 1321. The Macmillan Company, New York, 1952, p. 671. Publishing Co., Amsterdam and New York, 1962, Volume IC, p. 370. Green and Co. Ltd., London, 1951, p. 431. John Wiley and Sons Tnc., New York, 1959. Received December ISth, 1962

ISSN:0003-2654    

DOI:10.1039/AN9638800500

出版商:RSC    

年代:1963

数据来源: RSC

摘要:

506 NEWMAN AND WATSON DETERMINATION OF [Analyst, 1701. 88 The Determination of Magnesium in Calcium Salts BY E. J. NEWMAK AND C. A. WATSoK (Hopkin G Williams Ltd., Freshwater Road, Chadwell Heath, Essex) A method is described for determining magnesium in the range 0 to 50 pg in calcium salts. After preliminary extractions of the sample solution a t pH 1 1 (containing citrate and tartrate as masking agents) with 8-hydroxy- quinoline and chloroform to remove interfering metals, the magnesium is extracted with 8-hydroxyquinoline in the presence of n-butylamine. The extracted magnesium is then determined spectrophotometrically a t 380 mp. The results show that there is no interference from calcium. PUBLISHED specifications for analytical-reagent grade calcium salts commonly describe limit tests for magnesium based on its colour reaction with Titan Yellow (Clayton Yellow, Thiazole Yellow) .l ,2 9 3 7 4 This reagent is not, however, particularly satisfactory for routine determina- tions of small amounts of magnesium in calcium salts.I t forms a lake with magnesium, which will quickly precipitate unless stabilised with, for example, starch and glycerol. The colour reaction is performed in rather alkaline solution, so that the bulk of the calcium must be removed by a preliminary precipitation, for example, as calcium molybdate. The small amount of calcium remaining in the solution intensifies the colour of the magnesium - Titan Yellow lake, so that calcium should be added to each of the magnesium solutions used in the preparation of the calibration graph.Moreover, the quality of Titan Yellow varies from one supplier to another, and from batch to batch. I t is a necessary precaution, therefore, to prepare a fresh calibration graph with every batch of Titan Yellow used. Umland and Hoff mann5 described the extraction of magnesium with 8-hydroxj.quinoline (oxine) into chloroform in the presence of n-butylamine and tartrate. The magnesium in the organic extract was then determined directly by measuring its optical density at 380 mp. Calcium was not extracted by oxine and chloroform either in the absence or presence of butylamine. Umland and Hoffmann were thus able to determine the magnesium contents of calcium minerals in the range 0.28 to 18.2 per cent. of magnesium. The magnesium was extracted as the ion-association system /n-C4H,S H3] + [Mg(ox)3j-, where “ox” represents the anion of 8-hydroxyquinoline.In the absence of butylamine, the normal magnesium oxinate, Mg(ox),.2H20, could not be extracted into chloroform. A preliminary extraction with oxine and chloroform in the absence of butylamine served to remove other metals, such as iron, that would interfere in the extraction of magnesium. (The organic phase from this preliminary extraction could be used to determine iron.) Jankowski and FreiserG have used a similar procedure for determining microgram amounts of magnesium in the presence of milligram amounts of calcium, strontium and barium, with tartrate as a masking agent and tetra-alkyl ammonium salts to provide the cations of the extractable ion-pair.We have adapted the method of Umland and Hoffmann to determine amounts of magnesium within the range 0 to 50 pg in the presence of up to 80 mg of calcium. The method could probably be applied to somewhat larger amounts of calcium. EXPERIMENTAL Umland and Hoffmann5 extracted magnesium from a tartrate-containing solution having a pH between 10.5 and 11.5, which was obtained by adding ammonia solution. Tartrate was insufficiently effective to prevent precipitation of calcium with the amounts of calcium salts we wished to use. Precipitation was accompanied by low recoveries of added mag- nesium. Citrate proved to be an extremely effective masking agent, and in the presence of citrate it was found necessary to use sodium hydroxide to raise the pH of the solution to 11 or above.Tartrate was still added to the test solution, as we found that it prevented sudden large increases in pH, which occasionally caused localised precipitation in the vicinity of the added sodium hydroxide.July, 19631 MAGNESIUM IN CALCIUM SALTS 507 Preliminary experiments also showed that, under these conditions, two extractions with oxine and chloroform were necessary completely to remove more than about 10 pg of mag- nesium from the solution. A 2 per cent. solution of calcium chloride was treated with citrate and tartrate, and its pH was adjusted to 11 with sodium hydroxide solution. It was then extracted several times with oxine and chloroform in the presence of n-butylamine to remove any magnesium present in the salt. It was then washed several times with chloroform to remove any remain- ing oxine.Three portions of this solution, each containing 80mg of calcium, were treated with 20, 30 or 40 pg of magnesium. The magnesium was then extracted by the procedure des- cribed below. A parallel series of tests was carried out on solutions that contained no calcium. The optical densities obtained in the two series of tests are shown in Table I, and indicate that there is no interference caused by calcium. The optical densities of the organic extracts were measured at 380mp. TABLE I COMPARISON OF OPTICAL DENSITIES OF MAGNESIUM COMPLEX EXTRACTED IN THE PRESENCE AND ABSENCE OF CALCIUM Optical density of magnesium complex extracted from solutions containing- Xagnesium added, r - p p A p p 7 no calcium 80 mg of calcium PLg 20 0.380 0.380 30 0-565 0.575 40 0.765 0-760 METHOD REAGENTS- Analytical-reagent grade chemicals should be used when possible.8-Hydroxyquinoline (oxine) solution-Dissolve 0-15 g of oxine in 100 ml of chloroform. Butylamine solution-Dilute 20 ml of n-butylamine to 100 ml with water. Potassium sodium tartrate soladion, approximately M-Dissolve 28 g of potassium sodium Sodium citrate solution-Dissolve 20 g of trisodium citrate in 100 ml of water. Sodium hydroxide solution, approximately M-Dissolve 4 g of sodium hydroxide pellets in 100ml of water. Chloroform. PROCEDURE- Dissolve a suitable weight of the calcium salt, containing not more than 50 pg of mag- nesium, in 10 ml of water and 6 ml of sodium citrate solution. Add 3 ml of potassium sodium tartrate solution, and adjust the pH of the solution to between 11 and 12 (measured with a pH meter) with sodium hydroxide solution. Transfer to a separating funnel, add 5 ml of oxine solution, shake vigorously for 5 minutes, allow to separate, and discard the organic layer.Extract again with 5 ml of chloroform, and again discard the organic layer. Add 1 ml of butylamine solution, and extract by shaking for 5 minutes with 5 ml of oxine solution. Run the organic layer into a dry vessel, add a further 1 ml of butylamine solution, and repeat the extraction with 5 ml of oxine solution. Combine the organic extracts, and filter the solution through a small funnel fitted with a glass-wool plug supporting a little anhydrous sodium sulphate into a l-cm spectrophotometer cell.Measure the optical density at 380 mp of the test solution against the blank solution in the reference cell. Calculate the amount of magnesium in the sample from a previously prepared calibration graph. tartrate in 100ml of water. Perform a blank determination on the reagents by the same procedure. NOTES- solution. 1. Fifty milligrams of calcium sulphate, hydrated, dissolve easily in 6 ml of citrate508 NEWMAN AND WATSON : DETERMINATION OF [Analyst, Vol. 88 2 . Two hundred milligrams of calcium carbonate were dissolved in 5 ml of N hydro- chloric acid and boiled. After 5 ml of water, 6 ml of citrate solution and 3 ml of tartrate solution had been added, about 2 ml of sodium hydroxide solution were required to raise the pH to between 11 and 12. When testing the other calcium salts, which did not require treatment with acid, only about 0.2 ml of sodium hydroxide solution was needed.PREPARATION OF CALIBRATION GRAPH- Dissolve 1.014 g of magnesium sulphate heptahydrate in 5 ml of 5 N sulphuric acid and sufficient water to produce 1 litre. Dilute this solution ten-fold immediately before use to give a solution containing 10 pg of magnesium per ml. Measure from a burette 0-, 0.5-, l-O-, 2.0-, :3.0-, 4.0- and 5.0-ml portions of the magnesium solution into small beakers. Dilute each solution to 10 ml with water, add 6 ml of sodium citrate solution, and continue as described under “Procedure.” Plot a graph of optical density against micrograms of magnesium. The graph is linear and passes through the origin. The molar extinction coefficient of the magnesium complex is about 5000.RESULTS AND CONCLUSIOKS Known amounts of magnesium were added to solutions of calcium chloride (0.2 g of the hydrated salt) and calcium sulphate (50 mg of the hydrated salt). The magnesium was extracted and its optical density measured as described above, the original magnesium content of the salt being subtracted from the value obtained. The results of these experiments are shown in Table 11. TABLE I1 RECOVERIES OF MAGNESIUM FROM SOLUTIOIYS OF CALCIUM SALTS Sample Calcium chloride (Batch “D”) . . Calcium sulphate (Batch “ J”) Magnesium added, Yg 5.0 10.0 20.0 30.0 40.0 20.0 Magnesium found, Y*.g 5.0 9.5 19.2 29.0 38.0 20.0 Some typical results obtained by this method on AnalaR calcium salts are shown in Table 111, together with the stated limit for magnesium.TABLE 111 TYPICAL MAGNESIUM CONTENTS OF ANALAR CALCIUM SALTS Compound Calcium carbonate . . . . .. . . Batch “A” . . . . .. . . Batch “B” . . . . .. . . Batch “C” . . . . .. . . Batch “D” . . . . .. . . Batch “E” . . . . .. . . Batch “F” . . . . .. . . Calcium chloride, dried . . .. . . Batch “G” . . . . . . . . Batch “H” . . . . . . . . Calcium sulphate, hydrated . . . . Batch “ J” . . . . . . . . Calcium chloride, hydrated . . . . Specified limit for magnesium, Magnesium found, p.p.m. p.p.m. 65.5 45 100 15 10 12.5 200 200 400 1000 105 75 60 The procedure has been found suitable for determining small amounts of magnesium It could easily be modified to deal with smaller or larger amounts of We thank the Directors of Hopkin & Williams Ltd. for permission to publish this paper. in calcium salts. magnesium.Jrily, 1963 I . *' l ' r ~ i f r t ~ g tlcr t - / t t v ~ i s c h m Hrugrnzicn alrf Reiwhtit." Fifth FAition. E. 3lcrck. Darnrhtadt . IWV. 2. ".\n;ilaH Standards for laboratory C'hcmicals." Fifth Edition. nritish Drug HOIIWS i Ad.. I ' c w h . 3. "Itc*ry(cwt ('11c~iiiic.als." .I.('.!% Spcificationn, 1960. Applied I'ublications. .\iiwricwt ('lic-inic.;il 4 . Ittwin. J ., " Itc*wc*nt Chc-miwls and Standards," Foiirth Edition. 1). \'an Scwtmnd ('onrpnny I.tcl., .i. l'mlniid. I:.. and Iloffmann, IY., ..I,tnl. Chim. ..lclo. 1957. 17. 131. 6. J;iiikcw?rki, S. J.. and Frc*isc*r, H., .-InnZ. Chrnr.. 1961, 33, i76. I)orwt. and Hopkin & Williams Ltd., Chadwcll Heath, Esscx, 19Ai. Stwic-ty. \\'adlington. l>.t-., 1901. I'rince*tcm. X.J.. 1961. Itcwivtd .I/nn./t I WI. I !W

ISSN:0003-2654    

DOI:10.1039/AN9638800506

出版商:RSC    

年代:1963

数据来源: RSC

摘要:

510 WILSON AND LEWIS: APPLICATION OF THE URAKYL SALT METHOD TO [Analyst, V O l . 88 Application of the Uranyl Salt Method to the Determination of Arsenic by the Oxygen Flask Technique BY A. D. WILSON AND D. T. LEWIS (Department of Scientific and Industrial Research, Laboratory of the Government Chemist, Clement's Inn Passage, Strand, London, W.C.2) A modified oxygen flask technique is described for determining arsenic in organic compounds, the conventional platinum supports being replaced by aluminium spirals or steel gauzes, which resist attack by arsenic during the vigorous combustion. A favourable gravimetric factor is ensured by the quantitative precipi- tation of the arsenic as ammonium uranyl arsenate and subsequent ignition, under controlled conditions, to the black triuranium octaoxide, U,O,.Phos- phate and vanadate form similar insoluble ammonium uranyl salts, and the method is not directly applicable to organic compounds containing these elements without prior separation. SEVERAL procedures are available for determining arsenic ; titrimetric methods include the titration of tervalent arsenic by bromatel or iodate2 in hydrochloric acid solution. A useful separation of arsenic from other elements can be achieved by reducing arsenic salts to elemental arsenic with consequent precipitation from solution. Hypophosphorous acid is the preferred reductant, and a titrimetric finish is generally employed, the precipitated arsenic being dis- solved in standard i0dine~9~ or b r ~ m i n e . ~ The usual gravimetric method depends on the precipitation of quinquivalent arsenic as ammonium magnesium arsenate with subsequent ignition to the pyroarsenate.6 According to Duval and Duval,' this compound is stable over the temperature range 415" to 885" C, but some arsenate may be reduced to the tervalent state and lost.8 Ammonium uranyl arsenate has an extremely low solubility product9 (1.71 x 10-24); this, and a favourable gravimetric factor, has made it the basis for the determination of arsenate arsenic.lO,ll PullarlO precipitated arsenate with uranyl ion and obtained a fine gelatinous precipitate soluble in mineral acids and insoluble in acetic acid, the determination being completed by careful ignition of the precipitate to uranyl pyroarsenate.Dupuis and Duval12 do not consider this a suitable weighing form.Lewis and Davisll ignited the ammonium ursnyl arsenate precipitate to triuranium octaoxide, an established weighing form, the arsenic being completely volatilised. Tervalent arsenic compounds are volatile, and care has to be taken when destroying organic matter in organo-arsenical materials, this being usually accomplished by wet-oxidation rnethods.l1~l3 The sealed nature of the oxygen flask technique makes it an obvious alternative, and it was first proposed for this purpose by Corner14; after combustion, most of the arsenic is in the tervalent state.15 Belcher et al.15 and Tuckerman et aZ.16 do not consider the method adequate, as arsenic tends to alloy with the platinum supports and combustion is also unsatisfactory with silica spirals ; they recommend wet combustion. EXPERIMENTAL PRECIPITATION OF AMMONIUM URANYL ARSENATE- Lewis and Davisll precipitate ammonium uranyl arsenate a t the boiling-point by adding uranyl acetate to a solution of arsenate in an acetic acid - ammonium acetate buffer, and then age the fine precipitate for some hours before collection by filtration. In an attempt to produce a more compact and filterable precipitate we substituted ammonia derivatives for ammonia, adding them to boiling solutions of uranyl acetate in nitric - acetic acid.Precipitates were formed with diethylamine, ethylenediamine, guanidine carbonate and hydrazine hydrate, but these had no particular advantages over the am- monium derivative. However, slow precipitation with ammonia, generated by the hydrolysis of urea, under similar conditions yielded a compact coarsely crystalline and readily filterable precipitate, strikingly different from the ~nucli more intractable product of normal precipi- t ation.July, 19631 511 A suitable initial pH before the precipitation was found in the range 1.3 to 1-8; above pH 1.8 premature precipitation can occur, and below pH 1.3 the time of formation of the precipitate becomes too lengthy.In a study of the change in pH during the course of one precipitation, an initial drop in p1-I (from 1.7 to 0.9) was observed, the rate of decrease being most pronounced a t the point when the bulk of the precipitate came down, after which the pH steadily increased as more ammonia was produced. This indicates that initially the homogeneously generated ammonia reacts to form ammonium uranyl arsenate and does not raise the pH as expected.Sodium and hydrogen ions were present, but the hydrogen and sodium uranyl arsenates are not preferentially formed, their solubility products being known to be 3-17 x 10-l1 and 1.35 x respe~tively,~ i e . , much greater than the value for the corresponding ammonium salt. Spectroscopic analyses of the ammonium derivative have confirmed that contamination by sodium ion is negligible. In all the subsequent work, the precipitation was completed by continuing the hydrolysis until the pH was raised to 3.0 to 3.5 and then setting the solution aside for some hours before filtration at room temperature. DRYING AND IGNITION OF THE PRECIPITATE- Precipitates of h'H,.U02.As0,.nH20 prepared in the manner described, when collected on a porous porcelain crucible and dried in air, contained between 2-5 and 3.0 molecules of water.On drying these precipitates in an electric oven, most of this water was easily lost, the apparent degree of hydration depending on the temperature and time of drying. Gentle drying indicates the presence of a monohydrate, but this broke down and near137 anhydrous forms were always obtained. Thermogravimetric studies by Dupois and Duvall2 showed no stable plateau, and it is apparent that these hydrates are not good weighing forms. Their indefinite composition and hygroscopic nature make them of little analytical value, although they could be used for approximate work. Further ignition of these precipitates, in electric ovens and furnaces, resulted in slight but continual losses in weight up to 400" C.At 500" C there was a sharp decrease in weight apparently corresponding to the formation of uranyl pyroarsenate, the weighing f o rm used by Yullar, and this was accompanied by a change in colour of the precipitates from yellow to green. On raising the temperature to 800" C there was little, if any, change in weight, but the colour of the precipitates reverted to yellow. The weights of these ignited precipitates were always several per cent. below that calculated for (UO,),As,O,, and this weighing form can be of no use for the exact determination of arsenic. It seems probable, from these observations, that evolution of ammonia reduces the precipitate ; the colour changes indicate the presence of uraniumIV, and possibly arsenic is reduced to the tervalent state and so lost.Confirmation of this deduction was obtained by heating a sample of uranyl arsenate, previously dried at 220" C, in an ignition tube. Water and ammonia were expelled and a little white sublimate collected on the upper surface of the tube; at the same time the colour of the precipitate changed from yellow to green. The white sublimate was found to contain tervalent arsenic. The yellow uranyl pyroarsenate begins to decompose slowly when heated in an electric furnace at 850" to 900" C. This process is greatly accelerated by increasing the furnace temperature to 1200" C, but even at this temperature 16 hours' ignition is necessary to complete the decomposition to triuranium octaoxide.Petit and Kienbergerl' report prolonged ignition periods for the decomposition of uranium compounds to triuranium octaoxide, and, as this oxide loses a little oxygen at high temperatures (above 950" C), they recommend a final re-oxidation of the oxide by ignition at 850" C. We found it necessary to dissolve the ignited oxide in nitric acid, evaporate, and re-ignite at 850" C. An entirely different pattern of behaviour was observed when the precipitate was collected on filter-paper and ignited over a gas burner. On gentle ignition over a low flame, just sufficient to burn off carbonised paper, a residue remained that was mainly black triuranium octaoxide contaminated by a little yellow uranyl pyroarsenate. (These observations are similar to those previously made by Lewis and Davis.ll) It was found that the decomposition of the remaining amount of uranyl pyroarsenate was best effected by a further ignition in the reducing atmosphere of an enveloping gas flame from a large Amal burner.This was found to be a more satisfactory technique than the alternatives of adding powdered filter- paper or ammonium sulphate to the precipitate. As it is possible that this ignited oxide of uranium may contain some dioxide, complete oxidation to triuranium octaoxide was THE DETERMINATION OF ARSENIC BY THE OXYGEN FLASK TECHNIQUE512 TVILSOS ASD LEWIS: APPLICATION OF THE L-RASYL SALT METHOD TO TAnalyst, vol. 88 effected either by re-ignition in the oxidising conditions of an electric furnace (at 850" C) or bg- dissolving in nitric acid and re-igniting.(Care must be exercised when applying the latter technique, so that no loss occurs due to decrepitation.) I t was also found possible to carry out a gas-ignition procedure after collecting the precipitate in a porous porcelain crucible, the reducing gases from the burner permeating the porous base of the crucible and effecting reduction. It is clear from these observations that reducing conditions greatly facilitate the ignition of the precipitate to triuranium octaoxide, and presumably reduction of arsenate to arsenious oxide or arsine occurs, with the consequent loss of arsenic. The sensitivity of uranyl pyro- arsenate to a reducing atmosphere makes it an unreliable weighing form. Results of these tests of reliability and accuracy on the gravimetric procedures are tabulated in Tables I A and IB, and show that both uranyl pyroarsenate and ammonium uranyl arsenate as weighing forms have little analytical value, although the latter could be used for approximate work. Weighing the precipitate as triuranium octaoxide, after ignition by either of the described procedures, yielded good results.For this procedure it was found preferable to collect the precipitate on a So. 44 filter-paper and ignite in a small silica or recrystallised alumina crucible (10 ml) . The occasional low results when a porous porcelain crucible was used, were probably due to mechanical loss, the fine oxide being carried awa!. bj. the gas passing through the porous base. TABLE IA ,%PPAREST RECOYERI- OF ARSESIC FROM DIFFEREXT WEIGHISG FORMS ,%rsenious oxide was used as standard and the amount of arsenic present in each test was 25.00 nig -\rsexiic, wcighcd a.; Drying -Arsenic, weighcd as A4rsenic, weighed as s H, .I - 0, .\so,, teniperaturth, Drying tinic. ( uo2)2. .AS,O;, c-308, mg "C hours 111 g mg 23.3 110 2 24.1 24.99 25.4 110 16 24.1 - 25.2 110 I t i 2 4 i 9 24.99 25.5 110 1 CI 26.0 110 2 __ 26.3 110 16 - -- 24.97 25.2 125 16 24.7 24.81 24.8 125 16 24.2 24.88 25.8 125 2 24.6 __ 25.1 160 34.3 _. 25.3 160 2 24.5 _ _ - _ _ - > A TA4BLE I B F c RT H E R RE s u LT In a series of further experiments with 25-00 mg of arsenic (again as arsenious oxide) the results obtained b>- the uranyl acetate precipit a t ' loll method, with subsequent direct ignition to IT,O,, were-- .\rsenic found, nig .. . . 26.88, 24.93, 25-01, 23.00, 25.01, 23-OS, 24.1)s Mean, ing . . . . 24.!%i Coefficient of variation, . . 0.022 Standard deviation, mg . . . . 0.0555 FACTOR FOR THE COSVERSIOS OF T R I U R A S I ~ I OCTAOXIDE TO AKSEXIC- The sample of uranyl acetate used in this work was not of natural isotopic cornposition. The atomic weight of uranium of natural isotopic composition is 23847 and consists of a mixture of three isotopes, 238L!, 235U and 234U. The physical atomic weight of pure isotope, 2 3 8 r , is 238.12493, and uranyl compounds supplied by the manufacturers will be enriched in this isotope. In calculating a factor we look the atomic weight of uranium as 238.10, ;is the results below show that the effect on the factor is slight- *Atomic weight of uranium .. . . . . 238.0'7 238.10 238.12 Conversion factor of T',O, to arsenic . . . . 0.26683 0.26680 O-268iiJuly, 19831 513 The stoicheiometric formulation for the corresponding uranyl phosphate (prepared by the direct ignition of ammonium uranyl phosphate) as (UO,),.P,O, has recently been criticised by Wright, Hayes and Ryan.18 As a result of a detailed investigation they suggest that the more correct formula is (UO)(U0,).P20,, which contains both quinquivalent and sexavalent uranium. This criticism obviously does not affect the work described here, a5 we use the oxide as the final weighing form. THE DETERMIKATIOS OF ARSENIC BY THE OXYGEN FLASK TECHNQCE OXYGEN FLASK COMBUSTION- Previous workers using platinum or silica supports for the sample material have found the oxygen flask technique unsatisfactory for determining arsenic in organic compounds.Our preliminary observations confirmed this ; with a sample of o-arsanilic acid, we found un- mistakable signs of attack on platinum gauzes, which were rendered brittle. Substitution of glass spirals for the platinum gauze did not solve the problem, as combustions were smoky and carbon was deposited on both the spiral and the flask, and results were low. I t was decided to try other metallic supports in the hope that they might prove more satisfactory than platinum. For this purpose the stopper of the oxygen flask was fitted with a glass hook, so that spirals and gauzes of various materials could be hung from it. Oxidised copper spirals were first used in the hope that the oxide film would assist combustion and prevent reduction to arsenic, but, despite excellent combustions, poor results were obtained, again, presumably, because of alloy formation (there were some signs of attack on the copper).The use of steel gauzes and aluminium spirals gave much improved results, there were no signs of attack on these materials (although the aluminium tended to melt) and combustion was adequate. Steel spirals gave good combustions and aluminium less so; with aluminium it was necessary to use tissue rather than filter-paper and get the combustion going well initially. After dissolution of the white sublimate of arsenic oxides, which were formed on the sides of the flask, and subsequent acidification and oxidation to arsenate with bromine water, the arsenic was determined as previously described.The results of these preliminarg- investigations are tabulated in Tables I1 and 111. Further results obtained with steel and aluminium supports gave satisfactory recoveries of arsenic from o-arsanilic acid, arsenious oxide and acetarsol. METHOD 0xygel.t combustion $ask, 1 litre, jut-bottomed. Sample ssip$orts-(a) A 6-cm x 2-cm rectangle of mild-steel gauze (30 meshes to the lineal inch), folded across its length 2 cm from one end; ( b ) 1-mm diameter aluminium wire. Sample contuixers-A folded S o . 44 filter-paper or tissue paper, 2 cm square, with a tail to act as a touch paper. REAGENTS- Precipitating soZzztio?z-A solution containing 2 per cent. w/v of uranyl acetate dihydrate, 12.5 per cent. w/v of urea, 20 per cent.v/v of glacial acetic acid and 0.5 per cent. v/v of concentrated nitric acid. PROCEDURE- Either wrap it in a tissue paper and bind with aluminium wire or place in a shaped Xo. 44 filter-paper and secure in a mild-steel gauze. Attach to the hook fused on to the glass stopper of the oxygen flask. Fill the oxygen flask with oxygen (this can be ensured by displacement of water), and add 40 ml of 1.25 per cent. w/v sodium hydroxide solution. Ignite the touch paper, and quicklj- insert the stopper into the flask, carefully tilting the flask so that the arsenical sublimate condenses on the side of the flask rather than on the neck and stopper. Set the flask aside for 2 to 3 hours so that the mist formed during combustion is cleared. Wash down the spiral or gauze with water, and remove from the hooked stopper. Replace the stopper, and shake the flask for 5 minutes.Filter the contents into a 150-ml beaker, acidify with 4 K nitric acid to the red colour of thymol blue indicator, and oxidise arsenite to arsenate with bromine water, adding a 5-ml excess. Add 20 ml of precipitating solution, and raise the temperature of the solution to between 80" and 90" C. A precipitate should form within a few minutes. APPAKATUS- Weigh out about 40mg of the organic arsenical sample.514 WILSON AND LEWIS: APPLICATION OF THE URANYL SALT METHOD TO [ ~ n a ~ y s t , 1701. 88 Maintain the temperature until the pH of the solution reaches 3-0 to 3.5, as indicated by the green colour of the bromophenol blue indicator. Set aside for some hours a t room temperature.Filter through a No. 44 filter-paper, and wash the precipitate with 0.5 per cent. v/v acetic acid. Transfer the filter-paper to a 10-ml fused alumina crucible, carefully burn off the filter-paper over a bunsen burner, and then ignite in the enveloping flame of an Amal burner for 1 to 2 hours. Alternatively, place the crucible in a cold electric furnace, burn off the filter-paper at 600" C, and increase the temperature to 1200" C; maintain this temperature overnight. Dissolve the ignited oxides in a few drops of concentrated nitric acid, carefully evaporate to dryness, and re-ignite by raising the temperature to 850" C in an electric furnace. Weigh as U,O,. TABLE I1 RECOVERY OF ARSENIC AFTER OXYGEN FLASK COMBUSTION WITH VARIOUS o-Arsanilic acid was the organic compound used in these recovery experiments SUPPORTING MATERIALS Support material Arsenic added, Arsenic recovered, Observations mg mg Platinum gauze 11.16 7-52 Platinum pitted and swollen.Glass spiral 11.02 9.98 1 Poor combustion, smoke and Good combustions 14-65 13.26 I carbon deposits Copper spiral Copper gauze 17.32 14.03 14.01 Aluminium spiral 14.29 14.11 Stainless-steel gauze 16.02 13.92 14.04 Good combustions. Copper 13-75] me1 ts 14.28 Moderate combustions. 13-77) Aluminium melts 15-90] Good combustions TABL:E I11 RECOVERY OF ARSENIC FROM VARIOUS COMPOUSDS WHEN STEEL OR ALUMINIUM SUPPORTS \YERE USED Support material Arsenical compound Arsenic (calculated), Arsenic found, mg mg 16.08 { ;;::: 16.60 Aluminium spiral Arsenious oxide Stainless-steel gauze Arsenious oxide 15.76 15-88 Mild-steel gauze Arsenious oxidc 15.71 f 14.29 Aluminium spiral o-Arsanilic acid 113.87 16.02 13-92 14-04 14.49 14.13 { 15.63 Stainless-steel gauze o-Arsanilic acid Mild-steel gauze 0-~4rsanilic acid Aluminium spiral Acetarsol [ :;::3 11.48 Mild-steel gauze Acetarsol { :::% 15.74 14.28 13-77 14.10 13.70 15.90 13.53 13.74 14.49 13.85 15.23 11.28 12.04 11.04 11.18 10.81 CONCLUSION The work of Lewis and Daviesll is confirmed and extended.Arsenic can be accu- rately determined by controlled precipitation as ammonium uranyl arsenate and subsequent ignition to U,O, under the conditions described in this paper. Metals forming insolubleJuly, 19631 515 arsenates, hydroxides or acetates in the stipulated pH range will obviously interfere and must be removed by conventional methods.Phosphates and vanadates form similar am- monium uranyl salts, but on ignition give the corresponding pyrophosphate and pyrovanadate compounds, respectively. REFERENCES THE DETERMINATION OF ARSENIC BY THE OXYGEN FLASK TECHNIQUE 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. Gyory, S., 2. anal. Chern., 1893, 32, 415. Wilson, H. N., Analyst, 1943, 68, 361. Evans, B. S., Ibid., 1929, 54, 523. Haslam, J., and Wilkinson, N. T., Ibid., 1953, 78, 390. Sloviter, H. A., McNabb, W. M., and Wagner, E. C., Ind. Eng. Chem., Anal. Ed., 1942, 14, 516. Scott, W. W., and Furman, N. H., “Standard Methods of Chemical Analysis,” Sixth Edition, Duval, T., and Duval, C., Anal. Chinz. Acta, 1948, 2, 45. Mellor, J . W., “A Comprehensive Treatise on Inorganic and Theoretical Chemistry,” Longmans, Green & Co. Ltd., London, 1929, Volume IX, p. 177. Chukhlantsev, V. G., and Sharova, A. K., Zhur. Neorg. Khim., 1956, 1, 36. Pullar, R. E. O., 2. anal. Chem., 1871, 10, 72. Lewis, D. T., and Davis, V. E., J . Cheln. Soc., 1939, 284. Dupuis, T., and Duval, C., Anal. Chim. Acta, 1950, 4, 262. Cox, H. E., Analyst, 1925, 50, 3. Corner, M., Ibid., 1959, 84, 41. Belcher, R., Macdonald, A. M. G., and West, T. E., Talanta, 1958, 1, 408. Tuckerman, M. M., Hodecker, J. H., Southworth, B. C., and Fleischer, K. D., Anal. Chinz. Acta, Petit, G. S., and Kienberger, C. A., Ibid., 1961, 25, 579. Wright, J . S., Hayes, T. J., and Ryan, J. A., Nature, 1961, 190, 1188 and 191, 1290. D. Van Nostrand Co. Inc., New York, 1962, Volume I, p. 113. 1959, 21, 463. Received January 16th, 1963

ISSN:0003-2654    

DOI:10.1039/AN9638800510

出版商:RSC    

年代:1963

数据来源: RSC

摘要:

516 CROPTON AND JOY: DETERMINATION OF LOW CONCENTRATIONS [ k f " y S t , VOl. 88 Determination of Low Concentrations of Sodium Dodecylbenzenesulphonate BY R. W. G. CROPTON AND A. S. JOY (De$artment of Scientijk and Industrial Research, Warren Spring Laboratory, Stevenage, Herts.) A brief review is given of methods of analysis in which a complex formed between a dyestuff and a surface-active agent of opposite polarity is extracted into an organic liquid in which the dyestuff itself is insoluble. Two new methods are described in which this technique is used ; they are suitable for the determination of anionic surface-active agents at low con- centration. The first, in which an excess of magenta is added and the optical density of the organic layer is measured, can be used at concentrations down to 10-7 M.The second, in which magenta is titrated directly against surface- active agent, and either the visually assessed complete transfer of colour to the organic phase or the transfer to equal colour intensity in both phases is used as end-point, is satisfactory down to a concentration of 5 x 1 0 - 7 M. The influence of salting-out effects is discussed. DURING a recent investigation into the adsorption of surface-active agents on mineral powders it became necessary to develop a rapid and simple method for determining sodium dodecyl- benzenesulphonate in the concentration range M in solutions whose total ionic concentrations varied from lo-' to 10-1 M. A search of the literature revealed several possible methods, and they were examined to see if they were suitable or could by modifica- tion be made suitable for the purpose. Methods have been described1$2 in which a complex formed between a dyestuff and a surface-active agent of opposite polarity is extracted into an organic liquid in which the dyestuff itself is insoluble ; the extract is then examined spectro- scopically.3 A quicker and more convenient method is the direct titration of anionic - cationic surface-active agents proposed by Hartley and R~nnicles.~ Their method made use of the colour change of bromophenol blue to determine the end-point.Epton5 and Barr, Oliver and Stubbing9 introduced the idea of two-phase titration, which combined anionic - cationic titration with. extraction of the coloured complex. Epton's method consists in shaking the unknown anionic surface-active agent with methylene blue and chloroform.The cationic surface-active agent used as titrant reacts preferentially with the unknown anionic surface-active agent until the reaction is complete, after which further addition of titrant reforms the water-soluble salt of the dye. With this system three end- points are possible, viz.- to (a) the first appearance of dye in the aqueous phase; ( b ) the complete transfer of dye to the aqueous phase; (c) the partial transfer of dye to give equal coloration in both layers. If the titration is performed in the reverse manner, that is, with the anionic surface-active agent as titrant, two more end-points are possible, viz.- (d) the first appearance of colour in the chloroform; (e) the complete transfer of dye to the chloroform.Barr, Oliver and Stubbings found that bromophenol blue was a more suitable indicator for end-point (d). Edge,7 who used bromophenol blue, found that sodium dodecyl sulphate could be titrated successfully to end-point (b) with cetyltrimethylammonium bromide at concentrations down to 1 x The presence of electrolytes, however, caused a stable emulsion to be formed at the end-point. A disadvantage of end-points ( b ) , (c) and (e) is that the coloured complex transferred to the non-aqueous layer is formed from the surface-active agent and the dyestuff. Therefore, particularly at low concentrations, it is essential that the concentration of dye- stuff and the amount of it added in each titration should be exactly the same as those used in constructing the calibration curve.M.July, 19631 OF SODIUM DODECYLBENZENESULPHONATE 517 Edge,7 and later Kemp and Nutt (personal communication), used Barr, Oliver and Stubbing’s original system and end-point by titrating cetyltrimethylammonium bromide with sodium dodecyl sulphate and bromophenol blue as indicator. It was found that, after the solution had been shaken, the first colour to cross the interface was in the form of con- centrated blue wisps that were easily detectable. By taking the appearance of these wisps as the end-point, it was found possible to determine concentrations as low as 1 x 1 0 - 5 ~ . Further investigation of this system for application in the concentration range in which we were interested showed it to be unsatisfactory, since the wisps of colour were barely visible.Other investigators have used rosaniline and pararosaniline for the colorimetric deter- mination of long-chain alkyl sulphates. Karush and Sonenbergs used these dyes in 0.025 M phosphate buffer and extracted the dye - surface-active agent complex with a mixed 50 per cent. chloroform - 50 per cent. ethyl acetate solvent. Some difficulty was experienced, since transfer of dye to the organic phase occurred before the surface-active agent was added. The method was found to be sufficiently sensitive for determining concentrations of 5 x M with an error of about 2 per cent. Walling used basic fuchsin as the dye for the colorimetric determination of sodium dodecyl- benzene sulphate. Basic fuchsin consists of rosaniline and pararosaniline, but Wallin states that it is not believed that both components take part in the reaction.The dye is insoluble in chloroform, so that chloroform can be used as a blank. The fuchsin does, however, react to produce a chloroform-soluble complex with several salts, namely nitrates of alkali metals on the acid side, potassium iodide and bromide, and calcium and sodium chlorides. It appears that the transfer of dye noted by Karush and Sonenberg may have been due to their use of the phosphate buffer. Wallin was able to detect 0.5 mg of surface-active agent in 20ml of extractant. Preliminary experiments showed that the detection of the coloured wisps could be improved by selecting a dyestuff having an intense colour and a high interfacial activity.Dyestuffs that transferred from one phase to another during the shaking (emulsification) procedure were unsuitable for use in Barr, Oliver and Stubbings’s method. Several dyestuffs were tried, and it was finally decided to use magenta (rosaniline) for the determination of sodium dodecyl benzenesulphonat e. EXPERIMENTAL REAGENTS- The aqueous solution was prepared by triturating a portion of magenta powder with a little water in an agate mortar, and then washing it into a beaker and adjusting the volume to 200 ml. An equal volume of hot water was added, the solution was stirred and then set aside for 3 days. The supernatant liquor was decanted and again diluted with an equal volume of hot water. This solution was then stored as a stock solution. A working solution was prepared by accurately diluting a portion of the stock solution to one fifth of its con- cen tration .During the period of the investigation, it was found that the stock solution had deposited some material and formed a stable froth after it had been shaken. The deposited material was removed by siphoning off the liquid from below the surface. A further working solution was made up and was found to have the same concentration as the original working solution. M solutions of sodium dodecylbenzenesulphonate were normally made up weekly; weaker solutions were prepared daily by dilution of the 1 0 - 3 ~ solution, an Agla micrometer-syringe pipette being used when 0-5 ml or less was required. M solution of sodium dodecylbenzenesulphonate is stable for a fortnight, loss by decomposition and slow adsorption being negligible a t this concentration, whereas a M solution decomposes in a few days and a 10-7 M solution is apparently completely decomposed overnight.Standard A CALIBRATION OF SPECTROPHOTOMETER READINGS- Initially, the spectrophotometric technique was explored, the method used being similar to that described by Few and O t t e ~ i l l . ~ A calibration curve was obtained for the spectrophotometer (Unicam SPSOO) by the procedure described below. Analytical-reagent grade chloroform (5 ml) and 5 ml of standard surface-active agent solution were put by pipette into a glass-stoppered test-tube with an excess of saturated magenta solution; one drop from a dropper was usually sufficient. The518 CROPTON AND JOY: DETERMINATION OF LOW CONCENTRATIONS [Analyst, VOl.88 tube was shaken 50 times and then set aside for about 5 minutes to allow the phases to separate. The chloroform was then carefully transferred by pipette to a 1-cm cell, and the cell was placed in the spectrophotometer. A further 5 minutes were allowed for the water droplets produced during the transfer of the chloroform to disperse, and then the optical density was measured at 550 mp against chloroform. This, procedure was repeated a t different concen- trations of surface-active agent, and the curve shown in Fig. 1 was prepared. 0.01- I I I lo-’ I o-6 I( Concentration of dodecylbenzene sulptionate, M Fig. 1. Graph showing relationship between optical density and concentration of dodecyl- benzenesulphonate The method was compared with that of Longwell and Maniece2 at the Water Pollution Research Laboratory, and the results (personal communication from E.G. Eden) showed that in speed and simplicity it had advantages over Longwell and Maniece’s procedure, but it gave slightly higher values and was insensitive a t concentrations of surface-active agent above 5 p.p.m. No examination was made of the effect of impurities on the results by the method. AbbottlO has modified Longwell and Maniece’s method and has made a careful study of the effects of impurities at a constant ionic concentration of about 0.01 M. As shown later, the amount transferred between phases depends on ionic concentration, but the difference in emphasis of the analytical errors is entirely due to the different objects of the investigations.TWO-PHASE TITRATION- Further investigation into the possibilitjles of the two-phase anionic - cationic titration was carried out. The first appearance of wisps of colour in the chloroform was found to be unsuitable as an end-point. As the magenta itself reacts with the anionic surface-active agent to form the chloroform-soluble complex, it was decided to titrate solutions containing known volumes of the working magenta solution with standard surface-active agent. Titrations were carried out in 60-ml glass vials with po!lythene stoppers. Adsorption of the dye on the surface of the glass was reduced to a minimum by treatment with Silicone Fluid MS1107. Equal volumes (25 ml) of chloroform and water were used for the titrations, so that end- points (c) and (e) could be studied.Small portions of the working magenta solution were added from an Agla micrometer-syringe pipette, and the sodium dodecylbenzenesulphonate solution was added from a 5-ml burette. Normally, three titrations were cawied out for each different concentration of magenta, and., after each titration, the optical densities of the chloroform phase and aqueous phase were :measured. These results are shown in Table I as average values for each series of three determinations. It was found that the end-point showing equal colour in both phases was easily detected in daylight and in yellow-green artificial light from a microscope lamp, provided that more than 0.05ml of magenta solution was present. Titrations at this low concentration were successfully carried out in bright daylight only; even then the results showed moreJuly, 19631 OF SODIUM DODECYLBENZENESULPHONATE TABLE I TITRATIOK OF MAGENTA SOLUTION WITH SODIUM DODECYLBEXZENESULPHONATE 519 Volume of magenta solution, ml 0.25 0.25 0.15 0.15 0.05 0.05 0.25" 0.15" 0.26 0.15 0-15 0.05 Volume of 10-4 M surface-active agent, ml 0-79 0.86 0.62 0.51 0.128 0.155 0.79 0.50 1.89 1.14 1-07 0.293 Volume of surface-active agent per 0.1 ml of magenta solution, ml 0.316 0.344 0.347 0.340 0.256 0.310 0.316 0.333 0-756 0.760 0-713 0-586 Optical densities A r 7 Chloroform Aqueous phase phase End-point 7 - - 0.199 0.174 0-130 0.117 0.029 0-036 Partial transfer i;;;; I of dye Complete transfer of dye - - - - 0-387 0.013 0.231 0-009 0-012 0-234 0.065 0*008 * Results obtained with freshly diluted magenta solution.scatter than those obtained at higher concentrations. Complete transfer of the dye was a little more difficult, since the aqueous phase always appeared faintly pink, probably from scattered light from the underlying coloured chloroform phase. In this titration artificial light was found to be of no use. It can be seen from column 3 of Table I that the volume of surface-active agent solution required to the half-way end-point was about 0.3 ml per 0-1 ml of magenta solution. The concentration of the surface-active agent solution was M, and this was diluted to 25 ml for the titration. Thus surface-active agent concentrations of about 1 0 - 6 ~ were in fact successfully titrated. SPECTROPHOTOMETRIC INVESTIGATION OF THE TWO-PHASE METHOD- To check on the quantitative transfer of the dye complex into the chloroform phase, equal amounts of chloroform and water (25 ml) were shaken with 0.25 ml of magenta solution Volume of IO-'M dodecylbenzene sulphonate added, ml Fig.2. Graphs of titration of magenta solution with M dodecylbenzenesulphonate : curve A, chloroform solution; curve B, aqueous solution and measured amounts of M sodium dodecylbenzenesulphonate. After the extraction, the optical densities of the chloroform phase and of the aqueous phase were measured spectro- photometrically; the results are shown graphically in Fig. 2. No allowance has been made for the increase in volume of the aqueous phase. It can be seen that a linear relationship520 [Analyst, Vol.88 exists between the concentration of surface-active agent and the amount of dye transferred to the chloroform. The point of intersection of the lines, corresponding to 0.83 ml of lo4 M surface-active agent, agrees reasonably well with the titration figures obtained previously for the half-way end-point. Spectrophotometric determination of the end-point indicated by complete transfer of dye was always found to be slightly more than double the concen- tration at which optical densities of the phases were equal, and this was also usually so when the end-points were determined visually. CROPTON AND JOY: DETERMINATION OF LOW CONCENTRATIONS EFFECT OF NON-SURFACE-ACTIVE AGENTS- The effect of ionic concentration on the transfer of magenta from the aqueous to the organic phase was studied with potassium chloride.Equal volumes (25 ml) of chloroform and potassium chloride solution of known concentration were shaken with 0.25 ml of magenta solution. Considerable transfer of magenta into the chloroform was observed. In two tests, 2 x moles of sodium dodecylbenzenesulphonate were added; the optical densities are shown in Table 11. It appears that, for a constant ionic concentration, there would be a constant displacement of the optical density. and 4 x SALTING-OUT OF MAGENTA WITH POTASSIUM CHLORIDE Sodium dodecylbenzene- Concentration of sulphonate present, Optical density of potassium chloride, M moles x lo6 chloroform phase 1.0 0 0.900 0.2 0 0.386 0.1 0 0.252 0.1 2 0.376 0.1 4 0.544 Since Wallin had noted complex formation between basic fuchsin and several salts, it was considered necessary to investigate more fully the transfer of magenta in the presence of potassium chloride.Twenty-five millilitres of 0.1 M potassium chloride and 1 ml of 0.003 M magenta were repeatedly extracted with portions of chloroform until no further transfer of dye could be detected; the aqueous layer was faintly coloured with magenta. The concen- tration of dye was then made up to approximately its original value, and a further extraction with chloroform was carried out. The amounts of magenta transferred during this extraction and the first extraction were found to be the same. This shows that the transfer is not due to impurities in the magenta nor to small amounts of surface-active agent in the potassium chloride.A direct test was then made to distinguish between salting-out effect and complex formation by measuring the conductivities of dilute solutions of potassium chloride and magenta. The results were- (a) specific conductance of l O V 3 ~ magenta solution was 2.46 x (b) specific conductance of 3.84 x M potassium chloride was 5.51 x mho per cm; (c) specific conductance of saturated magenta solution in 3-84 x 1 0 - 3 ~ potassium chloride was 7-87 x 10-4mho per crn. Adding (a) and (b) gives 7.97 x mho per cm, which is about the expected value of (c) if there were no loss of ions due to complex formation. I t appears, therefore, that the transfer of magenta in the presence of potassium chloride is a salting-out effect and is not due to formation of a chloroform-soluble complex.The salting-out of some other dyes was checked by extracting the dye with chloroform in the presence of 0.1 M potassium chloride. Large amounts of malachite green were found to be transferred in the presence of the salt. Methylene blue showed some transfer, but with bromophenol blue the transfer was negligible. With rhodamine B the colour disappeared from both the aqueous and chloroform phases after shaking. mho per cm;July, 19631 OF SODIUM DODECYLBENZENESULPHONATE 521 CONCLUSIONS The appearance of the water - organic liquid interface has proved critical when assessing end-points in the determination of surface-active agents by the two-phase titration of anionic against cationic reagent with a dye as indicator that forms a complex with one reactant and can transfer from one phase to the other.Visual detection of these effects, however, becomes difficult at concentrations of surface-active agent of less than 10-5 M. In the two methods of analysis studied here the dye has been used as titrant rather than as an indicator. That is, the positively charged dye magenta has been used for determining an anionic surface-active agent, but still by the two-phase technique. These methods should be applicable to other anionic surface-active agents. The spectrophotometric technique, in which an excess of magenta is added and then the optical density of the organic phase is measured, can be used down to a concentration of lo-‘ M. The direct titration of magenta against surface-active agent, in which either the visually assessed complete transfer of colour to the chloroform or transfer to equal colour intensity in both phases can be used as end-point, is satisfactory down to 5 x lo-’ M.In this method the “half-way” end-point is the best. Magenta has a disadvantage in that its transfer is dependent on ionic concentration. In solutions of 0.01 M sodium fluoride the amount of magenta extracted into the chloroform phase in the absence of surface-active agent was just measurable spectrophotometrically. In the presence of higher concentrations of potassium chloride, 0-1 to 1.0 M, considerable amounts of dye were transferred to the chloroform before the addition of the surface-active agent. Although this point was not investigated, it would appear that, for solutions of high ionic concentration, the spectrophotometric method could still be used provided that measure- ment was made against a cell containing a blank solution prepared by shaking chloroform with the ionic solution containing magenta but no reagent. The transfer of some dyes, in the presence of different electrolytes, has been found to be due to the formation of chloroform-soluble complexes.It has been shown in this investi- gation that the effect of potassium chloride on a magenta solution is a salting-out of the dye. I t is noted from the literature, however, that magenta does form a complex with silver ions. This effect was noticed when attempts were made to determine concentrations of anionic surface-active agents in solutions containing approximately This work forms part of a research programme of Warren Spring Laboratory and is published by permission of the Director. M silver fluoride. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. REFERENCES Jones, J. H., J . Ass. 08. Agric. Chem., 1945, 28, 398. Longwell, J., and Maniece, W. D., Analyst, 1955, 80, 167. Few, A. V., and Ottewill, R. H., J . Colloid Sci., 1956, 11, 34. Hartley, G. S., and Runnicles, D. F., Proc. Roy. SOC. A , 1938, 168, 420. Epton, S. R., Tvans. Faraday SOC., 1948, 44, 226. Barr, T., Oliver, J., and Stubbings, V. W., J . SOC. Chem. Ind., 1948, 67 45. Edge, R. M., Ph.D. Thesis, University of Birmingham, 1959. Karush, F., and Sonenberg, M., AnaZ. Chem., 1950, 22, 175. U‘allin, G. R., Ibid., 1950, 22, 616. Abbott, D. C., Analyst, 1962, 87, 286. First received November Sth, 1961 Amended, February 27th, 1963

ISSN:0003-2654    

DOI:10.1039/AN9638800516

出版商:RSC    

年代:1963

数据来源: RSC

摘要:

522 GREEN, HESLOP AND WHITLEY: DETERMINATION OF SILICON, ZINC AND [Analyst, vol. 88 The Determination of Silicon, Zinc and Magnesium in Gallium Arsenide by Neutron-activation Analysis BY D. E. GREEX,* J. A. B. HESLOPT AND J. E. WHITLEU: ( A .E.I. Research Laboratory, Aldermaston Court, Berks.) The use of short-lived isotopes in the determination of trace amounts of impurities in gallium arsenide is deslxibed. Methods are presented for determining silicon, zinc and magnesium. The effect of thermal neutron self -shielding in the determination of silicon and magnesium has been investi- gated and shown to be insignificant. NEUTRON-activation analysis is a useful technique for the study and quantitative measurement of trace amounts of impurities in highly purified materials, such as those used in the semi- conductor fie1d.l When the host material consists of elements of high thermal neutron cross-section, a considerable amount of undesired activity is induced on irradiation.If the activities induced in the host material are short-lived compared with those induced in trace constituents, samples may be allowed to cool before the determinations of trace amounts of impurities are carried out, e.g., the determination of arsenic (76A~, tl = 26.7 hours) in silicon (31Si, t, = 2.62 hours).2 When the activities in the host material are longer lived than those produced in the trace constituents, considerable radiochemical purification (with high decontamination factor) may be required, and the early stages of such procedures may have to be performed in shielded facilities, with associated difficulties.When it is possible to measure the trace amounts of impurities by activation to short-lived species, the short irradiation times required bring a reduction in the activity induced in the host material, and with it a reduction in the radiation associated with handling the material. For the determination by neutron activation of trace amounts of impurities in gallium arsenide, the fast-transfer facility of the Merlin reactor3 has been used. The fast-transfer facility, a pneumatic tube for the transfer of samples from the laboratory to the reactor core, is a feature frequently incorporated in the design of research reactors. At a reactor power of 5 MW the thermal neutron flux at the core terminus of the fast-transfer facility was 4 x 1013 neutrons per sq.cm per second. Methods have been developed for determining silicon, zinc and magnesium in gallium arsenide. The half-lives of the active species measured were 2.62 hours, 52 minutes and 9.5 minutes, respectively. The separation of these species requires radiochemical puri- fication procedures that must be completed in a time similar to the half-life of the species separated. All chemical operations were performed in a fume cupboard with a shielding of 2 inches of lead. An upper limit of 100 millicuries of activity to be handled was set, for reasons of radiological protection. For the determination of silicon by irradiation, this limited the size of gallium arsenide samples to approximately 100 mg, which was the size usually taken for the determinations of zinc and magnesium.The final precipitates obtained in the chemical procedures were mounted on standard aluminium planchettes, and the decay of the radioactive species was followed with a 1-inch diameter Geiger - Miiller counter. A scintillation counter with a hundred-channel pulse- height analyser was used for the rapid identification of contaminating activities when the procedures were being developed. When the host material and trace impurity differ by only 1 or 2 units of 2, the effect of fast neutrons on the host material may lead to the emission of charged particles, with the production of other active species of the trace constituent. This has been found to occur in the determination of zinc in gallium arsenide, and a measurement of the magnitude of this interference is given under “Determination of Zinc,” p.524. The development of these procedures is described below. * Present address : Radley Research Institute, Reading, Berks. t Present address : Admiralty Materials Laboratory, Holton Heath, Poole, Dorset. 1 Present address : Scottish Research Reactor Centre, East Kilbride, nr. Glasgow.July, 19631 MAGNESIUM I N GALLIUM ARSENIDE BY NEUTRON-ACTIVATION ANALYSIS 523 When the neutron cross-sections of elements in the host material are large, the neutron flux through a specimen may be considerably attenuated giving a reduced activation of impurities inside the sample, and hence an erroneously low content of impurity. The effect of such neutron self-shielding for silicon and magnesium in gallium arsenide has been investigated.EXPERIMENTAL DETAILS OF IRRADIATIONS- Samples of gallium arsenide were contained in polythene ampoules, which were placed in an epoxy-resin carrier and transferred to the core of the reactor by the fast-transfer facility. The build-up of long-lived activities in the matrix material was kept to a minimum by limiting the length of irradiation to a time equal to the half-life of the species to be separated. After irradiation, the sample was returned to the laboratory by the fast-transfer facility (transit time 8 seconds) and transferred to a shielded cell for processing. DISSOLUTION OF SAMPLES- Dissolution of gallium arsenide in aqua regia has been found to take 10 to 15 minutes for a 100 mg specimen. When short-lived species are being determined, this time would result in a decrease in the sensitivity of the method.The use of electrolytic dissolution has been found to aid the solution of pure gallium in nitric acid.4 This method has been found applicable to gallium arsenide, and it was possible to dissolve 100-mg samples of gallium arsenide in nitric acid on passing a current of 1 to 2 amps a t 15 volts for 3 to 4 minutes. DETERMINATION OF SILICON It has been found possible completely to remove arsenic activity from silicon by precipi- tation of silica with ammonium carbonate and thorough washing of the precipitate. Gallium hydroxide is also precipitated by this procedure, and complete removal of gallium activities from silicon has been found difficult to achieve.With 14-1-hour gallium-72 as tracer, the effectiveness of different procedures in removing gallium from silicon was investigated. Although most of these methods led to separation factors of lo4 or better, when they were tried sequentially little increase in separation factor was obtained. This may be attributed to the association of gallium and silica as a radio-colloid that was not dissolved by the hydro- fluoric acid used in the intermediate treatment. Use of sodium hydroxide solution as the dissolving medium in the intermediate treatment has overcome this difficulty. Two gravimetric compounds for the determination of chemical yield have been investi- gated; silica and quinoline molybd~silicate.~ The latter is the easiest to form in stoicheio- metric proportions, but suffers from the disadvantage of having an extremely low silicon content.Investigations suggest that this leads to approximately 50 per cent. absorption of silicon-31 p- particles in the precipitate, which is not desirable. For these reasons, silica has been used as the final gravimetric compound for the determination of chemical yield and for counting, Sources of silica prepared from irradiated gallium arsenide were found to contain a contaminating activity with a half-life of 82 minutes. This was identified as germanium-75, produced by the fast-neutron reaction 75As ( v z , ~ ) 75Ge, and this impurity was eliminated by adding germanium carrier a t the dissolution stage and the removal of germanium with gallium in the subsequent procedure. The complete radiochemical procedure for separating and purifying silicon from gallium arsenide is described below.After dissolution of gallium arsenide in a solution of silicon and germanium carriers in hydrofluoric - nitric acid, silica, germanium hydroxide and gallium hydroxide were precipitated with ammonium carbonate. The precipitate was washed, and dissolved in sodium hydroxide solution. Silica was precipitated by heating to fumes with perchloric acid. Dissolution and precipitation of the silica was repeated, with the addition of holdback carriers for gallium and germanium. The silica was redissolved and converted to the yellow molybdosilicate complex, which was extracted into pentanol and back-extracted into sodium acetate solution.6 The complex was destroyed by heating with perchloric acid, and silica was precipitated.Molybdenum trioxide, which was also precipitated at this stage, was leached from the precipitate with ammonia solution. The precipitate was then washed with ammonia solution, water and alcohol, and mounted on an aluminium planchette for counting and for determining the chemical yield.524 GREEN, HESLOP AND WHITLEY: DETEKMINATION OF SILICON, ZINC AND [A'kdySt, VOl. 88 The processing of a single sample was found to take approximately 2 hours. The per- formance of replicate determinations side by side was considered inadvisable in view of the high activity levels associated with irradiated gallium arsenide. PROCEDURE- Transfer the irradiated gallium arsenide (up to 100 mg) to a dissolution cell containing 1 ml of concentrated nitric acid, 5 mg of silicon in hydrofluoric - nitric acid mixture and a few milligrams of germanium carrier.Dissolve the gallium arsenide by passing a current of 1 to 2 amps at 15 volts for 3 to 4 minutes. Transfer the solution to a 50-ml beaker containing 5 g of ammonium carbonate, 2 ml of ammonia solution and 5 ml of water. Rinse the cell with 2 ml of water, and add the rinsings to the contents of the beaker. ' Boil for 10 minutes to complete the precipitation of silica, gallium hydroxide and germanium h,ydroxide. Spin in a centrifuge, and discard the supernatant liquid. Wash the precipitate with a 10 per cent. w/v solution of ammonium carbonate, and discard the washings. Dissolve the precipitate by warming with 6 to 8 drops of 6 N sodium hydroxide, and transfer to a 10-ml beaker with the aid of 11111 of water.Add 0-5 ml of perchloric acid, and heat until fumes are evolved to precipitate silica. Digest the precipitate with 5 ml of 6 N hydrochloric acid, spin in a centrifuge, and discard the supernatant liquid. Wash the precipitate with a further 5 ml of 6 N hydrochloric acid, and discard the washings. Dissolve the precipitate in 6 N sodium hydroxide, add a few milligrams of carriers for gallium and germanium, and reprecipitate the silica with perchloric acid and leach with hydrochloric acid as described above. Dissolve the precipitate in 6 to 8 drops of 6 N sodium hydroxide, and add 6 ml of 10 per cent. w/v ammonium molybdate solution and then sulphuric acid dropwise until the yellow molybdosilicate colour appears.Add 5 ml of 9 N sulphuric acid, and extract the molybdo- silicate into 5 ml of pentanol. Discard the aqueous layer, and wash the organic layer successively with 5-ml portions of 9 N sulphuric acid and water. Back-extract the molybdo- silicate into 2 ml of 10 per cent. w/v sodium acetate solution, and discard the organic layer. Heat the solution with perchloric acid until fumes are evolved to precipitate silica and molybdenum trioxide. Dilute to 5 ml with wa.ter, and make alkaline with ammonia solution to dissolve molybdenum trioxide. Spin in a centrifuge, and discard the supernatant liquid. Wash the precipitated silica successively with 5-ml portions of ammonia solution, water and ethanol, and mount for counting and far determining the chemical yield.COMPARATIVE STANDARDS- A sample of a few milligrams of pure silicon was irradiated simultaneously with the gallium arsenide, to be used as a comparative standard. This was dissolved in hydrofluoric - nitric acid, and silica was precipitated with arnmonium carbonate. The silica was dissolved in sodium hydroxide solution, and reprecipitated for counting and for determining the chemical yield by heating with perchloric acid. Although it is usual to have comparative standards of a few micrograms of the element in solution in it silica ampoule irradiated simultaneously with the solid sample to be investigated, this is not practical for silicon owing to the chemical nature of the ampoule and the instability of dilute solutions of silicon. If 100 counts per minute is taken as the lowest limit of sensitivity, then for an irradiation of 24 hours at 5 MW and a decay time of 24 hours, the sensitivity of the method has been found to be 1.3 x g of silicon.DETERMINATION OF ZINC A method has been developed for separating and purifying zinc from irradiated gallium arsenide. As with silicon, decontamination from arsenic was achieved by precipitation of basic zinc carbonate from an alkaline carbonate solution. Gallium was removed by dis- solving the precipitate in hydrochloric acid and passing the solution through a column of anion-exchange resin. Basic zinc carbonate was reprecipitated and then redissolved, and zinc mercuri-thiocyanate was precipitated in the presence of oxalic acid. The precipitate was dissolved in the minimum amount of nitric acid, and the solution adjusted to pH 7 with ammonia solution.Zinc quinaldate was precipitated as the gravimetric compound forJuly, 19631 MAGNESIUM IN GALLIUM ARSENIDE BY NEUTKON-ACTIVATION ANALYSIS 525 determining the chemical yield and for counting. and acetone, and mounted on a planchette. is described below, took approximately 1 hour to complete. The precipitate was washed with water The complete radiochemical procedure, which PROCEDURE- Transfer the irradiated gallium arsenide to a dissolution cell containing 1 ml of con- centrated nitric acid and 10 mg of zinc carrier. Dissolve the sample by passing a current of 1 to 2 amps at 15 volts for 3 to 4 minutes. Transfer the solution and one washing to a beaker, and add solid sodium carbonate until the solution is alkaline to methyl orange.Spin in a centrifuge, and discard the super- natant liquid. Wash the precipitate with water, spin in a centrifuge, and discard the supernatant liquid. Dissolve the precipitate in the minimum amount of concentrated hydrochloric acid, make the solution 2 M in hydrochloric acid, and allow this solution to percolate through a column of anion-exchange resin (see Note 1). Wash the column with a further seven 2-ml portions of 2 M hydrochloric acid a t a flow rate of 1 to 2 ml per minute. Remove the zinc from the column with 6ml of hot water containing a few milligrams of potassium iodide (see Note 2). Add 1 mg each of gallium and arsenic carriers to this solution, and make alkaline with excess of sodium carbonate. Dissolve the precipitate in 2 ml of M nitric acid, and add 1 ml of saturated oxalic acid solution and 1 ml of mercuri-thiocyanate reagent (see Note 3).Stir for 2 to 3 minutes, then. spin in a centrifuge, and discard the supernatant liquid. Dissolve the zinc mercuri-thiocyanate in 0.2 ml of concentrated nitric acid (with care), and add 2 ml of water and chlorophenol red. Make the solution just alkaline with ammonia solution. Add glacial acetic acid until the solution is just acid, dilute to 8m1, and add 1 ml of 3 per cent. w/v quinaldic acid. Heat in a water bath, then spin in a centrifuge, and discard the supernatant liquid. Wash the precipitate with water and acetone, and transfer to a planchette for weighing and counting. Spin in a centrifuge, and discard the supernatant liquid.NOTES- The resin used was Amberlite CG-400, which was washed with 2 M hydrochloric acid and water and kept in 2 M hydrochloric acid. The exchange time Zn2+ (column) + Zn2+ (aqucous) is relatively long, SO a t least 10 minutes should bc allowed for this step. Mercuri-thiocyanate reagent was prepared by dissolving 27 g of mercuric chloride and 39 g of ammonium thiocyanate in 1 litre of water. 1 . The column was approximately 3 cm long and 0.7 cm in diameter. 2 . 3. INTERFERENCES- Sources of zinc quinaldate produced from irradiated gallium arsenide by this procedure were found to contain considerable zinc activities that have been identified as arising from the reactions of fast neutrons on gallium, viz.- 14 h 52 m 69Ga (n,?) 69Zn -+ 69Zn --+ 69Ga (stable) I.T.P 52 m 69Ga (n,$) 69Zn ----+ 69Ga (stable). P That the 14-hour activity on the sources was not due to the presence of 72Ga was demonstrated by the facts that no proportional 20-minute 70Ga was detected and that the 14-hour activity did not show a peak at 0.84 MeV in its gamma-ray spectrum. Thus the zinc activities pro- duced by the reaction of fast neutrons with gallium will interfere in the determination of zinc in gallium arsenide. Samples of zinc and gallium were irradiated in the reflector region of Merlin a t 100 kW, the neutron fluxes at this power being 1-5 x 10l2 thermal neutrons per sq. cm per second and 2.1 x 1011 A measurement of the magnitude of the interference has been carried out.526 GREEN, HESLOP AND WHITLEY: DETERMINATION OF SILICON, ZINC AND [Analyst, VOl.88 fast (>1 MeV) neutrons per sq. cm per second. from each sample, and the decay curves were resolved. Radiochemically pure zinc was separated The ratio obtained was- 69Zn, c.p.m. per g of gallium 69Zn, c.p.m. per pg of zinc 7-9 x lo6 1-5 x lo4 - Thus 0-5 g of gallium, which is approximately equivalent to 1.0 g of gallium arsenide, will given an apparent zinc content of 265 pg. The magnitude of this interference may be considerably reduced by irradiating samples in the thermal column where, at a reactor power of 5 MW the fluxes are 2.5 x loll thermal neutrons per sq. cm per second and 2.4 x lo8 fast neutrons per sq. cm per second. At these fluxes, the 69Zn produced by each method would be 2-5 x lo3 c.p.m. per pg of zinc from 68Zn ( n , ~ ) 69Zn and 9.0 x lo3 c.p.m.per g of gallium from 69Ga (n,@) "Zn. This gives an apparent zinc impurity of 1-8 pg per g of gallium arsenide. The contribution from the fast- neutron reaction can be determined by simultaineously irradiating samples with and without cadmium covers. The use of a cadmium cover reduces the slow-neutron reaction 68Zn ( n , ~ ) G9Zn by a considerable amount, and then by comparing the amounts of 6gZn produced in the covered and uncovered samples the amount of zinc present may be inferred by difference. If this method is used in the thermal column, a zinc impurity of 0.1 pg per g of gallium arsenide would give a 20 per cent. difference between cadmium-covered and uncovered samples. DETERMINATION OF MAGNESIUM The main activity produced during a short irradiation of pure gallium arsenide has a half-life similar to that of 27Mg ('OGa, t, = 21 minutes).A rapid separation procedure that removed all gallium activities without complete removal of the long-lived arsenic activity may have been acceptable in the interest of sensitivity, but it has been found possible to remove all gallium and arsenic activities in a procedure taking 20 minutes. After dissolution of irradiated gallium arsenide and equilibration with magnesium carrier, magnesium (and some gallium) was precipitated on addition of sodium hydroxide. The precipitate was dissolved in hydrochloric acid, and the solution was passed through a column of anion-exchange resin, on which the gallium was adsorbed. Magnesium hydroxide was precipitated from the effluent and then redissolved, and ammonium magnesium phosphate was precipitated from a citric acid solution and used for preparing sources and determining the chemical yield.PROCEDURE- Transfer the irradiated gallium arsenide t o a dissolution cell containing 1 ml of concen- trated nitric acid, 10 mg of magnesium carrier and a few milligrams of germanium and zinc carriers. Dissolve the sample by passing a current of 1 to 2 amps at 15 volts for 3 to 4 minutes. Boil for 1 minute, transfer to a centrifuge tube with washings of water, spin in a centrifuge, and discard the supernatant liquid. Wash the precipitate thoroughly with water, and discard the washings. Dissolve the precipitate in 1 ml of concentrated hydrochloric acid, and adjust to 6 N in hydrochloric acid. Pass the solution through a column of anion-exchange resin (see Note 1) at a flow rate of 1 to 2 ml per minute, and then pass 2 ml of 6 N hydrochloric acid through the column (see Note 2).Collect the effluent in a centrifuge tube. Add gallium and arsenic carriers to the solution, and precipitate magnesium hydroxide with solid sodium hydroxide as described above. Spin in a centrifuge, discard the supernatant liquid, wash the precipitate with water, and discard the washings. Add 0-1 g of citric acid, 2 or 3 drops of 10 per cent. w/v potassium iodide solution and 1 ml of 20 per cent. w/v ammonium dihydrogen orthophosphate solution, and make the solution alkaline by adding ammonia solution to precipitate ammonium magnesium phosphate. Spin in a centrifuge, and discard the supernatant liquid.Dissolve the precipitate in concentrated hydrochloric acid, and reprecipitate with am- monia solution as described above. Wash the precipitate successively with 10 per cent. v/v ammonia solution, water and ethanol, and mount for counting and determining the chemical yield. The complete procedure is described below. Transfer the solution to a beaker, and add excess of solid sodium hydroxide. Dissolve the precipitate in concentrated hydrochloric acid, and dilute to 10 ml.July, 19633 MAGKESIUM IN GALLIUM AKSENIDE BY NEUTKOS-ACTIVATION ANALYSIS 527 NOTES- lite CG-400 equilibrated with 6 N hydrochloric acid. should therefore be avoided. 1 . 2. The column was approximately 2 cm long and 0.7 cm in diameter. The resin used was Amber- Excessive washing of the column has been found t o introduce gallium into the effluent, and SENSITIVITY- If 100 counts per minute is taken as the lowest limit of sensitivity, then for an irradiation of 10 minutes at 5 MW and a decay time of 20 minutes, the sensitivity of the determination is 7.1 x 1O-Sg of magnesium.THERMAL NEUTRON SELF-SHIELDING IN GALLIUM ARSENIDE The magnitude of thermal neutron self-shielding is determined by the extent of over- lapping of neutron resonances in the host and trace materials. When this effect is expected, its magnitude must be investigated and, when necessary, suitable corrections applied to the results. Okada' gives an empirical formula for the self-shielding effect, and application of this formula to a weight of 1 g of gallium arsenide suggests that the self-shielding should be less than 10 per cent.The effect of self-shielding on the determination of silicon and magnesium in gallium arsenide has been investigated, the irradiations being carried out in the Merlin reactor. The effect of self-shielding on the determination of zinc has not been attempted, as the interference of fast-neutron reactions is considerable in this determination. EXPERIMENTAL- Small pieces of the element under investigation were irradiated with and without a gallium arsenide cover. The cover consisted of a cube of gallium arsenide with a 6-mm side and a cylindrical hole in the centre of diameter 2 mm and 4 mm deep. The top of the cube was covered by a 2-mm slice of gallium arsenide. A gallium arsenide covered specimen and an uncovered specimen were held in known positions relative to each other in a Perspex jig in an irradiation can.After irradiation the two samples were processed to bring them to a form suitable for weighing and counting. Each sample was checked for purity by measurement of half-life. A second irradiation was carried out with the same sample jig, but without the gallium arsenide cover. Each sample was processed in an identical way to the first pair, weighed, and counted. The difference in the specific activity of the latter pair represented the correction to be applied to take account of flux differences between the two positions. RESULTS- SiZicofl-Two types of irradiation facility were used for this experiment ; the blanking- element facility and the fast-transfer facility.The activities induced in samples of silicon covered with gallium arsenide and uncovered samples, corrected for flux variations as described above, are shown in Table I. TABLE I ACTIVITIES INDUCED IN SILICON WITH AND WITHOUT GALLIUM ARSENIDE COVER Activity, c.p.m. per mg of Si" Irradiation facility J Uncovered 8.95 x 103 * * 1 Covered 8-94 x 103 TJncovered 5.45 x 104 * . { Covered 5-55 x 104 Blanking element .. Fast transfer . . .. * For irradiations of 2 hours at a reactor power of 10 kW. I t is apparent that neutron self-shielding will not cause serious errors in the determination of silicon in gallium arsenide by thermal-neutron activation analysis.528 GREEN, HESLOP AND WH[TELY [Aizalyst, Vol, 88 i%,?agnesium-Similar experiments have been carried out for magnesium ; irradiation was for a short time in the fast-transfer facility.The results obtained, corrected for flux variation, are shown in Table IT. TABLE rr ACTIVITIES INDUCED IN MAGNESIUM WITH AND WITHOUT GALLIUM ARSENIDE COVER Uncovered Covered Uncovered Covered Activity, c.p.m. per mg of Mg* 6-20 x lo3 6.25 x LO3 6-34 x 103 6.5 x 103 * For irradiations of 10 minutes at a reactor power of 10 kW. Again it is apparent that neutron self-shieelding will not lead to serious error in the determination of magnesium in gallium arsenide. CONCLUSIONS Methods have been developed for determining silicon, zinc and magnesium in gallium arsenide, with the separation of short-lived species. The sensitivities for silicon and magnesium are 1.3 x 10-9g and 7.1 x 10-8g, respectively. For zinc, a fast-neutron reaction with gallium interferes, but 1 x lo-' g can be determined under specified conditions. The sensi- tivities for silicon and magnesium represent the limits of the method, but for a host material that does not interfere in the detection of zinc, 2.5 x 10-lo g of zinc could be detected. The effect of thermal neutron self-shielding by gallium arsenide in the detection of silicon and magnesium has been shown to be insignificant. Since the magnitude of this effect is dependent on the position of neutron absorption resonances in the host and trace constituents, this result is not directly applicable to the determination of other trace materials or other matrixes. We thank Dr. T. E. Allibone, C.B.E., F.R.S., Director, A.E.I. Research Laboratory, Aldermaston Court, for permission to publish this paper. REFERE:NCES 1. 2. 3. 4. 5. 6. 7. Jenkins, E. N., and Smales, A. A., Quart. Rev., 1966, 10, 83. James, J. A,, and Richards, D. H., Nature, 1955, 175, 769. illlibone, T. E. et al., Proceedings of the 2nd U.N. International Conference on Peaceful Uses of Foster, L. M., and Stumpf, H. C., J . A w m . Chem. SOL., 1951, 73, 1590. Armand, M., and Berthoux, J., Anal. Chzm. .4cta, 1953, 8, 510. Strickland, J . D. H., J . Amer. Chern. Soc., 1!352, 74, 862. Okada, M., Int. J . AppZ. Radiation and Isotofes, 1962, 13, 53. Atomic Energy, Geneva, 1-13th September, 1958, I'olume 10, p. 202. Rcccived Februavy 18th, 1963

ISSN:0003-2654    

DOI:10.1039/AN9638800522

出版商:RSC    

年代:1963

数据来源: RSC

摘要:

July, 19631 HAMPSON 529 Determination of Zirconium-95 and Niobium-95 in Seaweed and Sea Water BY B. L. HAMPSON (Ministry of 14gricult.tive, Fisheries and Food, Fisheries Radiobiological Lahorafovy, Hamilton Dock, Lowestoft, Suflolk) A method is described for determining zirconium-95 and niobium-95 in seaweed and sea water. Oxalic acid in nitric acid is used to complex both elements, and, after destruction of the oxalic acid with potassium chlorate, zirconium is precipi- tated as zirconium phosphate and niobium as niobic acid. After removal of rare-earth activities, zirconium is separated as barium fluorozirconate, decontaminated from other active isotopes, and finally ignited t o zirconium dioxide ; the niobium fraction is similarly decontaminated, and finally ignited to niobium pentoxide.The oxides are then determined by gamma-ray spectrometry. METHODS for measuring radionuclides in sea-water and marine organisms, including marine foodstuffs, are essential in the control of the discharge of radioactive effluent into the sea. Safe limits for the release of radioactivity from a nuclear installation are calculated from the results of preliminary studies on radionuclides discharged a t low release rates, and are then checked by regular monitoring during normal operation. An assay procedure is described here for zirconium-95 and niobium-95 in environmental materials. Both these nuclides are among the more important fission products in the effluent discharged from plant processing spent nuclear fuel. The two nuclides are normally present together in effluent, but the proportions released vary, and individual measurement is necessary since the maximum permissible concentrations in drinking water1 and in foodstuffs, which set the release limits, are different.Separate measurement is essential in the investigation of the distribution patterns of the two elements between sea water, biota and sediment. Further, correction for radioactive decay during handling of samples is facilitated by simultaneously determining both isotopes, as the transient parent - daugher relationship can then be resolved. Zirconium-95 and niobium-95 can be measured together by gamma-ray spectrometry, but cannot be readily distinguished because of close energy levels. Difficulty was experienced in applying the di-n-butyl phosphoric acid separation proposed by Scadden and Ballou2 to biological ash, owing to extraction of large amounts of inorganic constituents.A modification of Kahn's phenylarsonic acid procedure3 gave only 20 to 30 per cent. recovery of zirconium when applied to material having a high phosphate content. Separation of these nuclides by precipitation of the zirconium as barium fluorozirconate, as used by H ~ m e , ~ was adopted after they had been obtained from the sample by precipitation with a carrier as zirconium phosphate and niobic acid. Subsequently, a series of radiochemical purification steps was carried out, each designed to give the required degree of decontamination from particular groups of active or inactive elements present or expected to be present.Radio-assay was then carried out by gamma scintillation counting. METHOD REAGENTS- (obtainable from Johnson, Matthey & Co. Ltd.) in 250ml of water. 16 M nitric acid. crucibles and igniting a t 800" C for 3 hours; weigh the residue as zirconium dioxide. about 50 ml of water. Zirconium carrier solution, 10 mg per ml-Dissolve 7.5 g of high purity zirconium nitrate Add a few drops of Standardise the solution by evaporating 5-ml portions to dryness in silica Niobium carrier solution, 10 mg per ml-Dissolve 6-5 g of potassium hexaniobate in Add, slowly with stirring, 4 ml of 16 M nitric acid to the near-boiling530 HAMPSON DETERMINATION OF ZIRCOXIUM-95 [Analyst, Vol. 88 solution, and then heat, and stir for 5 minutes. Filter, and wash the precipitated niobic acid with a hot 2 per cent. solution of ammonium nitrate.Dissolve the washed precipitate in 50ml of a hot saturated solution of oxalic acid, filter, dilute the filtrate to 250m1, and store in a polythene container. To standardise the solution boil 5-ml portions with 30 ml of 6 M nitric acid and 1 g of potassium chlorate. Cool, and add 16 M ammonium hydroxide until the pH of the solution is 8 to 10. Collect -the precipitate on a Millipore HA membrane filter, and wash with hot water. Ignite at 800" C for 3 hours, and weigh the residue as niobium pentoxide. Standard xirconium-95 - niobium-95 solution, approximately 0-02 pC per ml-An old, i.e., near equilibrium, isotopic mixture standardised with respect to both isotopes is obtainable from the Radiochemical Centre, Amersham; it should be used only up to 6 months from the time of standardisation.Lanthanum carrier solution, 10 mg per ml-I)issolve 3.2 g of lanthanum nitrate in 100 ml of water. Ruthenium carrier solution, 10 mg per ml--Dissolve 6.3 g of ruthenium trichloride in 250 ml of 0.1 M hydrochloric acid. Cup ferron solution-Dissolve 6 g of analytical-reagent grade cupferron (having a sul- phated ash content (0.01 per cent.) in 100 ml of water. Store the solution in a refrigerator. Thioacetamide solution, 100 mg per ml. Mandelic acid solution, saturated. Thioglycollic acid, 20 per cent. v/v. APPARAT u S- well crystal protected with steel and lead, and a single-channel pulse-height analyser. Gamma counting equipment-A scintillation counter having a sodium iodide - thallium GZassware-Treated with dimethyldichloroc .a '1 ane.PROCEDURE- Preliminary isolation from seaweed-Carefully wash the seaweed with sea water during collection. Ash the seaweed quantitatively by drying at 140" C, and ignite at 500" C in silica; then heat at 800" C in platinum to remove carbon, and grind the residue to pass a 240- mesh sieve. Add 20 ml of M nitric acid - 0-02 M oxalic acid per g of ash, and heat, with stirring, to 70" C. Add 1.00 ml each of niobium and zi.rconium carrier solutions, and stir; then add 0-25 g of potassium chlorate per 20 ml of acid, and boil gently to precipitate zirconium phosphate and niobic acid. Cool, and remove the precipitate by spinning in a centrifuge; repeat with a further 1-00 ml of both carrier solutions. Combine the precipitates, and wash with 20 ml of 3 M hydrochloric acid.Digest the washed precipitate with 5 ml of 40 per cent. hydrofluoric acid for 1 hour at 100" C in a capped polypropylene tube. Spin in a centri- fuge, and repeat the digestion on any residue. Transfer the solution to a platinum dish, and evaporate to dryness with 20 ml of 16 M nitric acid. Carry out a second evaporation with further portions of hydrofluoric and nitric: acids, dissolve the residue in 2 ml of 40 per cent. hydrofluoric acid, transfer to a nitrocellulose centrifuge tube, and dilute to 10 ml. Preliminary isolation from sea water-Collect filtered- sea water in a polythene bottle. Measure the volume, and transfer the entire sample to a beaker, washing the bottle with 20 ml of 0-5 per cent. oxalic acid per litre of sea water.Add 1.00 ml each of niobium and zirconium carrier solutions, and then, for each litre of sea water, 100 ml of 16 M nitric acid, 1 ml of phosphoric acid, sp.gr. 1.75, and 1 g of potassium chlorate. Extract zirconium-95 and niobium-95 by heating to 90" C until a precipitate forms and coagulates sufficiently for complete collection by decanting and spinning in a centrifuge. Add a further 1.00 ml of each carrier solution, and repeat the scavenge. Combine the two precipitates, and transfer to a polypropylene centrifuge tube with the aid of 20 ml of 3 M hydrochloric acid. Dissolve the precipitate in 2 ml of 40 per cent. hydrofluoric acid, and heat in a boiling-water bath to ensure equilibration of the active isotopes with the carriers. Dilute to 10 ml, and transfer to a nitrocellulose centrifuge tube. Separation, purification and assay of xirco:~ium-95 and nio bium-95-Remove rare-earth activities by a double lanthanum fluoride scavenge, with 0-5 ml of lanthanum carrier solution each time.Separate zirconium by precipitation with 2 ml of saturated barium nitrate solu- tion. Suspend the barium fluorozirconate in 2 ml of saturated boric acid solution, andJuly, 19633 AND NIOBIUM-95 I N SEAWEED AND SEA WATER 53 1 redissolve in 1 ml of 16 M nitric acid and 10 ml of water, neglecting any insoluble barium sulphate; reprecipitate with 1 ml of 40 per cent. hydrofluoric acid. Combine the two solutions for the niobium fraction. For adequate decontamination of the zirconium fraction from radioactive niobium, carry out a third precipitation cycle; then dissolve the precipitate in 3 ml of 11 M hydrochloric acid, 1 ml of saturated boric acid solution and 26 ml of water, and precipitate the barium with 2 drops of sulphuric acid, sp.gr.1.84. Scavenge active ruthenium isotopes by adding 1 ml of ruthenium carrier solution and 1 ml of thioacetamide solution, and digesting at 90" C until coagulation is complete and the solution is colourless. Filter through a Millipore HA membrane filter. Finally, recover the zirconium by precipitation with 18 M ammonium hydroxide in the presence of 2 ml of 20 per cent. thioglycollic acid solution added to complex the iron; wash the precipitate with a 2 per cent. solution of ammonium nitrate. Dissolve the washed precipitate in 10ml of 6 M hydrochloric acid, and reprecipitate with 10ml of saturated mandelic acid solution by heating in a boiling-water bath.Collect the precipitate on a filter-paper, wash with boiling water, and ignite at 800" C for 3 hours. Weigh the zirconium dioxide into a gamma counting tube, count, and calculate the recovery of zirconium. Decontaminate the niobium fraction from radioactive zirconium by adding 1 ml of zirconium carrier solution and precipitating with 1 ml of saturated barium nitrate solution. Remove ruthenium isotopes as described above ; use a polytetrafluoroethylene or polypropyl- ene beaker or flask. To recover niobium from the hydrofluoric acid medium, mix the solution with 35 ml of saturated boric acid solution and 10 ml of 11 M hydrochloric acid. Extract the cupferron complex into chloroform by using 4ml of 6 per cent.cupferron solution and 20 ml of chloroform for a first 1-minute extraction and 2 ml of cupferron solution and 10 ml of chloroform for second and third 1-minute extractions. Cool all solutions to 10" C. Destroy organic matter by heating the chloroform phase with 3 ml of sulphuric acid, sp.gr. 1-84, and 20 ml of 16 M nitric acid until the solution is nearly colourless and about 3 ml in volume. Add 1 or 2 drops of 16 M nitric acid to the boiling solution to complete the oxidation. Pre- cipitate niobium in the presence of 2 ml of 20 per cent. thioglycollic acid by adding 16 M ammonium hydroxide until the pH is 10, and wash the precipitate twice by boiling with 20 ml of 1.2 M ammonium hydroxide - 0.6 M ammonium nitrate solution, separating the precipitate each time by spinning in a centrifuge. Immediately, dissolve the washed precipi- tate in 0.5 ml of 6 M nitric acid and 2 ml of saturated oxalic acid solution, and reprecipitate the niobium by heating with 10 ml of 16 M nitric acid and 0.5 g of potassium chlorate; digest in a boiling-water bath for 15 minutes.Separate the precipitate by spinning in a centrifuge, and suspend in 30 ml of 0-6 M nitric acid - 1.2 M ammonium nitrate solution. Boil, filter through a Millipore HA membrane filter, and wash the filter with 2 to 5 ml of the acidic wash solution. Ignite the precipitate at 800" C for 3 hours. Weigh the niobium pentoxide into a gamma-counting tube, count, and calculate the recovery of niobium. For radio-assay mount the sources in glass tubes having wall thickness a t least 200 mg per sq.cm to eliminate the beta radiation. Count the gamma radiation from zirconium-95 and niobium-95 through a 0.72- to 0-76-MeV window selected by setting up on a source containing the two isotopes in similar proportions. Calibrate the counter with a standard zirconium-95 - niobium-95 source at a counting rate within the linear response region of the instrument. CALCULATION- given by- The activity of the standard source, Ac, in picocuries (pC) a t the time of counting is where 2, = the activity due to zirconium-95 a t the time of standardisation (pC), N o = the activity due to niobium-95 at the time of standardisation (pC), t = the interval between standardisation and counting, = the disintegration constant for zirconium-95 and A2 = the disintegration constant for niobium-95.532 HAMPSON : DETERMINATION OF ZIRCONIUM-95 [APzaLvvst, Vol.88 The zirconium-95 activity at the time of sampling, Zs, is given by- and the niobium-95 activity at the time of sampling, N,, is given by- where Z c = the zirconium-95 activity at the time of counting, N c = the niobium-95 activity at the time of counting and T = the interval between sampling and counting. DISCUSSION OF THE METHOD The problem of collecting the two isotopes from bulky samples of natural materials is most easily solved by sample dissolution and carrier exchange in nitric acid containing oxalic acid as complexing agent for both elements. Subsequent destruction of the oxalic acid with chloric acid in the same acidic medium leads tcl the precipitation of zirconium as phosphate and niobium as niobic acid, with little occlusion of matrix constituents, and at the same time considerable decontamination from most other active isotopes.This self-scavenging pro- cedure is readily applicable to a wide range of samples. Hydrofluoric acid is undoubtedly a better complexing agent for both elements, but their recovery in isolation is more difficult, and the use of this reagent in large scale operations is to be avoided. When seaweed is analysed as a foodstuff, complete dissolution of the sample, including siliceous material, is required, as all may be ingested by man. Matter insoluble in nitric - oxalic acid mixture often amounts to about 1 per cent. of the ash.Dissolution of this com- ponent can be achieved either by fusion with potassium hydrogen sulphate and then extraction with oxalic acid or by digestion with hydrofluoric acid. Both gave about 98 per cent. recovery of zirconium-95 - niobium-95 from the ash of seaweed and detritus samples taken from the shore. After this treatment removal of silica, a t least down to 20 mg, is essential. After digestion with hydrofluoric acid, this can be accomplished by evaporation with nitric acid. Each stage of such an evaporation gives a ten-fold reduction in silica with negligible loss of volatile niobium pentafluoride, which would occur if sulphuric acid was used. When seaweed is analysed as a biological entity, most of the siliceous matter may be considered extraneous ; such matter, if finely divided, may adsorb considerable amounts of zirconium-95 and niobium-95 from sea water However, such directly adsorbed activity will be rendered refractory during ashing, and more relevant results may be obtained by rejecting the matter insoluble in nitric - oxalic acid mixture.Zirconium-95 - niobium-95 activity associated with the soluble salts in seaweed ash dissolves readily in this medium, as was shown by a trial experiment in which 98.7 per cent. of zirconium-95 - niobium-95 mixture, added to the ash immediately before the final treatment a t 800" C, was recovered, Geiger6 has shown that ruthenium can be precipitated by thioacetamide from 2 M hydrochloric acid containing 0-08 M hydrofluoric acid. Experiments carried out in con- nection with this procedure indicated that the operation can also be conducted in 2 M hydro- fluoric acid.Mean decontamination factors from ruthenium-106 for the entire procedure were 1 x lo5 for zirconium fractions and 5 x lo4 for niobium fractions. These are adequate for monitoring samples in which ruthenium-106 predominates, e.g., current samples from the Windscale area, the bias from this cause not exceeding +1 per cent. The method has so far been applied to the edible seaweed Porphyra used in the manu- facture of laverbread, to the more widely occurring Fucus species of seaweed often chosen in fallout and effluent dispersal studies, to fish flesh and to sea water. Chemical recovery is normally about 70 per cent. for zirconium and 80 per cent. for niobium. A test of the procedure with zirconium-95 - niobium-95 added to seaweed ash indicated a coefficient of variation of 2 per cent.(5 replicate determinations), with a bias of -2 per cent. Similar analytical variation was found for both isotopes on environmentally contaminated samples of seaweed. Gamma counting was adopted for routine measurements because of easy source preparation. When 20 g of seaweed ash or 5 litres of sea water were used, 0.2 pC of zirconium-95 or niobium-95 The sensitivity attainable depends on sample size and counting statistics.July, 1963: AND NIOBIUM-95 IN SEAWEED AND SEA WATER 533 per g of seaweed or 0.02 pC per g of sea water could be determined with an over-all coefficient of variation of 3 per cent. by the gAmma-counting method described. With these amounts, measurements of this accuracy can be made down to times the maximum permissible concentration in seaweed; for measurements at this level two sources and background can be counted in 24 hours if equipment having a counting efficiency of 9 per cent. and a background rate of 12 counts per minute is used. A ten-fold increase in sensitivity can be obtained by low-level soft beta-particle ~ounting,~ although this involves the spreading of uniformly thin sources and manipulation of gas-flow counters. However, this extra sensitivity is needed for accurate measurements of zirconium-95 and niobium-95 at the low concentrations found in fish flesh arising from fallout. REFERENCES 1. 2. Scadden, E. M., and Ballou, N. E., Anal. Chem., 1953, 25, 1602. 3. 4. Recommendations of the International Commission on Radiological Protection (I.C.R.P.), Report of Committee 11, Pergamon Press, New York and London, 1959. Kahn, B., U.S. Atomic Energy Commission Report ORNL 1951, Oak Ridge, Tennessee, 1955. Hume, D. N., in Coryell, C., and Sugarman, N., Editors, “Radiochemical Studies: The Fission Products,” National Nuclear Energy Series IV-9, Book 3, McGraw-Hill Book Co. Inc., New York, 1951, p. 1499. 5. 6. Sugihara, T. T., James, H. I., Troianello, E. J., and Bowen, V. T., Anal. Chew., 1969, 31, 44. Geiger, E. L., Ibid., 1959, 31, 806. Received November 26th, 1962

ISSN:0003-2654    

DOI:10.1039/AN9638800529

出版商:RSC    

年代:1963

数据来源: RSC

摘要:

534 SIVARAMA SASTRY, PADMANA-BAN, ADIGA AKD SARMA: [Analyst, Vol. 88 A Sensitive Microbiological Assay Procedure for determining Magnesium in Biological Materials BY K. SIVARAMA SASTRY,* G. PADMANABAN, P. R. ADIGA AND P. S. SARMA (Department of Biochemistry, Indian Institute of Scieizce, Bangalore 12, India) A simple microbiological assay procedure for determining magnesium in the range 1 to 8 pg has been developed, in which Neuvospova Gvassa Em 5297a is used as the test organism. For the range 1 to 4 pg of magnesium, an indirect assay procedure based on the determination of the mycelial content of the “phenol reagent - positive material’’ has also been evolved, the dry weight of mycelia being directly used as an index of response for higher levels of magnesium. The method has been found to be applicable to various biological tissues, the recoveries being about 95 to 105 per cent.ATTEMPTS to develop microbiological methods for determining metals have so far been mainly unsuccessful. Manganese has been assayed with Lacto bacillus arabinosus 17-5 in biological materials,l and Aspergillus niger has been used for determining several metals, including magnesium.2 However, the A . niger procedure for determining magnesium requires a sample containing a t least 50 pg of metal at a concentration of 1 pg per ml of culture medium. Macleod and SnelP studied the magpesium requirements of several Lactobacilli with a view to using them for the assay of magnesium, but found that, even with pre-treatment of a magnesium deficient medium with Lactobacilli, it was not possible to reduce growth in an unsupplemented medium.During earlier work with the mould Neurospora crassa4 it was found that the wild strain of this mould, Em 5297a, had a low requirement for mag- nesium, and growth was totally absent if pure salts were used to constitute the medium and magnesium was omitted. Advantage was taken of this observation, and a sensitive microbiological assay procedure is described here, in which this mould is used as the test organism. METHOD APPARATUS- The mould was grown in Pyrex-glass test-tubes (150 mm x 15mm) on a reciprocal shaker (stroke 10.8 cm; 84 cycles per minute) in a constant-temperature chamber kept at 30” 2 1” C; the test-tubes were kept inside the chamber in a test-tube-rack mounted so that the tubes were tilted, lips upwards, so as to spread the medium over a length of about 7 to 8 cm.After inoculation, the organism was grown with continuous shaking during the entire growth period. Pyrex glassware was used throughout ; it was cleaned with chromic - sulphuric acid, and then rinsed with distilled water, and, finally, with water distilled in an all-glass apparatus. REAGENTS- further purified by distillation before use. Farr and Randall.5 ASSAY MEDIUM AND GROWTH TECHNIQUE- The culture medium used in all assays was based on that used in earlier work4 with the difference that magnesium sulphate was replaced by sodium sulphate to provide an equivalent amount of sulphur. The percentage composition was : glucose, 2 ; potassium dihydrogen orthophosphate, 0.3 ; ammonium nitrate, 0.2; ammonium tartrate, 0.1 ; sodium sulphate, 0.025; sodium chloride, 0.01 ; calcium chloride, 0.01.A cids-Analytical-reagent grade acids were used ; the concentrated nitric acid was Folin and Ciocaltezds phenol reagent-Prepared as recommended by Lowry, Rosebrough , * Present address : Department of Biochemistry, University of Washington, School of Medicine, Seattle, Washington, U.S.A.July, 19631 DETERMINATION OF MAGNESIUM I N BIOLOGICAL MATERIALS 535 The trace elements supplied (in pg per 100 ml) were : zinc, 20 ; manganese, 20 ; copper, 8 ; iron, 2; molybdenum, 2. Biotin (0.5 pg per 100 ml) was added, and one drop of Tween 80 per 100 ml of medium was also added to inhibit sporulation.6 For assay purposes, the organism was grown on the above medium (adjusted to pH 4.8 to 5.0) for 72 hours a t 30” A 2.0-ml portion of double-strength medium and 2-0 ml of sample in aqueous solution adjusted to the pH of the medium were taken in each assay tube.STOCK CULTURE AND INOCULUM- It was maintained bj, transfer biweekly on agar slants of the above medium supplemented with 0.2 per cent. each of yeast and malt extracts and containing 2 per cent. of agar. For inoculation, spore from a 7-day-old culture was used, one drop of a suspension of spore having a transmission of approximately 90 per cent. in sterile water distilled in an all-glass apparatus being used in each tube. ASSESSMENT OF GROWTH REPONSE- At the end of the growth period, the mycelia were carefully removed, washed, and dried at 60” C overnight; the assay was then completed by determining the mycelial weight.In assaying magnesium in the range 1 to 4pg, the mycelia were carefully removed, rinsed on a filter-paper with water distilled in an all-glass apparatus, and transferred to Pyrex- glass test-tubes (100 mm x 10 mm). A 0.5-ml portion of 5 N sodium hydroxide was added to the contents of each tube, and the tubes were heated on a boiling-water bath for 30 minutes ; they were then cooled, and the contents were adjusted to 5 ml with distilled water. Portions (0.5 ml) were removed, after centrifugation if necessary, for determining the “phenol reagent - positive material” according to the method of Lowry et aZ.,5 a Klett-Summerson photo- electric colorimeter and a No. 66 red filter being used. PREPARATION OF SAMPLE- The biological sample was wet-ashed in a 50-ml Pyrex-glass conical flask with 10 ml of concentrated nitric acid and 1.0 ml of 66 per cent.perchloric acid (Merck G.R.) per gram of the sample on a sand-bath. After complete evaporation, the digest was treated with 5 ml of concentrated hydrochloric acid, and once again evaporated to dryness. To the residue dissolved in 5.0 ml of distilled water was added 0.1 ml of a 50 per cent. w/v solution of ammonium chloride, and then 0.05 ml of ammonia solution, sp.gr. 0.88, was added to make the solution alkaline. A 0-1-ml portion of a 0.5 per cent. solution of ferrous sulphate was added and hydrogen sulphide was bubbled through the solution for 15 to 20 minutes. After the solution had been set aside for 2 hours, the precipitated ferrous sulphide, together with sulphides of other heavy metals, if any, was removed by filtration through a fluted Whatman No.42 filter-paper, and washed with 15 to 20ml of distilled water. The filtrate and the washings were combined, and evaporated to dryness on a sand-bath, 1.0 ml of concentrated nitric acid being added during the later stages of the evaporation. The residue was dissolved in a known volume of distilled water, and adjusted to pH 4-8 to 5.0. Portions of this solution were used for the assay. RANGE OF ASSAY- Under the conditions of growth obtaining in the proposed procedure, when analytical- grade reagents are used and normal precautions are taken to eliminate magnesium as a contaminant, growth is negligible on the basal medium constituted without magnesium.The microbiological response curve obtained is shown in Fig. 1. It will be seen that the useful range of assay is from 1 to 8 ,ug of magnesium. The mycelial weights have been found to be highly reproducible, variation between duplicates never exceeding 0.2 mg over the entire range. For samples containing 1 to 4 pg of magnesium, the indirect method of estimating growth based on “phenol reagent - positive material” has been developed to ensure higher accuracy without involving elaborate manipulation. If the mycelia are pressed gen tljr to dryness before suspension in sodium hydroxide, the alkali digest can be directly adjusted to volume by adding 4.5 ml of water, and portions can be removed for assay. Under the 1” C with shaking as described above.Neurospora crassa Em 5297a was used as the assay organism. RESULTS536 SIVARAMA SASTRY, PADMANA13AN, ADIGA AND SARMA: [Analyst, vol. 88 conditions proposed, we have found the extraction of “phenol reagent - positive material” is complete in one step. In the range 1 to 4 pg of magnesium, the “phenol reagent - positive material” gives a linear standard graph. Preli-minary experiments showed that over 70 per cent. of the “phenol reagent - positive material” is protein. For samples containing less than 1-0 pg of magnesium, the microbiological assay procedure is not recommended, even with the indirect method based on “phenol reagent - positive material” of the mycelium. TABL:E I APPLICABILITY OF THE N . crassa ASSAY PROCEDURE FOR DETERMINING MAGNESIUM IN BIOLOGICAL TISSUES Portion taken ml Teconza stuns leaves 0.2 0.4 0-5 L.sativus seeds . . 0-2 0.3 0.4 Rat kidney . . 0.2 0.4 0.6 Rat blood . . . . 1.0 1-5 1-9 Rat spleen. . . . 0.2 0.4 0.6 Rat liver . . . . 0.2 0.4 0.6 Sample for assay, Magnesium found, Pg 2.50 4.95 6.20 3.00 4-35 5.90 2.25 4.60 6.70 1.20 1.80 2.45 1.65 3.40 5.00 2.10 4.30 7-00 Total magnesium content, pg per g fresh weight 625.0 618.7 620.0 1500 1450 1475 162-5 166.1 158-9 30.07 30.07 32.27 221.7 228.5 224.0 145-8 149.3 161.7 Magnesium recovered, * 3.50 5.90 7-10 3.95 5-30 6.90 3-20 5.55 7-80 2.20 2.75 3.40 2-70 4.40 6.00 3.15 5.25 7.9.5 * Amount recovered after 1 pg of magnesium had been added. f Expressed as pg per ml. 0 Concentration of tmagnesium. pg per 4 ml Fig. 1. Microbiological assay curve for magnesium with N .cyassa Magnesium recovered, 100 95 90 95 95 100 95 95 110 100 95 95 105 100 100 105 95 95 %July, 19631 DETEKMINATION OF MAGNESIUM IN BIOLOGICAL MATERIALS 537 APPLICABILITY OF THE METHOD TO BIOLOGICAL TISSUES The applicability of the proposed method has been studied for several biological tissues. Recoveries have been tested by adding magnesium at the l-0-pg level to three different portions of each biological tissue tested. The absolute values and the mean recoveries obtained are shown in Table I, and represent values derived from at least four independent analyses for each type of sample. It can be seen that the method is suitable for the type of samples investigated throughout the concentration range. INTERFERENCE FROM OTHER METALS The effect of other metals on the assay of magnesium by the proposed procedure has been studied.The metals mentioned below do not interfere in the assay up to the levels indicated (pg per tube) : sodium, 3000; potassium, 51,000; calcium, 2100; iron, 1000; zinc, 10; cobalt, 10; nickel, 5. DISCUSSION OF THE METHOD The only useful microbiological method for determining magnesium so far has been that of Nicholas2 in which A . niger is used. However, it involves the use of carefully purified media and a sample having a magnesium content of at least 50 pg. The considerably lower require- ment for magnesium of N . wassa and the fact that elaborate purification of the assay medium is unnecessary are distinct advantages in the proposed method. Further, by use of a simple procedure for the indirect determination of growth reponse, it is possible to work conveniently with good precision at levels as low as 1 pg.The proposed procedure is also specific for magnesium, since calcium, sodium and potassium do not interfere, except a t extremely high levels compared with the amount of magnesium present. Heavy metals, such as cobalt, nickel and zinc, which interfere with magnesium metabolism in the mould4 and are often present in biological material in various amounts, were found to be completely eliminated by the hydrogen sulphide treatment introduced during the preparation of the sample. Since uniformly good recoveries were obtained for all the sample sizes with the various types of biological materials, the proposed procedure can be adopted with advantage, particularly since the conventional complexometric titration of magnesium’ involves a correction for the presence of calcium and is of limited usefulness below 10 pug of magnesium. The financial assistance of the Council of Scientific and Industrial Research, New Delhi, and The Rockefeller Foundation, New York, is gratefully acknowledged. 1. 2. 3 . 4. 5. 6. 7. REFERENCES Bentley, 0. G., Snell, E. E., and Phillips, P. H., J . Biol. Chem., 1947, 147, 343. Nicholas, D. J. I)., Analyst, 1952, 77, 629. Macleod, R. A., and Snell, E. E., J . Biol. Chem., 1947, 147, 351. Sivarama Sastry, K., Adiga, P. R., Venkatasubramsnyam, V., and Sarma, P. S., Biochem. J . , Lowry, 0. H., Rosebrough, N. J., Farr, A. L., and Randall, R. J., J . Biol. Chem., 1951, 193, 265. Zalokar, M., Arch. Biochern. Biophys., 1954, 50, 71. Schwarzenbach, G., “Complexometric Titrations, ” translated by Irving, H., Intcrsciencc Pub- First received June 4th, 1962 Amended, February 19th, 1963 1962, 85, 486. lishers Inc., New York, 1957, p. 62.

ISSN:0003-2654    

DOI:10.1039/AN9638800534

出版商:RSC    

年代:1963

数据来源: RSC

摘要:

538 MITCHELL, BYRNE AND TRESADERN : PYRETHRUM ANALYSIS : [ A ~ ~ d y s t , Vol. 88 Pyrethrum Analysis: ,4n Infrared Method BY WILLIAM MITCHELL, J. H. N. BYRNE AND F. H. TRESADERN (Staffoord Allen & Sons Ltd., Wharf Road, London, iV.1) A method of pyrethrum analysis i:3 described, in which “pyrethrin I” is determined by infrared measurement. The results are similar to those obtained by the Association of Official Agricultural Chemists’ method, and it is shown that the Pyrethrum Board of Kenya’s method gives falsely high results for “pyrethrin I.” The determination of “pyrethrin 11” and the accuracy of pyrethrum analyses are dixussed. MANY methods have been proposed for the asrsay of pyrethrum, but in recent years com- mercial analyses have generally been conducted by one of two variants of the “mercury reduction” method originally devised by Wilcox on1 and modified by Holaday.2 One version of the method, recommended by the Association of Official Agricultural chemist^,^ is generally known as the A.O.A.C.method. The other (mainly used in the United Kingdom) is described in the British Pharmaceutical Codex,4 the Britis’h Veterinary Codex,5 and in leaflets published by the Pyrethrum Board of Kenya (P.B.K.)6 and by the East African Extract Corporation7; the P.B.K. method may be taken as typical of 1:hese four, which differ in only minor details. They also resemble the A.O.A.C. method in requiring hydrolysis of a purified extract of pyreth- rum flowers with alcoholic alkali and removal of extraneous acids from the hydrolysate, after dilution with water and removal of alcohol, by the addition of barium chloride. In the P.B.K.method, a filtered portion of the solution is acidified with hydrochloric acid, and the liberated chrysanthemic acid is extracted with light petroleum. The important dif- ference is that in the A.O.A.C. method sulphuric acid is used for this liberation, precipitating barium sulphate from the excess of barium chloride present. This precipitate must be filtered off before the solvent extraction is carried out. It is well known that the A.O.A.C. method gives results for chrysanthemic acid, and hence “pyrethrin I,”* that are some 10 to 15 per cent. lower than those given by the methods in which hydrochloric acid is used, and this has been ascribed to loss of chrysanthemic acid by occlusion on the barium sulphate precipitate.Mitchell* confirmed this, showing that chrysanthemic acid could be recovered from barium sulphate residues (previously leached with water until the washings contained no chrysanthemic acid). I t was also pointed out by Beckley and Hopkinsg that a group of amorphous acids was occluded on the barium sulphate residue, and they suggested that these in turn were responsible for the retention of chrysanthemic acid. It is empirical in nature and gives correct results only by strict adherence to the stated conditions of time, temperature, etc. Further, the factor used to calculate the results is an average value reached as a result of collaborative trial between several laboratories.1° Evidently, there is not a stoicheiometric relation between the reactants.An attempt to find a more specific means of determination prompted the work dcscribed here, which invokes measurement of the infrared absorption of chrysanthemic acid. This has given promise, and the results obtained are in fairly close agreement with those provided by the A.O.A.C. method. It is now clear that the methods in which hydrocliloric acid is used, despite previous claims, include and measure some extraneous acids as “chrysanthemic acid.” It was nevertheless a surprise to us that agreement between the infrared and A.O.A.C. results was so close, particularly as it is known that some chrysanthemic acid (though perhaps less than was formerly thought) is lost on the barium sulphate precipitate. It seems evident that the extraneous acids (possibly the amorphous acids claimed by Beckley and Hopkinsg) are mostly removed with the barium sulphate precipitate ; it seems likely, however, that a proportion is not lost, but is recorded as “chrysanthemic acid,” and that it is similar in amount to the true chrysanthemic acid that is lost.In other words, it may be that the A.O.A.C. method gives results in fair agreement with those by the infrared method by a summation of errors, positive and negative. have similar significance for “pyrethrin 11” and total “pyrethrins.” The “mercury reduction” reaction itself leaves much to be desired. * Comprising pyrethrin I and cincrin I, but calculated as pyrethrin I-hence the parentheses, whichJuly, 1963j AN INFRARED METHOD 539 EXPERIMENTAL The infrared spectrum (see Fig. 1) of pure D-trans-chrysanthemic acid (isolated after hydrolysis of pyrethrum extract; m.p.18.5" C; mercury reduction assay 99.9 per cent.), dissolved in carbon tetrachloride, shows an isolated peak at about 9.0 p, which we selected for purposes of measurement. Our procedure was to scan the range 8-0 to 9 6 p , draw a base line between the two troughs a t 8.6 and 9.2 p, measure the percentage transmission (To) from the radiation zero to this base line, and also the percentage transmission (T) from the radiation zero to the absorption peak. This method of measurement was adopted to discount, so far as possible, extraneous absorption caused by impurities in the acid. It was shown (see Table I) that the absorption of pure chrysanthemic acid obeyed the Beer - Lambert law over a considerable range of concentrations, the resulting graph being a straight line passing through the origin.The absorption (A) was given by log,, T,/T. 5 Fig. 1. Infrared spectrum of D-trans-chrysanthemic acid in carbon tetrachloride TABLE I INFRARED ABSORPTION AT 9 p OF NATURAL D-~YLWS-CHRYSANTHEMIC ACID (IN CARBON TETRACHLORIDE) RELATED TO CONCENTRATION f Concentration, g per 10 ml 0.0130 0-0234 0.0341 0.0422 0.0572 0.0656 0.0748 0.0886 Transmission To, T, % % 95.7 65.5 95.8 52.8 91.0 38.8 91.3 31-7 85.3 20.3 85.9 17-4 82-3 12.5 79.7 9.1 A 7 - Absorption 0.165 0.259 0.370 0-459 0.623 0.693 0.819 0.943 (log T o m To apply this method of measurement to the analysis of pyrethrum extract, we used the P.B.K. method6 to the stage where the chrysanthemic acid had been extracted with light petroleum (boiling range 40" to 60" C).This light petroleum solution was then extracted with four portions (10, 10, 5 and 5 ml) of 0.5 N aqueous sodium hydroxide. The combined alkaline extracts were neutralised to phenolphthalein with concentrated hydrochloric acid, and then a 0.5-ml excess of the acid was added. The acidified solution was then extracted with five portions (3, 2, 2, 1 and 1 ml) of analytical-reagent grade carbon tetrachloride. Each extract was successively strained through anhydrous magnesium sulphate (0.5 g sup- ported on a plug of cotton-wool in a small funnel) into a 10-ml calibrated flask; more solvent540 MITCHELL, BYRNE AND TRESADERN : PYRETHRUM ANALYSIS: [Analyst, VOl.88 was then passed through the magnesium sulphate and into the flask to give a total volume of 10ml. Infrared measurements were made on this solution as described above with a Perkin-Elmer “Infracord 137” infrared spectl-ophotometer and 1 -0-mm cells. From the measured absorption (A), the weight (x) of chrysanthemic acid contained in the 10 ml of solu- tion was then read off from the graph prepared from the measurements recorded in Table I. “pyrethrin I” was calculated by substitution in the formula- x x 41 x 5 x 100 P,I, yo = ____ 2 1 X 4 X W where x = weight of chrysanthemic acid, in grams, and w = weight of sample of extract, in grams. “Pyrethrin 11” was determined by the usual procedure, but with the additional modi- fication described by Mitchell and Tresadernl to reject extraneous water-insoluble acidic matter.The A.O.A.C. assays all included the prescribed preliminary chilling in light petroleum solution. With the P.B.K. method, comparisons were made with and without the pre- liminary treatment in light petroleum solution. Omission of the treatment means that the extract, as received, was subjected directly to hydrolysis. RESULTS AND DISCUSSION Table I1 shows the comparative results obtained for “pyrethrin I” on eleven samples of Kenya pyrethrum extract, ordinary grade. I t can be seen that those obtained by the recommended infrared method are significantly lower than those found by the P.B.K. method (mean difference is 11 per cent.), that omission of the preliminary treatment in light petroleum solution causes only slightly higher results by the recommended infrared method, and that the results by the A.O.A.C.method are fairly closely similar to, and generally slightly lower than, those obtained by the recommended infrared method. TABLE I1 COMPARATIVE RESULTS FOR “PYRETHRIN I” IN PYRETHRUM EXTRACTS “Pyrethrin I” found by- Recommended method P.B. K. With preliminary Sample method,6 treatment, A 13.6 11.9 B 13.2 11.6 C 13.5 12.1 D 13-3 12.0 E 13.2 11.3 F 13.8 12.0 G 13.3 11.5 H 14-8 13.5 I 13.4 12.2 14-5 12.5 13.6 11.7 K Mean . . 13-4 11.8 % % J * Not determined. Without preliminary treatment, 12.0 11.7 12.2 12.2 11.1 12.2 11.6 % -* __ * .-* 11.8 11.9 A. 0. A. C . r n e t h ~ d , ~ 11.8 11.7 11.9 11-8 11.1 12-1 11.5 % -* - * -* 11.6 11.7 Results of comparative “pyrethrin 11” assays on several of the same extracts are shown in Table 111.As the P.B.K. and A.O.A.C. methods give similar directions for determining “pyrethrin 11,” it is not surprising that the results obtained are also similar. This also applies when Mitchell and Tresadern’s modified methodll is included, but the results are some 6 per cent. lower because of the rejection of acids insoluble in hot water. Equally, this applies to the recommended method, except when the preliminary light petroleum treatment of the extract is omitted. In that case, the results are rather intermediate, being about 4 per cent. lower than those to which the modification has not been applied. When both the preliminary treatment and the modification to the “pyrethrin 11” method are omitted, the mean results for “pyrethrin 11” are some 5 per cent.higher than those by the P.B.K. method.July, 19631 AN INFRARED METHOD TABLE I11 COMPARATIVE RESULTS FOR “PYRETHRIN 11” IN PYRETHRUM EXTRACTS “Pyrethrin 11” found by- 541 , > Recommended Recommended P.B.K. method with method without A.O.A.C. method,s preliminary treatment preliminary treatment method ,3 r----h--7 7-p 7 7 * 7 Sample (i), (ii), ( 4 , (2) I (ii), ( 4 , (4, 5% Y O Y O % % % % A B C D E F G Mean . . 11.6 11.0 11.9 11.0 11.2 10.5 11.0 10.3 11.3 10.7 12.3 11.5 11-7 11-3 11.6 10.9 11.0 11.0 10.5 10.3 10.7 11.5 11.4 10.9 -* 10-9 -* 11.3 11.9 11.0 -* 10-7 11.8 11-0 12.8 11-5 12.4 11.6 12.2 11-1 -* -* 11.7 11.0 11-2 10.6 10.8 10.2 11.2 -* 12.2 -* 11.9 -* 11.5 10.6 (i) = Omitting Mitchell and Tresadern’s method.’l (ii) = Including Mitchell and Tresadern’s method.l1 * Not determined. From these results it is clear that the preliminary treatment has no particularly significant effect on the results for “pyrethrin I,” but its omission does give somewhat higher results for “pyrethrin 11,” even when steps have been taken to remove interfering water-insoluble acidic matter. To facilitate comparison, Table IV shows the figures for “total pyrethrins” given by the various methods, individually modified as before. On the whole, it would appear to us that inclusion of the preliminary treatment is not warranted. It occupies extra laboratory time; by involving extra manipulation it increases the risk of error; and it does not have a particularly marked effect on the results.Omission of this step would probably introduce some small positive errors; this would not seem to us to matter unduly in a method designed to control commercial transactions in pyrethrum extract. Without the preliminary treatment, but including the purification step in the “pyrethrin 11” deter- mination, the results are close to those given by the current A.O.A.C. method. It having been noted that the ratio of “pyrethrin I” to “pyrethrin 11” is close to unity, the idea (advanced some years ago by MitchelP) of standardising only in terms of “pyrethrin I” may be worth further consideration. Its adoption would certainly simplify and shorten the assay. TABLE IV COMPARATIVE RESULTS FOR TOTAL “PYRETHRINS” IN PYRETHRUM EXTRACTS Total “pyrethrins” found by- r \ Recommended Recommended P.B.K.method with method without A.O.A.C. methods preliminary treatment, preliminary treatment method3 7-7 Sample ( 2 ) . (ii), % % A 25.2 24.6 B 25.1 24.2 C 24.7 24.0 D 24.3 23.6 E 24.5 23.9 F 26.1 25.3 G 25.0 24.6 Mean . . 25.0 24.3 (ii), % 22.9 -* 22.6 -* 22.6 24- 1 22.3 -* 22.0 22.9 23.5 25.0 22.9 24.0 22.7 24.0 22.9 23.0 23.2 22.9 22.1 23.7 23.2 23.0 -* -* 23.4 22.7 23-1 22.5 22.6 22.0 22.3 -* 24.3 --* 23.4 -* 23.2 22.4 (i) = Omitting Mitchell and Tresadern’s method.ll (ii) = Including Mitchell and Tresadern’s method.‘l * Not determined. A parallel investigation of these problems has been made by Brierley and Brown.13 They also examined the possibilities of an infrared assay involving use of the absorption maximum at 9 p. Further, they examined a method involving isolation of chrysanthemic542 [Analyst, Vol.88 acid by steam-distillation along the lines of the Seil method,14 but under conditions that avoid hydration, and consequent loss, of part of the chrysanthemic a ~ i d . ~ ~ , ~ ~ y ~ ~ They also applied a silica gel column-chromatographic procedure that afforded them relatively pure chrysanthemic acid in one fraction, and in another a mixture of other acids that reduced the mercury reagent, but without the usual characteristic colour changes. Each of these last two methods gave them closely similar results on any given sample, results that were also in fairly close agreement with those given by the A.O.A.C. method, but the authors made no firm recommendation for the adoption of any of these methods.However, Brierley and Brown believe, probably correctly, that these results comprise not only the true active principles, but also any “false” pyrethrins and free chrysanthemic acid present in the original extract. To overcome this, they propose to use the preliminary chromatographic purification with alumina suggested some years ago by Brown, Phipers and Singleton.18 This further lowers the results, so that their average final results were some 10 per cent. lower than those by the A.O.A.C. method or 25 per cent. lower than those by the P.B.K. version. MITCHELL, BYRNE AND TRESADERN : PYRETHRUM ANALYSIS : TABLE V COMPARATIVE RESULTS ON A SAMPLE (B) OF PYRETHRUM EXTRACT Preliminary treatment in light petroleum (boiling range 40” to 60” C) was carried out in all tests Steam- distillation13 Infrared Chr~matography~~ A.0.A.C3 “Pyrethrin I,” yo ..11.5 11-6 ll.3t 11.7 “Pyrethrin 11,”” yo , . 11.1 11.0 11.1 11.0 Total “pyrethrins,” yo . . 22.6 22.6 22.4 22.7 * Purified by Mitchell and Tresadern’s method.ll t This represents only “fraction A” as defined in the method.l3 The acidic matter was Infrared measurement then showed 11.5 per cent., in good agreement In another similar experiment, the recovered material, recovered after titration. with the direct infrared measurement. measured by “mercury reduction,” also amounted t o 11-5 per cent. Using our Sample B, we confirmed that the steam-distillation method,13 the silica gel chromatographic method13 and the infrared method gave results in close agreement, in turn agreeing well with those obtained by the A.O.A.C. method (see Table V).We also obtained a useful cross check by measuring the material comprising “Fraction A” by the infrared and by the “mercury. reduction” methods. The results were all in good agreement (see Table V). However, we question the need for applying to extracts the initial chromatographic purification procedure (Brown, Phipers and SIingletonl8), for the reasons stated below- 1. It would extend an already lengthy analytical procedure (which we would propose to shorten by eliminating even the relatively simple preliminary treatment now applied). 2. Unless extremely carefully controlled, this chromatographic procedure can result in loss of pyrethrins, cinerins or both. This was observed during a collaborative trial of the method some years 3.It seems scarcely justified to strive after greater accuracy by this means, when one reflects that the factors used for cahlating the amounts of “pyrethrin I” and “pyrethrin 11” are derived from the molecular weights of the pure pyrethrins. In other words, no cognisance is taken of the presence of the two cinerins. Because it is not unlikely that the relative proportions of these four active principles vary in different batches of pyrethrum flowers or pyrethrum extracts, one can expect that the figures reported will not always be directly proportional to the over-all biological activity in any event. ?Ve consider that the method adopted need not strive for greater accuracy than is needed to control commercial transactions in pyrethrum.If we have to abandon the P.B.K. method -a step that would now seem to be justified-we would consider it a good opportunity to discard all variants of the methods incorporating the empirical “mercury reduction”July, 19631 AN INFRARED METHOD 543 procedure. Instead, we would like to substitute a method providing measurement of chrysan- themic acid by a more reliable procedure. We consider that the infrared method described here meets this requirement sufficiently well for the purpose. Application of the process to eliminate interfering water-insoluble acidic matter suffices, again in our opinion, to correct the “pyrethrin II” value sufficiently for practical purposes (though we have reason to believe that extraneous acids are still being measured, hence falsely raising the calculated results for “pyrethrin 11”).In an earlier publication20 one of us suggested that greater accuracy was being expected of the pyrethrum analyst than was reasonable. Because of the high value of pyrethrum, a relatively small difference in analytical results can represent a large sum of money as between buyer and seller. For example, if the buyer’s analyst returns 24 per cent. of total pyrethrins as his result on an extract offered as containing 25 per cent., the difference to the uninitiated appears considerable; it would almost certainly represent a lot of money in a commercial transaction, Yet such a difference is evidently only 4 per cent.-well within the usual limits of analytical error-particularly when a rather complex procedure is involved.It is only when inter-laboratory differences consistently occur in one direction that the position can really be regarded as unsatisfactory. Experienced pyrethrum analysts, with the methods currently in use, usually secure results agreeing within the limits of k5 per cent. In putting forward this proposed new method, we do not suggest for one moment that it is likely to lead to better inter-laboratory agreement than this. We only advance it because we think it gives results nearer the truth, and because it uses a more realistic measurement than the “mercury reduction” procedure. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. REFERENCES Wilcoxon, F., Contrib. Boyce Thompson Inst., 1936, 8, 175. Holaday, D. A., I n d .Eng. Chem., Anal. Ed., 1938, 10, 5 . hssociation of Official Agricultural Chemists, “Methods of Analysis, ” Ninth Edition, TT-ashington, “British Pharmaceutical Codex 1954,” The Pharmaceutical Press, London, 1954, p. 631. “British Veterinary Codex 1953” Supplement, 1959, The Pharmaceutical Press, London, 1959, “Determination of Pyrethrins (Extract) .” A printed leaflet published by the Pyrethrum Board of Kenya, September, 1954. “Method for the Determination of Pyrethrins (Extract) .” A printed leaflet published by the East African Extract Corporation, March, 1958. Mitchell, Wm., J . Sci. Food Agric., 1953, 4, 246. Beckley, V. A., and Hopkins, J., Soap. Sanit. Chemicals, 1954, 30 (l), 141, 143, 145, 147 and 173. Bray, G. T., Harper, S. H., Lord, K. A., Major, F., and Tresadern, F. H., J . SOC. Chem. I d . , 1947, Mitchell, Wm., and Tresadern, F. H., J . Sci. Food Agric., 1955, 6, 465. Mitchell, Wm., Ibid., 1953, 4, 278. Brierley, A., and Brown N. C., Soap. Chem. Specialties, 1962, 38 (lo), 105, 107, 109, 111 and 121. Seil, H. A., Soap, 1934, 10 (5), 89, 91 and 111. Mitchell, Wm., Tresadern, F. H., and Wood, S. A., Analyst, 1948, 73, 484. Crombie, L., Harper, S. H., and Thompson, R.A., J . Sci. Food Agric., 1951, 2, 421. Harper, S. H., and Thompson, R. A., Ibid., 1952, 3, 230. Brown, N. C., Phipers, R. F., and Singleton, K. G., Pyrethrum Post, 1954, 3 (3), 3; 1954, 3 (4), 23. Kelsey, D., J . Ass. Off. Agric. Chem., 1959, 42, 90; 1960, 43, 354. Mitchell, Wm., J . Sci. Food Agric., 1953, 4, 278; Chem. & Ind., 1960, 356. 1960, pp. 41 to 43. p. 53. 66, 275. Received November 19th, 1962

ISSN:0003-2654    

DOI:10.1039/AN9638800538

出版商:RSC    

年代:1963

数据来源: RSC