摘要:

Analyst, May, 1973, Vol. 98, pp. 347-350 347 A Critical Study of Brilliant Green as a Spectrophotometric Reagent : Extraction of Chloro-complexes of Antimony, Thallium, Gallium and Indium, and of Tetrabromoindate(III), and Improved Procedures for the Determination of Antimony and Thallium BY A. G. FOGG, C. BURGESS* AND D. THORBURN BURNS (Department of Chenzzstry, University of Technology, Loughborough, Leicestershzre, LE11 3T U ) Use of cerium(IV), instead of nitrite, to oxidise antimony(II1) improves the procedure for the determination of antimony with Brilliant green. Antimony(II1) is oxidised in 6 M hydrochloric acid solution, the excess of oxidant being reduced with hydroxylammonium chloride. Brilliant green hexachloroantimonate (V) is extracted from 2 M hydrochloric acid solution with two 10-ml portions of toluene, the Brilliant green reagent solution being added to the aqueous solution immediately before each extraction. When a pure sample of dye is used, and the extraction is made rapidly after the addition of Brilliant green, complete recovery of antimony is effected. The recommended procedure is precise and reliable.A similar procedure based on the extraction of tetrachlorothallate(II1) ions gives complete recoveries of thallium and is an improvement on existing procedures. Tetrachlorogallate(II1) and tetrachloroindate(II1) ions are not completely extracted by toluene, and procedures based on such extractions cannot be recommended. IN all systems so far studied in the present series, namely the extraction of antimony,l gold,2 perrhenate2 and per~hlorate,~ the molar absorptivity of the R+ form of Brilliant green has been found to be about lo5 1 mol-l cm-l at the wavelength of maximum absorbance, irre- spective of the anion associated with it and the solvent in which it is dissolved.Procedures for which apparent molar absorptivities significantly below this value are evident clearly give low recoveries of the anion to be determined, and are liable to be imprecise and unreliable from day to day. Further studies in the development of precise analytical procedures that involve the use of Brilliant green are now described. An improved procedure for determining antimony was developed earlier in this laboratory1 but several disadvantages remained. The procedure has now been improved further by using cerium(1V) to oxidise antimony(II1).Thallium, but not gallium or indium, can be determined precisely by an analogous procedure. DETERMINATION OF ANTIMONY- In the previous procedure1 for the determination of antimony with Brilliant green, nitrite was used to oxidise antimony(II1). Time periods for oxidation and removal of excess of oxidant had to be strictly observed, and all operations were carried out on solutions at temperatures of about 0 "C. Brilliant green was added after the extracting solvent, and the extraction was made rapidly as the R+ form of Brilliant green is protonated in 2 M hydro- chloric acid solution. The previous procedure gave a higher sensitivity and precision than procedures which then existed, but despite all the precautions taken reproducibility from day to day was often unsatisfactory.From this study it was not clear whether the rapid protonation of the R+ form of Brilliant green was the only cause of the unreliability of the procedure; the purer dye samples gave a higher apparent molar absorptivity (1.03 x lo5 1 mol-l cm-l) than less pure samples (0.92 x lo5 1 mol-l cm-l). Consequently, the extraction of perrhenate and tetrachloroaurate(II1) ions was studied. The former ions can be extracted from neutral solution and no oxidation step is required. On the other hand, the tetrachloroaurate(II1) * Present address : Chemistry Department, University of Southampton. @ SAC and the authors. EXPERIMENTAL348 FOGG et al. : BRILLIANT GREEN AS A SPECTROPHOTOMETRIC [Analyst, Vol. 98 ions are extracted from 0.5 M hydrochloric acid solution, but again, no oxidation step is required.Complete recovery of perrhenate was achieved even with impure samples of Brilliant green. Complete recovery of gold from acidic solution, however, could only be effected with pure samples of Brilliant green. A further factor, which is relevant to the present study, is that the recovery of gold was increased by making a second addition of Brilliant green after the first extraction step, and extracting again with toluene. This effect was particularly significant with the impure samples of Brilliant green. With a pure dye sample, 97 per cent. recoveries were obtained with a single addition of Brilliant green. By analogy with the procedure for the determination of gold, there would appear to be no reason why procedures based on the extraction of Brilliant green hexachloroanti- monate(V) should not be precise and reliable, provided that a pure sample of dye is used and the extraction is made rapidly.As our previous procedure was not reliable even when these precautions were taken, it follows that the oxidation step with nitrite is probably the main cause of its unreliability. Sandel14 has indicated that cerium(1V) is one of the few suitable reagents for the oxidation of antimony(II1). The following procedure involving the use of cerium(1V) was developed and found to be entirely satisfactory. REAGENTS- Hydrochloric acid, concentrated, sp. gr. 1.18-Analytical-reagent grade. Cerium(1V) sulphate solution, 0.1 M in 1 M sulphuric acid.Hydroxylammonium chloride solution, 1 per cent. mlV. Brilliant green solution, 0-5 per cent. m/V in ethanol-Use a pure sample of Brilliant green, Toluene-Analytical-reagent grade. Concentrated standard antimony solution, 200 pg ml-l of Sb-Dissolve 0.2743 g of potas- sium antimony tartrate in water and dilute the resulting solution to 500ml with water in a calibrated flask. Dilute standard antimony solution, 4 pg ml-l of Sb-Dilute 10.00 ml of concentrated stan- dard antimony solution to 500ml with water in a calibrated flask. This solution has been found to be stable for at least 1 month, but is probably best prepared freshly as required. PROCEDURE- Place an aliquot of the dilute standard antimony solution containing less than 20pg of antimony into a dry 100-ml separating funnel.Add an equal volume of concentrated hydrochloric acid and dilute the solution to 10 ml with 1 + 1 V/V hydrochloric acid solution. Add ten drops of cerium(1V) sulphate solution and mix the solution for 1 minute. Reduce the excess of cerium( IV) by dropwise addition of hydroxylammonium chloride solution until the yellow colour disappears. Without delay, dilute the mixture with 25ml of water, add 10ml of toluene and 0.5 ml of Brilliant green solution. Shake the mixture for 1 minute and then filter the toluene layer through a No. 31 Whatman filter-paper into a 25-ml calibrated flask. Add a further 10 ml of toluene and 0.5 ml of Brilliant green solution to the aqueous layer. Shake the mixture for 1 minute and filter the toluene layer through the filter-paper into the 25-ml calibrated flask.Wash the filter-paper with a small volume of toluene, adding the washings to the flask. Dilute to 25ml with toluene, and read the absorbance of the solution at 640 nm in a 1-cm glass cell against toluene, The mean of twenty determinations at the 20 pg of antimony level gave an apparent molar absorptivity of 0.98 x lo5 1 mol-l cm-1 with a coefficient of variation of less than 1 per cent. DETERMINATION OF THALLIUM- Ariel and Bach5 determined thallium by extracting the TlBr4- complex into toluene with Brilliant green. Their results indicate an apparent molar absorptivity of 7.2 x lo4 1 mol-1 cm-l for the ion-association complex, which in turn indicates that only about a 72 per cent. recovery of thallium was obtained. In the present work it was found that an almost complete recovery of thallium could be achieved by extracting the TlC14- complex into toluene with Brilliant green by a procedure analogous to that described above for antimony. REAGENTS- e.g., British Pharmacopoeia grade.Hydrochloric acid, concentrated, sp. gr. 1-18-Analytical-reagent grade. Hydrochloric acid, dilute (1 + 2 V/V)-To 66 ml of water, add 33 ml of concentrated hydrochloric acid.May, 19731 REAGENT DETERMINATION OF ANTIMONY AND THALLIUM 349 Cerium(IV) sulphate solution, 0.1 M in 1 M sulphuric acid. Hydroxylammonium chloride solution, 1 per cent. m/V. Brilliant p e e n solution, 0.5 per cent. m/V in ethanol-Use a pure sample of Brilliant green, Toluene-Analytical-reagent grade. Concentrated standard thallium(I) sulphate solution, 1000 pg ml-I of TI-Dissolve 0.124 g of Tl,SO, (analytical-reagent grade) in water containing 2 ml of concentrated sulphuric acid and dilute to 100 ml in a calibrated flask with water.Dilute standard thallium(I) sulphate solution, 10 p,g ml-I of Tl-Dilute 5.00 ml of concen- trated standard thallium(1) sulphate solution to 600 ml in a calibrated flask with 1 + 2 V/V hydrochloric acid solution. PROCEDURE- Place an aliquot of the dilute standard thallium solution containing less than 30 pg of thallium into a dry 100-ml separating funnel. Dilute the solution to 10 ml with 1 + 2 V/V hydrochloric acid solution. Add 10 drops of cerium(1V) sulphate solution and mix the solution for 1 minute. Reduce the excess of cerium(1V) by dropwise addition of hydroxyl- ammonium chloride solution until the yellow colour disappears. Without delay, add 10 ml of toluene and 1 ml of Brilliant green solution.Shake the mixture for 1 minute and then filter the toluene layer through a No. 31 Whatman filter-paper into a 25-ml calibrated flask. Add a further 10 ml of toluene and 1 ml of Brilliant green solution to the aqueous layer. Shake the mixture for 30 s and filter the toluene layer through the filter-paper into the 25-ml calibrated flask. Wash the filter-paper with a small volume of toluene, adding the washings to the flask. Dilute to 25 ml with toluene and read the absorbance of the solution at 640 nm in a l-cm glass cell against toluene. The mean of twenty determinations at the 20pg of thallium level gave an apparent molar absorptivity of 1.03 x lo5 1 mol-l cm-l with a coefficient of variation of less than 1 per cent.EXTRACTION OF TETRACHLOROGALLATE(III)- Armeanu and Costinescu6 have studied the extraction, with a single portion of solvent, of tetrachlorogallate( 111) ion with several basic dyes. With Brilliant green apparent molar absorptivities of 0.30 x lo5 and 0.19 x lo5 1 mol-l cm-l were obtained when extraction was made with benzene and toluene, respectively. Higher apparent molar absorptivities were obtained with several halogenated hydrocarbons. In the present work, a standard gallium solution (10 pg ml-l of gallium) in 1 + 1 V/V hydrochloric acid was prepared from the pure metal. Extractions were made by shaking together 2-00 ml of standard solution, 8 ml of 1 + 1 V/V hydrochloric acid, 1 ml of 0.5 per cent, Brilliant green solution in ethanol and 10 ml of solvent.The solvent layer was filtered and a second extraction of the aqueous phase was made with 10 ml of solvent plus a further 1 ml of Brilliant green solution in ethanol. The combined extracts were diluted to 25 ml in a calibrated flask and absorbance measurements were made at A,,,. (630 to 640 nm). The apparent molar absorptivities together with blank absorbances obtained with several solvents are shown in Table I. It is apparent that, in order to obtain almost complete recoveries of gallium, very high blanks have to be tolerated. This situation is unacceptable in a precise procedure. e.g., British Pharmacopoeia grade. Prepare when required. TABLE I EXTRACTION OF TETRACHLOROGALLATE(II1) WITH BRILLIANT GREEN Extracting solvent Apparent molar absorptivity* Blank (1 mol-l cm-l/106) absorbance Benzene .- . . .. .. .. . . .. .. .. .. 0.08 0-002 Chlorobenzene + carbon tetrachloride (4 + 1 V / V ) . . . . . . . . 0.70 0.065 1,2-Dichlorobenzene + carbon tetrachloride ( 2 + 1 V / V ) . . .. .. 0.92 0.352 Chlorobenzene + carbon tetrachloride + 1,2-dichlorobenzene (4 + 1 + 1 V / V / V ) 0.127 0.1 15 l,l,l-Trichloroethane . . . . .. .. .. .. .. .. 0.92 0.145 Chlorobenzene + carbon tetrachloride (1 + 1 V / V ) . . . . .. . . 0.15 0.002 0.90 0.93 Chlorobenzene + carbon tetrachloride + dichloromethane (4 + 1 + 1 V / V / V ) * After deduction of blank absorbance.350 FOGG, BURGESS AND THORBURN BURNS EXTRACTION OF TETRACHLORO- AND TETRABROMOINDATE(III)- A standard indium solution was prepared from indium( 111) sulphate.Although Schulfe’ has shown that the InC14- complex is formed in hydrochloric acid solution of concentration higher than 8 M, no extraction of this anion was observed in the present work with the solvents used above for extracting tetrachlorogallate(II1). Indium has been determined by Liteanu and Cord09 by extracting Brilliant green tetrabromoindate( 111). A double-extraction procedure similar to that described above for gallium was adopted in the present study. Extraction was made from an aqueous solution that had the following composition: 2.00 ml of standard indium solution (10 pg ml-1 of In), 5 ml of 25 per cent. V/V sulphuric acid solution and 3 ml of 4 M potassium bromide solution.The apparent molar absorptivities obtained with various solvents are given in Table 11. As with extractions of tetrachlorogallate(III), high blanks have to be tolerated in order to achieve almost complete extraction. TABLE I1 EXTRACTION OF TETRABROMOINDATE(III) WITH BRILLIANT GREEN Extracting solvent Apparent molar absorptivity* (1 mol-1 cir1/106) Benzene . . . . .. . . . . . . 0.74 Toluene . . . . . . . . . . . . 0-42 Benzene + chlorobenzene (4 + 1 V / V ) . . 0.93 Benzene + chlorobenzene (4 + 1 V / V ) t . . 0.75 Benzene + chlorobenzene (2 + 1 V / V ) t . . 0.83 , . Chlorobenzenet . . .. . . . . . . 0.82 Blank absorbance 0.135 0.166 0.132 0.035 0.073 0.083 * After deduction of blank absorbance. t By using 0.5 per cent. aqueous solution instead of ethanolic solution of Brilliant green. RESULTS AND DISCUSSION The recommended procedure described above for the determination of antimony differs from that previously recommended1 in that cerium(1V) is used to oxidise antimony(II1) and extraction is made with two portions of toluene (Brilliant green being added before each extraction with toluene). This modified procedure has been found to be more convenient and more reliable than the previous procedure.Certain precautions discussed in the previous paper, however, must still be taken. The sample of Brilliant green used must be pure; there must be no long delay after oxidising antimony(II1) as antimony(V) is rapidly hydrolysed in hydrochloric acid solution; and Brilliant green must be added after the solvent and rapid extraction made before the Brilliant green is completely protonated.In the modified procedure it was not found necessary to cool the solutions in an ice-bath, but previous workers9 have warned that low results are obtained when working at room temperatures higher than 25 “C. The recommended procedure given above for the determination of thallium is completely analogous to that for antimony, and similar precautions should be taken. It is considered that Brilliant green is not a suitable reagent for the determination of gallium or indium by extraction of tetrachlorogallate(III), tetrachloroindate(II1) or tetra- bromoindate( 111). With the solvent system used here complete extraction of the Brilliant green salt formed with these anions is possible only at the expense of a high blank absorbance, i.e., with considerable extraction of the reagent itself. 1. 2. 3. 4. 5. 6. 7. 8. 9. REFERENCES Fogg, A. G., Jillings, J., Marriott, D. R., and Burns, D. T., Analyst, 1969, 94, 768. Fogg, A. G., Burgess, C., and Burns, D. T., Ibid., 1970, 95, 1012. Sandell, E. B., “Colorimetric Metal Analysis,” Third Edition, Interscience Publishers Inc., New Ariel, M., and Bach, D., Analyst, 1963, 88, 30. Armeanu, V., and Costinescu, P., Talanta, 1967, 14, 699. Schulfe, E., J. Amer. Chem. SOG., 1951, 73, 1013. Liteanu, C., and Cordos, E., Bull. Inst. Politeh. IaSi, 1961, 7, 127; Awalyt. Abstr., 1963, 10, 567. Galliford, D. J. B., and Yardley, J. T., Analyst, 1963, 88, 653. Received November 21st, 1972 Accepted January 3rd, 1973 , , 1971, 96, 854. --- York, 1959.

ISSN:0003-2654    

DOI:10.1039/AN9739800347

出版商:RSC    

年代:1973

数据来源: RSC

摘要:

Analyst, May, 1973, Vol. 98, pp. 351-357 351. A Chemical Concentration Method for the Determination of Niobium, Zirconium and Tantalum in Carbon Steel by X-ray Fluorescence Spectrometry BY 2. KLIMA (Institute of Chemistry, Silesian Ufiiversity, Katowice, Poland) AND P. H. SCHOLES (BISRA-The Corporate Laboratories of the British Steel Corfioration, Hoyle Street, Shefleld, S3 7E Y ) A chemical concentration technique is described for the simultaneous determination of small added amounts of zirconium and niobium (less than 0.05 per cent.) and residual amounts of tantalum (less than 0.01 per cent.) in carbon steel by X-ray fluorescence spectrometry. Separation from iron and other constituents is effected by precipitation with phenylarsonic acid ; the precipitate is then collected on an organic membrane and X-ray measure- ment made either directly on the precipitate or after pelleting with poly(viny1 alcohol).The results obtained for niobium and Zirconium in the test samples compare favourably with published values by other techniques. THE literature of analytical chemistry continues to show evidence of a lively interest in the determination of small amounts of zirconium, niobium and tantalum in carbon and alloy steels by various techniques. Separation from iron and other elements is usually necessary and can be achieved by solvent extraction, ion exchange, precipitation or reversed-phase partition chromatography. For the determination of niobium, various modifications of the thiocyanate reaction in the presence of tin(I1) chloride have been used.This reaction forms the basis of the present British Standard method for the determination of niobium in carbon steel and carbon - manganese stee1.l Several other elements are known to form thiocyanates and may cause serious interference. Because of this disadvantage other spectrophotometric reagents have been recommended, for example, nitrosulphophenol S, bromopyrogallol and 4- (2-pyridy1azo)resorcinol. For the determination of zirconium, gravimetric procedures based upon precipitation with mandelic acid are commonly used.l Spectrophotometric procedures based upon reactions with chlorosulphophenol S, catechol, catechol violet, xylenol orange and arsenazo I11 have also been recommended. With tantalum, polyhydric phenols (e.g., pyrogallol, hydroquinone or pyrocatechol) are used for spectrophotometric measurement following complex separation procedures.The pyrogallol reaction forms the basis of the British Standard method1 for the determination of tantalum in a wide range of steels. Most of the spectrophotometric procedures have limitations when applied to the deter- mination of small amounts of niobium, zirconium and tantalum as multiple separations are necessary in order to remove interfering elements and the operational procedures are often time consuming. On the other hand, direct X-ray fluorescence and emission-spectrometric techniques are more rapid for the determination of niobium and zirconium in solid samples, but success with these methods is very dependent upon the availability of standard samples with accurately known contents.The amount of tantalum present in steel is generally too low to be determined by physical methods. Several reagents have been recommended for the gravimetric separation of niobium and zirconium, the best known of which being cupferron and mandelic and phenylarsonic acids and their derivatives. Both bromomandelic acid2 and phenylarsonic acid394 are more selective than cupferron, but the bromomandelic precipitate must be aged for 24 hours before filtration, especially when the zirconium content of the steel is low. Phenylarsonic acid was selected as being the most suitable precipitant for the separation of zirconium and niobium from iron. The concentration technique proposed in this work is a precipitation process. @ SAC and the authors.352 KLIMA AND SCHOLES: DETERMINATION OF NIOBIUM, ZIRCONIUM AND [Analyst, Vol.98 Two procedures are proposed: procedure A, for the determination of niobium and zirconium by using a 1-g sample with direct measurement of the precipitate; and procedure B, which involves the use of a 10-g sample to permit the determination of very small amounts of tantalum (less than 0.01 per cent.) together with niobium and zirconium by, in this instance, pelleting the precipitate with poly(viny1 alcohol) before measurement. EXPERIMENTAL SPECTROMETER- For niobium and zirconium, X-ray fluorescence measurements were made by using a Rank Precision Instruments Flurovac spectrometer, equipped with a tungsten-target X-ray tube operated at 40 kV and 20 mA. For tantalum, measurements were made by using a Philips 1220 spectrometer, equipped with a gold-target X-ray tube operated at 60 kV and 20 mA.Both spectrometers were equipped with a lithium fluoride crystal and a scintillation detector. The niobium Ka, zirconium Ka and tantalum Lp analysis peaks were at 21.32" 28, 22.51" 28 and 38.49'20, respectively, and their counting times were 20, 20 and 40 s, respectively. MEMBRANE AND HOLDER- The membrane was a Millipore Solvinert membrane, of 1.5 pm pore size. A holder for the membrane was made from a disc of nylon to fit the sample chamber of the Flurovac spectrometer. The disc was 5 cm in diameter and 17 mm thick, with a raised platform 3 mm high and 27 mm in diameter on one side. For fluorescence measurements, the membrane was placed on the platform and covered with a sheet of Mylar film held in place on the rim of the platform by a closely fitting split-ring made of copper.PELLETING MEDIUM- The pelleting medium was poly(viny1 alcohol), 100-mesh powder. SOLUTIONS REQUIRED- Tartaric acid solution, 15 per cent. m/V-Dissolve 150 g of tartaric acid in water and dilute to 1 litre. Ammonium oxalate solution, 4 per cent. m/V-Dissolve 40 g of ammonium oxalate in water and dilute to 1 litre. Phenylarsonic acid wash solution, 0.5 per cent. m/V-Dissolve 2.5 g of phenylarsonic acid in water and dilute to 500ml. Standard niobium solution, 100 pg ml-l-Fuse 50 mg of high-purity niobium metal powder with a minimum of potassium hydrogen sulphate in a platinum crucible. Extract the residue with 25 ml of tartaric acid solution, dilute the extract to 500 ml in a calibrated flask with tartaric acid solution and mix thoroughly.Standard niobium solation, 1 mg ml-f-Fuse 200 mg of high-purity niobium metal powder as described above and dilute the extract to 200ml in a calibrated flask. Standard zirconium solutiout, 100 pg ml-l-Dissolve 67.5 mg of high-purity zirconium oxide in 5 ml of 40 per cent. V/V hydrofluoric acid in a covered platinum crucible. Cautiously add 2ml of sulphuric acid, sp. gr. 1.84, and evaporate the solution to fumes to expel all of the hydrofluoric acid. Cool, transfer with water to a 500-ml calibrated flask, dilute to volume and mix thoroughly. Standard zirconium solution, 1 mg ml-l-Dissolve 270 mg of high-purity zirconium (IV) oxide as described above and dilute the solution to 200ml in a calibrated flask. Standard tantalum solution, 100 pg ml-l-Fuse 61 mg of high-purity tantalum(V) oxide with 5 g of potassium hydrogen sulphate in a platinum crucible.Extract the residue with ammonium oxalate solution, dilute it to 500 ml in a calibrated flask with ammonium oxalate solution and mix thoroughly. PROCEDURE A FOR THE DETERMINATION OF NIOBIUM AND ZIRCONIUM- Transfer 1 g of sample to a 125-ml beaker and dissolve it in 15 ml of hydrochloric acid, sp. gr. 1.18. Oxidise the sample with 10 ml of 100-volume hydrogen peroxide, added dropwise. Heat the solution to boiling and add 1 g of phenylarsonic acid dissolved in about 20 ml ofMay, 19731 TANTALUM I N CARBON STEEL BY X-RAY FLUORESCENCE SPECTROMETRY 353 hot water.Boil the mixture gently for 5 minutes and then maintain the temperature of the solution at about 90 "C for 1 hour. Cool the solution to room temperature and filter it by means of suction through a Solvinert membrane supported in a Millipore holder. Wash the beaker and precipitate with phenylarsonic acid wash solution, remove the Solvinert membrane from the holder, dry it in air and spray it with a poly(viny1 chloride) aerosol. Place the membrane on the platform of a plastics holder, cover it with a sheet of Mylar film and fix it tightly in position by using the split-ring as a clamp. Measure the X-ray fluorescence intensities of niobium and zirconium four times each, rotating the plastics holder through 90" after each measurement. Measure the background counts at the appropriate analysis peaks in a similar manner by using a sample of high-purity iron to which has been added 8mg of high-purity, finely powdered silica and which has been taken through the procedure.Take the average of the values, subtract the background counts and convert net counts into percentage content by reference to the appropriate calibration graph. CALIBRATION OF THE SPECTROMETER FOR PROCEDURE A- Transfer l-g portions of high-purity iron to 125-ml beakers and dissolve them in 15 ml of hydrochloric acid, sp. gr. 1.18. Add 0 to 5 ml each of niobium and zirconium solutions, concentration 100 pgml-1, to cover the range from 0 to 0.05 per cent. To each solution add approximately 8 mg of high-purity, finely powdered silica, oxidise and continue as described under procedure A.Prepare calibration graphs that relate net counts to percentage content. PROCEDURE B FOR THE DETERMINATION OF NIOBIUM, ZIRCONIUM AND TANTALUM- Transfer l o g of sample to a 250-ml beaker and dissolve it in 50 ml of hydrochloric acid, sp. gr. 1.18. Oxidise the sample with 25 ml of 100-volume hydrogen peroxide, added dropwise. Heat the solution to boiling, add 1 g of phenylarsonic acid dissolved in about 20 ml of hot water and continue as described under procedure A. Remove the membrane from the Millipore holder, dry it in air and carefully transfer the precipitate to an agate mortar by means of a spatula. Add 0.5 g of poly(viny1 alcohol) to the mortar and mix thoroughly by grinding with a pestle. Transfer the mixture to a steel die of 3 cm internal diameter and pelletise it by compression at a total loading of about 20 tons. Measure the X-ray fluorescence intensities of niobium, zirconium and tantalum. Subtract the background counts obtained by making measurements at the appropriate analysis peaks on a sample of high-purity iron treated as described above. Convert net counts into percentage content by reference to the appropriate calibration graph.CALIBRATION OF THE SPECTROMETER FOR PROCEDURE B- Transfer 10-g portions of high-purity iron to 250-ml beakers and dissolve them in 50 ml of hydrochloric acid, sp. gr. 1.18. Add 0 to 5 ml each of niobium and zirconium solutions, concentration 1 mg ml-l, and 0 to 10 ml of tantalum solution, concentration 100 pg ml-1, to cover ranges from 0 to 0.05 per cent. of niobium and zirconium, and from 0 to 0.1 per cent.of tantalum. Prepare calibration graphs that relate net counts to percentage content. Oxidise and continue as described under procedure B. TABLE I NIOBIUM CONTENT OF TEST SAMPLES-BUREAU OF ANALYSED SAMPLES AND PUBLISHED VALUES Values by chemical methods, per cent. \ Mean of results cited - B.C.S. White and by White and number B.A.S. Scholes4 Scholes4 Jenkins5 272 0.0055 0.006 0.006 0.008 274 0.050 0.0495 0.051 0.058 275 0.035 0.031 0.033 0.035 276 0-055 0.0495 0-051 0.053 277 0.02 1 0.019 0.016 0.016 Values by emission- spectrometric methods, R.A.S., per cent. 0.005 0.055 0-030 0.050 0.020354 KLIMA AND SCHOLES: DETERMINATION O F NIOBIUM, ZIRCONIUM AND [AndySt, VOl. 98 RESULTS Both procedures were tested by examining a series of mild steel standards, B.C.S.271 to 277. These standards do not have certificated results for niobium, zirconium and tantalum; the results (see Tables I and 11) given by the Bureau of Analysed Samples Ltd. are only tentative, and in the instance of tantalum are based on values from one laboratory only. Other published values5~6 for niobium and zirconium are also included in Tables I and 11. TABLE I1 ZIRCONIUM AND TANTALUM CONTENTS OF TEST SAMPLES-BUREAU OF ANALYSED SAMPLES AND PUBLISHED VALUES Zirconium, per cent. B.C.S. number 27 1 272 273 274 275 276 277 A I Values by chemical methods 1 6 A . S . Keller and Hennesen'j 0.045 0.044 0.030 0.031 0.005 0,012 0.015 0.02 1 0.005 0.008 0.050 0-051 - - 1 Values by Tantalum-Values by emission-spectrographic a chemical method, methods, B.A. S. B.A.S., per cent. 0.045 0-008 0.030 - - 0.02 0.010 - 0.020 0.003 0.005 0.0065 0.050 - The results obtained by use of procedure A for determining niobium and zirconium with the Flurovac spectrometer equipped with a tungsten-target X-ray tube are presented in Tables I11 and IV. Within the range 0.005 to 0.05 per cent. the standard deviations vary from 0.0007 to 0.0021 for both elements, with a mean of 0.0013 per cent. By comparison, the within-laboratory standard deviations for the determination of niobium and zirconium by British Standard methods1 at these levels are 0.001 and about 0.003 per cent., respectively. TABLE I11 RESULTS OBTAINED FOR NIOBIUM BY USING A TUNGSTEN-TARGET X-RAY TUBE Procedure A Procedure B B.C .S.------A-.,--- 7 r---&----- 7 sample Number Niobium, Standard Number Niobium, Standard number of tests per cent. deviation of tests per cent. deviation 272 8 0.005 0~0010 - - - 274 3 0.049 0.0019 - - - 275 5 0.033 0.0019 3 0.035 -0*0020 276 4 0.049 0*0006 3 0.044 -0.00 10 277 8 0.020 0.001 1 - - I The results obtained by using procedure B for determining niobium and zirconium with the Flurovac spectrometer equipped with a tungsten-target X-ray tube are given in Tables I11 and IV, and for tantalum with the Philips spectrometer with a gold-target X-ray tube, in Table V. Only a limited number of tests were carried out by use of procedure B, because of a shortage of standard samples, but based on these tests the standard deviations of results TABLE IV RESULTS OBTAINED FOR ZIRCONIUM BY USING A TUNGSTEN-TARGET X-RAY TUBE B.C.S.c sample Number number of tests 271 4 2 72 6 274 5 275 3 276 5 277 6 Procedure A Zirconium, per cent. 0.043 0.029 0.010 0.019 0.006 0.050 7 7 Standard Number deviation of tests 0.0021 5 0.00 11 - 0.001 1 - 0.0007 3 0.0007 3 0~0020 - Procedure B I Zirconium, Standard per cent. deviation 0.045 0*0020 - - - - 0.019 -0.0003 0.006 -0*0004 - -&fay, 19731 TANTALUM I N CARBON STEEL BY X-RAY FLUORESCENCE SPECTROMETRY 355 obtained for niobium and zirconium are similar to those obtained by using procedure A. The standard deviation of results for tantalum is about 0.0003 at the 0.002 per cent. level. By comparison, the within-laboratory standard deviation of the British Standard method1 for tantalum at this level is 0.000 13 per cent.RESULTS OBTAINED B.C.S. sample number 271 273 275 276 TABLE V FOR TANTALUM BY USING A GOLD-TARGET X-RAY TUBE Procedure B Tantalum, f A 1 Number of tests per cent. Standard deviation 5 0.0026 0.0004 3 0.0065 -0~0010 3 0.0022 -0.0002 3 0.0013 -0.0003 The sensitivity of an X-ray fluorescence method is normally expressed as counts per second for 1 per cent. of the element. In this instance, it can be expressed more appropriately as counts per second for 100 pg of the element, which, for a l-g amount of sample, is equivalent to 0.01 per cent. The sensitivities, background count-rates and theoretical limits of detection7 are given in Table VI. TABLE VI SENSITIVITIES AND DETECTION LIMITS Sensitivity, Theoretical limit of detection counts per I A \ second for 1OOpg Background, Percentage Percentage Element of element counts s-l pg on 1 g of sample on 10 g of sample By procedure A with tungsten target- Niobium .. . . I25 Zirconium . . 220 Niobium . . . . 340 Zirconium . . 600 Tantalum . . .. 190 By procedure B with gold target- Tantalum . . . . 360 By procedure B with tungsten target- 125 7 0.0007 122 4 0.0004 270 3 350 3 360 7 - 182 2 0.0003 0*0003 0.0007 0*0002 DISCUSSION White and co-workers3s4 found that it was necessary to add zirconium as a carrier in order to ensure complete precipitation of niobium with phenylarsonic acid. We were unable , however, to detect any significant change in the amount of niobium present in the precipitate when zirconium was added as a carrier. After separation, the precipitate on the membrane is dried in air and (in procedure A) is sprayed with poly(viny1 chloride) to prevent cracking.As a further precaution against powdering during measurement, the membrane is covered with Mylar film. Any variation in the fluorescence intensity due to surf ace roughness and heterogeneity of the precipitate can be minimised by taking repeat measurements with the membrane rotated through 90" each time. By comparison, pellets prepared (as in procedure B) by compressing the precipitate admixed with poly(viny1 alcohol) powder are more robust and can be stored without difficulty. Correction for background due to scattering of radiation by the membrane or poly(viny1 alcohol) pellets is not essential, but it, can nevertheless be readily made, if desired, by using the recommended procedure.The sensitivities and detection limits given in Table VI demonstrate the advantage of substituting a gold target for the tungsten target in the X-ray tube used in the determination of tantalum when the detection limit is improved by a factor of 34. The use of the more powerful gold-target tube should also lead to improved sensitivities and detection limits for niobium and zirconium, although this contention was not tested in the present work. The356 KLIMA AND SCHOLES: DETERMINATION O F NIOBIUM, ZIRCONIUM AND [AndySt, VOl. 98 mean count-rate for the determination of each of these two elements is also improved by the use of procedure B, but the high background count-rates prevent any significant gain in detection limits.The precipitate may be contaminated with iron, silicon, titanium and vanadium either by co-precipitation or by the presence of the element in the precipitate as a hydrolysis product or as undissolved nitride or carbide. The possible effects of each of these elements in pro- cedure A were considered. Tests showed that recoveries of zirconium and niobium were not influenced by the presence of a large amount of iron during precipitation. Moderate amounts of silicic acid are collected on the membrane, which causes minor discrepancies owing to the scattering of secondary X-ray emission. For this reason, it is advisable to add silica to the solutions used for calibration, and an addition of 8 mg, which is equivalent to an average silicon content in carbon steel of about 0.4 per cent., is recom- mended.From a consideration of mass absorption effects silicon might be expected to inter- fere in the measurement of emissions from niobium and zirconium, but tests (see Table VII) suggest that there is no effect at concentrations up to 1 per cent. TABLE VII INTERFERENCE OF SILICON, VANADIUM AND TITANIUM IN THE DETERMINATION 0.0250 per cent. of zirconium added to each test sample OF ZIRCONIUM BY PROCEDURE A Composition of test sample r- 3 Additive A r 7 Equivalent, Base material Element Mass/g per cent. Zirconium found, per cent. 1 g of iron 1 g of iron 1 g of iron 1 g of V - Fe alloy (0.89 per cent. V) 0.9 g of iron 0.8 g of iron 0-75 g of iron 0.5 g of iron None Silicon Silicon Vanadium Titanium Titanium Titanium Titanium - I 0.0245, 0.0260, 0.0260, 0.0260, 0.0240 10.5 mg of SiO, 0.5 0.0235, 0.0230, 0.0260 21 mg of SiO, 1.0 0.0245, 0-0265, 0-0240 0.89 0.0245, 0.0250, 0.0240, 0.0245 0.07 0.0245, 0.0250, 0.0240, 0.0240 0.15 0.0245, 0.0220, 0.0230 0-18 0-0215 0.37 0.0165 The effects of vanadium and titanium were tested by adding zirconium to iron alloys containing nitride.The results of one test on an alloy containing 0-89 per cent. of vanadium and 0.02 per cent. of nitrogen are given in Table VII and show the absence of any inter- ference effect. Titanium interference was examined by using an alloy containing 0.73 per cent. of titanium and 0.02 per cent. of nitrogen, diluted with high-purity iron to give test samples with various titanium contents. The results given in Table VII show that low values are obtained when the titanium content is 0.15 per cent.or greater. Similar tests, in which niobium was substituted for zirconium, showed that vanadium had no effect but that titanium gave rise to a negative interference at the 0.15 per cent. or greater level. The possibility of mutual interference between niobium and zirconium was also considered and found to be not significant under the conditions adopted. No attempts were made to measure possible interference effects in procedure B. In this instance, and with the type of steels examined, the use of poly(viny1 alcohol) as diluent would result in a considerable reduction in the effect observed in procedure A. CONCLUSIONS Precipitation with phenylarsonic acid forms the basis of a satisfactory X-ray fluorescence method for the simultaneous determination of niobium and zirconium present in carbon steels at levels below about 0.05 per cent.This principle can also be extended to the determination of tantalum provided that a large amount of sample is used and X-ray measurements are made under the most favourable conditions. The results obtained for niobium and zirconium agree well with other techniques and there is a considerable saving in time compared withMay, 19731 TANTALUM IN CARBON STEEL BY X-RAY FLUORESCENCE SPECTROMETRY 357 conventional chemical methods, particularly for tantalum. By use of the most modern spectrometers with close-coupled geometry, high-intensity X-ray tubes and curved analysing crystals, significant increases in sensitivity and detection limits can be anticipated. The authors thank their colleagues for helpful comments and, in particular, Mr. M. Baxter and Mr. W. Dalby for specialist assistance, Mr. G. M. Holmes of the London Scandinavian Metallurgical Co. for analysis facilities, including a Philips spectrometer, and the British Council for a grant towards this work. REFERENCES 1 . 2. 3. 4. 5. 6. 7. “Methods for the Sampling and Analysis of Iron, Steel and Other Ferrous Metals,” British Stan- dards Handbook No. 19, British Standards Institution, London, 1970. Kriege, 0. H., and Rudolph, J . S.. Talanta, 1963, 10, 215. Kidman, L., Danvent, G. L., and White G., Metallurgia, 1960, 62, 125. White, G., and Scholes, P. H., Ibid., 1964, 70, 197. Jenkins, N., Ibid., 1964, 70, 95. Keller, H., and Hennesen, K., Arch. EisenhiittWes., 1968, 39, 921. Jenkins, R., and de Vries, J, L., “Worked Examples in X-ray Analysis,” Philips Technical Library, Macmillan, London, 1971. Received October 17th, 1972 Accepted December 7th, 1972

ISSN:0003-2654    

DOI:10.1039/AN9739800351

出版商:RSC    

年代:1973

数据来源: RSC

摘要:

358 Analyst, May, 1973, Vol. 98, pj5. 358-363 The Determination of Plutonium-241 in Effluents* BY K. G. DARRALL, G. C . M. HAMMOND AND J. F. C. TYLER (Department of Trade and Industry, Laboratory of the Government Chemist, Cornwall House, Stamford Street, London, SEl 9NQ) A method is described for the simultaneous determination of pluton- iuxn-241 and plutonium alpha-activity. The plutonium is isolated by co- precipitation on barium sulphate followed by extraction into di(2-ethylhexy1)- phosphoric acid, which is incorporated in a liquid scintillator for counting in a liquid scintillation spectrometer. Interferences from alpha- and beta- emitting radionuclides are studied together with interferences from non- radioactive elements. The lower limit of detection is in the region of 1 pCi.PLUTONIUM radionuclides are produced from uranium in nuclear reactors in the following way, with half-lives and particles emitted as indicated- 24000 years 6500 years 13.2 years 380000 years 86 years u! P a a! cy The radionuclides of mass 238, 239, 240 and 242 are alpha-emitters and are highly toxic. The International Commission on Radiological Protection (I.C.R.P.) recommend a “maximum permissible body burden” of only 0-04 pCi for plutonium-239 and made a similar assessment for the other plutonium alpha-emitting radionuclides. Plutonium-241 is a beta-emitter with a very low maximum beta-energy of 0.022 MeV, similar to that of tritium (0-018 MeV). It decays to americium-241, an alpha-emitter with a half-life of 458 years, which makes a significant contribution to the plutonium-241 hazard.The I.C. R.P. recommended a maxi- mum permissible body burden for plutonium-241 of 0.9 pCi so that, in terms of radioactive emission, the plutonium-241 hazard is about one twenty-third of that of plutonium alpha- activity. The ratio of plutonium-241 activity to plutonium alpha-activity increases wit11 reactor burn-up, ie., with the length of time and intensity of neutron irradiation. At the design burn-up of 3000 MW d tonne-l of the Central Electricity Generating Board’s Magnox nuclear power stations, the ratio exceeds 23 : 1 so that the plutonium-241 is the greater hazard. The Advanced Gas Cooled Reactors, the first of which will become operational in 1973, have a design burn-up of 12 000 MW d tonne-l and the plutonium-241 hazard will be about five times the plutonium alpha-activity ha2ard.l These are minimum assessments of the hazard as Magnox fuels are now rated as high as 3800 MW d tonne-l and the Advanced Gas Cooled Reactor fuels will be rated at 18 000 MW d tonne-l in the near future.2 The counting of the low-energy beta-particles from 0-022-MeV plutonium-241 in a pure source of plutonium presents no difficulties and is readily accomplished by liquid scintillation counting as shown by Horrocks and Studier3 (Fig.1). Most methods for the determination of plutonium are concerned with the plutonium alpha-activity and because the zinc sulphide scintillation counters and silicon surface-barrier detectors used to measure the alpha-activity are specific for alpha-particles, decontamination of the plutonium source from /3-emitters is not important; however, it is of great importance in the determination of plutonium-241.Few papers have been published on the determination of plutonium-241. Ludwick4 deter- mined it in urine by co-precipitating the plutonium on lanthanum fluoride and then extracting it into dibutyl phosphate, in which form it was incorporated in the scintillator. Eakins and Lally5 recommended isolating plutonium by any suitable method such as ion-exchange followed by co-precipitation on the white iron complex of diphosphatoferric acid and gel counting. No assessment of interference was made for either method. * Presented at the Third SAC Conference, Durham, July 12th to 16th, 1971. @ SAC; Crown Copyright Reserved.DARRALL, HAMMOND AND TYLER 359 Effluents from nuclear installations generally contain a wide range of beta-emitting radionuclides, and the alpha-activity constitutes only a small proportion of the total activity.They frequently contain small amounts of suspended matter, and any method for the determination of plutonium should include an effective digestion procedure. t 5 C * B Fig. 1. Liquid scintillation spectrum of a mixture of plutonium-241 and plutonium-239. Ratio of plutonium-241 to plutonium-239 is 40: 1. The spectrum is shown a t a single gain setting. For counting in the three channels different gains are used : channel A is the plutonium-241 counting channel ; channel B the channel ratio; and channel C the plutonium-239 counting channel Plutonium tends to deposit from solution at high pH, not only at vessel - solution inter- faces, but also at air - solution interfaces where polyhydroxides are formedJ6 so that a method in which a low pH is used is to be preferred.It appeared that the work of Sill7 and Sill and Williamsg might be applied to the determination of plutonium-241, especially as it could form part of a scheme for isolating and determining uranium and transuranic elements. Sill' found that these elements would co-precipitate on barium sulphate provided that they were in the tervalent or quadrivalent state and that the co-precipitation was carried out in the presence of potassium ions. By co-precipitation after selective oxidation the elements can be separated, as illustrated in Table I. The following method depends upon two of these oxidation stages for the separation of plutonium.The plutonium is oxidised with potassium permanganate to the sexavalent state and curium, americium, radium, etc., are removed by co-precipitation on barium sulphate. The solution is then reduced with sulphur dioxide and re-oxidised with potassium dichromate and the plutonium, which remains in the tervalent TABLE I CO-PRECIPITATION OF TRANSURANIC ELEMENTS IN BARIUM SULPHATE Oxidation after reduct ion to the tervalent state. Oxidant Transuranic elements r A 1 Uranium Neptunium Plutonium Americium Curium Peroxodisulphate . . + Permanganate . . .. + + Dichromate . . . . + + + No oxidation . . .. + + + + 4- denotes element co-precipitated on barium sulphate.360 DARRALL, HAMMOND AND TYLER: THE DETERMINATION OF [AutalySt, VOl.98 state, is co-precipitated on barium sulphate. The latter is then dissolved in 2 N nitric acid and the plutonium is extracted into di(2-ethylhexy1)phosphoric acid, which is incorporated in a scintillator for counting. METHOD APPARATUS- A liquid scintillation spectrometer was used. REAGENTS- Nitric acid, concentrated, sp. gr. 1.42. Sulphuric acid, concentrated, sp. gr. 1.84. Perchloric acid, 60 per cent., sp. gr. 1.54. Potassium sulfihate. Potassium permanganate solution, 0.5 per cent. m/V. Barium rtitrate [Ba(NO,),J solution, 0.48 per cent. m/V. Sulphur dioxide. Potassium dichromate solution, 1 per cent. m/V. Sulphuric acid solution, 0-5 per cent. VlV. Nitric acid, 2 N. Scintillator solution-Dissolve 4 g of terphenyl (p-diphenylbenzene) and 0.1 g of 1,4-di- 2-(5-phenyloxazolyl)benzene (POPOP) in 1 litre of toluene (scintillator grade).Extraction solution-Dissolve 20 ml of di(2-ethylhexy1)phosphoric acid in 80 ml of scin t illat or solution. Standard solution of plutonizcm aZPha-activity-Use a T.R.C. plutonium-239 standard* and calculate total plutonium alpha-activity, including plutonium-238 and plutonium-240 activities. Standard plutonium-241 solution-Dissolve a specimen of mass-analysed plutonium dioxide containing about 0.1 per cent. of plutonium-241 in concentrated nitric acid. The plutonium alpha-activity is assessed by 4n liquid scintillation counting and the plutonium-241 activity of the solution is derived from this result and the mass analysis. PROCEDURE- Transfer a sample volume of not more than 50 ml into a 100-ml conical flask, add 3 g of potassium sulphate, 3 ml of concentrated sulphuric acid, five drops of concentrated nitric acid and a few drops of perchloric acid (60 per cent.) as necessary to complete the oxidation of organic matter, taking appropriate precautions.Finally, evaporate the mixture to fumes and heat it strongly over a high-temperature burner (such as a Meker or Amal-type burner) until the excess of acid has been evaporated off and then heat for a further 2 minutes. Cool, add 0.5 ml of concentrated sulphuric acid and dissolve the pyrosulphate cake by boiling it for 5 minutes with 35 ml of water. Dilute the solution to about 40 ml, add 1 ml of 0.5 per cent. potassium permanganate solution, and boil for 10 minutes at constant volume to oxidise the plutonium to the sexavalent state.Carry out two scavenges as follows. ( a ) Add 1 ml of 0.48 per cent. barium nitrate solution to the boiling solution at a rate of about one drop every 2 s with swirling, boil for 1 minute, repeat the addition of 1 ml of barium nitrate solution, boil the solution for 1 minute and cool it for 10 minutes in cold water. Centrifuge, reject the barium sulphate precipitate and return the solution to the 100-ml conical flask. (b) Evaporate the solution to 30 ml and repeat procedure ( a ) , rejecting the barium sulphate and returning the solution again to the conical flask. Add 2ml of concentrated sulphuric acid and bubble sulphur dioxide gas through the solution for 1 minute to reduce plutonium to the tervalent state, add 2 m! of 0.48 per cent.barium nitrate solution and evaporate to fumes, when barium sulphate will dissolve, cool for 1 minute and add 1 ml of 1 per cent. potassium dichromate solution to oxidise neptunium; allow the solution to cool spontaneously, add 15 ml of water, heat to dissolve the resultant cake, leaving the plutonium on the barium sulphate, boil for 1 minute and cool in running water. Transfer the mixture to a centrifuge tube and use 5 ml of 0.5 per cent. sulphuric acid solution to transfer the residual barium sulphate. Centrifuge, reject the solution, wash the * The Radiochemical Centre, Amersham, Bucks. 2 ml of solution 3 10 mg of barium sulphate.May, 19731 PLUTONIUM-241 IN EFFLUENTS 361 precipitate with 10 ml of 0.5 per cent. sulphuric acid and again reject the washings.Dissolve the barium sulphate in 15 ml of 2 N nitric acid by placing the tube in a boiling water bath for 10 minutes, stirring occasionally. Do not delay carrying out the steps between dis- solution of the cake and the dissolution of the barium sulphate. Transfer the solution to a 50-ml separating funnel, add 5 ml of extraction solution and shake the funnel for 4 minutes. Reject the lower aqueous layer, wash the organic layer with 10 ml of 2 N nitric acid solution, transfer the organic layer into a 20-ml liquid scintillation vial, add 10 ml of scintillator solution, mix and count in three channels of the liquid scintillation spectrometer: one to count plutonium-241 activity, one to count plutonium alpha-activity and one to give a channels ratio for the plutonium-241 (see Fig.1). Prepare plutonium-241 and plutonium-alpha standards in 15 ml of 2 N nitric acid and treat them as in the above procedure starting from “add 5 ml of extraction solution. . . .” Calculate the result for plutonium-241 as follows- a plutonium-241 activity of sample = 2 x - x 1000 pci 1-1 x u where x and y are the numbers of counts per minute recorded in the plutonium-241 channel for the standard and sample, respectively, a is the activity of the added standard in picocuries and z, is the volume of sample in millilitres. Calculate the plutonium alpha-activity in a similar way by using the number of counts per minute recorded in the plutonium alpha-particle channel. 1001 I I I 1 I I 10 20 30 40 50 6aSO I /mg Fig.2. Dependence of plutonium recovery on mass of barium sulphate precipitated DISCUSSION Freshly precipitated barium sulphate is readily soluble in 2 N nitric acid and the amount produced in the final precipitation (10 mg) can be dissolved in 15 ml of 2 N nitric acid under the conditions described. On ageing, the precipitate becomes progressively more difficult to dissolve and after 2 hours it is not possible to dissolve it completely. The dissolution TABLE I1 DECONTAMINATION FACTORS FOR BETA-EMITTERS Isotope x 104 x 105 Isotope x 104 x 105 Plutonium-24 1 Flu tonium-239 Plutonium-241 Plutonium-239 3H 1.6 10 losRu 25 2-5 14c 5.6 3.3 llOmA g 8.7 8.3 3 5 s 4.8 5.3 124Sb 0.89 0.28 45Ca 2 7 2-5 la5Sb 1.1 2.0 S4Mn 4.5 10 134Cs 1-3 10 G5Fe 0-45 3.6 137cs 6.8 5.0 BOCO 8.9 9.1 le4Ce 1.1 1.7 63Ni 13 5-0 la7Pm 50 5.0 652n 25.5 2.5 Wr 14 39 E9Sr 4.0 10 9 5 5 - 0.0082 0.0021 OOSr 14 1.2362 DARRALL, HAMMOND AND TYLER: THE DETERMINATION OF [Analyst, Vol.98 should therefore take place immediately after its formation. A series of experiments starting at the final precipitation stage of the method shows how the recovery of plutonium in the final stages varies with the amount of barium sulphate precipitated (Fig. 2). With amounts of less than 10mg, plutonium is incompletely co-precipitated: a further 6.5 per cent. of plutonium was obtained in a 10-mg scavenge following a 5-mg precipitation. With amounts in excess of 10 mg the barium sulphate is not completely dissolved. TABLE I11 DECONTAMINATION FACTORS FOR ALPHA-EMITTERS Isotope Plutonium-241 Plutonium-239 22*Th .. . . .. .. . . 4.5 x 103 1.9 x 103 226Ra . . .. . . . . . . 5.0 x 104 7.7 x 103 Natural uranium (10 mg) . . . . 7.7 x 102 1.1 x 103 (250mg) . . . . 3.6 x lo1 1.1 x 102 2a7Np . . . . .. . . . . 8-1 x lo1 1-9 x 108 241Am . . .. . . . . . . 1.7 x los 1.3 x 104 242Cm . . .. .. .. . . 1.4 x 10s (30) * * The curium-242 contained about 3 per cent. of the daughter 8ssPu activity The compositions of the extractor and scintillator solutions are based on those used by Keough and Powers0 for the determination of plutonium-239. The beta-particle spectrum of the plutonium-241 and alpha-particle spectrum of the plutonium-239 are well resolved, even when the plutonium-241 beta-activity is forty times the plutonium alpha-activity (Fig.1 .) The approximate positions of the counting windows are shown and these should be adjusted so as to give the maximum value for (efficiency)2/background, as in normal practice. The approximate position of the channels ratio window is also shown. The channels ratio is given by B/A. The efficiency and background for our instrument are- Plutonium-24 1 .. .. .. 28 15 Efficiency, per cent. Background/counts min-' Plutonium alpha-activity . . . . 94 4 and from these values the following limits of detection are derived for a 100-minute counting period and for a significance of twice the standard deviation of the background : plutonium-241, 1.3 pCi (0.78 counts min-l); and plutonium alpha-activity, 0.21 pCi (0.40 counts min-l); but it is probably better to take 2 counts min-l as the least significant difference and the limits of detection are then 3.5 and 1.1 pCi, respectively.Interferences from beta-emitters are recorded in Table I1 as decontamination factors defined in terms of picocuries of interfering radionuclide appearing as 1 pCi of plutonium-241 (or plutonium alpha-activity). There is little interference except from zirconium-95 (including its daughter niobium-95 in about an equal amount). This inter- ference can be detected, for the channels ratio is 0.57 for zirconium-95 and 0.22 for pluton- ium-241. The decontamination factors for alpha-emitters are shown in Table 111. TABLE IV INTERFERENCE FROM ELEMENTS Plutonium Element Amount/mg yield, per cent. Aluminium . . 1 91 10 67 Calcium . . 5 86 10 87 20 74 30 55 40 35 0-5 88 2 71 6 35 Cerium .. Iron . . .. Element Amountlmg Magnesium . . 10 40 Phosphorus . . 40 Silicon .. 0.5 5 Samarium . . 0-4 4 Yttrium .. 0.4 4 Plutonium yield, per cent. 90 86 89 84 86 81 78 78 78 1 91 Zirconium . . 1 83 5 90 6 63 (hydrogen peroxide reduction)May, 19731 363; PLUTONIUM-241 I N EFFLUENTS TABLE V RECOVERY OF PLUTONIUM FROM WATER AND EFFLUENTS Recovery of plutonium from water, per cent. Recovery of plutonium from effluents, per cent. Total solids in aliquot of effluent taken for analysislmg 89-4 82-5 75 91.8 75-0 95 91.2 86-2 67 86.9 85-8 59 92.1 88.6 51 79.7 86.7 58 80.1 88.2 58 90.7 87.6 13 87.5 85.7 25 88-5 88.4 210 - 89-3 420 Average f standard deviation 87.8 & 4.3 85.8 f 3.9 Any co-precipitation process is of course susceptible to interference from many substances and Sill7 and Sill and Williamss have shown that the barium sulphate procedure is no exception.It is important, therefore, to keep the amount of interfering substances at a low level by using the smallest volume of sample that is consistent with the required sensitivity and to measure the recovery of added plutonium if the sample is of unknown composition. Sulphur dioxide is used as a reducing agent rather than hydrogen peroxide because iron interferes with the reduction of plutonium by hydrogen peroxide.lU The extent to which some elements interfere is shown in Table IV. The over-all yield of plutonium from de-ionised water was 88 per cent. and the average yield from a selection of effluents from nuclear installations was 86 per cent.(Table V). The mean ratio of plutonium-241 to plutonium alpha-activity in several effluents from C.E.G.B. Magnox Nuclear Power Stations was 38 (Table VI), thus confirming the importance of the determination of plutonium-241. TABLE VI PLUTONIUM IN EFFLUENTS FROM C.E.G.B. MAGNOX NUCLEAR POWER STATIONS Sample No. 1 2 3* 4 5 6 7 8 Plutonium alpha-activity/nCi 1-1 1.80 0-41 0.45 0.72 8.61 6.97 3-42 1-59 Plutonium-24 1 activity/nCi 1-’ 73.7 15.1 32.5 7.32 378 269 132 69.0 Ratio of plutonium-241 activity to plutonium alpha-activity 41.0 37.2 16.3 45.4 43.9 38.6 38.6 43.3 Mean . . 38.0 * Effluent from a station that had not reached fuel equilibrium. We thank the Government Chemist for permission to publish this paper. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. REFERENCES Dalton, J. C., McDonald, B. J., and Barnes, V., “Biological Monitoring for Plutonium-241,” in “Radioisotope Sample Measurement Techniques in Medicine and Biology,” International Atomic Energy Agency, Vienna, 1965, p. 351. Central Electricity Generating Board, Annual Report and Accounts, 1971-72, H.M. Stationery Office, London, p. 22. Horrocks, D. L., and Studier, M. H., Analyt. Chem., 1958, 30, 1747. Ludwick, J. D., Hlth Phys., 1961, 6, 63. Eakins, J. D., and Lally, A. E., Re$. U.K. Atom. Energy Auth., AERE-R6640, 1970. Robertson, R. B., Ibid., AWRE-0-37/70, 1970. Sill, C. W., Hlth Phys., 1969, 17, 89. Sill, C. W., and Williams, R. L., Analyt. Chem., 1969, 41, 1624. Keough, R. F., and Powers, G. J., Ibid., 1970, 42, 419. Sill, C. W., Percival, D. R., and Williams, R. L., Ibid., 1970, 42, 1273. Received J d y 12th, 1972 Accepted December 28tk, 1972

ISSN:0003-2654    

DOI:10.1039/AN9739800358

出版商:RSC    

年代:1973

数据来源: RSC

摘要:

364 Analyst, May, 1973, Vol. 98, @. 364-371 A Reproducible Pyrolysis Gas-chromatographic System for the Analysis of Paints and Plastics* BY R. W. MAY, E. F. PEARSON, J. PORTER AND M. D. SCOTHERN (Home Ofice Central Research Establishment, AIdermaston, nr. Reading, Berkshire) A system is described that will permit inter-laboratory comparisons of pyrograms, a method rarely used in the past owing to difficulty in obtaining satisfactory reproducibility. After investigation of various pyrolysers and column packings, a Curie point pyrolyser and a solid-phase packing, Pora- pak Q, were chosen as being the most suitable. INTER-LABORATORY comparison of pyrograms (chromatograms obtained by pyrolysis gas chromatography) is rarely carried out because of the difficulty of obtaining sufficient repro- ducibility.Consequently there are no commercial collections of reference pyrograms as there are of infrared, ultraviolet and nuclear magnetic resonance spectra, etc. Pyrolysis gas chromatography has in the past been used mainly for the identification and quantitative analysis of very restricted ranges of polymers or for differentiation of polymers without identification. Although many polymer types have been investigated by using this method, few workers have investigated wide ranges of polymers.1-3 The aims of this work were to find a system of pyrolysis gas chromatography that would require a sample size of less than 20pg and could be used in the identification of a range of polymers, with particular emphasis on those commonly found in paints, and that would also be sufficiently reproducible so that pyrograms could be duplicated in other laboratories.Seven liquid phase and three solid phase packings were investigated in this work and only one packing, Porapak Q, was found to be satisfactory in terms of reproducibility from column to column, useful lifetime and ability to discriminate between polymer pyrolysates. The major part of this work was therefore restricted to pyrolyses with Porapak Q as the column packing and with fixed gas-chromatographic operating conditions. The literature on pyrolysis gas chromatography is expanding and there have recently been several reviews on this Three main types of pyrolyser can be used for solid samples, namely, hot filament, Curie point and furnace'; other forms of heating include electric discharge* and laser light.Q After testing a hot filament and a Curie point pyrolyser, the latter was chosen and three units of this type were used as part of the study on repro- ducibility.This Curie point unit gives rapid and reproducible heating to the Curie point of the pyrolysis wire. If pyrolysis wires of a fixed composition are used in different pyrolysers, identical pyrolysis temperatures are obtained that give highly reproducible pyrograms. A temperature of 610 "C was chosen, as it is in the region in which most polymers give charac- teristic fragmentation patterns. A linked pyrolysis gas-chromatographic - mass spectrometric system was used for the identification of the peaks in all the pyrograms, as this method gives rapid unequivocal identifications. Most of the substances thus identified were injected under standard conditions in order to check that their retention times matched those in the pyrograms. The pyrograms included in this paper are examples taken from a reference collection that has been prepared and that will be published in a monograph.1° The pyrolysis gas- chromatographic method described here is sufficiently reproducible to allow identification of a pyrogram by reference to this collection. EXPERIMENTAL The following column packings were investigated: Carbowax 20M, OV17 and Apiezon L, each 10 per cent.on acid-washed, silanised Chromosorb G; dimethanol succinate and SE-30, both 10 per cent. on acid-washed, silanised Chromosorb W; E301, 10 per cent. on acid-washed Celite C ; Durapak (Carbowax 400); and the solid phases Porasil C, Porapak Q and Chromo- sorb 102.@ SAC; Crown Copyright Reserved.MAY, PEARSON, PORTER AND SCOTHERN 385 Each of these packings was tested to ascertain its ability to discriminate between a series of alkyd paints; of these packings only the Carbowax 20M, Durapak, Porasil C, Porapak Q and Chromosorb 102 had good discriminating power, although temperature programmes were used in each instance. In the manner used, the Carbowax 20M and Durapak had limited lifetimes; in this context, lifetime is taken to indicate the time in use before significant changes in retention times or resolution of peaks, or both, are noticeable. These packings were therefore rejected. We were unable to obtain reproducibility from column to column with Porasil C, so this packing was also rejected.Porapak Q and Chromosorb 102 gave similar results but Chromosorb 102 required a longer ageing period and the batch used gave slightly inferior peak resolution compared with the Porapak Q, so the latter was chosen as the most suitable column packing. APPARATUS AND CONDITIONS- The following gas-chromatographic equipment and conditions were used. Pyrolyser-Pye Unicam Curie point, attached directly to the column ; pyrolysis tem- Gas chromatograph-Pye 104, Model 64 (dual columns, twin flame-ionisation detectors). Columns-Standard Pye, 5 feet long x 4mm id., glass (silanised). Carrier gas-Nitrogen, with a flow-rate of approximately 60 ml min-l. Column packing-Porapak Q, 50 to 80 mesh or 80 to 100 mesh.Temperature programme-From 100 to about 200 "C at 8 "C min-l (the temperature was maintained at 200 "C for up to 25 minutes, and the programmer started on completion of 10 s pyrolysis). The columns were silanised before being packed, by passing a 5 per cent. solution of dichlorodimethylsilane in toluene through the columns, and were dried at 100 "C. Packing was facilitated by applying suction from the detector end of the column, accompanied by gentle tapping. The columns were aged by heating at 250 "C for 48 hours, with a nitrogen flow of 60mlmin-l. The hydrogen and air flows were adjusted to give approximately the maximum sensitivity . The flow-rate required for pyrolysis was determined by injecting through the pyrolyser head 1 p1 of headspace gas from a retention time standard comprising methanol - n-propanol (50 + 50), with the column temperature at 100 "C, and immediately programming at 8 "C min-1.The flow-rate of nitrogen was adjusted to give retention times of 2-6 and 9.1 minutes for methanol and n-propanol, respectively. After fixing the flow-rate of carrier gas, the upper temperature limit was adjusted to give a retention time of 4.2 minutes for benzene or 4.4 minutes for cyclohexane when these compounds were injected at the upper temperature limit. These two compounds have reten- tion times of 13.4 minutes (benzene) and 14.0 minutes (cyclohexane) when injected at 100 "C and the oven is programmed as described. The pyrolysis wire was prepared in the following way. A 5 to 10-mm length at one end of the wire was first flattened until the flattened portion was approximately 1 mm wide.A "hook" was then made by doubling over about 2 mm of the flattened end. At this stage, the hook end plus a few centimetres of the wire adjacent to it were heated to red heat in a bunsen flame so as to remove contaminants. The sample was placed in the elbow of the hook, care being taken not to touch or otherwise contaminate that end of the wire. The sample was then pressed into close contact with the wire surface by crimping the hook. This method was used with all samples. perature, 610 "C maintained for 10 s. RESULTS AND DISCUSSION The polymers were not pre-treated in any way and they were pyrolysed in the solid phase as received. The pyrograms reproduced here are those for a linseed pentaerythritol o-phthalate alkyd (Fig.l), an acrylic emulsion (Fig. 2) and a vinyl chloride (95 per cent.) - vinyl acetate (5 per cent .) copolymer (Fig. 3). The two main criteria are the reproducibility of the system and the effectiveness of the method as a means of discriminating between polymers.366 MAY et al. : A REPRODUCIBLE PYROLYSIS GAS-CHROMATOGRAPHIC [Analyst, VOl. 98 9 11 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 20 15 10 5 0 Time/minutes Fig. 1. Pyrogram for a linseed pentaerythritol o-phthalate alkyd. Peak 1, methane; 2, ethylene; 3, ethane; 4, propylene; 5, propane; 6, acetaldehyde; 7, n-butene; 8, n-butane; 9, acrolein; 10, ally1 alcohol; and 11, methacrolein. Amount of pigmented paint pyrolysed 14 pg REPROD UCIBILITY- An important reason for choosing the polymer bead packing Porapak Q is that it can be used without preparation and has a long life under the conditions used in this work. Reproducibility from column to column is therefore mainly dependent on the manufacturer’s quality control.Although the work was restricted to the instrumental combination of the Pye Unicam Curie point pyrolyser and the Pye 104, Model 64, chromatograph, three completely indepen- dent systems of this type were used. Possible sources of instrument error are the mass flow controllers, the temperature programmers and the column oven temperature. The mass flow controllers were found to be acceptable, although checks on the retention times of the standard methanol - n-propanol mixture should be made at least daily. The time of mechanical rotation 2 7 6 I 4 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 20 15 10 5 0 Time/minutes Fig.2. Pyrogram for an acrylic emulsion. Peak 1, methane; 2, ethylene; 3, ethane; 4, methanol; 5, acetaldehyde; 6, ethanol; and 7, methyl methacrylate. Amount pyrolysed 7May, 19731 SYSTEM FOR THE ANALYSIS OF PAINTS AND PLASTICS 367 of the programmer dials from 100 to 200 "C at 8 "C min-l varied by &6 s (mean 12 minutes 30 s), which is equivalent to k0.8 "C. This variation was probably due to the difficulty in accurately fixing the initial and final temperature pegs, rather than to differences in the programming rates. The temperature, measured by a mercury-in-glass thermometer, con- tinued to increase after the end of the programme, increasing by 2 "C in the first 3 minutes of the final period and a further 1 "C in the following 7 minutes (at 200 "C).This error was constant on the three chromatographs. I 10 Time/minu tes Fig. 3. Pyrogram for vinyl chloride (95 per cent.) - vinyl acetate (5 per cent.) copolymer. Peak 1, methane; 2, ethylene; 3, ethane; 4, propylene; 5, propane; 6, n-butene; 7, n-butane; 8, acetic acid; 9, benzene; and 10, toluene. Amount pyrolysed 44 Pg The three pyrolysis gas-chromatographic instruments were each used with the same analysing column and ten pyrograms were produced from a single alkyd paint on each instrument. The pyrograms contain the same peaks, as can be seen in pyrograms (a) and ( b ) , Fig. 4. The first three peaks have retention times of less than 1 minute and are of unreliable relative peak height.The heights of peaks 4 to 13 were measured and the individual heights compared with the sum total of the heights of these peaks. Table I shows the mean relative peak heights and their coefficients of variation for each of the sets of ten pyrograms obtained with two instruments. All pyrograms given in Tables I to V are of the same common alkyd paint. I I 4 15 10 5 0 15 10 5 Time/minutes Fig. 4. Guide to acceptable limits of resolution [see text for (a) and (b)] I368 MAY et al. : A REPRODUCIBLE PYROLYSIS GAS-CHROMATOGRAPHIC [Analyst, Vol. 98 The third instrument produced a trace with a high base-line drift, which could not be corrected in the time that this machine was available for use. The resulting errors in peak height estimates are reflected in the coefficients of variation given in Table I1 for a set of ten pyrograms, although the mean relative heights are similar to those shown in Table I.TABLE I VARIATION OF THE HEIGHTS OF PEAKS IN PYROGRAMS OF AN ALKYD PAINT BETWEEN TWO PYROLYSIS GAS-CHROMATOGRAPHIC INSTRUMENTS BY USING THE SAME ANALYSING COLUMN Instrument 1 Instrument 2 r 3 I 3 A A Coefficient of Coefficient of Peak number Mean relative height variation, per cent. Mean relative height variation, per cent. 4 14.2 2.0 11.9 12.0 5 8.4 1.9 8.1 8.6 6 6.5 1.8 6.2 6-1 7 9.5 0.9 9-6 2.8 8 5.1 3.5 6.0 5.8 9 22.8 6.5 24.1 6.6 10 4.8 2.9 5.0 10.4 11 7.5 8.1 7.3 7.0 12 12.8 5.2 13.8 7.4 13 7.8 11-4 7.6 4.5 The effect on relative peak heights of using two glass columns packed with the same batch of Porapak Q is shown for a set of ten pyrograms in Table 111.The results given in Table IV show the effect of using different batches of Porapak Q for the analysis. Four batches, two of 50 to 80 mesh and two of 80 to 100 mesh were used. Ten pyrograms of an alkyd paint were produced from each batch. The combined error due to differences from pyrolysis to pyrolysis, instrument to instru- ment, column to column and batch to batch of packing is shown in the standard deviations derived from the ninety results given in Tables I to IV. These standard deviations and coefficients of variation are given in Table V. TABLE I1 VARIATION OF THE HEIGHTS OF PEAKS IN PYROGRAMS OF AN ALKYD PAINT DEVELOPED ON AN INSTRUMENT WITH A DRIFTING BASE-LINE Instrument 3 r-t of Peak number Mean relative height variation, per cent.4 5 6 7 8 9 10 11 12 13 14.2 9.0 7.2 9.2 5.4 21.9 5.6 7.1 13.1 5.5 12.3 21.7 14.2 16.6 23-1 11.7 23.3 26.3 13.2 32.9 With complex pyrograms such as those from which Tables I to IV were derived, peak height measurements were found to be more precise than peak area measurements obtained from an automatic base-line following, shoulder-sensing, electronic integrator (an Infotronics CRS 104 was used). This effect can be seen by comparing the coefficients of variation for the two peak “size” parameters (Table VI). These results are in accord with those of Deans,ll who found that peak-height measurements were more reliable than integrated area measure- ments in the quantitative analysis of a six-component mixture by gas chromatography in which a degree of peak overlap was present.May, 19731 SYSTEM FOR THE ANALYSIS OF PAINTS AND PLASTICS TABLE I11 VARIATION BETWEEN TWO COLUMNS PACKED WITH THE SAME BATCH OF PORAPAK Q (BATCH 411, 80 TO 100 MESH) FOR PYROGRAMS OF AN ALKYD PAINT 369 Peak number 4 5 6 7 8 9 10 11 12 13 Column 1 A f 1 Coefficient of Variation, per cent. Mean relative height 15.5 8.0 8.4 4.0 6.8 7.5 10.2 5.7 5.1 3.3 22.3 5.0 5.7 7.9 7.2 5.7 12.3 5.8 5.9 6.6 Column 2 f A 3 Coefficient of variation, per cent. Mean relative height 14.0 5.1 8.0 4.5 6.2 5.6 9-6 5.9 4-5 5.1 20-7 4.5 7.9 15.4 9.2 3.6 13.6 6.2 6.2 12.4 RESOLUTION- The peak-height measurements on which the tables are based are affected by the resolu- tion of each column, which is largely determined by the properties of the batch of Porapak Q used.The recognition of a pattern of peaks that form a pyrogram of an unknown as being that pattern found in one of the pyrograms in the collectionlo is also highly dependent on the degree of resolution of the analyser column. A guide to the acceptable limits of resolution is found in pyrograms (a) and (b), Fig. 4. The overlapping peaks with retention times of 2 and 2 i minutes (propylene and propane, respectively), which elute immediately before the discrete acetaldehyde peak (retention time 3.4 minutes), and the overlapping peaks n-butene and n-butane with retention times of 5 and 54 minutes, respectively, which appear between acetaldehyde and acrolein (retention time 6.8 minutes) , are used as resolution standards. The degree of overlap of these two pairs of peaks should fall between that found in pyrogram (a) and that in pyrogram (b), as judged by visual inspection, when a gloss house- hold alkyd paint is pyrolysed.A column with resolution beyond these limits may give pyrograms with patterns the successful comparison of which with the collectionlo will be difficult or impossible. Errors in retention times of &lo per cent. are allowed for in both the reference collection and unknown pyrograms when the retention time is more than 5 minutes and A15 per cent. when it is less than 5 minutes. These limits of error are adequate for the retention times of the compounds reported.12 It is recommended that 50 to 80-mesh Porapak Q be used, but if the resolution proves poorer than that in pyro- gram ( b ) , then the use of a finer mesh size such as 80 to 100 mesh will improve it.The C, and C, alkene - alkane pairs that can be seen in these two pyrograms are common to other polymers with long hydrocarbon chains. However, the use of a gloss household alkyd paint TABLE IV Peak number 4 5 6 7 8 9 10 11 12 13 VARIATION OF PEAK HEIGHT BY USING Batch 348 (50 to 80 mesh) M-ent relative of variation, height per cent. 15-5 8.0 8.4 4.1 6.8 7.5 10.2 5.7 5.1 3.3 22.3 4.9 5.7 7.9 7.2 5.7 12.3 5.9 5.9 6.6 Batch 827 (50 to 80 mesh) M-ent relative of variation, height per cent. 13.6 12.7 9-3 9.8 6.6 12-4 10.4 7.6 8.7 7.0 24.0 11.2 2.0 27.0 7.4 5.0 13.2 13-0 4.5 19.1 FOUR BATCHES OF PORAPAK Q Batch 411 (80 to 100 mesh) Mean Coefficient relative of variation, height per cent.14.0 5.1 8.0 4.5 6.2 5.7 9.6 6.9 4.5 5.1 20.7 4.5 7.9 15.4 9.2 3.6 13.6 6.2 6.2 12.4 - Batch 821 (80 to 100 mesh) M-ent relative of variation, height per cent. 10.9 17.2 9.0 15.6 7.1 9.7 9.1 8.2 6.5 10.5 25.2 12.1 4.2 7.6 8.3 5.3 12.6 13.7 7.1 12.4370 MAY et al. : A REPRODUCIBLE PYROLYSIS GAS-CHROMATOGRAPHIC [Analyst, Vol. 98 is essential in testing the resolution of a column compared with pyrograms (a) and (b), as the relative proportions of these C, and C, peaks will differ in other types of polymer. The pyrograms (a) and (b) are visual representations of the variations in resolution that lead to differences in relative peak heights. The resultant slight variations in the appearance of pyrograms would have to be considered when using the reference collection referred to previously.l* It must be remembered that more efficient discrimination is possible when one column is used to produce two pyrograms for comparison; thus, although two different polymers can be identified only as being of the same general class by using this reference collection, it is often possible to distinguish between them by carrying out consecutive pyrolyses of the polymers on one column.This can be seen in the lower coefficients of variation of the results expressed in Tables I, I11 and IV compared with those in Table V. TABLE V MEAN RELATIVE HEIGHTS, STANDARD DEVIATIONS AND COEFFICIENTS OF IN TABLES I TO IV WERE DERIVED VARIATION FOR THE NINETY PYROGRAMS FROM WHICH THE RESULTS Peak number 4 6 6 7 8 9 10 11 12 13 Mean relative height 13.7 8.4 6.6 9.6 5.6 22.0 5.6 7.7 13.1 6.4 Standard deviation 2.0 1.4 0.7 1.0 1.1 2.2 1-4 1.1 1.3 1.4 Coefficient of variation, per cent.14.6 16.3 10.4 10.0 19.6 18.7 24-8 13-8 9.7 21.5 DISCRIMINATING POWER- All of the polymers investigated gave a pyrogram consisting of at least one peak, but each peak is not necessarily a distinguishing characteristic of a single polymer. This is especially true of the peaks with very low retention times, for example, those of the hydro- carbons of low relative molecular mass that are evolved by the pyrolysis of many polymers. When only one significant peak is present in a pyrogram it frequently represents a monomer. In these instances the reliability of identification is clearly related to that attained when a volatile compound is identified from one retention time on one column packing; however, as only a limited group of polymers pyrolyses entirely to the monomer, this is itself a dis- tinguishing characteristic.TABLE VI COMPARISON OF THE PRECISION OF RELATIVE HEIGHT AND RELATIVE AREA MEASUREMENTS I N PYROGRAMS I N WHICH MOST PEAKS ARE INCOMPLETELY RESOLVED Peak Mean Coefficient of Mean Coefficient of number relative height variation, per cent. relative area variation, per cent. 4 14.2 2.0 6-6 19.1 6 8.4 1.9 3.3 10.9 6 6.5 1.8 3-6 11.6 7 9.6 8 6.1 9 22.8 10 4.8 11 7.5 12 12.8 13 7.8 0.9 3.5 6.5 2-9 8.1 5.2 11.4 8-0 3.6 21-6 5.2 9.2 15.9 22-6 6.7 10.5 3.5 15.0 8.1 4.6 12-4 Overlapping peaksMay, 19731 37 1 The discriminating power (and sensitivity) of the method is dependent on the type of detector as well as the column used.Some polymers with a high hetero-atom content yield compounds that cause little or no response in a flame-ionisation detector. These products usually have very small molecules, such as the nitrogen oxides, and consequently are of limited diagnostic value. The pyrogram patterns are not completely independent of the mass of polymer pyrolysed, but this mass dependence is reduced and, for our purposes, becomes unimportant when small amounts are pyrolysed; 5 to 10 pg is a useful mass range that gives good reproducibility of pyrograms (a low concentration of pyrolysate vapour in the pyrolyser reduces the chance of secondary reactions). A much wider range of masses can be used with polymers that pyrolyse by a process of chain depolymerisation from one end (unzipping), thus giving high yields of monomers. SYSTEM FOR THE ANALYSIS OF PAINTS AND PLASTICS 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. REFERENCES Nelson, D. F., Yea, J . L., and Kirk, P. L., Microchem. J., 1962, 6, 225. Groten, B., Analyt. Chem., 1964, 36, 1206. Coupe, N. B., Jones, C. E. R., and Perry, S. C., J . Chromat., 1970, 47, 291. Levy, R. L., Chromat. Rev., 1966, 8, 48. Perry, S. G., Adv. Chromat., 1969, 7, 221. Brauer, G. M., in Slade, P. A., jun., and Jenkins, L. T., Editors, “Techniques and Methods of Perry, S. G., J . Chromat. Sci., 1969, 7, 193. Sternberg, J . C., and Little, R. L., Analyt. Chem., 1966, 38, 321. Folmer, 0. F., jun., and Azarraga, L. V., J . Chromat. Sci., 1969, 7, 665. May. R. W., Pearson, E. F., and Scothern, M. D., “Pyrolysis Gas Chromatography,” Analytical Deans, D. R., Chromatographia, 1968, 1, 187. Dressler, M., Vespalec, R., and Janak, J., J . Chromat., 1971 ,59 ,423. Polymer Evaluation,” Volume 11, Marcel Dekker Inc., New York, 1970, Sciences Monograph, Society for Analytical Chemistry, London, in the press. Received June 13th, 1972 Accepted July 81st, 1972

ISSN:0003-2654    

DOI:10.1039/AN9739800364

出版商:RSC    

年代:1973

数据来源: RSC

摘要:

372 Analyst, May, 1973, Vol. 98, pp. 372-375 Titrimetric Determination of Amphetamine Sulphate in Tablets BY N. VELGHE AND A. CLAEYS (Laboratory of Analytical Chemistry, Faculty of Pharmacy, University of Ghent, Ghent, Belgium) A titrimetric method is described for the determination of amphetamine sulphate in solutions, which can be applied to the analysis of pharmaceutical products. The analysis, which is based on bromination of the primary amino group followed by iodimetric determination, can be carried out by ampero- metric titration or titration by use of an indicator. The principal advantages of the method are that it is rapid and relatively easy to perform, that it has high sensitivity, which permits the analysis of individual tablets, and that inexpensive apparatus can be used.THE methods described for the determination of amphetamine sulphate in tabletsl-5 are rather time consuming because they require its separation by filtration, extraction, distillation or chromatography before the final analysis can be carried out. During an earlier study6 we developed a rapid amperometric method for the determination of amphetamine sulphate by using hypobromite. Further investigation showed that this method was suitable only for pure aqueous amphetamine sulphate solutions while in the analysis of tablets positive errors occurred owing to interfering reactions with other substances. The purpose of the present investigation was to evolve a modified method in which these interferences are eliminated. EXPERIMENTAL APPARATUS- Amperometric titrations were carried out with a Metrohm E261 R polarograph with a rotating platinum micro-electrode and a silver - silver chloride - saturated potassium chloride electrode.Potentiometric acid - base titrations were carried out with a Metrohm E510 pH meter, with a glass electrode and a saturated calomel electrode. REAGENTS- B u f e r solution, p H 6.4-Prepare by dissolving 138 g of sodium dihydrogen orthophos- phate monohydrate in water, adding 50 ml of 10 M sodium hydroxide solution and diluting the mixture to 1 litre. Sodium arsenite solution, 0.025 M, p H 4-8-Standardise by amperometric titration with a standard potassium bromate solution. Iodine in 2 per cent. potassium iodide solution, 0.025 M. Sodium hydroxide solution, 0.1 M-Standardise by potentiometric titration with standard Hydrochloric acid, 0.1 M.Amphetamine sulphate solutions, aqueous, of various concentrations-These solutions were as described in the Belgian Pharmacopoeia. Bromine solution, aqueous, 0.025 M. Potassium iodide solution, 20 per cent. Starch solution, 1 per cent. Bromocresol green solution, alkaline, 0.1 per cent. oxalic acid solution. METHOD The method of analysis is based on the bromination of the primary amino group with bromine whereby the dibromoamphetamine formed is determined by iodimetric titration ; bromine is used as the brominating reagent instead of hypobromite6 because of the absence of interfering bromite. The course of the titration is followed amperometrically or by use of an indicator. @ SAC and the authors.VELGHE AND CLAEYS 373 The reactions that occur during the determination can be summarised as follows- (i) Addition of a small excess of bromine: fast R-NH, + 2Br, --+ R-NBr, + 2H+ + 2 B r (ii) Titration of the excess of bromine with arsenite solution: fast Br, + AsO,~- + H,O --+ 2Br- +:AS~,~-~-+ 2H+ (iii) Addition of arsenite in excess with respect to the dibromoamine, immediately slow followed by an excess of iodide: a.R-NBr, + 2 A ~ 0 , ~ - + 2H,O --+ R-NH, + 2As0,” + 2H+ + 2 B r b. R-NBr, + 4KI + 2H,O - R-NH, + 21, + 4K+ + 2 B r + 20H- c. I, + fast fast + H,O + 21- + AsOd- + 2H+ (iu) Titration of the excess of arsenite with iodine solution, reaction (i;;) c. PROCEDURES- 1. Amperometric titvation-Amphetamine sulphate (2 to 10 mg) (or one pulverised tablet) dissolved in 25 ml of water plus 4 ml of buffer solution are transferred into a dark-coloured beaker.The beaker is raised into position under the electrodes and a potential E of +O-7 V is applied. Bromine solution is added until a rapid increase in the reduction current occurs. The excess of bromine is titrated with arsenite solution, the current being recorded im- mediately after each addition. Then an amount of arsenite, which is at least equivalent to the amount of dibromoamphetamine present (about 0-5 ml of 0.025 M arsenite per milligram of amphetamine sulphate), and 1 ml of 20 per cent. potassium iodide solution are added consecutively; the arsenite solution is added before the addition of the iodide solution in order to prevent the presence in the solution of free iodine, which reacts with some tablet ingredients.The potential is shifted to E = +0.2 V and the excess of arsenite is titrated with the iodine solution; with tablets the iodine solution is added in very small volumes near the equivalence point because of the deflection from linearity and the instability of the iodine current, which is caused by reaction of iodine with the tablet ingredients. Titration curves obtained in the analysis of pure aqueous amphetamine sulphate solutions are given in Fig. 1. A rse n i te/m I I od ine/m I Fig. 1. Titration graphs obtained in the amperometric determination of amphetamine sul- phate. (a), Titration of excess of bromine with arsenite solution, and ( b ) , titration of excess of arsenite with iodine solution374 VELGHE AND CLAEYS : TITRIMETRIC DETERMINATION OF [Analyst, Vol.98 2. Titration by use of an indicator-Amphetamine sulphate (10 mg) (or two pulverised tablets) dissolved in 25 ml of water, 4 ml of buffer solution and 0-1 ml of bromocresol green indicator are transferred into a clear-glass beaker. Bromine solution is added until about 0.1 ml is present in excess of that required to reach the equivalence point, which is indicated by a change of colour of the solution to yellow. After the addition of a second 0-1-ml volume of bromocresol green indicator, arsenite solution is added in volumes of about 0.02 ml, The equivalence point is reached when the blue colour of the solution has developed its maximum intensity. Then 5 ml of arsenite solution and 1 ml of 20 per cent. potassium iodide solution are added consecutively; in the analysis of tablets the addition of potassium iodide causes a colour change from blue to green.After the addition of 1 ml of 1 per cent. starch solution, the excess of arsenite is titrated with the iodine solution and a colour change to mauve (pure amphetamine sulphate) or grey (tablets), which is difficult to define, occurs at the equivalence point. The titration is carried out under artificial light. 3. In the analysis of pure aqueous amphetamine solutions a more simple procedure can be applied, whereby the titration with iodine solution is superfluous: after bromination and reduction of the excess of bromine, an excess of iodide solution is added and the iodine formed is titrated with arsenite solution.This procedure cannot be used in the analysis of the pharmaceutical products studied because of the interfering reactions of iodine with some tablet ingredients previously mentioned. The above three procedures must be carried out rapidly in order to minimise errors caused by the instability of dibromoamphetamine. The amount of amphetamine sulphate is calculated from the titration results as follows- Amphetamine sulphate (mg) = 92.13 x (v$fAs9+ - V,MI~) (procedures 1 and 2) where Vl ml is the excess volume of arsenite solution added after titrating the excess of bromine, V2 ml the volume of iodine solution used for titration of residual arsenite, M the molarity and V ml the volume of arsenite solution used for titration of the iodine formed. RESULTS AND DISCUSSION Pure amphetamine sulphate solutions were analysed by using the procedures described for the amperometric titration and titration by use of an indicator.The amphetamine in analogous but more concentrated solutions was determined by acid - base titration after steam distillation from alkaline medium. The titration results, summarised in Table I, demonstrated the good agreement obtained by the bromination methods; the difference between these results and those obtained by using the classical acid - base titration technique was less than 1 per cent. For the proposed methods the reproducibility, defined as the relative standard deviation on the average result of ten titrations, was about 0.1 per cent. = 92.13 x VMA,S+ (procedure 3) TABLE I DETERMINATION OF PERCENTAGE PURITY OF AMPHETAMINE SULPHATE Titration by use of an BY By Purity, By By Purity, By By Purity, - Acid - base titration indicator Amperometric titration after distillation r r weigh- titra- per weigh- titra- per weigh- titra- per ing/mg tionlmg cent.ing/mg tionlmg cent. ing/mg tionlmg cent. Excess of 4.230 4.204 99.4 2.049 2.029 99.0 arsenite 5.104 5.070 99.3 5.105 5.070 99.3 6.114 6.081 99.5 10.01 9.943 99.3 No excess of 4.230 4.204 99.4 - - I arsenite 5.104 5-070 99.3 - - - 6.114 6.075 99-4 - I I 398.9 399.7 100.2 405.0 405.5 100.1 The amphetamine sulphate content of three commercial products was determined. About fifty tablets were pulverised and an amount of the powdered material containing about 5 mg (amperometric titration) or 10 mg (titration by use of anindicator) of amphetamineMay, 19731 AMPHETAMINE SULPHATE IN TABLETS 375 sulphate was determined after dissolution in water ; also, some amperometric determinations were carried out after steam distillation from alkaline medium of the amphetamine from an amount of powder containing 25 mg of amphetamine sulphate.The results, summarised in Table 11, indicate that the amount of amphetamine sulphate found is in close agreement with that indicated by the label; the small difference that occurs between the results by the two direct methods and that by the method after steam distillation demonstrates that the tablet ingredients of the products analysed do not have much influence on the results. CONCLUSIONS The proposed bromination procedure can be used for the determination of amphetamine sulphate in pure aqueous solutions as well as in pharmaceutical products.The principal advantages of the method are that it is easily and rapidly performed, that it is sensitive, thus permitting the analysis of individual tablets, and that inexpensive apparatus can be used. From the satisfactory results obtained for the three commercial products with different compositions studied, it is reasonable to expect that other drugs that contain amphetamine can also be analysed by the described procedure. TABLE I1 DETERMINATION OF AMPHETAMINE SULPHATE IN DRUGS Commercial product Per cent. of amount indicated by the label f A 1 By direct determination By determination P after distillation, Titration by use of Amperometric amperometric an indicator titration titration Dexedrine tablets, Recherche et Industrie Dexamyl tablets, Recherche et Industrie Adiparthrol dragdes (Will-Pharma, Brus- Th6rapeutique . . .. .. . . 101.7 f 0.4 101.8 & 0-6 100-6 Thdrapeutique . . .. .. . . 99.7 f 0.4 98.8 f 0.3 97.7 sels) . . .. .. .. . . 100-0 f 0.4 100.8 f 1.1 98.1 REFERENCES 1. 2. 3. 4. 5. 6. Stewart, J. T., and Lotti, D. M., J . Pharm. Sci., 1971, 60, 461. Moody, J. E., Ibid., 1963, 52, 791. Fabrizio De Fabrizio, Ibid., 1972, 61, 101. “The United States Pharmacopoeia,” Eighteenth Revision, Mack Publishing Co., Easton, Pa., “The British Pharmacopoeia,” The Pharmaceutical Press, London, 1968, pp. 45 and 298. Velghe, N., Ph.D. thesis, Faculty of Pharmacy, University of Ghent, 1970, pp. 198 and 199. 1970, p. 179. Received August llth, 1972 Accepted January loth, 1973

ISSN:0003-2654    

DOI:10.1039/AN9739800372

出版商:RSC    

年代:1973

数据来源: RSC

摘要:

376 Book Reviews Analyst, May, 1973 ANALYTICAL ATOMIC ABSORPTION SPECTROMETRY. By W. J. PRICE. Pp. xiv + 239. London, The past decade has seen an almost exponential growth in the application of atomic-absorption spectrometry to trace-metal analysis. During this period, several text-books concerned exclusively with the technique have been written. This book, written by a well known and respected analytical spectroscopist, is a welcome and refreshing addition to the existing texts. The author states in his preface that the book is written specifically for the analytical chemist a t his bench or the student of analytical chemistry, who now finds it difficult to select between the many procedures published and referred to in other texts, and who may consider the numerous sources of theory and background information to be confusing rather than enlightening. The text is therefore set out in a way in which the author believes the analytical chemist would wish to develop a knowledge of the subject.The introductory chapters provide a brief survey of the basic principles of the technique. The exploitation of these principles in the necessary instrumenta- tion for atomic-absorption spectrometry is then discussed in Chapter 3. Chapter 4 deals in general terms with analytical techniques and procedures ; instrument operation, calibration, sample preparation, interference effects, trace and major element determination and method development are each discussed in separate sections, Chapter 5, which occupies approximately 30 per cent. of the length of the book, is devoted to an extremely useful and authoritative account of the applications of atomic-absorption spectrometry in metallurgical, biological, pathological and medical analysis.This chapter also contains an account of indirect atomic-absorption methods and the uses of atomic-fluorescence spectroscopy. The atomic-absorption characteristics of different elements and tabulation of manufacturers of atomic-absorption and related equipment are dealt with in two appendices. This volume is well written and the text accurately reflects the present status of the technique. It should find a place among those texts on subjects of current interest, which the practising analyst prefers to keep a t hand in his laboratory rather than in his library. New York and Rheine: Heyden & Son Ltd.1972. Price L5.80; $15.25; DM52. G. F. KIRKBKIGHT RECENT TOPICS IN MASS SPECTROMETRY. Edited by ROWLAND IVOR REED. Pp. x + 357. New York, London and Pans: Gordon and Breach Science Publishers. 1971. Price k12.25; $29.40. This book is based on the proceedings at a NATO Study Institute of Mass Spectrometry held in Lisbon in August, 1969. The success of this Study Course, as with those held previously, was due, in the main, to the inspiration of the organiser, Dr. Ivor Reed. This collection of articles, gathered together under his editorship, is an appropriate record of his work and that of the Institute. Each chapter is an authoritative discussion of a particular aspect of mass spectrometry. The articles give good coverage of their particular topics while the references are sufficient without being exhaustive.The topics covered are: Multiply charged ions (M. E. Wacks and W. M. Scott) ; Energetics of ionisation and dissociation processes by electron impact (M. A. A. Ferreira) ; Meta- stable ions (M. T. Robert-Lopes) ; Metastable ions and some mechanistic interpretations (G. R. Lester); Mass spectrometry of ferrocenes and related complexes (G. A. Junk and H. J. Svec); Application of high-resolution mass spectrometry to complex mixtures (A. G. Sharkey, Jr.) ; Photo-ionisation and photoelectron spectroscopy (J, E. Collin) ; Study of rearrangement reactions by field ionisation mass spectrometry (H. D. Beckey and K. Levson); Combined gas chromato- graphy and mass spectrometry (C. Memtt, Jr.) ; Nuclear measurements by mass spectrometry (N. R. Daly and N. J. Freeman); Industrial application of mass spectrometry (A. Quayle); and Some considerations of the na'ive analysis of structure (R. I. Reed and D. H. Robertson). As the NATO Institute course was held in 1969 and this book was not published until 1971, the title appears to be somewhat misleading. Nevertheless, the book is a valuable addition to the mass spectrometry library. The contents will serve as useful introductory reviews to the areas covered. The book is well produced and pleasing to read. D. PRICE

ISSN:0003-2654    

DOI:10.1039/AN9739800376

出版商:RSC    

年代:1973

数据来源: RSC