摘要:

652 Analyst, October, 1966, Vol. 91, Pp. 652-653 SHORT PAPERS An Automatic Method for Determining Orthophosphate in Sewage and Highly Polluted Waters BY A. HENRIKSEN* (Norwegian Institute for Water Research, Oslo 3, Norway) IN a previous paper1 an automatic method for determining low levels of phosphates in fresh and saline water by using the Technicon AutoAnalyzer was described. The principle of this method is as follows: phosphate ions react with molybdate in a sulphuric acid medium to form a yellow complex that is extracted into isobutanol and reduced to molybdenum blue with tin(I1) chloride in the organic layer. However, this method was too sensitive for analysing sewage and highly polluted waters where concentrations from 0.5 to 10 mg of phosphate phosphorus per litre are usual.The yellow phosphomolybdate complex is widely used for determining phosphate in higher concentrations. Nitric acid is most commonly used to provide an acid medium, and sensitivity is increased by the addition of vanadate. From a practical point of view it was desirable to use the flow scheme of the method previously described, with only minor modifications in the reagents. It was found that merely by omitting the reduction step in the procedure described above, a reliable and reproducible method for determining higher concentrations of phosphorus was obtained. The flow scheme and reagents used are the same as those described earlier,l except that the reducing mixture is replaced by isobutanol. A linear relationship between optical density and phosphate concentration is obtained in the range from 0 to 6mg of phosphate phosphorus per litre.A tendency for tailing is shown on the AutoAnalyzer traces. Fifteen samples are analysed per hour, with 1$ minutes allowed for sampling and 24 minutes for washing out between samples. The absorption spectrum of the yellow phosphomolybdate formed under the experimental conditions of this method showed a maximum a t 345 mp, The optical density at 420 mp is only about one-seventh of the density a t 345 mp. Twenty-two samples taken from sewage effluents and polluted rivers were analysed in duplicate for orthophosphate content, and the yellow colour formed was measured a t both 340 and 420 mp. A phototube colorimeter was used. The concentration range was from 0-40 to 5-50 mg of phos- phate phosphorus per litre.The results from these experiments are given in Table I. The standard deviations of the method a t both wavelengths are nearly equal, and both the F-test andt-test show that identical results are obtained with either wavelength. A less sensitive, although equally specific, method was therefore needed. TABLE I DETERMINATION OF ORTHOPHOSPHATE AT 340 AND 420 mp Range: 0.40 to 5.50 mg of phosphate phosphorus per litre Standard deviation, Mean value, Wavelength, mg of phosphate mg of phosphate mCL phosphorus per litre phosphorus per litre Variation F-value t-value 340 0.078 2.55 3.97 420 0.075 2-56 4.04 1.02 0.016 - - Sewage effluents and polluted rivers contain significant amounts of coloured organic matter, which may be extracted into isobutanol and thereby give rise to an additional absorption a t the low wavelengths used.The samples were, therefore, analysed for a second time a t both wavelengths by replacing the acid molybdate reagent with 4 N sulphuric acid. The mean values of the additional absorption due to extractable matter in the samples were 0-15 and 0.20 mg of phosphate phos- phorus per litre a t 340 and 420 m,u, respectively. The corresponding deviations from the mean values were 0.06 and 0.05 mg of phosphate phosphorus per litre. These deviations are lower * Present address : The Institute of Paper Chemistry, Appleton, TVisconsin, U.S.A.SHORT PAPERS 653 than the standard deviation of the method itself. may be used without introducing any significant error in the results. the correction factor should be determined on representative samples. copper per litre and 20 mg of silica per litre do not interfere in the method described. Consequently, a constant correction factor It is recommended that Also, concentrations of up to 10 mg of iron(I1) per litre, 10 mg of iron(II1) per litre, 5 mg of REFERENCE 1. Henriksen, A., Analyst, 1965, 90, 29. Received November 26th, 1965

ISSN:0003-2654    

DOI:10.1039/AN9669100652

出版商:RSC    

年代:1966

数据来源: RSC

摘要:

SHORT PAPERS 653 The Determination of 4Aminobiphenyl in Aromatic Amines BY 0. NORMAN (Staveley Chemicals Ltd., Staveley Works, Chesterfield) AND G. A. VAUGHAN (The Coal Tar Research Association, Govnersal, Leeds) STAGG and Reed1 described, in 1957, a chromatographic method for determining 4-aminobiphenyl in technical diphenylamine. It was brought to our attention that this method had been applied to technical aniline and a content of 0.15 per cent. of 4-aminobiphenyl had been found. This figure was confirmed by both our laboratories. The Stagg and Reed method is based on the extraction with acid of 1 g of the technical di- phenylamine which normally contains a small amount of primary aromatic amines (usually less than 1 per cent. and often consisting principally of aniline). The extract containing the primary aromatic amines is diazotised and coupled with R-salt in alkaline solution to give a soluble-dye mixture which is chromatographed on filter-paper.The chromatogram shows an outer orange - red band due to aniline, if this impurity is present, and an inner bluish-red band due to 4-amino- biphenyl. It can therefore be assumed that this method is directly applicable to the determination of 4-aminobiphenyl in aniline, provided that not more than 0.01 g of aniline is taken for test. EXPERIMENTAL In an investigation into the source of the contamination of the aniline, it became apparent that in some way 4-aminobiphenyl, or a material having similar chromatographic properties, was being produced in the test. It was shown, for example, that the fore-runnings of a comniercial aniline batch-distillation gave the same content of 0.15 per cent.as the final distillate, and pure aniline, doubly-fractionated and chromatographed in a preparative-scale gas - liquid chromato- graphic apparatus, still gave a similar content. Comparison of the separated 4-aminobiphenyl dye obtained from pure 4-aminobiphenyl and of the inner bluish-red dye band from pure aniline showed that they had similar visible spectra. Saunders2 suggests that when benzene diazotates decompose in alkaline solution phenyl radicals are formed- The benzene diazotate can then be attacked to produce p-biphenyl diazotates- It is thus possible for biphenyl compounds to be formed in the coupling reactions of diazotised anilines. In an investigation of the simpler reaction of diazotised aniline with potassium iodide to form i~dobenzene,~ the infrared spectrum of the crude product showed unequivocally the presence of 4-iodobiphenyl.If, therefore, 4-aminobiphenyl is being produced a t the coupling stage of the test, it will be affected by the concentration of the R-salt coupling reagent. Table I shows the content of 4-aminobiphenyl found at various R-salt concentrations when using 0.01 g of the purified aniline. TABLE I C,H,-NH-OH = C,H,* + N, + OH* C6H5" + C,H,-NN-OH = C,HS-C6H*-NN-OH + H" RESULTS OBTAINED IN TESTS FOR 4-AMINOBIPHENYL CONTENT O F ANILINE R-salt concentration, 4-Aminobiphenyl content, molar per cent. 0.025 0.17 0.05 0.15 0.10 0.05 0.25 < 0.002 0.50 < 0.002654 SHORT PAPERS [Analyst, 1’01.91 The effect of the change from the recomrncnded R-salt reagent concentration of 0.05 N to the nearly saturated 0-50 M, is thus to reduce the apparent 4-aminobiphenyl content of the purified aniline to the “not detectable” level of less than 0-002 per cent. It was shown that this high R-salt concentration did not affect the determination, as satisfactory calibration graphs were obtained by adding 4-aminobiphenyl to the purified aniline in the range 0.005 to 0-05 per cent. Tests were made with the 0.50 M R-salt on a number of commercial aniline samples of lsboratory-produced anilines and also of the fore-runnings and final distillates of a number of distillation runs. An investigation into the use of the method to separate impurities in pure sulphanilic acid and l-naphthylamine also showed that inner chromatographic bands that were found with low concentration R-salt were absent with the high concentration R-salt.In no instance was 4-aminobiphenyl detected in the samples. CONCLUSION The Stagg and Reed chromatographic method for determining 4-aminobiphenyl in technical diphenylamine has been shown to produce small amounts of 4-aminobiphenyl, if aniline is present, owing to the free-radical formation at the R-salt coupling stage with the diazotised aniline. The error introduced by this side-reaction is small if the method is applied to technical diphenylamine containing not more than 1 per cent. of aniline, as it will not be greater than 0.0015 per cent. However, the error becomes important if the aniline content is high, or if the method is applied to aniline alone or used in the analysis of other primary aromatic amines. In these circumstances 0.5 M R-salt must be used in the coupling stage to reduce the side-reaction to a minimum. REFERENCES 1. 2. 3. Stagg, H. E., and Reed, R. H., Analyst, 1957, 82, 503. Saunders, I<. H., “The Aromatic Diazo Compounds,” Second Edition, Edward Arnold and Co., Vogel, A. I., “A Textbook of Practical Organic Chemistry,” Third Edition, Longmans, Green Received September 3rd, 1965 London, 1949, p. 112. and Co., London, 1962, p. 578.

ISSN:0003-2654    

DOI:10.1039/AN9669100653

出版商:RSC    

年代:1966

数据来源: RSC

摘要:

654 SHORT PAPERS [Analyst, 1'01. 91 Specific Spot Tests for Silver Cyanide BY F. FEIGL (Laboratorio da Produpdo Mineyal, Ministe'rio das Minus e Energia, Rio de Janeiro) AND A. CALDAS (EscoZa Nacional de Quimica, Uiziversidade do B r a d ) HITHERTO, no reactions of silver cyanide, associated with substances used to dissolve it, that lead to the formation of coloured products permitting application in qualitative analysis have been known. To identify this compound when alone or mixed with other water-insoluble, acid-resistant siIver halides, decomposition with zinc and dilute sulphuric acid was necessary to produce hydrogen cyanide, which can be detected by means of appropriate sensitive colour tests.l.2 The application of this method is rather difficult when only small amounts of the sample are available.It must also be taken into account that the hydrogen cyanide formed is partly reduced to methylamine by nascent hydrogen. A further difficulty is that silver thiocyanate also splits off hydrogen cyanide (besides hydrogen sulphide) when treated with zinc and acid. Again, reductive cleavage of hydrogen cyanide likewise occurs with silver ferricyanide, silver ferrocyanide and silver cyanate. The spot tests described here overcome the interferences mentioned; they are based on the behaviour of silver cyanide towards mercury ions. I. DETECTION OF SILVER CYANIDE THROUGH ITS CONVERSION INTO SILVER CHROMATE As stated by Korenman,3 all silver halides are soluble in a nitric acid solution of mercury(I1) nitrate owing to the formation of water-soluble undissociated mercury halide- ..* - (1) 2AgHal + Hg2+ -+ HgHal, + 2Ag+ . . . .October, 19661 SHORT PAPERS 655 We have found that a solution of mercury(I1) acetate (or nitrate) that contains sodium acetate and acetic acid dissolves only silver cyanide, whereas silver halides remain completely unchanged- Obviously, owing to the formation of complex mercury(I1) acetate, the Hg2+ concentration in the reagent is diminished to such an extent that (I) does not occur when (2) takes place. This assumption is confirmed by the fact that the acetate-containing solution of mercury(I1) acetate (or nitrate) does not react with alkali chromate to precipitate yellow mercury(I1) chromate. Likewise, no precipitation occurs when ammonia, sodium arsenite or sodium arsenate is added.Therefore, a solution can be prepared that contains complex mercury( 11) acetate and alkali chromate. This solution reacts immediately with silver cyanide to produce red - brown silver chromate- 2AgCN + Hg2+ -+ Hg(CN), + 2Ag+ . . .. .. * * (2) 2AgCN + Hg2+ + CrOd2- --f Ag2Cr04 + Hg(CN), . . . . . . (3) The conversion of silver cyanide into silver chromate permits the specific detection of silver cyanide, because other water-insoluble, acid-resistant silver salts such as silver chloride, silver bromide, silver iodide, silver thiocyanate, silver iodate, silver cyanate, silver ferrocyanide and silver ferricyanide do not react. Reaction (3) also occurs with the residue obtained by evaporating an ammoniacal solution of silver cyanide. Treatment of the silver salts with ammonia solution permits the separation of silver cyanide (together with silver chloride, silver iodate, and silver thiocyanate) from silver iodide and partially from silver bromide.This previous separation procedure should be applied if small amounts of silver cyanide are to be detected in the presence of large amounts of silver iodide or silver bromide. METHOD REAGENT- hfevczwy( I I ) acetate (oil niti/ate) - potassium chromate solution-Prepare by acidifying a few milli- litres of a 5 per cent. w/v aqueous solution of mercury(I1) acetate or mercury(I1) nitrate with acetic acid, adding 1 to 2 g of sodium acetate and several drops of potassium chromate solution. If slight precipitation occurs, filter or spin the solution in a centrifuge.The yellow solution thus obtained is stable. PROCEDURE- Place a minute amount of the sample on a filter-paper or in a depression of a spot plate and add a drop of the reagent. The immediate formation of a red - brown silver chromate precipitate indicates the presence of silver cyanide. The mercury(I1) acetate (or nitrate) - potassium chromate reagent also precipitates silver chromate from solutions that contain complex [Ag(CN) ,] - ions- 2[Ag(CN),]- + 2Hg2+ + Cr04,- -+ Ag,CrO, + 2Hg(CN), The presence of alkali cyanide does not interfere because CN- ions are eliminated through the formation of mercury(I1) cyanide- 2CN- + Hg2+ + Hg(CN), The de-masking of Ag+ ions from [Ag(CN),j- ions in the presence of CN- ions permits a rapid preliminary test for silver halides because Ag+ ions easily dissolve in alkali cyanide to form [Ag(CN),]- ions.A drop of the solution is tested as previously described. DETECTION OF TRACES OF SILVER CYANIDE- A solution of potassium cyanide was prepared sufficiently dilute so that no visible precipitation of silver cyanide was produced when silver nitrate was added to 10 ml of it. To 3 ml of this solution 1 drop of a 1 per cent. solution of sodium chloride was added, followed by excess of 1 per cent. silver nitrate solution. The precipitate containing silver chloride and silver cyanide was collected by spinning in a centrifuge. I t was carefully washed to remove the excess of Ag+ ions and then treated with the mercury(I1) acetate (or nitrate) - potassium chromate solution. The precipitate became distinctly brown.It therefore appears that traces of silver cyanide entrained by the precipitation of silver chloride can be detected in this way.656 SHORT PAPERS [Analyst, VOl. 91 11. DETECTION OF SILVER CYANIDE THROUGH THE FORMATION OF MERCURY(I1) CYANIDE Silver cyanide can be detected through the detection of mercury(I1) cyanide which is formed as a result of the reaction of this silver salt with mercury(I1) acetate or nitrate (equation 2). A rapid spot test characteristic for mercury(I1) cyanide was recently described.* It is based on the fact that mercury(I1) cyanide solution reacts with arsenic trisulphide to give a black mixture of elemental- mercury and mercury sulphide. The test can be performed by spotting filter-paper that contains finely divided arsenic trisulphide impregnated in its pores with a drop of a neutral mercury(I1) cyanide solution; a black stain appears immediately (limit of identification: 1 pg of Hg(CN),).The stable reagent-paper is prepared by bathing strips of filter-paper (S & S 410 or 576) in a saturated solution of arsenic trisulphide in concentrated ammonia, followed by drying. The ammoniacal solution contains (NH,) ,ASSO, and (NH,),AsS,O. On drying the moistened paper for some time, the ammonia volatilises and yellow arsenic trisulphide remains. To detect silver cyanide through the formation of mercury(I1) cyanide, place a small amount of the sample on the yellow arsenic trisulphide paper and add a drop of mercury(I1) acetate (or nitrate) solution. The mercury(I1) cyanide formed reacts a t once with the arsenic trisulphide. The black stain thus obtained is not affected when spotted with ammonia or exposed to ammonia vapour, whereas the unspotted yellow paper becomes white owing to the formation of ammonium sulpho-arsenites. We thank the Conselho Nacional de Pesquisas (Rio de Janeiro) for financial support. REFERENCES No other water-soluble mercury salt reacts in this way. 1. 2. 3. 4. Sieverts, A., and Hermsdorf, A., 2. angew. Chem., 1921, 34, 3. Feigl, F., and Anger, V., Analyst, 1966, 91, 282. Korenman, I. M., and Potenkina, V. G., Chem. Abstr., 1949, 43, 5701. Feigl, F., and Caldas, A., Chemist-Analyst, 1966, 55, in the press. Received February 18th, 1966.

ISSN:0003-2654    

DOI:10.1039/AN9669100654

出版商:RSC    

年代:1966

数据来源: RSC

摘要:

656 SHORT PAPERS [Analyst, Vol. 91 The Determination of Benzylpenicillin Traces in Propyliodone Injection B.P. BY MRS. DOROTHY J. SIMMONS AND J. P. JEFFERIES (Glaxo Laboratories Ltd., Greenford, Middlesex) THE contamination of other drugs with traces of penicillin has been investigated recently by the United States Food and Drug Administration. This investigation resulted in the publication of methods for detecting penicillin in drugs1 and the adoption of standards for the maximum per- missible penicillin content of drugs distributed in the United States.2 The prescribed limit for a parenteral injection is a penicillin content equivalent to 0.05 unit of benzylpenicillin per dose. This limit is particularly severe for Propyliodone Injection, the adult dose of which can be as large as 18 ml, so that it is equivalent to a sodium benzylpenicillin content of 0.0016pg per ml.S o minute a level of permissible contamination can pose difficult problems of both production and analysis. A study of the extraction methods proposed by the Food and Drug Administration indicated that they were not sufficiently sensitive for such testing of Propyliodone Injection. It proved difficult to detect contamination with benzylpenicillin a t the maximum permitted level of 0.003 unit per ml, even when the viscous supernatant liquid from the suspension was used without dilution for the microbiological assay. Some method of concentrating the benzylpenicillin appeared essential to enable the required sensitivity to be achieved. EXPERIMENTAL Preliminary experiments showed that dilute solutions of benzylpenicillin could be concen- trated with some success by modifying the Food and Drug Administration Sample Preparation Methodl Number 2.In the original method, 10 ml of aqueous solution a t pH 2.5 is shaken with two successive 20-ml volumes of chloroform, and the penicillin in the combined chloroform solutions is then extracted into 10ml of pH 6 phosphate buffer solution for microbiological analysis. The penicillin concentration in the final extract can be increased about 6-fold by using only 1.6 mlOctober, 19661 SHORT PAPERS 657 (instead of 10 ml) of pH 6 phosphate buffer solution to extract the penicillin from the chloroform. With the modified procedure, it is also necessary to double the volumes of sample, chloroform and buffer solution used in the extraction t o provide sufficient solution for the analysis.When accurately prepared solutions of benzylpenicillin were assayed by this technique, the recovery was about 70 per cent. It seemed possible that the low recovery was caused by incomplete extraction of the benzyl- penicillin from the chloroform and that the recovery might be improved by using a buffer solution of higher pH. The effect of pH on the partition of benzylpenicillin between the chloroform and the buffer solution was therefore investigated by shaking 3.0 ml of phosphate buffer solution containing 0.12 unit of benzylpenicillin with 80 ml of chloroform that had been equilibrated with pH 2.5 phosphate buffer solution, as in the extraction procedure. Experiments were carried out with buffer solutions of pH 6.0, 7-0, 7.5 and 8.0, and with both analytical-reagent grade chloroform and alcohol-free chloroform, with the results given in Table I.Each recovery is the mean of four independent observations. TABLE I EFFECT OF pH ON THE PARTITION OF BENZYLPENICILLIN BETWEEN pH of buffer solution . . .. . . .. . . 6-0 7.0 7.5 8.0 solution, per cent. (analytical-reagent grade chloroform) 77 83 94 94 solution, per cent. (alcohol-free chloroform) . . .. 91 103 103 95 CHLOROFORM AND BUFFER SOLUTION Mean recovery of residual benzylpenicillin in the buffer Mean recovery of residual benzylpenicillin in the buffer The use of alcohol-free chloroform and pH 7.5 phosphate buffer solution gave a satisfactory partition of benzylpenicillin into the aqueous phase; when these reagents were used in the assay of accurately prepared solutions of benzylpenicillin, the mean recovery was 96 per cent.(Table 11). Application of the concentration technique to the supernatant liquid from Propyliodone Injection enabled the material to be tested for compliance with the requirements of the Food and Drug Administration. TABLE I1 RECOVERY OF BENZYLPENICILLIN FROM AQUEOUS SOLUTION Prepared strength of benzylpenicillin, unit per ml 0*0084 0*0082 0.0079 0.0079 0-0042 0.0041 0.0039 0.0039 0.002 1 0.002 1 0*0020 0.0020 Recovery of benzylpenicillin, per cent. 96 10s 100 101 103 90 104 100 90 87 89 89 METHOD SPECIAL PRECAUTIONS- These are essentially as described by the Food and Drug Administration. Carry out the assay in an atmosphere as far removed as possible from the likelihood of penicillin contamination.The operator should avoid visiting areas that contain penicillin dust. Swab down benches with penicillin-free water before use and cover them with clean waxed paper or similar material. Thoroughly clean all glassware and equipment. Immediately before use, rinse the apparatus with penicillin-free water and penicillin-free acetone, and dry rapidly in a current of air. Alterna- tively, the apparatus may be heated in an oven at 200" to 220" C for 4 hours to destroy penicillin traces, as recommended by the Food and Drug Administration. Carry out the extraction quickly, and do not expose samples or solutions to the air for longer than is necessary to perform the operation. Sample solutions should be assayed as soon as possible658 SHORT PAPERS [Analyst, Vol.91 after preparation to minimise loss of penicillin ; satisfactory results have been obtained, however, with sample solutions and benzylpenicillin standard solutions that had been stored a t 0" C for several hours before microbiological analysis. Carry out a blank determination with each set of samples as a check on chance contamination, and carry out regular recovery experiments to confirm satisfactory performance. REAGENTS- Buffer solution, p H 7-5-Dissolve 9.78 g of analytical-reagent grade disodium hydrogen orthophosphate (anhydrous) and 1.85 g of analytical-reagent grade potassium dihydrogen ortho- phosphate in sufficient water to produce 1 litre. Sterilise in an autoclave a t 121" C (15 p.s.i,) for 20 minutes.Chlovo form, alcohol-free-Shake analytical-reagent grade chloroform with an equal volume of penicillin-free water and reject the aqueous extract. Repeat the extraction with three further volumes of penicillin-free water. Prepare immediately before use. EXTRACTION OF THE PENICILLIN- Transfer 80 ml of the Propyliodone Injection to a 100-ml centrifuge tube, cover the tube with aluminium foil, and spin it in a centrifuge for 20 minutes a t 3000 g. Transfer 20-0 ml of the super- natant liquid to a 100-ml separator, and add 40 ml of alcohol-free chloroform and 2.6 ml of N hydrochloric acid (the pH of the solution should be below 3; check the pH with indicator paper after the extraction with chloroform has been completed). Immediately shake the separator for 15 seconds, set aside to separate, and run the chloroform into a second 100-1n1 separator.Do not allow any of the aqueous phase to run through with the chloroform. Repeat the extraction with 40 ml of alcohol-free chloroform. Combine the chloroform extracts, and immediately shake the chloroform solution for 30 seconds with 3.0 ml of pH 7.5 buffer solution. Transfer the aqueous solution to a suitable tube and seal the tube with aluminium foil. DETERMINATION OF THE PENICILLIN- Determine the penicillin content of the extract microbiologically by cylinder plate assay with a double-layer medium and Savcina lutea ATCC 9341 as test organism, as describedl in the Food and Drug Administration Assay 11. Large plates3 9 4 may be used instead of Petri dishes, with the cylinders, double-layer medium and test organism as recommended by the Food and Drug Administration.Suitable volumes of base layer and seed layer are 125 ml and 100 ml, respectively, for plates measuring 12 by 12 inches. Solutions containing 0.0125, 0.025, 0.05 and 0.1 unit of benzylpenicillin per ml are suitable standards for large-plate assay of Propyliodone Injection containing up to 0-2 unit of benzyl- penicillin per 18 ml. CALCULATION OF THE PENICILLIN CONTENT- Calculate the penicillin content of the Propyliodone Injection from a calibration graph prepared with the solutions of known benzylpenicillin concentration: 18.0 ml of 50 per cent. w/v Propyliodone Injection contains 13.8 ml of supernatant liquid. RESULTS The method was evaluated with 8 x 8 Latin square large-plate assays accommodating standards a t four levels and four samples a t one level; the mean recovery of benzylpenicillin added to penicillin-free Propyliodone Injection was 79 per cent.(Table 111). The possibility of some loss occurring during the extraction of penicillin from drugs has been recognised by the Food and Drug Administration,l and the recovery is considered satisfactory in view of the minute amounts of benzylpenicillin involved. The presence of penicillin in a sample can be confirmed by penicillinase inactivation of the extract if desired,l but this is not necessary in the routine examination of samples that comply with the Food and Drug Aministration requirements. The laboratory in which the work was carried out had been used for the analysis of penicillin products for several years.The ceiling and walls were thoroughly washed down before work was begun, and penicillin products were subsequently excluded as far as possible. The chances of contamination appeared to be reduced to acceptable proportions by all the precautions taken, as over fifty blank determinations were carried out during the investigation, and only one of them showed penicillin activity.October, 19661 SHORT PAPERS 659 TABLE I11 RECOVERY OF BENZYLPENICILLIN FROM PROPYLIODONE INJECTION Prepared strength of benzylpenicillin, unit per 18 ml 0.1 15 0.1 15 0-115 0.107 0.057 0.056 0.056 0,055 0-028 0.028 0.027 0.027 Recovery of benzylpenicillin, per cent. 74 73 72 87 73 69 53 9s 78 81 82 82 The use of a “clean” room and stringent precautions have been advocated for the determina- t i ~ n .~ This may prove to be the ideal solution to the problem of chance contamination when large numbers of assays are involved, but our experience suggests that it is possible to carry out this type of work in an ordinary laboratory if suitable techniques are used. We thank Mr. S. Varsanyi and Mrs. M. Arnold for carrying out the microbiological assays. REFERENCES 1. Wilner, J , , A4rret, B., Selzer, G. B., and Kirshbaum, A., “Tentative Procedures for Detecting and Measuring Penicillin Contamination in Drugs, ” Department of Health, Education and Welfare, U.S. Food and Drug Administration, Washington, D.C., February, 1965; Addendum 11, March, 1965; Addendum 111, April, 1965; Addendum IV, June, 1965. 2. 3. 4. 5. Larrick, G. P., Federal Register, 1965, Title 21-Food and Drugs, 932. Lees, I<. -4., and Tootill, J. P. R., Analyst, 1955, 80, 95. Lightbown, J. W., and Sulitzeanu, D., Bull. Wld Hlth Org., 1957, 17, 553. Elias, W., “Text Book on Quality Control in Drug Production from FDA Industry Papers during University of Wisconsin Seminar,” F-D-C Reports, Inc., 1152 National Press Building, Wash- ington, D.C., 1965, p. 63. Received April 21st, 1966

ISSN:0003-2654    

DOI:10.1039/AN9669100656

出版商:RSC    

年代:1966

数据来源: RSC

摘要:

October, 19661 SHORT PAPERS 659 The Spectrophotometric Determination of Trace Amounts of Tantalum BY JOHN H. HILL (Law,rence Radiation Laboratoyy, University of CaEifor.Pzia, Livernzore, California) TANTALUM is now being used in many alloys. Because materials which make contact with such alloys may be contaminated by the tantalum, there is a demand for a method to determine this element in the 0 to 50-pg range. The lower limit for the spectrophotometric determination has been about 50 pg because colorimetric reagents for tantalum were insensitive below that level. Luke1 recently found that 2,3,7-trihydroxy-9-phenyl-6-fluorone (phenylfluorone) is a very sensitive reagent for tantalum. The spectrophotonietric method which he developed is satisfactory for the determination of 4Opg or more of the element.For smaller amounts, his method is inadequate because tantalum is lost during the evaporation and fuming steps that are required to expel the isobutyl methyl ketone and excess of fluorides. Luke’s work showed that phenylfluorone is sensitive enough to determine tantalum in the 0 to 50-,ug range. He also showed that small amounts of tantalum can be separated from many other elements by extracting with isobutyl methyl ketone. Further investigation in this laboratory of the isobutyl methyl ketone extraction procedure led to the development of this method for determining 2 to 50pg of tantalum. By this developed method, tantalum is extracted from hydrochloric acid - hydrofluoric acid solution into isobutyl methyl ketone and back-extracted into buffered ethylenediaminetetra- acetate (EDTA) solution.Phenyl The excess of fluoride is complexed with aluminium(II1).660 SHORT PAPERS [Analyst, VOl. 91 fluorone is then added to form the coloured tantalum - phenylfluorone complex, and its absorbance is measured at 530mp on a spectrophotometer. EXPERIMENTAL APPARATUS- Teflon plug stopcocks. solutions in these flasks. Extraction flasks were made from 25 x 150-mm Pyrex test-tubes fitted with close-coupled Electric stirrers fitted with glass stirring rods were used to mix the REAGENTS- Aluminium chloride solution-Dissolve 70 g of aluminium chloride (AlCl,. 6H,O) in water and dilute to 1 litre with water. ED TA solution-Dissolve 100 g of disodium ethylenediaminetetra-acetate in 1 litre of water and filter the solution. Bufler solution-Mix 500 g of ammonium acetate, 700 ml of glacial acetic acid and 500 ml of water.Add a solution of 2 g of benzoic acid in 20 ml of methanol and dilute to 2 litres. Dissolve 1 g of gelatin in 1 litre of hot water, cool and mix the two solutions. Adjust the pH to 4.5 with ammonia solution or acetic acid. Phenylfluorone solzition-Dissolve 0.010 g of phenylfluorone in 4 to 5 drops of 12 M hydro- chloric acid and 10 ml of ethanol. Dilute to 100 ml with ethanol. Isobutyl methyl ketone-Add a few pellets of sodium hydroxide to isobutyl methyl ketone and re-distil, discarding low-boiling and high-boiling fractions. Hydrofluoric acid solution, 4.0 M-Dilute 16.26 g of 48 per cent. hydrofluoric acid to 100 ml with water. Store the solution in a polyethylene container. Standard tantalum solution-Add 10 ml of hydrofluoric acid to 250 mg of tantalum in a Teflon beaker.Add 30ml of hydrofluoric acid and dilute to about 200 ml with water. Transfer to a 250-ml calibrated flask and dilute to volume. The final concentration of hydrofluoric acid is about 4 M. Store in a poly- ethylene container. Prepare solutions more dilute as required. The more dilute solutions should be about 0 . 4 ~ with respect to hydrofluoric acid. Prepare daily. Add nitric acid dropwise and warm gently until solution is complete. PROCEDURE- Prepare the sample in hydrochloric acid or hydrochloric acid - hydrofluoric acid solution. Transfer an aliquot containing 2 to 50 pg of tantalum to an extraction flask. Add 4.0 M hydro- fluoric acid and 12 M hydrochloric acid and dilute to 10 ml with water.The final concentration should be 0-40 XI with respect to hydrofluoric acid and 6.0 M with respect to hydrochloric acid. Add 5.00 ml of isobutyl methyl ketone and mix for 5 minutes. Allow the phases to separate. Drain off and discard the aqueous phase. Add 0.50 ml of 4.0 M hydrofluoric acid and 2.5 ml of 1 2 ~ hydrochloric acid, and dilute the aqueous phase to 5-Oml. Stir for 15 to 20 seconds and allow the phases to separate. Drain off and discard the aqueous phase. Add 2 ml of water, drain and discard ; repeat twice to rinse the acid out of the stopcock. Add 5.0 ml of buffer solution and stir for 5 minutes. Add about 1 ml of ethanol to break up the emulsion and allow the phases to separate. Collect the aqueous phase in a 25-ml calibrated flask.Wash the isobutyl methyl ketone twice with small portions of buffer solution and add the washings to the calibrated flask. Then add by pipette 3-00 ml of aluminium chloride solution and 5-00 ml of phenylfluorone solution to the calibrated flask, mixing thoroughly after each addition. Dilute to volume with buffer solution, mix, and allow to stand for 1 hour. Transfer to a 1-cm absorption cell, and measure the absorbance at 530mp on a spectrophotometer, with water as a reference. Add the reagents in the order given and carry reagent blanks and standards through the entire procedure. Clean the absorption cells immediately after use to prevent etching by the hydrofluoric acid. Transfer by pipette 5-00 ml of EDTA solution to the extraction flask.DISCUSSION The values for the graph were obtained by analysing aliquots of a standard tantalum solution that contained 2.5, 10, 30 and 50,ug of tantalum. Results for 15 determinations of each amount are shown A calibration graph was prepared by using the procedure described above.October, 19661 SHORT PAPERS 661 in Table I. These results indicate the precision of the method. In addition, direct colour develop- ment of aliquots of the standard tantalum solution gave absorbance values that lie on the calibra- tion graph. Therefore, the tantalum extractions are essentially complete and the accuracy of the method is comparable to its precision. TABLE I PRECISION OF DETERMINATION OF TANTALUM llg Absorbance* per cent. Tantal um , Relative standard deviation, 2.48 0.054 7.2 9.93 0.175 4.9 29.83 0.477 2.9 49.65 0.756 2.7 * Average of 15 determinations.The formation of the coloured tantalum - phenylfluorone complex is dependent on the con- centration of the reagents. Therefore, the calibration graph may be displaced when a new reagent solution is used. A change in the buffer solution has the greatest effect. All the results given above were obtained with the same solutions, except phenylfluorone. Phenylfluorone solution was prepared daily because the sensitivity of the reagent decreases with prolonged standing. The use of fresh phenylfluorone had no effect on the calibration graph. These competing equilibria probably account for the concavity of the absorption curve in the range from 2 to 50 pg of tantalum.To obtain the best results, calculations should be based on standards carried through the entire procedure, along with the unknown samples. However, results based on a calibration graph will be satisfactory for most purposes, provided that samples are analysed with the same reagent solutions as those used to obtain the graph. Phenylfluorone competes with EDTA, fluoride and acetate for the tantalum ions. DETERMINATION Metal salt Antimony(II1) chloride Boric acid . . . . Calcium chloride . . TABLE I1 O F TANTALUM I N THE PRESENCE OF FOREIGN IONS .. .. .. Cerium@) hydrogen suIphate Caesium chloride . . .. Copper(I1) chloride . . .. Mercury(I1) chloride . . .. Niobium(v) fluoride . . .. Molybdenum(v1) chloride . . Potassium chloride . . .. Silver nitrate . ... .. Sodium chloride . . .. Tin@) chloride. . .. . . Titanium(Ir1) chloride . . .. Tungsten(v1) fluoride . . . . Uranium hexachloride (UCl,) Vanadium oxychloride (VOCl,) Tantalum taken, 9-93 pg .. .. .. . . . I .. .. .. .. .. .. .. .. .. .. .. .. f Weight of metal ion added, mi3 20 1 9 10 1 25 20 10 1 50 100 0.1 0.250 0.050 0.0 10 100 10 10 100 10 100 100 50 50 100 50 50 0.1 1 Tantalum recovered, Difference, CLg CLg 13.18 + 3.25 9.90 - 0.03 10.43 + 0.50 9-16 - 0.77 9.75 -0.18 9-73 - 0.20 10-33 + 0.40 10.63 + 0.70 9-75 -0.18 10.22 + 0.29 absorbance too high to read 9-93 0.00 14.50 + 4.57 9.93 0.00 10.36 + 0.43 8.82 - 1-11 9.87 - 0.06 9.93 0.00 9.09 - 0.84 10.15 3.0.22 absorbance too high to read 0.21 + 0-28 10.80 + 0.87 10.32 + 0.39 10.41 + 0.48 9-87 - 0.06 10.22 + 0.29 9.65 - 0.28662 SHORT PAPERS [Analyst, VOl.91 INTERFERENCES- The behaviour of 33 foreign ions was determined by adding known amounts of each ion to 1Opg of Ta(V) and analysing as above. No interference was encountered from 100-mg amounts of the following 15 cations: As(III), Ba, Cd, Cr(III), Co(II), In(III), La, Pb(II), Li, Mg, Mn(II), Ni(II), S, Zn and Zr. Although 100 mg of Fe(II1) caused interference, it was easily eliminated by reduction with hydroxylammonium chloride before colour development. Only Mo(VI), Nb(V) and Sn(I1) interfere in amounts of less than 1 mg. There is no interference from 100 pg of Mo(V1) or Sn(I1) and niobium does not interfere in 50-pg amounts. In larger amounts, Mo(VI), Sn(I1) and Nb(V) extract and form coloured phenylfluorone complexes that absorb at 530mp. The extent of interference encountered with Mo(VI), Sn(II), Nb(V) and various concentrations of 14 other ions is shown in Table 11, and can be evaluated statistically by reference to Table I. This is a rapid analytical procedure for the determination of 2 pg or more of tantalum. For 2-5 to 50 pg of tantalum, the coefficient of variation ranges from 7.9 to 2.7 per cent. An experienced analyst can perform about 30 determinations in 8 hours. The method should be applicable to at least 16 different matrices. This work was performed under the auspices of the U.S. Atomic Energy Commission. REFERENCE 1. Luke, C . , Analyt. Chem., 1959, 31, 904. Received July 5th, 1965

ISSN:0003-2654    

DOI:10.1039/AN9669100659

出版商:RSC    

年代:1966

数据来源: RSC

摘要:

662 SHORT PAPERS [Analyst, VOl. 91 The Organic-phase Spectrophotometric Determination of Iron with Thioc yanate BY E. CERRAI AND G. GHERSINI (Laboratori C.I.S.E., Casella Postale 3986, Milano, Italy) AT the S.A.C. Conference (Nottingham, 1965) we gave a lecture1 on the organic-phase spectrophoto- metric determination of iron( 111) with thiocyanate after extraction with di-( 2-ethylhexy1)ortho- phosphoric acid. In this procedure iron is extracted from a 0.1 M hydrochloric acid solution with a 0.5 M solution of di-(2-ethylhexyl) orthophosphoric acid in cyclohexane, and colour is developed in the organic phase by adding an equal volume of a solution containing 50 mg of potassium thiocyanate per ml of 95 per cent. ethanol. The optical density of the red colour is measured at 477 mp.As the molar extinction coefficient of the iron - thiocyanate complex thus obtained is 10,400, this method appears to be one of the most sensitive for the colorimetric determination of iron(III), and is a t the same time rapid and simple. As reported in the published paper, the presence of 47 foreign cations in the initial aqueous solution was studied, but in the commentary presented a t the Conference, the effect of 22 addi- tional cations was also discussed, and a comparison was drawn between this method and the normal one based on the aqueous phase iron - thiocyanate colorimetry. These results are reported in this paper. Table I gives the results that were obtained when 10 micro-equivalents per ml of each of the 22 foreign cations were present in the aqueous phase from which iron was extracted.The standard chloride medium procedure was followed, the only deviations from this method being for silver(1) and mercury(I), when the iron extraction was carried out from a 0.1 M nitric acid solution. The extraction step is e ~ p e c t e d ~ 3 ~ to separate iron(II1) from vanadium(III), vanadium(I\’), rhenium(VI), ruthenium(III), rhodium(III), iridium(IV), platinum(IV), silver(I), mercury(I), mercury(II), germanium(IV), arsenic(III), arsenic(V), selenium(1V) and tellurium(V1). If a fl per cent. deviation is considered to be within experimental error, most of these cations should not interfere. This actually happens except for ruthenium(III), rhodium(II1) and mer- cury(l), which probably interfere because traces of them pass into the organic phase.Furthermore, platinum(1V) was found to interfere, which it did not in the aqueous phase medium determination, and this could not be explained.October, 19661 THE DETERMINATION OF Metallic ion added Scandium(rI1) Vanadium (111) Vanadium(1v) Niobium(v) . . Rhenium(v11) Ruthenium ( 111) Rhodium(II1) , . Osmium(vII1) Iridium(1v) . . Platinum(1v) . . Silver(I)* . . Mercury(I)* . . Mercury(I1) . . Germanium(1v) Lead(@ . , Arsenic(II1) . . Arsenic(v) . . Antimony(v) . . Bismuth(II1) . . Selenium(1v) . . Selenium(v1) . . Tellurium(v1) .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. * . .. .. .. .. .. * . SHORT PAPERS TABLE I IRON I N THE PRESENCE OF FOREIGN IONS Weight ratio of Error per cent. foreign ion to iron in iron before extraction determination .... .. .. .. .. .. .. .. .. .. .. .. .. a . .. .. * . .. .. .. .. 12 14 11 16 22 25 28 20 40 41 90 167 83 15 86 21 12 20 58 17 11 17 - 33 + 0.7 + 0-6 - 77 - 0.2 + 23 - 13 + 26 - 0.5 - 15 - 0.1 - 7 + 0.7 - 1.2 + 9 0 0 + 0.8 + 21 - 0.7 + 12 + 0.2 663 * In nitrate medium. The initial solution was 10 ml of 0.1 M hydrochloric acid containing between 11.9 and 14.3 pg of iron per ml and 10 micro-equivalents per ml of foreign ion. The extraction was made with 10 ml of 0.5 M di- (2-ethylhexyl)orthophosphoric acid in cyclohexane, and the iron was determined by a standard colorimetric procedure in 5 ml of the organic solution. The metals were added as the chlorides. Silver(1) and mercury(1) interferences were determined after extraction from 0.1 M nitric acid.All other elements listed in Table I are partially, or completely, extracted by di-(2-ethyl- hexyl) orthophosphoric acid in the conditions considered2 '3; among them, only antimony(V) does not interfere in the determination of iron. Below is a list of the cations, from those considered in both this paper and the previous lecture, that interfere in the aqueous medium colorimetry of iron(II1) as the thiocyanate complex. The interference of some of them is reported by Sandell*; for the others, determinations were carried out in our laboratory, following the procedure described by the same a ~ t h o r . ~ A cation was considered to interfere when its presence in the amount of 200 micro-equivalents per ml affected the determination of 9.9 pg per ml of iron with an error higher than 5 per cent.The bold printed cations are those that do not interfere when iron(II1) is first extracted with di-(2-ethylhexyl)- orthophosphoric acid and its colorimetric determination is carried out in the organic phase. Be(II), Ca(II), Sr(II), Sc(III), Y(III), Th(IV), U(VI), Ti(IV), Zr(IV), Hf(IV), Nb(V), Cr(III), Mo(VI), W(VI), Mn(II), Re(VI), Co(II), Ru(III), Rh(III), Pd(II), Os(VIII), Ir(IV), Cu(II), Ag(I), Zn(II), Cd(II), Hg(I), Hg(II), In(III), Ge(IV), Pb(II), As(III), As(V), Sb(V), Bi(III), Te(V1). From this list it can be seen how this organic-phase colorimetric determination of iron(II1) not only makes the direct colorimetric determination of this element in the extraction processes possible, but also enables many of the interferences in the conventional potassium thiocyanate method to be avoided. The determination of iron in an analogous organic phase by the use of bathophenanthroline after the reduction of Fe(II1) to Fe(I1) (as suggested in Nottingham by Mr. B. Braithwaite) is now under investigation in our laboratory, with favourable results, and it is hoped that it will be the subject of a future publication. Rare earths from Pr(II1) to Lu(II1).664 SHORT PAPERS [Analyst, VOl. 91 REFERENCES 1. 2. Kimura, K., Bull. Chem. SOC. Jafian, 1960, 33, 1038. 3. Cerrai, E., and Ghersini, G., J . Chromat., in the press. 4. Sandell, E. B., “Colorimetric Determination of Traces of Metals,” Interscience Publishers, New 6. Cerrai, E., and Ghersini, G., in Shallis, P. W., Editor, “Proceedings of the S.A.C. Conference, Nottingham 1965,” W. Heffer & Sons Ltd., Cambridge, 1965, p. 462. Yorlr and London, 1959, p. 530. - , op. cit., p. 534. Received February 2nd, 1965

ISSN:0003-2654    

DOI:10.1039/AN9669100662

出版商:RSC    

年代:1966

数据来源: RSC

摘要:

664 SHORT PAPERS [Analyst, VOl. 91 A Direct Photometric Procedure for the Determination of Boron in Nickel BY T. R. ANDREW AND P. N. R. NICHOLS (Central Materials Laboratory, The Mullard Radio Valve Company, New Road, Mitcham Junction, Surrey) NICKEL supplied for use in the manufacture of electronic devices is required to have a very low boron content. (Sometimes a maximum of 5 p.p.m. may be specified.) The two well known published procedures for this determination both require the separation of boron from the nickel; that of Chirnside,l by mercury-cathode electrolysis; and that of by distillation. Both procedures have a lower limit of determination of about 0.5 to 1 p.p.ni. of boron in the sample, (100 to 500 mg). With the publication of the reappraisal of the boron - curcumin reaction by Hayes and Met~alfe,~ it seemed to us that i t should be possible to eliminate these separation stages, provided that the sample could be presented in a suitable non-oxidising solution containing less than 0.25 ml of water.Nickel requires an oxidising acid for rapid solution, and we decided to work with perchloric acid which, although strongly oxidising when hot, is not an oxidising agent a t room temperature, except when in contact with strongly reducing substances. Preliminary calculations suggested, and experiments confirmed, that 100 mg of nickel could be readily dissolved in 0.5 ml of 72 per cent. perchloric acid (0.14 ml of water). The investigation was, therefore, divided into 3 sections- (i) establishing that the procedure of Hayes and Metcalfe3 can be used in the presence (ii) establishing that the presence of 100mg of nickel has no effect; and of 0.5 ml of 72 per cent.perchloric acid; (iii) establishing that no significant amount of boron is lost during the solution procedure. Hayes and Metcalfe’s procedure3 requires that the boron is in a neutral, or almost neutral, solution containing not more than 0.25 ml of water. To this is added a solution of curcumin in glacial acetic acid, followed by a mixture of sulphuric and glacial acetic acids. Under these conditions both the boron - curcumin complex (rosocyanin) and protonated curcumin are formed, the former taking a few minutes to form, the latter being formed immediately. On dilution with ethanol the protonated curcumin is destroyed leaving the rosocyanin unaffected.It was possible that the presence of high concentrations of perchloric acid when the curcumin reagent was added might affect the reaction. To test this, experiments were made in which amounts of up to 5 pg of boron, as boric acid, contained in 1 ml of glacial acetic acid were added to 0.5 ml of 72 per cent. perchloric acid and treated with curcumin - acetic acid and sulphuric acid - acetic acid as described by Hayes and Met~alfe.~ The absorption spectrum of the final solution agreed with that given by Hayes and Metcalfe, as did the sensitivity of the calibration graph at 555 mp. It was, therefore, concluded that the addition of 0.5 ml of 72 per cent. perchloric acid did not adversely affect the procedure. To minimise loss of boron during solution, i t was proposed to use as a container in which to dissolve the sample, a long, narrow-bore, silica tube that would also act as an air condenser.When 100-mg amounts of nickel were dissolved in 0-5 ml of 72 per cent. perchloric acid in such an apparatus it was found that the clear green liquid rapidly solidified on removal from the source ofOctober, 19661 SHORT PAPERS 665 heat. The addition of 1 ml of glacial acetic acid to this solid mass, followed by gentle heating of the mixture usually gave a clear solution, but sometimes heating of the nickel perchlorate gave some decomposition to the oxide and, consequently, a turbid solution. When the solution tube was inverted on removal from the source of heat, the nickel solution solidified in a thin layer along the tube, and dissolved readily in 1 ml of glacial acetic acid with no decomposition to oxide. When such solutions were added to 0.5 ml of 72 per cent.perchloric acid containing amounts of up to 5 p g of boron, the recoveries obtained were very low, and the red protonated curcumin complex rapidly disappeared. This attack on the protonated curcumin suggested some chemical reaction, rather than a masking of the boron - curcumin reaction, and we considered possible chemical interferences that could have been generated during the solution procedure. As Hayes and Metcalfe3 refer specifically to the effect of oxidants, we reviewed the possibility of generation of chlorine while forming the solution. This could occur by a reaction between the nickel and hot perchloric acid to give some nickel chloride, and the formation of chlorine from reaction between this generated chloride and hot perchloric acid.Several reagents were tried in an attempt to overcome this interference, and the most satis- factory of those studied was a solution of phenol in acetic acid, used in place of the glacial acetic acid to dissolve the cooled nickel salts. With this modification good recovery was found for up to 5 p g of added boron. Having established that the presence of nickel and perchloric acid was not detrimental to the performance of the procedure, there remained the need to verify the recovery of boron carried right through the procedure. In the absence of samples of nickel of known boron content, a series of tests was carried out in which aliquots of a solution of boric acid in 72 per cent.perchloric acid were added to 100-mg portions of nickel, and 72 per cent. perchloric acid added to a total volume of 0-5 ml. TABLE I EFFECT OF REACTIOK TIME ON COLOUR FORMATION 100 mg of nickel and 2 pg of boron Reaction time, Optical density, minutes I-cm cells, 555 mp 5 0.1 14 10 0.205 15 0.288 30 0.400 60 0.400 Itecoveries of boron were excellent by this procedure, and the method was felt to be acceptable for trial. The procedure described by Hayes and Metcalfe3 was known to be sensitive to smallchanges in water content, too high a content giving a more sluggish reaction and finally resulting in poor recovery. It was, therefore, thought desirable to ascertain if the standing time of 30 minutes, used in the preliminary studies above, was adequate for full development, or if it could be reduced to any worthwhile degree.The results of this study are shown in Table I. From these results, it can be seen that 30 minutes’ standing time is adequate, and could not be reduced to any marked extent. As the formation of the coloured rosocyanin takes place only under strongly acid conditions, it was possible to evaluate the extent of the contribution of nickel salts to the over-all absorption at 555 nip by adding most of the ethanol before adding the curcumin - acetic acid reagent, under which conditions no colour is produced by any boron that may be present. With 4-cm cells, the optical density with no nickel present is 0.060, and with 100 mg of nickel present is 0.088.The absorption arising from 100 mg of nickel is equivalent to that of 0.01 pg of boron or 0-1 p.p.m. of boron on the nickel sample, and may, therefore, be neglected for most purposes. A series of trials was carried out on two “standard” samples for which approximate boron figures were known, a tungsten - nickel and a sample of cathode nickel from current production.666 SHORT PAPERS [Analyst, Vol. 91 The results are shown in Table 11. TABLE I1 RESULTS ON NICKEL SAMPLES Sample Nominal boron value, Boron found p.p.m. A . . .. .. . . 25 25.6; 26.2 B . . . . . . . . 5 8.2; 8-6 C (4 per. cent. tungsten) . . - 1.5; 1.8 Current cathode nickel .. - 1.1; 1.3 The reproducibility of the method is clearly adequate and its application to tungsten - nickel was checked by confirming that the calibration graphs in the presence of 100 mg of nickel and 100 nig of 4 per cent.tungsten - nickel were identical, after removal of tungsten(V1) oxide by filtration of the final ethanol solution through a dry Whatman No. 540 filter-paper. The blank values found by this procedure are very low (about 0-005 to 0.01 pg of boron), possibly due to the absence of any alkaline stage, and ordinary AnalaR chemicals have been found satisfactory. METHOD REAGENTS- Yerchloric acid, 72 per cent. Phenol, 10 per cent. w/v in glacial acetic acid-Prepare daily, and store in boron-free glassware. Cuycwnin, 0.125 per cent. in glacial acetic acid-Prepare daily, and store in boron-free glassware. Sulphuvic acid - acetic acid (1 + 1)-Mix 50 ml of sulphuric acid (sp.gr.1.84) and 50 ml of glacial acetic acid and allow the solution to cool. Prepare daily, and store in boron-free glassware. Indztstrial methylated spirit. APPARATUS- Silica lubes-7 to 8 mm x 300 mm, closed at one end. Graduated pask, 100 ml. Spectrophotometer OY filter photometer-For use a t 555 mp. Transfer 100 ml of the sample to a clean, dry, silica tube and add 0.50 ml of perchloric acid. Support the tube loosely in a vertical position with the closed end resting in a bed of sand When the sample is dissolved remove the tube from the source of heat and invert in a 100-ml It will be found that the sample solution will flow down the inside of the tube Reverse the position of the tube in the flask and add, from a pipette, 1.0 ml of 10 per cent. Remove the tube from the flask and dissolve the nickel salts in the acetic acid - phenol mixture PROCEDURE- contained in a nickel crucible heated by a bunsen burner.graduated flask. and solidify before reaching the open end. phenol in acetic acid. by very gently warming the mixture over a bunsen burner. NOTE- No explosion hazard exists, but the heating should not be strong, nor so prolonged, as to cause any blackening of the green solution. When the sample is dissolved, transfer the solution to the 100-ml flask by inversion and allow the tube to drain. After cooling the tube for 5 to 10 minutes, add to the tube 3-0 ml of 0.125 per cent. curcumin in acetic acid. Transfer the solution to the flask as before, and add to the tube 3.0 ml of the acetic acid - sulphuric acid mixture.Transfer the solution to the graduated flask, swirl it well and leave it to stand for 30 minutes. Dilute the solution to volume with industrial methy- lated spirit, mix it well and measure the optical density at 555 mp in l or 4-cm cells against water. A blank, omitting nickel, should be carried through the procedure. For tungsten-containing samples the ethanol solution should be filtered through a dry Whatman No. 540 filter-paper before measurement. cALIBR.4TION- Dissolve 0-114 g of boric acid in 10 ml of water and dilute the mixture to 100 ml with 72 per cent. perchloric acid. Transfer by pipette 5 ml of this solution into a 100-ml graduated flask and dilute to volume with 72 per cent. perchloric acid. (1 ml of solution = 10 pg of boron.)October, 19661 SHORT PAPERS 667 Into each of 6 sample tube:; weigh 100 mg of low-boron nickel, and add 0, 0.1, 0.2, 0.3, 0.4 and 0.5 ml of dilute boron solution. Add to each solution sufficient perchloric acid to make the total addition 0-5 ml, and carry through the procedure. For work at very low levels of boron content, a calibration with 0 to 1 pg of boron with 4-CIn cells is desirable. Preliminary studies indicate that this procedure is applicable to other materials soluble in perchloric acid. We thank Mr. C. H. R. Gentry, Head of the Central Materials Laboratory, and the Directors of Mullard Limited for permission to publish this paper. REFERENCES 1. 2. 3. Chirnside, R. C., Cluley, H. J., and Proffitt, P. M. C., Analyst, 1957, 82, 18. Luke, C. L., Analyt. Chem., 1958, 30, 1405. Hayes, M. R., and Metcalfe, J., Analyst, 1962, 87, 956. Received May Zdth, 1966

ISSN:0003-2654    

DOI:10.1039/AN9669100664

出版商:RSC    

年代:1966

数据来源: RSC

摘要:

October, 19661 SHORT PAPERS 667 A Modified Potentiostat for Controlled Potential Analysis BY J. GRIMSHAW AND R. K. QUIGG (Department of Chemistry, Queen’s University, Beljast, Northern Ireland) PALMER and Vogel’s apparatus1 seemed to us to have some advantages over other p o t e n t i o ~ t a t s , ~ ~ ~ particularly when large power outputs may be required. However, the instrument has some unsatisfactory operating features that we have tried to remove. The marketing of an improved design of variable voltage transformer has made i t possible to eliminate one undesirable feature, and we have designed an indicating circuit to show when the apparatus has adjusted the electrode potential to its pre-selected value. An instrument incorporating these two features and a few other minor modifications to the original circuit has functioned satisfactorily in our laboratories for 2 years.It was designed to deliver a maximum current of 10 amps, with a response time of 54 seconds for complete traverse of the variable voltage transformer. These two specifications can easily be varied over wide limits by choosing appropriate components for the d.c. power supply and the appropriate model of variable voltage transformer. The response time we chose is sufficient to give a tolerance of potential control of &0.003 volt during an experiment, once the correct cathode potential has been established. MODIFICATIONS The sections below list our modifications to Palmer and Vogel’s design. Their component numbers have been retained and extended to identify any additions that we have made.VARIABLE VOLTAGE TRANSFORMER- The need for switch S,, the 6-pin plug P and the motor-direction control unit has been eliminated by fitting a motor-driven variable-voltage transformer incorporating a slipping clutch between the motor and the transformer and having an extended shaft terminating in a knob on the front panel of the instrument for manual setting. This model, a standard product, is also fitted with 4 microswitches grouped around the transformer shaft; 2 are positioned to be operated a t 5 volts above zero and below maximum volts, and are wired.so that when operated they dis- connect power to the motor. The remaining 2 microswitches are mounted so that they are operated a t 10 volts above zero and below maximum volts, and are wired to 2 warning lamps, labelled LOWER VOLTS and RAISE VOLTS, respectively, which draw the operator’s attention to any need for re-adjustment of the coarse voltage control by means of manually selected taps on the secondary of T,.AMPLIFIER- The 2 twin valves V,, LT5 run rather warm even when the motor is not being driven, and the presence of the common cathode resistor Rd3 limits the maximum drive to the motor. To reduce the standing current in the motor-drive valves and thus minimise the heat generated while a t668 SHORT PAPERS [Analyst, Vol. 91 the same time permitting a heavy drive current to be passed to the motor so that it can respond rapidly when needed, we have introduced the following changes- V,, V, are now type EL86; they are given a negative d.c.bias (derived from the heater chain) of -25 volts and are, therefore, operating a t a low cathode current in the absence of a drive signal: R,, is changed to 100 ohms; it is retained only as a convenient monitoring point. Several improvements have been made to increase the amplifier gain. The anodes of V4 and V, are supplied by T, with alternately positive and negative voltage. When one pair is a t a maximum positive voltage i t is desirable, so as to give maximum drive to the motor, that the corresponding grids are a t the positive peak of the a.c. sine wave produced by the Carpenter chopper relay. There is no adjustment for this state in Palmer and Vogel's amplifier and, in fact, these voltages are about 20" out of phase because of the lag in the relay and phase shift in the amplifier.To overcome this we have, in our amplifier, driven the Carpenter relay from the 230-volt a.c. mains via a 0.04-pF capacitor, and made small adjustments to the coupling capacitors C,, C, and Cl0. These capacitors are selected while observing the grid and anode voltages of the output valves simultaneously on a double-beam oscilloscope. The adjustment is stable and not critical. We have also altered the values of R,, and R,, to 5600 ohms, giving improvement in gain and linearity to the amplifier. BALANCE INDICATING CIRCUIT- To balance the servo-amplifier and to standardise the reference voltage supply, Palmer and Vogel observed the cessation of rotation of the motor-driven variable transformer. We did not find this satisfactory, and we have provided in our apparatus a sensitive visual null indicator (Fig.1) involving the use of a miniature cathode ray tube. This cathode ray tube has its X-plates driven by a simple Miller time-base of 25 cyles per second locked to the a.c. mains. The Y-plates are fed from the anode of V3. The amplifier, a t maximum gain, is balanced by setting the trace on the cathode ray tube to a horizontal line by means of the balance control. Calibration of the reference voltage supply is easily carried out, even during a run, by switching quickly to calibrate and setting the trace horizontal, if necessary, by adjusting the set calibration control. The visual indication of balance is helpful during a run for checking that the potentiostat is holding the electrode to within a millivolt or so of the selected potential.An additional advantage is that the state of adjustment of the Carpenter relay is immediately evident. A well adjusted relay will give an approximately rectangular wave with equal on and off times. To display this wave form a small voltage is injected into the servo-amplifier by, for example, unbalancing t h e calibration control. + 300V + 5oov nhr Fig. 1. Visual balance indicator circuitOctober, 19661 SHORT PAPERS Appendix LIST OF COMPONENTS USED IN THE CONSTRUCTION OF THE APPARATUS Unless otherwise stated, all fixed resistors are 1-watt metal oxide type and all capacitors are polystyrene foil type 500 volt, d.c., working. Cathode ray tube = DH3-91 (Mullard cathode ray tube) Vl, = ECC83 valve v15 = EF91 valve R 5 8 p R6S = 33,000-ohm resistors R59 = 3.3-megohm resistor R60 = l-megohm potentiometer R 6 2 = 47,000-ohm resistor Rf34 = 680,000-ohm resistor Re6, R,,, R,, = 2.2-megohm resistors C,, = 0-03-pF capacitor c 2 7 = 0.001-pF capacitor C28 = 4700-pF capacitor c,,, c31 = 0.25-pF capacitors c30 = 0-l-pF capacitor Variable voltage transformer = 2200-ohm resistor = Berco Regavolt transformer, model M41A. 669 REFERENCES 1. 2. 3. Palmer, J. F., and Vogel, A. I., Analyst, 1953, 78, 428. Herringshaw, J. F., and Halfhide, P. F., Ibid., 1960, 85, 69. Wadsworth, N. J., Ibid., 1960, 85, 637. Received February 23rd, 1966

ISSN:0003-2654    

DOI:10.1039/AN9669100667

出版商:RSC    

年代:1966

数据来源: RSC

摘要:

October, 19661 SHORT PAPERS 669 Determination of Specific Gravity of Glass Particles by a Density Gradient Method BY S. S. KIND AND L. SUMMERSCALES (Home Office Forensic Science Laboratory, Haddon Lodge, 32 Rutland Drive, Harrogate, Yorkshire) THE traditional method of comparing the specific gravities of two glass particles in forensic science is to immerse them in a two-component liquid of uniform specific gravity and add the lighter or denser component until the two fragments suspend together, or until one floats and the other sinks. The same effect can be brought about by uniformly heating or cooling the entire bulk of the liquid. However, the “suspension” of glass particles in a uniform liquid tends to be an unstable equilibrium and, in practice, the glass particles will always end by floating or sinking, although the procedure will take considerable time.These techniques are capable of a high degree of accuracy, but they require a good deal of attention. Proposals have been made for the preparation of density gradients where the glass particles can be compared with each other and with coloured glasses of known specific gravity by noting the level to which a piece of glass sinks in a suitable specific gravity gradient. Such density gradient tubes can be prepared by layering mixtures of liquids of different specific gravities in a long vertically clamped tube.l Density gradients of this kind are tedious to prepare and suffer from the disadvantage that diffusion gradually destroys the gradient. We have lately used a simple method of preparing stable density gradients simply by irradiating a uniform column of liquid from the top with a tungsten filament lamp.A similar but much more complicated method has recently been proposed by Green and Burd.2 EXPERIMENTAL The apparatus and materials used consist of a 10-ml cylindrical measure containing a mixture of 85 per cent. bromoform and 15 per cent. benzyl alcohol, warmed from the top by a 100-watt bulb in an Anglepoise lamp. The system, although crude, is capable of swiftIy generating a stable density gradient, and if the initial specific gravity is adjusted by trial and error to a point at which the denser glass particle under examination just floats, then merely by switching on the670 SHORT PAPERS [Analyst, Vol. 91 lamp the particles will migrate downwards to take up positions that are related vertically to their specific gravities at the temperature of the particular layer.As expected, the gradient thus generated is not a linear one owing to the fact that the rise in temperature at any layer is due to the amount of heat absorbed at that layer, which follows a logarithmic rule. Thus, near the top where a steep gradient persists, glasses of different specific gravities will appear to suspend together, whereas further down the tube the gradient becomes sufficiently shallow for very slight differences in specific gravities to be detected. If coloured glasses of known specific gravities are added to the tube, then the specific gravity of an unknown sample suspended at the same time can be read off by interpolation. The positioning of the glass particles in the gradient can be varied by increasing or decreasing the distance of the light from the liquid surface.Distillation and concomitant change in specific gravity is not a problem because of the relatively non-volatile nature of the component liquids. Fig. 1. Positions of glass fragments in heat generated density gradient. Ordinate numbered in l-ml graduation marks of a 10-ml cylindrical measure Fig. gravities relate to samples of known specific 1 shows a logarithmic plot of the positions of four “marker” in one particular experiment at different time intervals. readings that were noted later than those for the steepest slope. The shallower sloping curves By using a 100-watt bulb and a 10-cylindrical measure (about 8 mm per ml) the average variation in specific gravity gradient can be 0-25 per cm for a position 1 mm below the surface to 0.0025 per cm at 6cm below the surface.The advantages of preparing a stable density gradient by this method are- (i) The gradient can be prepared when required merely by throwing a switch. (ii) No careful watching for a compensation point is required. (iii) A disturbed gradient, e.g., by removal of a glass particle, is automatically restored to (iv) The glass particles can be positioned on a steep or shallow part of the gradient merely by On switching off the lamp, the glass particles migrate upwards and the gradient changes to a more linear one. A more linear gradient can also be generated by a slower heating process. We have used the above system in case examinations over the past 18 months and have found it to be of great value, especially in the preliminary screening of glass particles, in enabling us to decide whether one or more types of glass are present in a sample, and to provide a swift method for comparing ‘‘control” and “crime” glass particles. a uniform one. changing the position of the lamp. REFERENCES 1. 2. Kirk, P. L., “Density and Refractive Index: Their Application in Criminal Identification,” C. C. Green, R. S., and Burd, D. Q., J . Foren. Sci., 1965, 10, 52. Received April 22924 1966 Thomas, Springfield, Illinois, 1951.

ISSN:0003-2654    

DOI:10.1039/AN9669100669

出版商:RSC    

年代:1966

数据来源: RSC

摘要:

October, 19661 SHORT PAPERS 67 1 An Injection Tap for Gaseous Samples in Gas Chromatography BY F. G. STAXFORD (The Radiochenzical Ceiztve, U.K. A fovlzic Enevgy Authority, Aazevshai~z, Rurlzs.) XKALYSIS of small monoxide, carbon device that could column Fig. 1. Injection tap samples of gases (0.2 to 1.0 ml), including carbon-14 labelled methane, carbon dioxide and acetylene, by gas chromatography has required an injection be filled in a laboratory remote from the analytical department, and then brought to the chromatograph (Pye Panchroniatograph). The unit now described has proved successful and consists of a spring-loaded double oblique tap. Various sizes of it may be used and quantitative comparisons are made by using the same tap for both the sample and standard gases, which are loaded a t the same pressure. The sample is introduced into the right-hand keyway by connection to a gas cylinder, a gas-handling vacuum manifold, or a storage bulb fitted with a mercury manometer. The key is then turned to the position shown, and the tap connected to the top of the gas-chromatographic column. When the gas-chromatographic recorder base-line has steadied, the sample is injected by rotating the key back t o the “load” position. Edwards silicone vacuum grease is a suitable lubricant. The design of the tap enables the carrier gas to sweep out all “dead” spaces. Received May 20th, 1966

ISSN:0003-2654    

DOI:10.1039/AN9669100671

出版商:RSC    

年代:1966

数据来源: RSC