摘要:

258 SUSIC AND P J E S ~ I C : POLAROGRAPHIC DETERMINATION [Analyst, Vol. 91 Polarographic Determination of Arsenic in Steel BY MILENKO V. SUsIc AND MILJAN G. PJEsCIc {Department of Physical Chemistry, Faculty of Science, University of Belgrade, Yugoslavia) A method for the polarographic determination of arsenic in steel is described. It consists of the reduction of arsenic(v) to arsenic(rr1) in 9.5 N hydrochloric acid solution by means of potassium iodide in the presence of ascorbic acid, and of the extraction of arsenic(II1) with chloroform. Arsenic is then re-extracted from chloroform with 0.2 M ascorbic acid and determined polarographically with the same ascorbic acid solution as supporting electrolyte. The method was checked on NBS steel standards and synthetic mixtures. Arsenic contents greater than 0.01 per cent.can thus be determined easily without concentrating in the re-extraction step. MANY results are recorded in the literature on the polarographic behaviour of arsenic in different supporting electrolytes. Arsenic must be in the three-valence state, since arsenic(v) does not give any analytically useful polarographic wave. Bayerle,l Kacirkova2 and Lingane3 have shown that arsenic(II1) is reduced at the dropping-mercury electrode in the presence of mineral acids as supporting electrolytes, usually producing two waves with half-wave potentials ranging from -0.3 to -1 volt against the S.C.E., depending upon the acid concentra- tion. According to K r y ~ k o v a , ~ arsenic( 111) is reduced with hydrochloric acid solution giving four waves which are dependent on the hydrochloric acid concentration.Kryukova has also shown that the reduction of arsenic(II1) with tartaric and lactic acid solution gives two well- defined waves that can be used for the polarographic determination of arsenic. However, in alkaline solutions arsenic(II1) is oxidised to arsenic(v), giving an anodic wave, the half-wave potential of which is -0-26 volt against the S.C.E.5 S u W was the first to establish that arsenic(II1) can be polarographically determined at different pH values with ascorbic acid as the supporting electrolyte. EXPERIMENTAL All attempts to determine arsenic in steel by a direct polarographic method were un- successful. Although ascorbic acid reduces ferric ions to ferrous ions and some other ions to their lower valence state, it cannot eliminate the influence of some interfering ions, such as molybdenum and vanadium, for steel.In addition, high concentrations of iron in the solution increase the residual current so that small amounts of arsenic cannot be determined accurately. Therefore, we decided to reduce arsenic(v) to arsenic(II1) with the potassium iodide, and then to extract arsenic(II1) with chloroform. The iodimetric method for the determination of arsenic, based on the reduction of arsenic(v) to arsenic(m) with potassium iodide in concentrated hydrochloric acid solution, is well known. Milayev and Varasnina7 have also described a method for the separation of arsenic from some other elements by reducing arsenic(v) to arsenic(m) with potassium iodide, and extracting arsenic(II1) with carbon tetrachloride from concentrated hydrochloric acid solution.Our choice of chloroform as the extractant instead of carbon tetrachloride was based on the polarographic behaviour of chloroform. In fact, chloroform is reduced at the dropping-mercury electrode at a more negative potential than arsenic, and therefore does not interfere with the polarographic wave of arsenic. This is important as, in the re-extraction step, small amounts of organic solvent pass over into the aqueous phase and may interfere with the determination of arsenic if they are reduced at a potential close to the reduction potential of arsenic, as occurs with carbon tetrachloride. The re-extraction of arsenic(II1) from the organic into the aqueous phase occurs rapidly and quantitatively, so that the extracted arsenic can be re-extracted easily with the ascorbic acid solution that was used as the supporting electrolyte for the polaro- graphic determination of arsenic.In the reduction of arsenic(v) to arsenic(II1) and extraction of arsenic(II1) in the presence of ferric ions, elementary iodine is liberated as the reaction product of ferric ions and arsenic(v) with potassium iodide. Most of the free iodine is liberated from the reaction of potassiumApril, 19661 OF ARSENIC IN STEEL 259 iodide and ferric ions if the ferric ion concentration is higher than that of arsenic. Therefore the reduction and the extraction of arsenic from steel samples was carried out in the presence of ascorbic acid. The rdle of ascorbic acid is to suppress the liberation and extraction of elementary iodine.The amount of ascorbic acid used is sufficient to reduce about ten times the amount of ferric ions present to ferrous ions. The reduction of arsenic(v) is ensured by the presence of 0.01 M potassium iodide, which is about 20 to 50 times the equivalent concentration of arsenic(v), and is almost unconsumed in the reactions. REAGENTS- The ascorbic acid solution should be freshly prepared, and the chloroform should not contain more than 0-2 per cent. of alcohol. All materials used should be of analytical-reagent grade. Solid ascorbic acid and 0-2 M ascorbic acid solzttion. Hydrochloric acid, 9.5 M. Nitric acid, sp.gr. 1.4. Sulphuric acid, sp.gr. 1 - 8 4 . Potassium iodide, 0.1 M.Chloroform. PROCEDURE- Weigh 0.2 to 0.5g of steel sample, depending upon the arsenic content, and dissolve it in 10 to 20 ml of a mixture of nitric acid, sulphuric acid and water (4 to 1 to 4 ) . Evaporate to dryness and dissolve the residue in 9.5 M hydrochloric acid. Transfer the solution to a 25-ml volumetric flask and fill it up to the mark with 9-5 M hydrochloric acid solution. Place 2 to 5 ml of this solution in a separating funnel, add 0-2 to 0.5 g of solid ascorbic acid and 0.1 to 0.3 ml of 0.1 M potassium iodide solution and extract twice with an equal volume of chloroform by vigorously shaking the funnel for 10 minutes. Then re-extract the arsenic from the organic phase into 2 to 5 ml of 0.2 M ascorbic acid solution in the polarographic vessel, stirring the two phases by bubbling nitrogen through the solutions for a few minutes; without separating the layers determine the arsenic polarographically in the aqueous phase.The half-wave potential of arsenic is between 0-8 and 0.9 volt against the S.C.E. For a small arsenic content in steel, the arsenic should be concentrated in the re-extraction step of the procedure. RESULTS Following the above procedure we have carried out the determination of arsenic in synthetic mixtures prepared from standard solutions of iron, arsenic and some other elements. The ratio of iron to arsenic was approximately the same as that usually met in steel samples. The composition of synthetic mixtures and the amount of arsenic determined are given in Table I. The results obtained show that the agreement between taken and found amounts of arsenic is satisfactory and is within the limits of the accuracy of the polarographic method.TABLE I (a) DETERMINATION OF ARSENIC IN SYNTHETIC MIXTURES Arsenic determined x 1 0 4 ~ Determina- tion No: Mixture A B C A B C 1 2 3 4 5 6 7 Taken arsenic, Average x l O 4 ~ - 1.60 1-61 1 *65 1-58 - - 1.62 1.60 0.808 0.782 0.845 0.808 0.750 0.831 0.831 0.808 0.800 0.377 0.420 0.421 0.377 0.426 0.392 0.382 0.399 0.400 TABLE I (b) COMPOSITION OF MIXTURES Iron, Chromium, Manganese, Molybdenum, Vanadium, Arsenic, gperml gperml x lo4 gperml x lo4 gperml x 104 gperml x 104 x ~ O * M 0.01 117 1.37 2.75 3.84 2.50 1-60 0.02234 3.42 2.75 3-84 2.50 0.800 0.02234 3.42 2-75 3.84 2.50 0.400260 SUSIC AND PJESCIC [Analyst, Vol.91 Each determination was performed with 2 ml of the solution of the corresponding mixture and no concentration was made in the re-extraction step. The recommended method was further checked by determining arsenic in four standard NBS steel samples, which contained from 0.1 to 0-018 per cent. of arsenic. The results obtained are given in Table I1 and are found to be in good agreement with the certificate values for arsenic content in steel standards. The polarographic waves of arsenic obtained in the determination of arsenic in NBS steel sample No. 466D are shown in Fig. 1. There are four curves; two of them represent the arsenic content in the sample after re-extraction, and the other two the content of arsenic in the same solution after the addition of a standard.Fig. 1. Polarographic determination of arsenic in 0.4000 g of NBS steel, No. 464-D. Curves A and B, arsenic content of sample; curves C and D, arsenic found after addition of standard. Polarographic conditions : Radiometer PO 3h polarograph, V = 2 ml (from 25 ml), Vs = 0.05 ml, C, = 1.5 x 10-4 g per ml, sensitivity = 1 x 100 TABLE I1 DETERMINATION OF ARSENIC IN NBS STEEL NBS number: 461-A (0.028 per 462-B (0.046 per cent. of arsenic), cent. of arsenic), g taken g taken & & 0.3075 0*3000 0.2878 0.2144 Determina- Arsenic found, per cent. tion No. f A -I 1 0.0294 0.0297 0.0461 0.0446 2 0.0316 0.0289 0.0436 0.0456 3 0.0266 0.0279 0-0469 0.0442 4 0.0287 0.0296 0.0438 0.0458 5 0.0273 0-0317 0.0460 0.0471 6 0.0291 0.0269 0.0456 - Average 0,0288 0.0291 0.0453 0.0443 463-C (0-10 per cent. of arsenic), 464-D (0.018 per cent. of arsenic), g taken g taken & & 0.2020 0.2181, 0-4000 0.4111 Arsenic found, per cent. I 0.108 0.098 0.0229 0.0193 0.101 0.104 0.0182 0.0172 0.095 0.099 0-0182 0.0182 0.094 0.098 0.0181 0.0162 0.096 0.102 0.0170 - 0.099 0.100 0.0190 0.0177 - - - - REFERENCES 1. 2. Kacirkova, K., Colln Czech. Chem. Commun., 1929, 1, 477. 3. 4. 5 . 6. 7. Bayerle, V., Recl. Trav. Chim. Puys-Bas Belg., 1925, 44, 514. Lingane, J , J., I n d . Engng Chem. Anulyt. Edn, 1943, 15, 583. Kryukova, T. A., Zav. Lab., 1937, 6, 1385; 1938, 7, 273; 1940, 9 , 950. Kolthoff, I. M., and Probst, R. L., Analyt. Chem., 1949, 21, 753. SuSid, M. V., Bull. Inst. Nucl. Sci. Boris KidriE, 1955, 5 , 65 Milayev, S. M., and Vorasnina, K. P., Zav. Lab., 1963, 4, 410. Received November 2nd, 1964

ISSN:0003-2654    

DOI:10.1039/AN9669100258

出版商:RSC    

年代:1966

数据来源: RSC

摘要:

April, 19661 SELIM AND MESSIHA 261 A Siniple Chromatographic Method for Determining the Basic Amino-acids in Protein Hydrolysates BY AHMED S. M. SELIM AND NAG1 N. MESSIHA (Department of Biochemistry, Abbassia Faculty of Medicine, Abbassia, Cairo, U.A.R.) A new micro method utilising charcoal chromatography has been de- veloped for the determination of the individual basic amino-acids in protein hydrolysates. I t is based on the fact that both the aromatic amino-acids and basic amino-acid picrates are strongly adsorbed by active-charcoal columns, and that the latter can be freed from the aromatic amino-acids by treatment of the adsorbent with aqueous ethyl acetate solution. The separa- tion of the individual basic amino-acids depends on differential decomposition of their picrates on the columns by various eluents.The method can be used for the determination of histidine in the presence of other imidazole derivatives. THE earlier chemical methods that were commonly used in the determination of the basic amino-acids in protein hydrolysates suffered from the disadvantages of being tedious and non-specific. The problem was not entirely solved by the use of the time-consuming micro- biological and enzymic assays. In recent years, many ion-exchange chromatographic and electrophoretic methods have been introduced for the analysis of the basic amino-acids. Although the new procedures have overcome most of the difficulties inherent in the older methods, they require highly specific conditions and involve the use of special equipments and materials. The present investigation was therefore undertaken with the object of de- veloping a simple analytical micro procedure for the determination of histidine, lysine and arginine in protein hydrolysates.Charcoal adsorption has been used by several workers for the quantitative removal of the aromatic amino-acids from protein hydrolysates, prior to the separation of the other amino-acids.l y 2 9 3 9 4 The quantitative separation of the basic amino-acids from protein hydroly- sates can be efficiently accomplished with the method described by Robson and Se1im.l These authors have shown that the basic amino-acid picrates are strongly retained by active charcoal. Later, charcoal chromatography was used by Selim et ~ 1 . ~ 9 ~ for the quantitative removal of the aromatic and basic amino-acids in a single operation.Therefore, by treating the protein hydrolysate with picric acid and applying the mixture to a column of charcoal, it is possible to wash out all of the aliphatic amino-acids with dilute acetic acid, while the aromatic amino-acids and basic amino-acid picrates remain strongly adsorbed on the charcoal. In the present study, further work on charcoal chromatography has revealed that treatment of the charcoal column, containing the aromatic amino-acids and the basic amino-acid picrates, with aqueous ethyl acetate results in the complete elution of the aromatic amino-acids, whereas the picrates of the basic amino-acids remain firmly retained by the adsorbent. I t was also found that when the charcoal column containing the adsorbed picrates is treated with a dilute aqueous pyridine solution, the histidine dipicrate is readily decomposed and all of the histidine appears in the eluate, while the picric acid remains in the column.Both lysine and arginine picrates are unaffected by the pyridine treatment and remain strongly retained by charcoal. I t has been shown that aqueous ammonia readily decomposes lysine picrate on charcoal, but has no effect on the adsorbed arginine picrate. Accordingly, further treatment of the column with dilute aqueous ammonia leads to elution of lysine, whereas the picric acid and arginine picrate remain firmly held by the adsorbent. Arginine can then be recovered quantitatively from the column by passing a dilute hydrochloric acid solution through the column.This decomposes the arginine picrate, and arginine passes out with the eluent leaving the picric acid strongly adsorbed on the charcoal. All these results furnish the basis of the present method. By the new technique the basic amino-acid can be recovered from the charcoal column and collected in three separate fractions. The first, or pyridine fraction, consists of a pure solution of histidine and is entirely free from tyrosine and other basic amino-acids. Histidine can therefore be262 SELIM AND MESSIHA: SIMPLE CHROMATOGRAPHIC METHOD FOR [Analyst, Vol. 91 readily determined in this fraction colorimetrically, by the Pauly reaction. The second, or ammonia fraction, contains no amino-acids except lysine, which can therefore be determined in this fraction, after removal of ammonia, by the colorimetric ninhydrin method.The third, or hydrochloric acid fraction, contains all the arginine and as the latter is not con- taminated by any other amino-acid, it can be readily determined in the acid eluate, after being completely freed from ammonia, by applying the colorimetric ninhydrin method. The applicability and the adequacy of the method have been demonstrated by analysing micro amounts of various proteins, not exceeding 2 to 3 mg, and the results are reproducible and comparable with those obtained by using other methods. The high recoveries achieved by analysing both pure standard solution of the individual basic amino-acid, as well as protein hydrolysates to which known amounts of the basic amino acids were added, reveal the validity and accuracy of the method.Also, the specificity of the present technique for the determination of the basic amino-acids in protein hydrolysates has been adequately examined. This procedure has the advantage that it can also be used for determining histidine in the presence of other imidazole derivatives. I t is simple, rapid and economical in reagents and does not involve the need for special equipment. EXPERIMENTAL REAGENTS- Pyridine was re-distilled before use. The nitrogen content of proteins was determined by the standard micro-Kj eldahl procedure. Duplicate determinations were made on each sample and the values presented are the average of these determinations. Charcoal-The charcoal used in all experiments was a B.D.H. Ltd. active de-colourising charcoal.It was purified by the procedure previously rep0rted.l The washed charcoal was further treated before use with 5 per cent. acetic acid. Egg nlbumi-E. Merck Darmstadt (nitrogen, 15.01 per cent. of moisture-free and ash-free protein; moisture, 12.01 per cent.; ash, 4.3 per cent.). Blood dbumin (bovine)-B.D.H. Ltd. (nitrogen, 15.15 per cent. of moisture-free and ash-free protein; moisture, 10.95 per cent.; ash, 8.8 per cent.). Gelatin-Coignet's gold label leaf gelatin (nitrogen, 17-09 per cent. of moisture-free and ash-free protein ; moisture, 14.3 per cent. ; ash, 4.05 per cent.). Edestin-B.D.H. Ltd. (nitrogen, 14-68 per cent. of moisture-free and ash-free protein ; moisture, 11.2 per cent.; ash, 6.1 per cent.). All reagents used were of the highest purity obtainable commercially.APPARATUS- internal diameter and 15 cm in effective length). end of each tube. model SP500, with 1-cm quartz cuvettes. Micro chromatographic, Pyrex tubes were used for the charcoal adsorption (7 mm in Small funnels are incorporated at the The colorimetric determinations were carried out with a Unicam spectrophotometer, PROCEDURE- Hydrolysis of proteins-Weigh 2 t o 3 mg of the protein into a 2-inch length of glass tubing (4 mm in internal diameter) sealed at one end. Add to the protein, by pipette, 0-1 ml of 6 N hydrochloric acid, and seal off the tube. Place the tube in an oven a t 110" C for 24 hours; then open the seal and remove the acid by evaporating the solution to dryness on the steam- bath. Adsorption of the aromatic amino-acids and basic amino-acid picrates on active charcoaL Transfer the buffered hydrolysate quantitatively to a small flask and make up the volume of the solution to approximately 30 ml with water.To this solution add 1 ml of a standard aqueous picric acid solution (containing 4 mg of picric acid), and shake the mixture thoroughly for 1 minute. Pass the resulting solution through a micro-chromatographic tube packed with 150 mg of active charcoal (already washed in the column with 10 ml of 5 per cent. acetic acid) at about 2 ml per minute. Then treat the column with 20 ml of aqueous 5 per cent. acetic acid to remove any retained free amino-acids other than the aromatic ones. EZzltion of the aromatic amino-acids-Apply 40 ml of a freshly prepared aqueous solution of 5 per cent.ethyl acetate to the column at 2 ml per minute. We found that the first 20 ml Dissolve the dried residue in 1 ml of 0.2 M acetate buffer at pH 5.6.April, 19661 DETERMINING THE BASIC AMINO-ACIDS IN PROTEIN HYDROLYSATES 263 of the ethyl acetate eluate contained practically all of the phenylalanine and tyrosine, and that the last portion of the eluate gave, when tested, a completely negative ninhydrin reaction. Elution of histidine and its determination-Treat the charcoal column, still containing the basic amino-acid picrates, with 25ml of a 2 per cent. aqueous solution of pyridine at 2 ml per minute. The first 10 to 15 ml of the eluate, we found, contained all of the histidine as indicated by the Pauly colour reaction. Evaporate the resulting eluate in vacuo, and make up the volume of the concentrate to 10ml with water.This histidine fraction was shown to be entirely free from tyrosine by the 1-nitroso-2-naphthol colour reaction and by paper chromatography. Introduce 4-ml aliquots from the histidine fraction into 25-ml measuring flasks, and neutralise the pyridine present by adding 0-5 ml of N hydrochloric acid. Add to these solu- tions 5-5 ml of water, and determine the histidine content of the final mixtures colorirnetrically by following exactly the Macpherson modification of the Pauly reaction.' Read the optical density of the resulting red coloured solution a t 530 m p in the spectrophotometer against water. Calculate the histidine concentration by using a standard solution of the amino-acid, prepared at the same time by the following method.Introduce 3 ml of an aqueous solution containing 0.03 mg of histidine into the flask and treat the solution with 4 ml of a 2 per cent. pyridine solution. Add, successively, 0.5 ml of N hydrochloric acid and 2.5 ml of water. The colour was then developed in the same manner as with the unknown, and its absorbance measured in the spectrophotometer. EZution of lysine and its determination-Apply 25 ml of an aqueous solution of 0.1 N ammonia to the adsorbent at 2 ml per minute. We found that the last fraction of the percolate, when tested after the complete removal of ammonia, gave an entirely negative ninhydrin reaction. Treat the colourless lysine fraction with sufficient dilute alcoholic potassium hydroxide solution until the fraction is distinctly alkaline to phenolphthalein.Remove the ammonia by evaporating the lysine solution to dryness in vucz~o. Repeat the evaporation procedure once again, after the addition of water and ethanol. The air bubbled into the distillation flask was already freed from any ammonia by its passage through a trap containing dilute sulphuric acid. Add a little water to the ammonia-free residue, and neutralise the resulting solution with N hydrochloric acid arid make the solution up to 5 ml. This lysine solution was shown, by paper chromatography, to be free from other amino-acids. Determine the lysine in 1-ml aliquots of solution by the modified ninhydrin colour reaction, according to Rosen8 Measure the absorbance of the final coloured solution at 570 mp in the spectro- photometer against a reagent blank that has been treated in the same manner.Calculate the lysine concentrations from the optical densities with the aid of a calibration curve prepared from a graphical plot of optical densities against pmoles of leucine (0-025 to 0.175). The pmoles of leucine obtained from the curve were converted into pmoles of lysine by using a conversion factor of 103 which corresponds to the percentage yield of lysine, based on leucine as 100 per cent., in the same way as demonstrated by Rosen.s Elution of arginine aad its determination-Treat the charcoal column, still containing the arginine, with 25 ml of N hydrochloric acid. Make the colourless eluate distinctly alkaline to phenolphthalein by adding a dilute alcoholic solution of potassium hydroxide, and then aerate the solution at room temperature for 1 hour. Free the air that is used from any possible contamination by ammonia with a dilute sulphuric acid trap.Then make the arginine solution neutral to litmus with N hydrochloric acid and evaporate the solution to dryness in vacuo. Dissolve the resulting residue in a little water and make the solution up to 5 ml. We found on examining of this solution by paper chromatography, that it was completely free from any other amino-acids. Determine the arginine, with 1-ml aliquots, by the same colorimetric ninhydrin method described above for lysine. The pmoles of leucine obtained from the calibration curve were converted into pmoles of arginine by using a conversion factor of 100.RESULTS The applicability of the method was determined by analysing four different proteins, vix., egg albumin, serum albumin, gelatin and edestin, The analyses were run in triplicate and the results were found to be reproducible. In all experiments duplicate samples of the individual basic amino-acid fractions, obtained from the column, were analysed. In Table I are shown the values of the basic amino-acids of the above proteins, as determined by the264 SELIM AND MESSIHA: SIMPLE CHROMATOGRAPHIC METHOD FOR [Analyst, VOl. 91 TABLE I BASIC AMINO-ACID CONTENTS OF PROTEINS EXPRESSED ON 16 GRAMS OF PROTEIN NITROGEN Amino- Protein acid Egg histidine albumin lysine arginine Serum histidine arginine Gelatin histidine lysinc arginine Edestin histidine lysine arginine albumin lysine Present procedure Ammo-acid, Mean, - g g 2.5, 2.6, 2.58 2.6 6.45, 6.52, 6.6 6.5 6.3, 6.22, 6.2 6.2 3.3, 3.37, 3.4 3.4 8.7, 8-66, 8.7 8.7 5.96, 5.9, 6-0 6.0 0.85, 0.8, 0.82 0.8 5.46, 5.48, 5.5 5*5* 8.3, 8.2, 8.2 8.2 2.3, 2.25, 2-23 2.3 2.75, 2.8, 2.78 2.8 14.8, 15.01, 14.9 14.9 Other methods 2.5, Miller et 2.6, Long13 6.5, Long13 6-5, Block and Weiss14 6.2, Long13 6.3, Miller et uZ.l2 3.4, Block and Weiss14 3.6, Keller and Blockl6 8-4, Block and Weiss14 8.6, Keller and Block15 5.4, Keller and Biock15 5.9, Block and Weiss14 0.8, Block and Weiss14 0.8, Long13 4-5, LongI3 4.8, Block and ?Veiss14 7.8, Block and Weiss14 8-1, Eastoe16 2.5, Block and Weiss14 2.5, Kimmel and Srnithl7 2.5, Kimmel and Smith17 2.8, Block and Weiss14 14.9, Kimmel and Smith1' 15.4, Yemm and Folkesls * This value includes hydroxylysine.present method and compared with previously reported figures. As indicated in Table I, the results obtained are in good agreement with those results cited by other workers. The accuracy of the new procedure was demonstrated by running recovery experiments simul- taneously with the unknowns. These were performed by the addition of known amounts of histidine, lysine and arginine, in two concentration ranges, to each protein hydrolysate before application of the latter to the charcoal column. Analyses were then made for the recovery of the added amino-acids. The results of these experiments are given in Table 11, and they show that high recoveries, ranging from 97 to 99 per cent. were achieved by the present technique.The time required for the analyses, run in duplicate, of the three basic amino-acids in protein hydrolysates is 3 hours, and 4 to 5 hours for analysing 6 samples simultaneously. TABLE I1 RECOVERY OF THE INDIVIDUAL BASIC AMINO-ACIDS ADDED TO ALIQUOTS OF PROTEIN HYDROLYSATES EQUIVALENT TO 2 TO 3mg OF PROTEIN Protein Egg albumin . . Blood albumin Gelatin .. Edestin .. Amino-acid .. . . histidine lysine arginine .. . . histidine lysine arginine .. . . histidine lysine arginine .. . . histidine lysine arginine Amino-acid added, pmole 0.15 0.32 0.2 0.34 0.27 0.55 0.25 0.32 0.2 0.34 0-17 0.57 0-1 0.25 0.2 0.68 0-57 0.85 0.15 0.25 0.25 0.34 0.75 1.20 Amino-acid recovered, pmole 0.149 0.3 11- 0.196 0.337 0.265 0.535 0-245 0.3 18 0.198 0.330 0.168 0-566 0-098 0.243 0.194 0.670 0-564 0.83 0.148 0.242 0.244 0.336 0-737 1-18 Recovery, percentage 99 97 98 9Y 98 97 98 99 99 97 99 99 98 97 97 97 99 98 99 97 98 99 98 98April, 19661 DETERMINING THE BASIC AMINO-ACIDS IN PROTEIN HYDROLYSATES 265 CAPACITY OF CHARCOAL- The use of micro columns packed with 150 mg of active charcoal was found to be quite sufficient to hold, as picrates, all the amounts of basic amino-acids that are present in aliquots of protein hydrolysates which are equivalent to 3 mg of proteins.Such columns have enough room to retain the excess free picric acid and to permit the separation of the basic amino-acids in the manner previously described. Robson and Seliml have shown that while amounts of picric acid in slight excess of those equivalent to the arginine and histidine present, in the molar ratio of 1 to 1 for the former and 2 to 1 for the latter, are enough for the complete removal of these amino-acids by charcoal, picric acid in amounts four to five times that required for the formation of lysine picrate should be present to secure the complete retention of lysine by charcoal.On this basis, it was found that treatment of the protein hydrolysate (2 to 3 mg of protein) with 4 mg of picric acid is quite sufficient for the conversion of all of the basic amino-acids present into their picrates, and for their quantitative adsorption on to charcoal. RECOVERY- In order to test the validity and specificity of the method for the determination of the basic amino-acids in protein hydrolysates, aliquots of pure standard solutions of histidine, lysine monohydrochloride and arginine monohydrochloride were each treated with 4 mg of picric acid and applied separately to micro columns of charcoal (150 mg).Each time, the adsorbent was treated with 20 ml of 5 per cent. acetic acid followed by 40 ml of 5 per cent. ethyl acetate. Histidine, lysine and arginine were further eluted from their columns in exactly the same way as previously described. The percentage of the basic amino-acids recovered was determined as before. These experiments were then repeated, but with two known amino-acid mixtures. Mixture (1) consisted of the three basic amino-acids, while mixture (2) was prepared from the three basic amino-acids, phenylalanine and tyrosine and a number of the aliphatic amino-acids. The latter included glycine, alanine and glutamic acid, all of which are weakly adsorbed by charcoal, as well as isoleucine, leucine and methionine which are relatively strongly retained by charcoal.In each experiment, the pyridine, ammonia and hydrochloric acid fractions were collected from the columns and analysed for their basic amino-acid contents. The results of these analyses, run in duplicate, are given in Table 111, and they indicate that the recoveries of the basic amino-acids were high and average from 97 to 99 per cent. From the figures given in Table 111, it would appear that the presence of the aliphatic and aromatic amino-acids does not interfere with the determination of the basic amino-acids by the present technique. TABLE rIr RECOVERY OF THE INDIVIDUAL BASIC AMINO-ACIDS FROM PURE SOLUTIONS AND AMINO-ACID MIXTURES Material analysed, volume 25 ml Pure histidine solution Pure lysine solution .. Pure arginine solution Mixture (1)- Histidine . . . . Lysine . . .. Arginine . . . . Lysine . . ,. Arginine . . .. Tyrosine . . . . Glycine . . .. Mixture (2)- Histidine . . .. Phenylalanine . . Glutamic acid . . Alanine . . .. Leucine . . .. Isoleucine . . .. Methionine * . .. .. .. . . .. .. .. . . .. .. .. .. . . .. .. .. .. Amino-acid Amino-acid recovered, added, average value, pmole pmole .. 0.75 0.740 . . 0.75 0.728 .. 0.75 0.742 .. 0.4 0.395 . . 0-4 0.300 .. 0.4 0.391 .. .. .. .. .. .. . . .. .. .. .. 0.5 0.486 0.5 0.493 0.5 0.485 1.0 1-0 4.0 - 2.0 - 3.0 - 3.0 - 2.0 - 0.5 - - - Recovery, percentage 99 97 99 99 98 98 97 98 97 - - - - - - - -266 SELIM AND MESSIHA: SIMPLE CHROMATOGRAPHIC METHOD FOR [AndySt, VOl.91 SPECIFICITY- The reliability and specificity of the method were also demonstrated by examining the pyridine, ammonia and hydrochloric acid fractions obtained in the above experiments, as well as in those carried out on egg-albumin hydrolysate, for the presence of any contaminating amino-acids. Each time the histidine fraction was found to be completely free from tyrosine, as indicated by the l-nitroso-2-naphthol colour reaction, according to the method of Cerriotti and Spandri~.~ The behaviour of the above fractions, on a two-dimensional chromatogram, with water-saturated 2,4,6-collidine and water-saturated phenol - ammonia was also examined.The results showed that the final histidine, lysine and arginine solutions were not contaminated with any other amino-acids. However, 6-hydroxylysine picrate will be adsorbed by charcoal together with the lysine picrate, and therefore, the presence of hydroxylysine in protein hydrolysates will interfere with the yuantiative assay of lysine. This, however, does not present a difficulty as the limited distribution of hydroxylysine in proteins greatly minimises the incidence of this interference. DISCUSSION Charcoal chromatography, which has proved its value as a convenient procedure for the quantitative removal of the aromatic and basic amino-acids from protein hydrolysates,5,6 910911 was selected as an effective tool for the micro determination of the individual basic amino- acids in the present investigation.In agreement with the work of Schramm and Primosigh,2 it has been found that the acetic acid treatment of charcoal, before and after the percolation of the hydrolysate, is necessary for facilitating the complete removal of the adsorbed amino-acids. No attempts were made in this work to prevent, or minimise, the catalytic action of active charcoal on amino-acids by poisoning the adsorbent with hydrocyanic acid solution, or hydrogen sulphide- saturated water. However, under the experimental conditions used, it was found that the absence of such poisoning of charcoal does not affect the results to any appreciable extent. It appears that if the whole process of charcoal chromatography is carried out rapidly a t high rates of liquid flow, with the active charcoal purified and washed, as previously described, there will be hardly any risk of the catalytic effect of charcoal playing an important r6le.The irreversible adsorption of amino-acids will also be avoided. The use of micro and very short columns of purified and acetic acid pre-treated charcoal has greatly helped to meet the above requirements. In addition, if any traces of aliphatic amino-acids escape removal from the column when the latter is washed with 5 per cent. acetic acid, these will be eluted, together with the aromatic amino-acids, by treatment with ethyl acetate. The present chromatographic method has the advantage that many of the imidazole derivatives do not interfere with the histidine determinations, and therefore the method will be of value in the determination of histidine in the presence of these compounds.From the results of preliminary experiments it was found that the imidazole derivatives that fail to form dipicrates are not strongly retained by charcoal. So imidazole, imadazole acetic acid and imidazole propionic acids, in the presence of picric acid, are readily eluted from charcoal columns by 5 per cent, acetic acid or an aqueous solution of ethyl acetate. Hista- mine, however, is readily adsorbed as the dipicrate on charcoal and is therefore not washed out from the column by acetic acid or ethyl acetate. Unlike histidine, it is unaffected by the pyridine treatment and accordingly its presence with histidine does not interfere with the determination of the latter.The adsorbed histamine can be readily eluted from the column with 0.1 N ammonia. CONCLUSION This method is simple, rapid, accurate and specific for the determination of the individual basic amino-acids in protein hydrolysates. It gives results that are reproducible and com- parable with those previously reported. It has also the advantage that it can be used for the determination of histidine in the presence of other imidazole derivatives. REFERENCES 1. 2. 3. 4. Robson, W., and Selim, A. S. M., Biochem. J . , 1953, 53, 431. Schramm, G., and Primosigh, J., Bey. chem. Ges. Frankfurt, 1943, 76, 373. Tiselius, A., Drake, B., and Hagdahl, L., Expera’entia, 1947, 3, 21. Partridge, S. M., Biochem. J., 1949, 44, 523.April, 19661 DETERMINING THE BASIC AMINO-ACIDS I N PROTEIN HYDROLYSATES 267 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. Selim, A. S. M., Ramadan, M. E. A., and El-Sadr, M. M., J . Biol. Chem., 1957, 227, 871. Macpherson, H. T., Biochem. J., 1946, 40, 470. Rosen, H., Archs Biochem. Biophys., 1957, 67, 10. Cerriotti, G., and Spandrio, L., Biochem. J., 1957, 66, 607. Shimizu, W., Fujita, M., and Endo, K., Bull. J a p . SOC. Scient. Fish., 1954, 20, 305. Greenstein, J . P., and Winitz, M., “Chemistry of the Amino Acids,” Volume 2, John Wiley and Miller, S., Ruttinger, V. E., Kovach, E:S., and Macy, I. G., Pan-Amer. Med. Wom. J., 1952,59, 9. Long, C., “Biochemist’s Handbook,” E. and F. N. Spon Ltd., London, 1961, 765. Block, R. J., and Weiss, K. W., ‘‘Amino Acid Handbook,” Charles C. Thomas, Springfield, Ill., ICeller, S., and Block, R. J., Archs Biochem. Biophys., 1959, 85, 366. Eastoe, J. E., Biochem. J., 1955, 61, 589. Kimmel, J. R., and Smith, E. L., Bull. SOC. Chim. Biol., 1958, 40, 2049. Yemm, E. W., and Folkes, B. F., Biochem. J., 1953, 55, 700. I , Ibid., 1957, 229, 547. - - _ _ Sons Inc., New York, 1961, 1451. 252, 274, 316, 341 and 343. Received February 25th, 1965

ISSN:0003-2654    

DOI:10.1039/AN9669100261

出版商:RSC    

年代:1966

数据来源: RSC

摘要:

268 BAAR: MICRO DETERMINATION OF CYANIDE [Analyst, VOl. 91 Its The Micro Determination of Cyanide: Application to the Analysis of Whole Blood BY S. BAAR (M.R.C. Industrial Injuries and Burns Research Unit, Birmingham Accident Hospital, Bath Row, Birmingham 15) An introduction to the existing methods for the analysis of cyanide is given and some of the limitations to the methods are pointed out. A modification of the Epstein method is described, in which Cavett blood- alcohol flasks are used. This method can be applied to small samples of 2 ml volume containing 0.2 pg of cyanide. By strict control of the conditions it is shown that a high degree of accuracy can be achieved. Interference by heavy metal ions is avoided by using 2 mg of the disodium salt of EDTA per ml of blood.It is therefore suggested that this anti-coagulant should be used when the blood is collected. Experiments on the partition of cyanide in whole blood showed that 5 minutes’ exposure resulted in more than 70 per cent. of cyanide being bound to haemoglobin. This value remaincd unchanged in the presence of a transport inhibitor. THE necessity to determine small amounts of cyanide formed by Ps. aerugiizosa in small samples of whole blood or tissue led to the development of the method that is described. Techniques suitable for determining low concentrations of cyanide are based on the Konigl reaction for the synthesis of pyridine dyes. The method developed by Aldridge2 involves the coupling of cyanogen bromide with a pyridine - benzidine mixture. The carcinogenic nature of the benzidine has led to its substitution by other amines, notably 9-phenylenediamine (Bark and Higson3).Even so, relatively large samples are necessary to carry out this technique. The original Aldridge method has been applied by Tompsett4 to biological material. But although it is suitable for determining low concentrations of cyanide, the samples to be analysed must contain at least 5 pg of hydrogen cyanide. The Epstein5 method is more sensitive and therefore appeared to be more suitable for the determination of small samples of low cyanide concentration. In this method cyanogen is converted t o cyanogen chloride, which is then converted to glutaconaldehyde by chloramine T. The excess of chloramine T is reduced by 3-methyl-1 -phenyl-2-pyrazolin-5-one.The presence of pyridine in the reagent prevents the reduction of the cyanogen chloride that is formed. A small amount of 4,4’-bis-(3-methyl-1-phenyl-i!-pyrazolin-5-one) in the reagent causes the formation of a blue dyestuff suitable for quantitative determination. The exact nature of the reaction is unknown. Ihe principle of this reaction has been applied by some workers to the determination of cyanide in small volumes of blood. Feldstein and Klendshoj6 used the modification des- cribed by Boxer and Kickards7 and adapted it for micro diffusion analysis by using Conway No. 1 units. In the absence of a special oscillating table it was found that recovery never exceeded 70 per cent., even at a dilution of 1 in 4. Feldstein and Klendshoj6 also found that the apparent recovery of the cyanide, added to whole blood, decreased on standing for a few hours, but reached the expected levels after 24 hours’ storage.An enzymt. system, within the red cell, capable of removing cyanide was suggested. The method described below offers a simple modification of a commercially available apparatus and avoids the use of an oscillating table. The factors leading to incomplete cyanide recovery from whole blood were also investigated and allowance was made for these. P. METHOD APPARATUS- For the micro diffusion work a modification of the Cavett flask was used,* as shown in Fig. 1. On the floor of the flask a small cup was mounted with Evostik. This was prepared from a polystyrene specimen tube (external measurements 3.8 cm x 1-5 cm). The tube was * Supplied by Quickfit & Quartz Ltd., Staffs., England.April, 19661 269 sawn off to permit the insertion of a wall, 0.5 cm high.In this wall a slot of 1.5-mm width was sawn. The capacity of such a cup is approximately 0 6 m l and, because it is "non- wettable," fluid does not readily leave it. The flask must be washed with detergent and water only and dried with warm air. The Evostik attachment has withstood frequent use, but the more permanent Araldite is also suitable. ITS APPLICATION TO THE ANALYSIS OF WHOLE BLOOD dob -Cup A Fig. 1. Modification of the Cavett flask usetl for micro diffusion work (quarter scale) REAGENTS- Pyrazolone reagen-This must be prepared freshly on the day of use, as otherwise the colour yield may be variable. Dissolve 240 mg of %methyl- 1 -phenyl-2-pyrazolin-5-one in 100 ml of water at 75" C ; this requires about 14 hours.Dissolve 20 mg of 4,4'-bis-(3-methyl- l-phenyl-2-pyrazolin-5-one)* in 20 ml of pyridine by shaking them together. Just before use, mix these two solutions. Sodiztm hydroxide, 0.1 N. Sodium dilzydrogeiz phosphate, M. Sulphuric acid, 15 per ceiat. v/v. Chloranzine T solutio~a, 0.1 per cent., aqueous-Keep solution refrigerated. Just before use, add 1 volume of this solution to 3 volumes of M sodium dihydrogen phosphate, and cool the resulting solution to about 5" C. Whole blood-Out-dated, citrated blood-bank material. Stnndard cyanide solution-Dissolve 500 mg of potassium cyanide in 1000 ml of 0.1 N Suitable dilutions of this stock solution with de-ionised water are sodium hydroxide.prepared before use. PROCEDURE- Introduce 1.2 ml of 0.1 N sodium hydroxide into cup A. Tilt the flask about 25" by resting it on a suitable support, Introduce 4 ml of water on to the floor of the flask, then add to the water either 2 ml of aqueous standard solution containing 200 mg of the disodium salt of EDTA per 100 ml, or whole blood containing 2 mg o f the disodium salt of EDTA per ml of blood. Iill cup B with 0.5ml of 15 per cent. sulphuric acid and replace the stopper, righting the flask. Attach the springs, and again tilt the flask to about the same angle as before. Tap the floor of the flask slightly to allow the acid to flow out. After about 30 seconds rotate the flask twice on a flat surface, and allow diffusion to proceed for 2 hours.At the end of this period withdraw a 1-ml sample from cup A, cool it to approximately 5" C and add 0-2 ml of buffered chloramine T solution. After 2 minutes, add 3 ml of pyrazolone reagent rapidly. Allow the colour to develop for 45 minutes, then add 2 ml of n-butanol to extract the coloured product. Determine the absorbancy of the supernatant liquid, obtained after spinning the solution in a centrifuge, at 615 mp and I-cm light path. It is found that Beer's law is obeyed for concentrations from 10 to 100 pg per 100 ml. The colour is stable for at least another 15 minutes. Moisten the ground-glass neck of the flask (see Fig. 1) with distilled water. Mix the solutions thoroughly. * Supplied by Koch-Light Laboratories, Colnbrook, England.270 BAAR: MICRO DETERMINATION OF CYANIDE [Analyst, Vol.91 RESULTS Two types of substances may be reasonably expected to interfere. The presence of sulphydryl groups could lower the apparent values. Free sulphydryl groups are known to react with cyanide, and the in vitro formation of thiocyanate from proteins has already been demonstrated by Pascheles.8 Catsimpoolas and Woodg carefully investigated the reaction between cyanide and albumin and showed that thiocyanate was only formed at pH 8 or higher. To investigate possible small changes in apparent cyanide levels, artificial cyanide solu- tions in plasma were prepared in the presence or absence of 5 x M sodium salt of p-chloromercuribenzoate. This concentration of the mercuric salt had been shown previously not to complex with the cyanide, but was sufficient to block the free sulphydryl groups in plasma.Even after standing for 24 hours both sets gave the same values. I t was therefore concluded that the presence of a blocking agent was not necessary. TABLE I THE RECOVERY OF ADDED CYANIDE FROM WHOLE BLOOD I N THE PRESENCE OF THE DISODIUM SALT OF EDTA Cyanide, pg per cent. 10 10 10 20 20 20 30 30 30 2 mg of the disodium salt of EDTA per ml of whole blood With EDTA Without EDTA P A r \ Percentage increase EE:Tm, Average pg per cent. E’ 615 cm rn, Average with EDTA Cyanide, 0.105 1 10 0.085 7 ::::: 1 10 0.090 J 0.210 20 0.100 ::2 i’ 20 0.193 J 0.340 0.342 0.340 0.103 10 0.088 > 0.088 14 0.212 20 0.195 ) 0-193 9 30 0.300 30 0.341 30 0-304 11 0.302 ) 0*302 Heavy metal ions are the second group of interfering materials and would also produce a negative error as they readily form cyanide complexes that are not easily dissociated. Whole blood contains small concentrations of ionic iron and copper.The disodium salt of EDTA was added in sufficient concentration to complex these and also to act as an anti- coagulant. Table I shows the results of such an experiment, which clearly demonstrate the advantage of using EDTA. The apparently greater difference at 10 pg per cent. may well be with significance, as the absorbancy values are rather low. TABLE I1 REPRODUCIBILITY 20 determinations in 3 batches Whole blood, 20 pg per cent. of hydrogen cyanide, 2 mg per ml of the disodium salt of EDTA No. 1 0.215 11 2 0.216 12 3 0.218 1 3 4 0.215 14 5 0.214 15 6 0.215 16 7 0.217 17 8 0.2 13 18 9 0.2 13 19 10 0.214 20 El C m No.615 1”1* Mean = 0.2152 = 20 pg per cent. 2u = &0.0014 = *0*26 pg per cent. Standard error of mean = ,-t0-031 pg per c E;:;:lIqL 0.215 0.214 0.2 17 0.318 0-215 0.215 0.214 0-2 16 0.214 0.214 :ent. As the addition of EDTA had been established to be advantageous, all further experi- ments were carried out in its presence. Table I1 shows the reproducibility attainable withApril, 19661 ITS APPLICATIOY TO THE ANALYSIS OF WHOLE BLOOD TABLE I11 RATE OF COLOUR DEVELOPMENT 27 1 Time, minutes E;Ammv Percentage of maximum 10 0-509 84.5 20 0.564 93.7 30 0.600 99.9 40 0.602 100 50 0,602 100 one cyanide concentration only. For 95 per cent. confidence limits at 20 pg per cent., the value was k0.26 pg per cent. or 1.3 per cent.within the optimal absorbancy range. Various modifications of the Epstein method utilise differing times for maximal colour development. This was therefore investigated with the method described above. As shown in Table 111, 45 minutes or more gave maximal values. TABLE IV THE EFFECT OF WHOLE BLOOD DILUTION ON CYANIDE RECOVERY Whole blood, Water, Cyanide added, Percentage of ml ml pg per cent. of hydrogen cyanide* El 615m cm v aqueous standard 1 5 50 0.246 100,o 2 4 40 0.243 99.0 3 3 30 0.2 10 85.3 4 3 20 0.151 61.1 5 1 10 0.137 55.7 * Mean of two individual determinations. As the rate of gas liberation is affected by the viscosity of the medium, the effect of using various dilutions of whole blood was determined. Table IV demonstrates that gas liberation is incomplete if dilutions of less than 1 in 3 are used.The geometry of the diffusion chamber significantly affects the time required for absorp- tion efficiency to approach 100 per cent. This has been shown in detail by Conwaylo for the chambers developed by him. TABLE V THE EFFECT OF DIFFUSION TIME ON ABSORPTION EFFICIENCY 1 ml = 0.2 pg of hydrogen cyanide Potassium cyanide in 0.1 N sodium hydroxide Time, hours 0.5 0-5 0.5 1.0 1.0 1.0 2.0 2.0 2.0 3.0 3.0 3.0 EtlK Average Percentage absorbed 0-177 0.173 0.176 82-8 0.178 0.196 0-196 0.196 92.0 0-195 0.210 0.212 0-21 1 99.3 0.210 0-213 0.215 0.2 13 100 0.2 11 This factor was therefore investigated for the proposed modification, and Table V shows that after 2 hours’ diffusion 99 per cent. of the 3-hour value is given. The 2-hour diffusion time was adopted for convenience.Because of earlier statements that the cyanide disappears from whole blood and that plasma could also be used, it seemed important to obtain information on the partition of added cyanide within the blood elements. Table VI shows that after 5 minutes’ exposure, over 70 per cent. of the cyanide was bound to the haemoglobin. This value should probably be somewhat higher as the stroma fraction was not washed, and was therefore slightly contaminated with haemoglobin. It272 BAAR [Analyst, VOl. 91 TABLE VI PARTITION OF CYANIDE ADDED TO WHOLE BLOOD Time of exposure = 5 minutes; packed cell volume = 34 per cent. Material El 613 C l h ~~ Percentage recovery of cyanide added Whole blood . . .... 0-323 100 Whole blood . . .. .. 0.326 100 Aqueous standard . . .. 0.324 I00 73.6 Cells . . .. .. .. 0.236 Cells . . .. * . .. 0.232 Plasma-undiluted . . .. 0.062 12.9 Stroma (4 ml blood) ,. 0.086 13-6 Recovery = 98.5 per cent. * The contents of the cells were adjusted with saline to the original volume of blood, plasma and stroma ahsorbancy were corrected to appropriate volumes. remained to be seen whether the cyanide reached the cell contents by a pure-diffusion process, or whether active transport was involved. For this purpose ouabain, an inhibitor to active cation transport, was added in high concentration. I t can be seen from Table VII that the results were similar to those obtained without the use of an inhibitor, and it is reasonable to assume that cyanide freely diffuses and presumably forms cyanmethaemoglobin.TABLE vrr THE EFFECT OF OUABAIN 1 0 - 3 ~ ON THE CYANIDE PARTITION IN WHOLE BLOOD Time of exposure = 5 minutes; packed cell volume = 34 per cent. Material* El c13 cm ml* Percentage recovery of cyanide added Whole blood . . .. .. 0-348 100 Whole blood . . .. .. 0.348 100 Cells . . .. . . .. 0.265 76.4 Cells . . . . . . .. 0.266 Plasma-undiluted . . .. 0.096 0.4 Plasma-undiluted . . .. 0.007 Stroma (4 ml blood) .. 0.0 12 14-6 Recovery = 100-4 per cent. * The contents of the cells were adjusted to the original volume of blood, plasma and stroma absorbancy werc corrected to appropriate volumes. It becomes clear from these experiments that plasma must never be used for analysis and, provided whole blood is used, the proposed modification can be adopted with accuracy for cyanide levels above 10 pg per cent. Levels of about 5 pg per 100 ml can be determined with similar accuracy if 4 ml of sample are available. The stopper, after the first absorption, is inserted into a second flask containing a fresh 2-ml sample of the same specimen to which 4 ml of water are added, and the hydrogen cyanide is liberated by the addition of 0.5 ml of 15 per cent. v/v sulphuric acid. 1 . 2. 3 . 4. 5. 6. 7. S. 9. 10. REFERENCES Konig, TV., J . prakt. Chein., 1904, 69, 105. Aldridge, W. N., A nalvst, 1944, 69, 262. Bark, L. S., and Higson, H. G., Talanta, 1964, 11, 471. Tompsett, S. L., CEh. Chern., 1959, 5, 587. Epstein, J., Analyt. Client., 1947, 19, 272. Peldstein, M., and Klendshoj, N., J . Lab. Clin. Med., 1954, 44, 166. Boxer, G. E., and Iiickards, J. C., Archs Riochem., 1952, 39, 292. Pascheles, W., Arch. exp. Path. Pharmak., 1894, 34, 28. Catsimpoolas, N., and Wood, J. L., J . Biol. Chem., 1964, 239, 4132. Conway, E. J ., “Microdiff usion. Analysis and Volumetric Error,” Fourth Edition, Crosby Lockwood Received July 29th, 1965 & Son Ltd., London, 1957, p. 20.

ISSN:0003-2654    

DOI:10.1039/AN9669100268

出版商:RSC    

年代:1966

数据来源: RSC

摘要:

April, 19661 VESSEY AND KENDALL 273 Determination of Particulate Matter in Intravenous Fluids BY I. VESSEY AND C. E. KENDALL (National Biological Standards Laboratory, Canberra, A .C.T., Australia) Methods for the examination of intravenous solutions for suspended particles have been investigated. A visual inspection method has been devised, which enables samples to be graded by comparison with reference materials. For quantitative work, techniques for particle-size analysis with a Coulter Counter have been developed. A survey of all the solutions com- mercially available in Australia has been made, and limits for the acceptance of such solutions have been proposed. ATTENTION has recently been drawn to the hazards associated with particulate matter present in intravenous fluids,1,2 and there is clearly a need for control of this type of contamination. No such control can be effected unless the particles can be sized and counted with reasonable accuracy.In addition, it may be necessary to distinguish different types of particles, particularly to recognise bast and other fibres,lP2 and to establish limits for them. As a first step, however, it appears that a considerable improvement would result if limits for total particulate matter were established, without discrimination of type. Many methods of examination of solutions are possible, of which the following have been considered in this work- ( a ) Visual examination in diffuse light. This enables larger particles to be detected, the lower size limit being about 50 microns. This method is used by many manufacturers for inspection of finished products, but leads to considerable operator fatigue, and is not quantitative.Garvan and Gunner1 used apparatus of this type, the light beam being introduced through the bottom of the bottle and the particles viewed from the horizontal position. As, however, the interest is mainly in particles that are large compared with a wavelength of visible light, this method gives insufficient discrimination of particle types. Particles over about 1 micron in diameter give most of their scatter at small angles to the direct emergent beam, and relatively little scatter in the 90" direction. Colloidal particles, z.e., those small compared with a wavelength of visible light, scatter equally about the normal. Thus large particles are less readily visible, and are seen against a diffuse background if any colloidal matter is present.Small angle viewing eliminates the colloidal background almost completely, and is much more satisfactory for detection of particles over about a micron in diameter. An apparatus for this purpose is described below. (c) Filtration through a membrane filter, followed by microscopic examination of the retained particles. This method is of value for the examination of many products, but in our experience is not ideally suited to solutions having relatively low counts, such as the best of the commercial intravenous solutions. Apart from the tedious nature of the counting process, there is a danger that particles having a similar refractive index to the filter may be missed, especially as they are viewed against the granular background of the filter. (d) Resistance counters (Coulter, Ljungberg, Sansar).This type of instrument has been found to be most suitable for particle-size analysis, and has been used for the survey which occupies the remainder of this paper. Only electrolytes have been examined so far, but methods of handling non-electrolytes by dilution with a suitable electrolyte are being de- veloped. One previous publication3 has discussed the use of a resistance counter for examina- tion of normal saline solution, a single count being made of particles greater than 1.3 microns. The visual examination methods, (a) and ( b ) , are carried out on unopened bottles, while the last two, (c) and ( d ) , are destructive, i.e., they have to be carried out by opening the containers; both types of test are necessary so that all bottles can be checked semi-quantita- tively and more accurate counts taken on representative samples.(b) Visual examination by light scattering (Tyndall effect).274 VESSEY AND KEXDALL: DETERMINATION OF [Analyst, VOl. 91 EXPERIMENTAL (a) LIGHT SCATTERING APPARATUS- This apparatus produces a sharply defined ribbon-shaped beam of light, projected horizontally into the bottle. The optical scheme is shown in Fig. 1. The light source, A, is an 8-volt, 50-watt iodine - quartz projector lamp. An image of the filament is focused by the condenser, B, on to an aperture, D, (the aperture is made equal in size to the image). A stop, C, immediately in front of the condenser defines the shape of the final beam; a vertical slit has been found to be the most satisfactory.A B C D E F A = Light source B = Condenser D = Aperture E = Lens F = Bottle c = stop Fig. 1. Light-scattering apparatus for the visual detection of small particles The lens, E, of approximately 8 to 10-cm focal length, projects an image of the slit, C, into the centre of the bottle. The bottle itself acts as a cylindrical lens and helps to keep the beam fairly uniform in cross-section. The apparatus fits into a box about 14 inches long and 8 inches wide, divided into compartments by baffles that support the optical elements. An extra baffle in front of the lamp carries a heat filter, and a final baffle is situated just in front of the sample; the whole interior is painted matt black. The final beam is a ribbon about 1-5 cm high and 1 to 2 mm wide.The end of the box is open, and the sample is viewed at an angle of 20” to 30” from the direct transmitted beam. With a little practice it is possible to form some judgment of the sizes of particles present. Particles over about 5 microns diameter and of moderately high refractive index are seen as twinkling spots of light, often appearing to change colour as the angle of viewing changes. Small particles, over about 1 micron, can be seen as rather faint but definite spots of light. ( b ) RESISTANCE COUNTER- A Coulter Counter model A (Medical) was used, with a 100-micron aperture. The threshold and threshold zero potentiometers were replaced by 10-turn “Helipots” fitted with “Duodials” for easier and more accurate setting.The counter was checked at intervals by placing a 10,000-ohm load resistor across the electrodes and injecting pulses obtained by clipping and differentiating SO-C.P.S. ax. mains. A switch was fitted to over-ride the mercury contacts so that the counter could be re-set, started and stopped. Counts of mains pulses were made for 1 minute; the count was in- variably within the accuracy of reaction time when the counting period was controlled by manual switching while observing a stop-watch. To avoid contamination from airborne particles, solution administration sets were used for sampling, and the sample container was handled in the down-draft from a Vokes absolute filter. Instead of the beaker normally used to hold samples in the counter, a tube of 2.5-cm diameter was used as the sample container.This tube had a side arm into which the tube of the administration set could be inserted. A drain tube on this sample container led to a T-piece, one arm of which led to the inner electrode vessel, so that the two electrode com- partments could be flushed through in series to avoid waste of sample. In use, the outer tube was filled and flushed through to the inner tube several times. The inner compartment needs only to have the same electrolyte concentration as the outer, and does not need to have the same particulate concentration, as the flow is inwards only. After electrolyte balanceApril, 19661 PARTICULATE MATTER I N INTRAVENOUS FLUIDS 275 had been reached, the outer vessel was drained and two or three changes of solution made in the outer sample container.I t was necessary to ensure that there was no continuous column of liquid between the sample container and the tubing of the administration set (see Fig. 2). > tube Fig. 2 . Sample vessel for Coulter Counter The counter was calibrated with a Dow polyvinyltoluene latex of 3.49-micron particle diameter. The calibration method given in the instructions supplied with the instrument is adequate when the electrolyte remains constant, but separate calibration for each electrolyte would be inconvenient. From theory4 the resistance across the aperture is- For a particle of high specific resistivity and diameter less than about 10 per cent. of the aperture diameter, the resistance change due to passage of a particle is- 4pd3 .... - (2) . . - * (3) Ar = - .. .. 1.5 7rD4 . ' It therefore follows that- ar a 3 r 1-5 D21 . ' . . . . .. -- -- where d is particle diameter, D the aperture diameter, I the effective aperture length and p the specific resistivity of the solution. For a given aperture- A l - = ad3 r * (4) . . .. .. .. .. where a is a constant for the aperture.276 VESSEY AND KENDALL: DETERMINATION OF [Autazyst, VOl. 91 The voltage pulse AE due to passage of a particle of effective diameter d is- . . E,Ar AE = OY Eoad3 AE = y (R + Y + c) d* * . .. .. . * (6) .. .. . . (5a) where R is the aperture current resistor (from 50,000 ohms to 6.4 megohms), x the shunt impedance of the input circuit4 (100,000 ohms), E, the H.T.voltage (300 volts), y = ('/I? + '/Y + l/G) . . . . . . . . * * (6) G is the input resistor of the first amplifier stage (1 megohm), and c is a 15,000-ohms resistor in series with R and 7. We can then put d3 = K't'y (R + Y + C)JV = K't' (R + I + C) d m - . . .. .. - - (7) or if 7 is less than about 30,000 ohms, d3 = K't' (1/R + l / ~ ) (R + Y + C) . . . . . . .. - (8) where K' is a constant which does not depend on 7 , and t' is the threshold reading on a 0 to 100 scale. The factors given by Coulter to correct readings at each aperture current setting to those of aperture current 1, involve cancellation of the Y term in equation (5) so that the calibration ~onstant,~ as given by Coulter, is not independent of Y. In use, it was found convenient to measure the voltage between the electrodes for an aperture current setting of 7 or 8 (mean of both polarities), and calculate 7 from the relationship- E (R + c) r = E o - E and then to calculate the threshold setting for a series of particle diameters by substituting for 7 and suitable aperture current (R) values in equation (7) or (8).Counts at 2.0, 2.5, 3-5, 5.0, 7.5, 10-0, 15-0 and 20.0-micron particle diameter were determined for each solution, means of several readings being taken for each setting. The container was gently shaken and allowed to stand until air bubbles had dispersed. When opening a new container, suc- cessive counts at a single size setting were taken until consistent counts were obtained to eliminate errors due to contamination by sample tubes or particles liberated by piercing the closure.It is possible that other models of resistance counter could be used for this type of work; it is essential, however, that the counter used should be capable of operating over a wide range of resistance, and some form of aperture current control and facilities for calibration with varying aperture currents is essential. RESULTS Samples, representing all the solutions commercially available in Australia, were obtained from seven manufacturers and examined; four were of Australian manufacture, the others being of U.K. or German origin. TABLE I COUNTS FOR SELECTED PARTICLE DIAMETERS Manu- of - / - - - - - 2-p particle 3.5-p particle 5-p particle 10-p particle Number diameter diameter diameter diameter facturer samples mean best worst mean best worst mean best worst mean best worst A1 A2 B C D E 'F G 29 11760 16 2255 11 2212 22 185 25 1275 9 4863 7 245 18 24 1 2262 38182 378 6896 520 8880 64 844 418 3362 1260 22292 208 562 22 814 340 1 586 592 50 296 1092 73 48 376 14434 120 1468 84 2498 20 172 44 816 228 4582 20 222 (6) 102 1286 126 186 38 199 32 (y; $1 363 76 27 (10) (18) - 5148 346 912 38 778 1416 88 40April, 19661 PARTICULATE MATTER I N INTRAVENOUS FLUIDS 277 For each batch of solution, a set of 6 to 10 containers was examined visually with the low-angle light-scattering apparatus described above.When all samples appeared to be similar, one or two bottles were taken at random for testing, but when appreciable differences were noticed, the best and worst bottles were taken.Table I shows the counts for selected particle diameters, the figures in the body of the table being the number of particles per ml of solution greater than the stated size. Solutions of different composition showed no systematic trends as far as could be judged from the small number of batches of each which have been examined, so the results on each manufacturer’s products have been characterised by taking the mean of all samples, supplemented by the best and worst counts from the manufacturer concerned. The results in Table I appear to come from a skew distribution. Examination of the detailed results confirm this suggestion ; comparisons of the mid-range (mean of highest and lowest counts) with the median and mean for the counts for the 3.5-p particle diameter are shown in Table 11.Table I11 shows the COMPARISON OF MID-RANGE Manufacturer A1 A2 B C D E F CT TABLE I1 VALUES FOR THE COUNTS OF THE 3.5-p PARTICLE DIAMETER Mid-range Median Mean 7384 2420 3401 794 510 586 1291 470 692 96 37 50 430 252 296 2405 356 1092 121 35 73 54 51 48 The presence of a tail on the high side of the distribution is most marked in the A1 and E series. results for the best and worst bottles from a given sample in cases where visual examination showed differences. Since the number of bottles for each batch was small, the differences do not fully indicate the range within a batch. TABLE I11 VARIATIONS WITHIN A BATCH SHOWN BY COMPARISON BETWEEN BEST AXD WORST BOTTLES 2-11 particle 3-5-11 particle 5-p particle 10-11 particle diameter diameter diameter diameter Manu- & & & & facturer best worst best worst best worst best worst A 2826 B 1326 D 958 418 1740 614 E 1454 1266 G 130 5294 1780 2404 698 2880 1072 4092 1986 308 562 310 126 94 418 198 356 230 44 902 576 208 178 598 210 1316 348 92 192 98 36 39 104 68 162 86 (14) 234 202 44 86 202 80 332 102 28 Two sets of samples were obtained from manufacturer A, the second being of solutions made early in 1965. The second set gave, on the whole, lower counts than the first, and the differences justified listing the two sets separately.No significant trends were apparent within the range of samples from other manufacturers, but qualitative comparisons of earlier samples showed that some manufacturer’s products had improved appreciably over a period of 2 or 3 years before these samples were taken.On the other hand, products made by manufacturer C in 1961 were very similar to current production. Very low counts suffer from large errors because there are a few spurious pulses, including one or two due to switching. Counts less than 20 have been placed in parenthesis to show that they are of doubtful accuracy. An investigation of these spurious pulses is being made; if they are randomly distributed in time, increasing the volume counted will not by itself increase the accuracy. Errors only become appreciable with the larger particles in the better quality solutions.278 VESSEY AND KENDALL: DETERMINATION OF [Analyst, VOl. 91 The results shown in Tables I and I1 indicate a wide variation between samples from any one manufacturer (the standard deviation- being comparable to the mean values).It is, however, clear that there is a considerable difference in over-all quality of solutions from different manufacturers in respect of particulate contamination. SOURCES OF PARTICLES The particles may arise from many sources-inadequate filtration, particles detached from the filter, particles in the bottles as received, contamination by the wash-water and particles derived from the closures. Garvan and Gunnerl92 consider that rubber bungs are the main source. Some manufacturers are using rubber bungs coated with a flexible lacquer. In one manufacturer’s products, lacquered bungs were used in 6 batches scattered through the production dates, and in the absence of other systematic trends could be used as a comparison of the two types. TABLE I V COMPARISON OF LACQUERED AND NORMAL BUNGS 2-p particle 3.5-p particle 5-p particle 10-p particle Number diameter diameter diameter diameter of &&&& Bungs samples mean best worst mean best worst mean best worst mean best worst Normal 23 13850 4622 38182 4120 1596 14434 1580 422 5148 183 42 424 Lacquered 6 4939 2262 7824 1049 376 1856 315 126 708 52 14 120 The mean and extreme results are shown in Table IV.The means show about a %fold improvement in the counts for bottles with lacquered bungs, although there was appreciable overlap between the individual results. Samples of these two types of bungs were supplied by the manufacturer concerned. One bung of each type was rinsed several times with water previously filtered through a 0-45-p Millipore filter, then autoclaved in a flask containing 100 ml of 0.9 per cent.sodium chloride solution. A batch of solution treated similarly, without a bung, was used as a control. It is clear that the lacquered bungs can be The counts are shown in Table V. TABLE V COUNTS ON SOLUTIONS iiUTOCLAVED IN CONTACT WITH RUBBER BUNGS Bungs 2-p particle diameter 3.5-p particle diameter 5-p particle diameter Control (no bung) . . 1812 346 126 Lacquered bung . . . . 1992 426 150 Unlacquered bung . . 75354 9252 1500 Lacquered bung : control 680 Unlacquered bung : control 74042 80 8906 24 1374 made almost particle-free by simple rinsing. The comparison does not, of course, show that all lacquered bungs are better than unlacquered ; there are considerable differences in composi- tion of the rubber compound, especially in the types of filler used, finish of mould surfaces, etc., and a great deal will also depend on the properties of the lacquer used.I t must be strongly adherent and flexible enough not to crack where the bung is deformed or abraded. Solutions in plastic containers showed the lowest counts (manufacturers C, F and G). Two manufacturers used heat-sealed polythene bottles and one, PVC bags with rubber closures. None of the foregoing, of course, necessarily shows that rubber-closed glass con- tainers cannot be made with low counts, only that the products in plastic containers are, on the whole, more free from contamination than those in glass. There is an overlap between counts for the worst plastic containers and the best glass bottles.PARTICLE SIZE DISTRIBUTION Most of the samples examined give a reasonable approximation to a straight line when Very low counts, which are almost certainly too high, log count is plotted against log size. deviate from this. -4 few results are shown in Fig. 3.April, 19661 PARTICULATE MATTER I N INTRAVENOUS FLUIDS 279 CONCLUSIONS It is clear that there is a wide range in the degree of particulate contamination of solutions at present on sale. Samples from each manufacturer show a considerable spread, but there are significant differences between the means for various manufacturers. A particle size distribution made with a resistance-type counter appears to be the most convenient and accurate means of quantitative analysis of particulate matter.The choice of limits for an acceptable solution must be an arbitrary one, and will need to be adjusted as manufacturing practice improves and medical evidence on the harmful effects accumulates. A t present, it is suggested that a solution is acceptable if the counts per ml are less than 1000 particles at 2-0 p, 250 particles at 3.5 p, 100 particles a t 5 p and 25 particles at 10 p. A 5-fold reduction in these limits may be a reasonable ultimate objective. Improvements in the technique for examination of solutions are needed to cover two deficiencies- (a) low counts are not accurate, and some modification of the sampling and counting system will be necessary to increase accuracy. This will become more important as the solution quality improves; to count particles in dextrose solutions and other non-electrolytes it will be necessary to add an electrolyte in such a way that no uncontrolled contamination is introduced. Techniques for doing this are now being studied. Further investigation is needed to provide supplementary tests for specific particle types, e.g., fibres, but the quantitative assessment of over-all contamination is an essential first step to better control of solution quality. (b) I I I l l I I I I I 0.3 0 4 0.5 0 6 0.7 0 8 0 9 1.0 I 1 1.2 I Log of particle size Fig. 3 . Particle-size distribution for samples of: curve A,, 0.9 per cent. sodium chloride solution; curvc A,, potassium-free maintance fluid; curve B, 4.3 per cent. dextrose in 0.18 per cent. sodium chloride solution; curve C , Darrow's solution; curve I), tissue repair solution, No. 1 ; curve E, 5-0 per cent. dextrose in 0.9 .per cent. sodium chloride solution; curve F, 0.9 per cent. sodium chloride solution REFERENCES 1. 2 . -,- , Ibid., 1964, 2 , 1 . 3. 4. Garvan, J . M., and Gunner, R. W., Med. J . Azist., 1963, 2, 140. Groves, M. J . , and Major, J . F. (;., Pharm. J . , 1964, 193, 227. Coulter Electronics I,td., instructions supplied with Coulter counter. Received September 15th, 1965

ISSN:0003-2654    

DOI:10.1039/AN9669100273

出版商:RSC    

年代:1966

数据来源: RSC

摘要:

280 SHORT PAPERS [Analyst, Vol. 91 SHORT PAPERS The Amperometric Titration of Submillinormal Concentrations of Iodine with Mercury(1) Perchlorate BY JOHN T. STOCK (Department of Chemistry, The University of Connecticut, Storrs, Connecticut, U.S.A .) IODIR’E can be determined by mercury(1) titration to a visual or a potentiometric end-point.l The present work concerns the amperometric mercury(1) titration of submillinormal concentrations of iodine by methods similar to those used for iron(III)2 and copper(^^).^ EXPERIMENTAL REAGENTS- Use analytical-grade reagents and distilled or de-mineralised water throughout. Mercury(1) perchlorate, 0.01 to 0.02 N in N perchloric acid-Prepare as described by Berka Standardise by et al.,l but suitably dilute with N perchloric acid. titration of iodine as indicated below.Potassium iodate, 0*01000 N (= 0.001667 M). Sulphuric acid, approximately 0-1 N. Potassium iodide, approximately N in boiled-out water. Store over metallic mercury. APPARATUS- Use conventional apparatus for amperometric titration a t a rotating platinum electrode.3 Since titrations are carried out at zero potential with respect to the saturated calomel electrode, 0 0.06 c.12 0.18 0.24 Volume of titrant, ml Fig. 1. Pre-addition titration of iodine in sulphuric acid and 0-02N potassium iodide. P and R are the pre-addition and residual current lines, respectively. Iodine, sulphuric acid and approximate titrant norm- alities are, respectively: curve I, 2.6 x 0.1, and 0.002 (left ordinate and upper abscissa); curve 11, 5 x 5.0, and 0.02 (right ordinate and lower abscissa)April, 19661 SHORT PAPERS 281 a polarising device is not needed.At the end of each day clean the electrode by the method described previously.3 At the beginning of each day, pre-condition the platinum electrode by running a preliminary titration of iodine until the titrant is present in slight excess. PROCEDURE- Insert the platinum electrode and salt bridge and de-oxygenate with a stream of nitrogen. Add 1 ml of N potassium iodide, stop the gas stream and inject sufficient 0.01 N potassium iodate to produce iodine, equal to about 20 per cent. of that contained in the sample. After 1 minute, note the current reading, P, then a t once inject the sample solution. Read the current after a further minute, then titrate with approximately 0.02 N mercury(1) perchlorate until the current has fallen to, or near, zero. Allow an interval of 1 minute between a titrant addition and the reading of the current.Find the end-point graphically as the intersection of the linear portion of the titration curve and the line: current = P. Note the residual current, R, then a t once inject the sample solution and titrate as in ( A ) . Find the end-point graphically by producing the linear portion of the titration curve to cut the line: current = R. ( A ) Place 50 ml of 0.1 N sulphuric acid in the titration cell. (B) Proceed as in ( A ) up to the stopping of the gas stream. RESULTS AND DISCUSSION In all experiments, iodine was produced by the addition of a controlled excess of potassium iodide and an aliquot of standard potassium iodate solution to 50 ml of an acid medium through which nitrogen had been bubbled for 25 to 35 minutes.The results given in Table I are expressed as the apparent normality of the titrant, and were obtained a t room temperature in the range 21° to 24" C. End-points were located by procedures ( A ) and (B) and, when possible, by a con- ventional method involving the extrapolation of the arms of an L-shaped titration curve.* Currents in the post-equivalence region were anodic and tended to increase with time.2 p 3 All observations in this region were therefore completed within a total time of about 4 minutes. Not recommended because of this current instability, the L-curve method failed completely when the iodine con- centration was less than about N (Fig.1, curve I). TABLE I TITRATION OF IODINE IN ACIDIFIED 0.02 N POTASSIUM IODIDE WITH APPROXIMATELY 0-02 N MERCURY (I) PERCHLORATE Iodine* concentration, PN 50 50 50 50 50 50 50 50 50 50 26 8 2.6 0-8 Apparent mercury( I) millinormality A 7 -7 Acid medium Procedure A L-curve Procedure B 0.3 N potassium thiocyanate - 0.02 N per- 0.3 N potassium thiocyanate - 0.02 N per- 0.02 N perchloric acidt . . . . . . 19.9 19.2 . 19.6 0-1 N perchloric acid . . . . . . 21.0 20.4 20.8 chloric acid . . . . . . .. 20.8 20.3 20.5 chloric acidt . . . . . . .. 20.0 19.5 19.9 0.1 N sulphuric acid . . . . . . 21.0: 21.0$ 21.lf 0.5 N sulphuric acid . . . . . . 20.8 20.9 21.0 4.0 N sulphuric acid . . . . . . 18.9 19.1 19.1 1.0 N sulphuric acid . . . . . . 20.5 20.4 20.4 3.0 N sulphuric acid .. .. . . 19.8 19.7 19-7 5.0 N sulphuric acid . . . . . . 18.2, 1'7.2, 16.0 17.9, 17.3, 15.5 18.0, 17-4, 15-6 0.1 N sulphuric acid . . . . . . 20.3 20.6 20.8 0.1 N sulphuric acid . . . . . . 20.8 failed 21.9 0.1 N sulphuric acid . . . . . . 2.055 failed 2.035 0.1 N sulphuric acid . . . . . . 2.105 failed 2.205 * Introduced as potassium iodate. t 0.1 N potassium iodide. j. Means of 7 runs; standard deviations rtO.27, kO.42 and hO.36 for procedure (A), L-curve, Q Titrant diluted 10-fold. and procedure (B) , respectively. The 0.3 N potassium thiocyanate - 0.02 N perchloric acid - 0.02 N potassium iodide medium used previously3 was satisfactory, and the omission of potassium thiocyanate had no significant effect. The concentration of potassium iodide is not critical, but should be kept quite low t o avoid positive errors in the determination of iodine.These errors may be due to slight oxidation of iodide by residual traces of dissolved oxygen.282 SHORT PAPERS [Analyst, VOl. 91 Titration is also possible in a sulphuric acid - potassium iodide medium. The acidity is not critical, and the residual current a t a properly cleaned and pre-conditioned platinum electrode is small (typically less than 0.05 pA at an electrode of sensitivity 146 p A per millimole of silver(1) per litre).3 In 0.1 K sulphuric acid - 0.02 N potassium iodide, titrations of 5 x 1 0 - 5 N iodine by procedures ( A ) and ( B ) are precise to within 1.5 per cent. and 2 per cent., respectively. N iodine were obtained a t an acidity as great as 5 N (Fig. 1, curve 11), but the results differed by nearly 20 per cent. from those obtained in 0.1 N acid. These results in 5 N sulphuric acid were reproducible to within about 5 per cent., so that this medium is fairly satisfactory if it is also used in the standardisation of the titrant. This work was carried out with the partial support of the United States Atomic Energy Commission (Contract AT(30 1)-1977), and was completed a t the Imperial College of Science and ‘Technology, London. The facilities afforded by the College authorities, in particular by Professors R. M. Barrer and T. S. West, are gratefully acknowledged. Normal titration curves of 5 x REFERENCES 1. 2. 3. 4. - , Berka, A., Vulterin, J., and Z$ka, J., Chemist Analyst, 1963, 52, 122. Stock, J. T., and Heath, P., Analyst, 1965, 90, 403. Stock,,!. T., Ibid., 1966, 91, 2’7. Amperometric Titrations,” Interscience Publishers Inc., N e w York, 1965, chapter 1. Keceived October lst, 1965

ISSN:0003-2654    

DOI:10.1039/AN9669100280

出版商:RSC    

年代:1966

数据来源: RSC

摘要:

282 SHORT PAPERS [Analyst, VOl. 91 Replacement of Benzidine by Copper Ethylacetoacetate and Tetra Base as Spot-test Reagent for Hydrogen Cyanide and Cyanogen BY F. FEIGL (Laboratorio da Produpdo iwineral, Ministerio das Minus e Energia, Rio de Janeiro) AND V. ANGER (Research Laboratory, Lobachemie, Vienna) RECENTLY the Society for Analytical Chemistry issued a warning1 against the use of benzidine, which has been recognised as a carcinogenic compound. The Ministry of Labour is therefore proposing a general prohibition of its manufacture, handling, storage and use. As the authorities in other countries are expected to follow suit, chemical analysis and especially spot-test analysis will be greatly affected. About sixty spot-test procedures are known in which benzidine is used as a reagent.The appeal made by the Society for Analytical Chemistry to replace benzidine by alternative non-toxic reagents is therefore of the utmost importance. The use of benzidine as a reagent in spot-test analysis depends almost entirely on its ability to form a blue quinoidic oxidation product from the base,2 as a result of direct or induced redox reactions mainly carried out in an acetic acid medium. The following applications come into this category- INORGANIC SPOT-TEST ANALYSIS3- Detection of water-soluble oxidants. Tests for metal ions that give water-insoluble higher oxides. Tests for acids and salts that reduce mangancse dioxide. Tests for the metallic and non-metallic elements that form heteropolyacids with molybdenum Tests for normal and complex cyanides by the generation of volatile hydrogen cyanide.Tests for mercuric cyanide through the pyrolytic release of cyanogen. Tests for elemental sulphur, selenium and tellurium by dry heating with mercuric cyanide.4 trioxide. ORGANIC SPOT-TEST ANALYSIS'- Preliminary tests of ignition residues containing manganese, lead, phosphorus pentoxide and Detection of pyridine and its derivatives via glutaconicaldehydc, and of the aldehydes that form Tests for compounds that reduce manganese dioxide, e.g., ascorbic acid. Detection of the pyrolytic release of hydrogen cyanide and/or cyanogen from certain nitrogen- Detection of the functional groups that release hydrogen cyanide by the action of nascent hydrogen. Detection of the functional groups that react pyrolytically with mercuric cyanide to produce arsenic pentoxide.coloured Schiff bases with benzidine. containing compounds. volatile hydrogen cyanide or sulphur dicyanide.April, 19661 SHORT PAPERS 283 In all of these tests benzidine can be replaced by o-tolidine, with the exception of those for hydrogen cyanide, cyanogen and the dicyanides of sulphur, selenium and tellurium. Tetra base (4,4'-tetramethyldiaminodiphenylmethane) , already recommended by Trillats for detecting traces of lead dioxide, has also been shown to be a suitable reagent for detecting the presence of higher metal oxides as well as the formation of phosphomolybdates. There are certain difficulties to be overcome in replacing benzidine in the well known test for hydrogen cyanide gas described by Sieverts and Hermsdorf.? This test is based on the appearance of a blue colour (“benzidine blue”) when hydrogen cyanide or cyanogen gas comes into contact with filter-paper moistened with an aqueous solution of copper acetate and benzidine acetate.The replacement of benzidine by o-tolidine, dianisidine, diphenylbenzidine or by tetra base does not produce a satisfactory result. We have examined the behaviour of mixtures of copper salts with free tetra base in non- aqueous solutions, and found that mixtures of tetra base with copper oleate or copper naphthenate, show a distinct blue colour when exposed to hydrogen cyanide or cyanogen gas. The deep colour of these copper salts is, however, a disadvantage in detecting the blue colour produced in the cyanide test. The behaviour of mixtures of tetra base and inner-complex copper salts, dissolved in chloro- form, was surprising. It has been stated that copper hydroxyquinolinate, copper salicyldoxime, copper cupferron, copper neocupferron, copper picolinate and copper acetylacetone do not give this reaction with tetra base.In contrast to these salts, however, the inner-complex copper ethyl- acetoacetate was found to give perfect results. Obviously the chelation of copper in this salt is not as stable as in the other inner-complex salts mentioned. This non-toxic reagent can therefore be recommended for use in the detection of hydrogen cyanide and cyanogen. The colour-test described below is based on the formation of an oxidation product of tetra base (I) ; it contains the quinodic or carbonium cationa (11) or (IIa). ( C H 3 ) 2 N ~ ~ - C H 2 , ~ ~ ( c H 3 ) 2 + Lo] + H+ * (11) -H2O { ( c H 3 ) 2 N - r \ CH ~ > - - N ( C H ~ ) ~ (1) u- \J (114 [Cl,H2IN,I+ I C,,H2%N2 This redox reaction occurs in the presence of higher metal oxides and also with chlorine, bromine or iodine vapours.As a contrast, the weak oxidants copper ethylacetoacetate (CuAcJ and cyanogen have no effect. Obviously the equilibria of the redox reactions SCuAc, + C17H2,N2 + 2CuAc + [C17H2,NJAc + HAc .. - - (1) (CN)2 f C17H22N2 [CI~H~~N~ICN + HCN .. .. * - (2) lies so far to the left that no colour appears. However, when the condition of (1) or (2) is realised, in the presence of copper ions or cyanide ions, the formation of the coloured oxidation product is suitable for detecting hydrogen cyanide and cyanogen. The net reactions are- 2CuAc2 + C17H2,N2 + 2HCN +- 2CuCN + [C17H2,N2]Ac + 3HAc * * (3) ZCuAc, + 2C1,H2,N2 + (CN), -+ 2CuCN + 2[C1,H2,N2]Ac + 2HAc .. (4) The necessary VVhen exposed to ammonia vapours the colourless It must be emphasised that only salts with the cations (11) and (IIa) are blue. acid is formed as in equations (3) and (4). carbinol base is formed. With the addition of acetic acid the blue colour reappears. METHOD REAGENT- Dissolve about 5 mg each of copper ethylacetoacetate* and tetra base in 1 to 2 ml of chloroform Although this slightly green solution is stable for some days when stored in closed bottles, it is recommended that a freshly prepared soh tion should be used. * The pure salt is manufactured by, and available from, Lobachemie, Vienna IXX, Austria.284 SHORT PAPERS [Analyst, Vol.91 PROCEDURE- Liberate the hydrogen cyanide or cyanogen, by wet destruction or pyrolysis, in a micro test- tube covered with filter-paper moistened with a drop of reagent. Previous evaporation of the solvent is not necessary. The appearance of a blue circular fleck on the almost colourless reagent paper indicates a positive result. The blue colour disappears when exposed to ammonia vapour. The limit of identification is 1 microgram of hydrogen cyanide. DISCUSSION The advantages of this test are that the reagent, whose components are stable, is easily prepared and that the blue fleck formed on the reagent paper is stable for some days. This is not so when the aqueous, unstable, benzidine-containing reagent of Sieverts and Hermsdorf is used; the latter forms black copper sulphide in the presence of the hydrogen sulphide which is evolved simultaneously with the hydrogen cyanide.In contrast to this, the new reagent is effec- tive even in the presence of some hydrogen sulphide with the hydrogen cyanide. The colour test is characteristic for volatile hydrogen cyanide and cyanogen and also sulphur, selenium and tellurium dicyanides. Cyanogen can be identified in the presence of hydrogen cyanide by a colour test described elsewhereQ that does not involve the use of benzidine. 1. 2. 3. 4. 5. 6. 7. 8. 9. REFERENCES Proc. SOC. Analyt. Chem., 1068, 2, 69. Schlenk, W., Justus Liebigs Annln Clcem., 19OS, 363, 313. Feigl, F., “Spot Tests in Inorganic Analysis,” Fifth Edition, Elsevier Publishing Co., Amsterdam ; Cleaver-Hume Press Ltd., London; D. Van Nostrand Co. Inc., h’ew York and Toronto, 1956. Feigl, F., and Del’Aqua, A., Chemist-Analyst, 1965, 54, in the press. Feigl, F., “Spot Tests in Organic Analysis,” Elscvier Publishing Co., Amsterdam, London, New Trillant, A., C.R. Lebd. Skanc. Acad. Sci., Paris, 1903, 136, 1205. Sieverts, A., and Hermsdorf, A., 2. angew. Chem., 1921, 34, 3. Wizinger, R., “Organische Farbstoffe,” Berlin, 1933, p. 11. Feigl, F., and Hainberger, L., Analyst. 1055, 80, 807. York and Princeton, 1960. Received A ugzd 12th, 1965

ISSN:0003-2654    

DOI:10.1039/AN9669100282

出版商:RSC    

年代:1966

数据来源: RSC

摘要:

284 SHORT PAPERS [Analyst, Vol. 91 A Projection Method for Inspection of Ampoules BY C. E. KENDALL (National Biological Standards Laboratory, Canberra, A .C. T., Australia) A MODIFIED 35-mm slide projector has been used to project an enlarged image of the contents of ampoules, vials, etc., for detection of particulate matter. The key point is that the ampoule is immersed in a rectangular glass tank (about 2 inches from front to back) filled with a liquid having a similar refractive index to the contents; water is suitable for most inspections. An ampoule in air gives a very poor image because of reflections that result in most of the image being dark; poor results are also obtained if the immersion liquid matches the refractive index of the glass. The use of water as immersion fluid, however, gives a clear image with the glass walls apparently seen in section, and any particles of adequate size can be detected even if not sharply in focus.A magnification of about 26 is recommended, enabling particles over about 40 microns to be detected and sizes to be estimated readily; a t this magnification, the whole contents of ampoules up to about 10ml can be seen at once. There are so many types of projector that it is difficult to give much detail on the modifications needed. A projector of fairly open design is easier to adapt ; some have a fair amount of open space between the slide carrier and projection lens, so removal of the slide carrier would provide enough space for a suitable tank. With other designs part of the casing can be cut away, or the lens mounting can be fitted to a bracket to bring it 1 to 2 inches farther away from the condenser t o allow space for the tank.A projector with a 100-mm lens gives the required magnification a t a distance of about 6 feet; a 150 to 200-watt lamp in a projector with an f/3.5-lens gives enough illumination for use in a moderately darkened room, but a lamp of higher wattage allows the lens to be stopped down to about f/5.6, or less, and this increases the depth of focus so that all particles are in reasonably sharp focus.April, 19661 SHORT PAPERS 285 In use, the ampoule is shaken, preferably with a swirling motion so that the particles move in a spiral, and quickly placed in a position in the tank that brings the side walls sharply into focus.Air bubbles quickly clear, apparently falling because the image is inverted; they can readily be distinguished from large and heavy particles, while smaller particles usually remain suspended for an appreciable time. The swirling motion of the contents brings all particles into focus twice in each rotation, even if the depth of focus is not very great. Particles present in the tank would appear at first to be a nuisance, but they are normally well out of focus and can be readily detected by their movement relative to the ampoule. Passing a stream of water, filtered if necessary, through the tank keeps the number of these particles to a negligible level. Continuous replacement is necessary because fresh particles are introduced with each ampoule. The projection method places the examination on a reasonably quantitative basis and makes it possible to propose a standard for acceptance of injectable solutions. It is suggested that a batch of ampoules should be considered satisfactory if, in any 10 ampoules taken a t random, (a) total counts do not exceed 10 particles over 50 microns and (b) there are no particles over 1 mm diameter. There is much less operator fatigue than in visual inspection, so the method should be of value for inspection on a production line, as well as for examination of check samples. Received September 14th, 1965

ISSN:0003-2654    

DOI:10.1039/AN9669100284

出版商:RSC    

年代:1966

数据来源: RSC

摘要:

April, 19661 SHORT PAPERS 285 The Determination of Water in Beryllium Oxide BY L. E. SMYTHE AND T. L. WHATELEY (Australian Atomic Enevgy Commission Researcla Establishment, Sydney, N.S. W., Aujtvalin) THE water content of various beryllium oxide powders is of interest for correlation with sintering studies. By thermal analysis and infrared studies,l it has been shown that water chemisorbed on beryllium oxide is firmly held, and is removed by heating in vacuum only a t 800" C. Therefore methods dependent upon the complete removal of all chemisorbed water (e.g., weight loss on heating to 800" C, or heating to 800" C in a stream of dry gas and determination of water by absorption) although possible, are complicated by factors such as the simultaneous loss of sulphate and surface carbonate present.The Karl Fisclier method2 was found to be unsatisfactory because it is im- possible to obtain a reproducible end-point for titration, even by using a null-point conductivity technique with the oxide stirred in methanol. This is due either to a slow reaction between the Karl Fischer reagent and the chemisorbed hydroxyl groups, or to the reaction of the reagent with beryllium oxide, similar to that which occurs with calcium oxide, magnesium oxide, zinc oxide, aluminium oxide, mercuric oxide and cuprous oxide.3 We have investigated three methods for determining the water content of beryllium oxide: (.I) infrared spectrophotometry of the oxide as a mull; (ii) exchange with deuterium oxide; (iii) reaction with 2,2-dimethoxypropane. All these methods have an infrared finish.The recommended procedures are given here, together with essential details of the experimental work. EXPERIMENTAL The infrared spectrophotometry was carried out with a model 21, Perltin Elmer, infrared spectrophotometer. Conventional equipment was used for the sample-mulling techniques which were carried out in a glove-box flushed with dry nitrogen. The method was established by examining mulls prepared under different conditions and containing various proportions of beryllium oxide (from different sources), mixed with potassium cyanide and hexachlorobutadiene. Standard samples for calibration were prepared from mixtures containing known weights of beryllium hydroxide. In the experiments based on exchange, precautions were taken to "condition" all the glass- ware that came into contact with the deuterium oxide.All operations involving the contact of deuterium oxide with the atmosphere were performed in a glove-box flushed with dry nitrogen. METHODS INFRARED SPECTROPHOTOMETRY OF THE OXIDE AS A MWLL- A mixture of approximately equal weights of sample and potassium cyanide as an internal standard (dried in vacuum a t 80" C for several hours) is accurately weighed out. This is thoroughly286 SHORT PAPERS [Autalyst, VOl. 91 mixed by grinding in a large agate mortar; a sample of this mixture is then placed in a small agate mortar and ground vigorously for 2 minutes. (This and the previous operation are carried out in a glove-box flushed with dry nitrogen.) Two or three drops of hexachlorobutadiene are added and the grinding continued for a further 2 minutes.It may be necessary to add a further drop or two of hexachlorobutadiene during this period to obtain a satisfactory mull. Bradley and Potts4 indicate that care is necessary in the preparation of mulls for quantitative work and we confirm this. The mull is examined between rock salt plates with a Perkin Elmer, model 21 spectrophotometer under the following conditions : resolution 980; gain 4.5; speed 0.5 p minute-l; response 1; suppression 0. The 5000 cm-1 to 1600 cm-l region is scanned twice in order to reduce errors in absorbance measurements. A 5-fold ordinate scale expansion is used if necessary. The absorbance at 3300 cm-1 and 2050 cm-1 is measured by the base-line density m e t h ~ d . ~ I t was found impossible, with the above procedure, to obtain mulls of potassium cyanide alone that did not show a weak absorption band at 3300cm-l.To correct for this, a curve is drawn up in which absorbance at 2050 cm-1 ( A 2050) is plotted against the absorbance at 3300 cm-l (A,,,,) for a series of potassium cyanide mulls prepared by using the above procedure. The correction required to A corresponding to a measured A 2050 in a beryllium oxide - potassium cyanide mixture is determined from the curve and subtracted from the measurement to give A’3300. The ratio A’3300/A2050 is corrected for the amount of sample and internal standard in the mixture to give the ratio 2- A ’3300 A 2050 Weight of sample Weight of internal standard z=- X From the ratio 2, the hydroxyl content of the sample is determined from the standard calibration curve, constructed by preparing mixtures of one given sample of beryllium oxide with beryllium hydroxide (hydrated) ; the hydroxyl content being determined by beryllium assay.Within the range 0.5 to 5 per cent. of hydroxyl content, Lambert - Beer’s law is observed. EXCHANGE WITH DEUTERIUM OXIDE- A known weight of the oxide (about 1 g) is added to a known weight of deuterium oxide (about log, 99.781 per cent. w/w) in a glass-stoppered centrifuge tube. All apparatus coming into contact with deuterium oxide is first washed with deuterium oxide, baked at 120” C for several hours and cooled in a desiccator before use. All operations involving contact of deuterium oxide with the atmosphere are performed in a glove-box flushed with dry nitrogen. The tube is shaken for 5 minutes, spun in a centrifuge, and a sample of the liquid phase is analysed for deuterium oxide by Gaunt’s method.6 Hence the weight of water originally present in the sample is readily calculated.REACTION WITH 2,2-DIMETHOXYPROPANE- The procedure used is similar to that given by Bishop and Critchfield.7 A 5-ml portion of 0.1 N methanesulphonic acid, 2 ml of 2,2-dimethoxypropane and 20 ml, of carbon tetrachloride are transferred by pipette into a dried stoppered conical flask and a known weight of oxide (about 1 g) is added. The flask is shaken periodically for 8 hours and allowed to stand for a further 16 hours. With beryllium oxide the supernatant liquor is quite clear at this stage, and a sample is withdrawn and its infrared spectrum recorded in the region 1600 to 1900 cm-l by the use of 0.1-mm calcium fluoride cells and standard instrumental conditions.A blank determination is perfonned simultaneously. The water in the sample is determined from the acetone absorbance at 1710 cm-l from a calibration curve. RESULTS AND DISCUSSION Some results are shown in Table I. TABLE I PERCENTAGE OF WATER IN THREE BERYLLIUM OXIDE SAMPLES BY THREE METHODS Method (i), Method (ii), Method (iii), Sample infrared deuterium oxide reaction with number spectrophotometry exchange ZJ2-dimethoxypropane 1 0.9 1.02 1.1 2 1.2 1-13 1-1 3 0.3 0.36 0.3April, 19661 SHORT PAPERS 287 All three methods gave satisfactory results. The time taken for a determination by means of method (i) is approximately 30 minutes.An advantage of this method is that a surface reaction or exchange is not involved. Method (ii) is the most accurate of the three methods, and also the most sensitive to small amounts of water. Method (iii) is the most time-consuming, but it is possible to process a number of samples without a great deal of attention. The slow reaction of 2,2-dimethoxypropane is no doubt due to steric hindrance involved in reaction with chemisorbed hydroxyl groups ; an initial rapid reaction being due to reaction with the physically absorbed water. Three different samples of beryllium oxide of commercial origin were examined. REFERENCES 1. 3. Stuart, W. I., Smythe, L. E., Price, G., and Lawrance, J. J., Paper presented at the International Conference on Beryllium Oxide, Sydney, Oct. 21st to 25th, 1963. Roman, W., and Hurst, A., “The Application of the Fischer Reagent to the Determination of Moisture in Gases,” a note contributed in a discussion on the paper by Jones, A. G., Analyst, 1951, 76, 10. Mitchell, J., and Smith, D. M., “Aquametry,” Interscience Publishers Inc., New York, 1948, 247. Bradley, N. B., and Potts, W. J., Appl. Spectrosc., 1958, 12, 77. Beavan, G. H., Johnson, E. A., Miller, R. G. J., and Willis, H. A., “Molecular Spectroscopy,’* Gaunt, J., Analyst, 1954, 79, 580. Bishop, E. T., and Critchfield, F. E., Analyt. Chenz., 1961, 33, 1034. 3. 4. 5. 6. 7. Heywood and Co., London, 1961, 286. Received May loth, 1965

ISSN:0003-2654    

DOI:10.1039/AN9669100285

出版商:RSC    

年代:1966

数据来源: RSC

摘要:

April, 19661 SHORT PAPERS 287 The Determination of Quinizarin in Hydrocarbon Oil BY K. FIELD AND E. W. GODLY (Ministry of Technology, Laboratory of the Government Chemist, Cornwall House, Stamford Street, London, S.E. 1) THE analytical method used for the detection, determination and identification of one of the prescribed markers, quinizarin ( 1,4-dihydroxyanthraquinone), in “marked” gas oils and marker concentrates has been described by Harrison and Heaysman.’ The references in this paper to the use of recrystallised quinizarin as standard and, later, to the use of commercial-quality quini- zarin for marking gas oils, with typical average optical densities of about 0.243 for 1 cm of 0.001 per cent. w/v solutions of such quinizarin in gas oil, have now been shown to be misleading, in that analysts have been basing calculations on the use of commercial quinizarin as standard.As the marking regulations refer to a concentration of 1,4-dihydroxyanthraquinone, they can be inter- preted as referring only to pure quinizarin. Consequently, all results should be based on the use of pure recrystallised quinizarin as standard. While investigating the effect of using quinizarin samples of different degrees of purity, we have shown that the repeatability of the method is influenced by lighting conditions, particularly when the alkali is present, any illumination leading to a fading in colour of the blue-violet quinizarin - sodium hydroxide complex with a reduced quinizarin recovery. The stated repeatability of 1 per cent. claimed by Harrison and Heaysmanl cannot therefore be achieved unless the lighting conditions are controlled. Consequently the analytical method has been re-examined and a stricter procedure, des- cribed later, has been developed.By using this new procedure and introducing varying times for the sodium hydroxide extraction, the effects of extraction time and of light conditions on the analysis of mixtures of 10 ml of cyclohexane (containing 10 p.p.m. of quinizarin) and 40 ml of gas oil have been examined. Table I gives typical recovery values under three general types of lighting conditions, vix., total darkness, ordinary laboratory conditions (fluorescent strip-lighting), which varied slightly according to the scattered daylight prevailing, and direct sunlight, varying with haze and occasional cloud interruption.288 SHORT PAPERS [Analyst, Vol.91 RESULTS TABLE I RECOVERY OF QUINIZARIN FROM DERV Percentage of recovery after extraction times of- A f 7 -- Light conditions 10 minutes 20 minutes 25 minutes 30 minutes 40 minutes 60 minutes Total darkness . . . . 9 6 4 96.6 - 93.3 88.0 63.6 9G.8 96.6 - 91.9 84.7 82.0 95.1 95.2 - 84.0 - - 93.1 87.6 75.1 69.4 59.9 42-6 Ordinary conditions* . . 95.8 93.8 - 81.5 - - Direct sunlight . . . . 85.8 88.6 74.8 66.9 62.6 35.0 * The levels in illumination in ordinary conditions and in direct sunlight may vary by 1 t o 100. As indicated by these results, any exposure to direct sunlight is undesirable, and even in total darkness losses result if contact with the the alkali is unduly prolonged. Under ordinary daylight conditions, with 10 minutes in the alkaline stage of each extraction, the average recovery of 95.5 per cent.compares fairly well with 96-6 per cent. under the ideal conditions of total darkness, and 10 minutes should suffice for adequate layer-separation. We therefore recommend that the following precautions should be observed- The duration of the extractions with sodium hydroxide should be minimised. (10 minutes should suffice to ensure an adequate separation of the oil and alkaline layers.) Bright illumination conditions, especially direct sunlight, should be avoided. The alkaline extract should be transferred to a second separating funnel containing hydrochloric acid. (6 ml of 1 to 1 acid should suffice for three alkali extractions.) Adequate time should be allowed for the separation of the final cyclohexane extract from the aqueous parent layer. Short separation periods sometimes result in the presence of a spray of fine aqueous droplets in the cyclohexane extract, giving rise to optical-density readings which are up to 25 per cent.too high. It is essential that the peak maximum be located for each sample of quinizarin examined. It has been noted that the indicated wavelength of the maximum of the absorption peak at 520 mp can vary from 619 to 521 nip. PROCEDURE Shake 50 ml of the gas oil, suspected to contain quinizarin, with 5 ml each of sodium hydroxide solution (5 per cent. W/V) and butanol in a 100-ml separating funnel for 45 seconds. When the two layers have separated (this may take up to 10 minutes) run the aqueous phase into a second separating funnel containing 6 ml of hydrochloric acid (1 to l ) , and wash it through with about 2 ml of water.Extract the oil with a further 5 ml of sodium hydroxide solution by shaking the funnel for 45 seconds. Allow the layers to separate and run the aqueous layer into the second separating funnel. Kepeat this operation with a further 5 ml of sodium hydroxide solution. If the third extract is not colourless, extract with further 5-ml portions of sodium hydroxide, if necessary introducing some hydrochloric acid into the second separating funnel. Add 10 nil of the spectroscopically pure cyclohexane to the acidificcl extract and shake the funnel for 30 seconds. After separation of the two layers, run the aqueous layer to waste.Use a dry pipette to transfer the amber liquid to a l-cm cell of a suitable spectro- photometer. With purc cyclohexane as blank, determine the position of the absorption peak around 519 to 521 mp and read the optical density. This is the most pronounced peak and it is also the farthest from any possible interference from components of the gas oil. A standard for comparison purposes can be prepared as follows-10 ml of a standard quinizarin solution in cyclohexane (10 p.p.m.) are added to 40 ml of quinizarin-free gas-oil. The revised procedure, previously described, is then applied to this mixture and the optical density at the absorption maximum is determined and compared with the optical density of the original standard solution in cyclohexane. Recovery, R, should be at least 96 per cent. The results of the quinizarin analyses of a batch of samples may then be adjusted by a factor of lOO/R. Wash thc extract through with water as before.April, 19661 SHORT PAPERS 289 The repeatability of the revised method is found to be less than 1 per cent. on solutions of quinizarin in unmarked gas oil. However, recent experience in analyses of marked gas oils has provided a mean difference between 13 pairs of duplicate samples of 2.5 per cent. and a maximum difference of 5-8 per cent. Where identification of quinizarin is required, the absorption spectrum from 420 to 540 mp is plotted with a suitable automatic-recording spectrophotometer. This note is published with the permission of the Government Chemist. REFERENCE 1. Harrison, R. B., and Heaysman, L. T., Analyst, 1961, 86, 506. Received June 2nd, 1965

ISSN:0003-2654    

DOI:10.1039/AN9669100287

出版商:RSC    

年代:1966

数据来源: RSC

摘要:

April, 19661 SHORT PAPERS 289 The Effect of Particle Size on Back-scattered X-ray Correction Methods in On-stream X-ray Fluorescence Analysis BY K. G. CARR-BRION (Warren Spring Laboratory, Stevenage, Herts.) THE measurement of back-scattered fluorescent X-rays from a block containing a suitable element, placed on the side of the slurry flow remote from the X-ray source has been described.' It was used as a means of correcting for variations in the fluorescent X-ray intensity from the slurry, caused by changes in the composition and content of the solid component of the slurry. If this method is used for on-stream analysis without considering the effect of changes in particle size, further errors could occur. It can be predicted quite easily from basic X-ray absorption theory2 that this back-scattered fluorescent X-ray intensity should increase as the particle size in a slurry of constant Composition increases.However, it has also been shown3 that the fluorescent X-ray intensity from particles in an aqueous slurry will decrease as the particle size in a slurry of constant composition increases. Percentage of solids in slurry Fig. 1. Effect of change in particle size on back-scattered cadmium I<, intensity : curve A, particles from 100 to 150 mesh; curve B, particles less than 300 mesh Percentage of iron in slurry Fig. 2. Effect of change in particle size on iron I<, intensity from the slurry: curve A, particles from 100 to 150 mesh; curve B, particles less than 300 mesh I t was decided to test these predictions experimentally, with a Phillips PW 1540 X-ray spectrograph adapted to take an inverted slurry presenter.It was found that the fluorescent intensity from the reference block-in this case made of cadmium-increased with increasing particle size in the haematite - water slurry (see Fig. l), while the fluorescent iron Ka intensity from the particles in the slurry decreased (see Fig. 2). Similar effects would be expected with all elements capable of being determined on-stream by X-ray fluorescence analysis.290 SHORT PAPERS [Analyst, VOl. 91 A decrease in the size of the X-ray absorbing particles will therefore cause a marked increase in the apparent solid content (see Fig. 1). This solid content is used, either directly or indirectly, to obtain the concentration of the element being determined in the solid component, since the fluorescent X-ray intensity is a function of the concentration of the element in the slurry as a whole.Even if separate means of correcting for real changes in solid content are incorporated, the use of back-scattered radiation will in no way compensate for particle-size effects. The measured value of the X-ray mass-absorption coefficient will accentuate these effects to a degree dependent on the system being examined. Thus if the method is used to correct for variations in the com- position and content of the solid component of the slurry, a correspondingly greater control of particle-size variations is necessary. It should be mentioned that this effect has also been observed in heterogeneous solid samples when back-scattered or transmitted methods of correcting for matrix effects are used,4 but that under these circumstances, sufficiently close control of particle size makes it possible to obtain accurate quantitative results. REFERENCES 1. 2. Croke, J., and Deichert, R. W., Norelco R e p . , 1964, 11, 116. Liebhafsky, H. A., Pfeiffer, H. G., Winslow, E. H., and Zemany, P. D., i n “X-ray Absorption and Emission in Analytical Chemistry,” John Wiley & Sons Inc., New York and London, 1960, p. 14. Claisse, F., and Samson, C . , A d v . i n X-ray Analysis, 1962, 5, 336. Carr-Brion, K. G., Analyst, 1965, 90, 9. 3. 4. Received July 12th, 1965

ISSN:0003-2654    

DOI:10.1039/AN9669100289

出版商:RSC    

年代:1966

数据来源: RSC