June, 19611 NOTES 41 1 Notes THE COMBUSTION OF ORGANIC COMPOUNDS BY IGNITION IN OXYGEN: THE DETERMINATION OF CARBON AND HYDROGEN THIS Note is a preliminary report of an investigation in which the flask combustion technique was used to determine carbon and hydrogen in organic substances. The first serious attempt to determine carbon and hydrogen involved igniting a mixture of the weighed sample with a known weight of potassium chlorate. This ignition process, de- veloped by Gay-Lussac and Thenard in 1810, was carried out by dropping the pelleted mixture into a closed tube held in the vertical position and heated a t its lower end by the flame of a spirit lamp. When the pellets came into contact with the hot end of the tube, ignition occurred, with the production of oxygen, carbon dioxide and water vapour.These products were collected over mercury, and the amounts of carbon and hydrogen in the sample were determined by gasometric analysis of the mixture. The obvious drawback to this method was the crude way in which the weights of products were obtained, and it was by no means certain that the carbon dioxide had been quantitatively collected. Because of this, the principle of combustion now in common use was evolved by Liebig from earlier work by Berzelius and has not changed, even with its application by Pregl to the combustion of a few milligrams of material. Over the years, the simple operation of direct ignition has been lost, and the method has become one of a rather complex nature, in which com- plicated and expensive apparatus is used to burn the few milligrams of substance involved.On the micro scale, great effort has been, and still is being, directed towards hastening the method of analysis and making the combustion operation less tedious; rapid combustion,l the use of catalysts, such as thermally decomposed silver permanganate,2 and automation of the combustion procedure are typical examples of the means used to achieve these aims. At the turn of the century, some workers attempted to establish a method in which the sample was ignited in a Berthelot bomb under pressure of oxygen (20 atmospheres) by means of an electrically heated platinum wire. After combustion, the mixture of carbon dioxide and water vapour was swept out with a current of purified air and analysed. This method did not become established because it had some faults; for example, the presence of nitrogen oxides and oxidation products of the halogens and sulphur caused complications. In view of the success of Hempel's oxygen flask combustion technique, a s adapted to the micro scale by Schoniger3 and now an established procedure, an attempt has been made to adapt this technique to the determination of carbon and hydrogen.Some preliminary experiments have established a simple and rapid method of determining carbon and hydrogen on the micro scale, and this method has overcome the need for using a baffle-chamber combustion tube.1 In the first instance, the weighed sample (about 3 mg) was ignited by passing an electric current through a platinum-wire coil placed round the sample boat held in a spherical 200-ml flask filled with oxygen.After ignition, the carbon dioxide and water were swept out by passing a stream of purified oxygen at the rate of 30 to 35 ml per minute and determined gravimetrically. The results were not satisfactory, being low and inconsistent; the losses were not caused by leakage from the flask, but by incomplete combustion. Malissa* recently described a method in which the sample was decomposed in a fast flow of oxygen a t a high temperature. The carbon dioxide formed was absorbed in sodium hydroxide solution and determined by means of a conductivity measurement. In the light of this work, further experiments were carried out in which the sample was introduced directly into a combustion chamber heated by a furnace (at about 750" C), oxygen being allowed to flow through the chamber a t 35ml per minute.After continued sweeping for 15 minutes, the carbon dioxide and water collected in weighed absorption tubes were determined. In these experiments, the results for hydrogen were satisfactory, but those for carbon were low by about 2 per cent. Finally, it was established that correct results for carbon and hydrogen could be obtained with the required accuracy by introducing the sample into the oxygen-filled chamber, but without maintaining the flow of oxygen during combustion. As soon as the sample had burned, a matter of a few seconds after its introduction into the hot zone, oxygen was passed into the chamber to sweep out the water and carbon dioxide, which were absorbed in the conventional manner for subsequent gravi- metric determination.412 NOTES [Vol.86 DESCRIPTION OF COMBUSTION APPARATUS The combustion tube had the shape shown in Fig. 1 and was constructed from quartz tubing having a wall-thickness of 1.8 to 2 mm. The chamber, A, was 150 mm long and about 33 mm in internal diameter; it had a volume of about 130 ml at room temperature and contained about 50ml of oxygen at its operating temperature. The length of tubing from the chamber to the ground joint was 150 mm, and its bore was about 14mm. Oxygen was admitted via a side- arm, B, 150mm in length, sealed to the tube just below the ground joint. The exit from the chamber consisted of a 90-mm length of tubing, C (8 to 9 mm bore), provided with a "beak" 30 mm long and 3-6 mm in diameter.A roll of silver gauze was placed in the exit tube to retain halogen and sulphur oxidation products. The chamber was heated by an electric furnace at 720" to 750" C, and the temperature of the silver gauze was maintained a t about 550" C by another electric furnace. Water and carbon dioxide were absorbed in Flaschentrager tubes packed with suitable reagents ; when nitrogen-containing compounds were analysed, the oxides of nitrogen were retained by manganese dioxide in an absorber1 placed between the two Flaschentrager tubes. A = Combustion chamber B = Side-arm F = Platinum-wire coil C = Exit tube containing silver gauze D = Flaschentrager tube for absorbing water E = Sample holder G = Borosilicate-glass sleeve Fig. 1. Diagram of combustion apparatus The oxygen was purified by passing it through a pre-heater and a large U-tube containing soda asbestos and anhydrone and was passed into the combustion tube via a stopcock attached with rubber tubing to side-arm B.The flow of gas was controlled a t 50 ml per minute by means of a precision screw-clip and a pressure regulator. The flow was measured with a Rotameter reading to 200 ml of oxygen per minute. The sample was introduced into the combustion chamber by means of a special holder, E. This consisted of a quartz tube, 330 inm long and 6 mm external diameter, having a coil of platinum wire, F, sealed into one end. The coil was made from a piece of 16 s.w.g. wire about 170 mm in length formed into a spiral (about 25 mm long) by winding it round a piece of glass rod 8 to 8-5 mm in diameter.The other end of the quartz tube was provided with an iron core of length 70mm kept in position by an indentation in the wall of the tube. The holder was placed inside an elongated borosilicate- glass sleeve, G, sealed into the ground-joint cap closing the mouth of the combustion tube. The sleeve projected 20 mm beyond the edge of the joint, as shown in Fig. 1. Tube E fitted into sleeve G in a way such that it would slide in and out easily, without friction. The sample was moved into the centre of the combustion chamber in one operation by application of a magnet to the iron core. In preliminary experiments, the holder was pushed into the chamber by hand, the tube being fixed in a rubber sleeve attached to the ground cap. It was found, however, that high blank values were obtained after the apparatus had been used because of a leak through the rubber sleeve.The sample boat was placed in the coil for introduction into the chamber. METHOD The absorption tubes were conditioned as for a rapid combustion1 for 10 minutes, with oxygen Blank determinations were carried out, and it was established that flowing a t 50 ml per minute. the blank value for water was 0.05 mg; the blank r d u e €or carbon dioxide xvas negative.June, 19611 NOTES 413 COMBUSTION- The weighed absorption tubes were attached to the combustion tube, and the ground cap was carefully removed from the mouth of the tube. Oxygen was allowed to escape via the open tube, and then the sample boat was placed in the platinum spiral of the holder, which was still completely in the sleeve.The holder was replaced, and the joint was secured with springs. Oxygen was passed through the apparatus at 50 ml per minute for 1 minute, the stopcock was turned off, and, after a few seconds, the inlet tap - stopper of the water-absorption tube was closed. With the aid of a magnet, the sample holder was then pushed, in one movement, into the centre of the combustion chamber. As soon as the sample had burned, the stopcock connected to the side-arm was opened to permit oxygen to enter the combustion tube, the inlet tap - stopper of the water-absorption tube was opened, and the apparatus was swept out with oxygen for 13 minutes at the rate of 50ml per minute. The absorption tubes were then disconnected as described for the rapid-combustion meth0d.l TABLE I RESULTS FOR CARBON AND HYDROGEF BY THE PROPOSED METHOD Compound Anthracene .. Naphthalene . . Benzene . . Octan-2-01 . . Benzoic acid . . Succinic acid . . Sucrose . . * . Phenacetin . . p-Nitroaniline . . p-Chlorobenzoic acid Sulphonal . . Triphen ylphosphine Weight of s a m p 1 e , mg 2.127 2.200 3.745 3.545 4.283 3.675 3.900 2.481 3.154 3.227 3-230 3.896 4.678 3.748 3.886 3-313 3.498 3-061 3.493 2.684 3.573 3.642 4.685 3.184 3.847 Carbon content found, 94.62 94-46 94-25 93.62 93-84 92.35 92.18 73.70 73.97 68.75 69.02 40.87 40.50 42.01 42.20 67.09 67-21 52-24 52-26 53.76 53.75 36.92 36.74 82-60 82.52 Yo Hydrogen content found, % 5.78 5-59 5-54 6.30 6-22 7-88 7-92 13.75 13.97 5.06 4-95 5.09 5.03 6-38 6.36 7-29 7-37 4.36 4.59 3.32 3.28 7.13 7-10 5.82 5.70 Calculated carbon content, % 94.36 93-75 92.26 73.78 68-90 40.68 42.10 67.04 52.13 53.68 36.81 82.41 Calculated hydrogen content, % 5-66 6-26 7.74 13-93 4-96 5.08 6.48 7.26 4-38 3-22 7.06 5-76 DISCUSSION OF THE METHOD The method has been tested with several types of organic compounds, and typical results are shown in Table I.Samples containing carbon, hydrogen, oxygen and nitrogen (the last-named present as NH,, NO or NO,) have been analysed satisfactorily by using a combustion tube without the exit tube containing silver gauze attached; the chamber was heated with the flame of a large bunsen burner, giving an operating temperature of 725" to 75OOC. In these experiments, the chamber was enclosed in a sheath of nickel foil, the ends of which were encased in plates of thick asbestos sheet.No difficulty was experienced in analysing liquids, e.g., benzene. Liquid samples were weighed in the usual form of capillary tube and placed in a short piece of quartz tubing 65 mm long and 6.5 to 6 mm in bore. The quartz sleeve was placed in the platinum spiral for introduction into the combustion chamber.414 NOTES pol. 86 Triphenylphosphine has been analysed satisfactorily by covering the sample in the boat with tungstic oxide: and the analysis of organo-metallic compounds is being investigated. Promising results have been obtained with fluorine-containing substances, e.g., m-trifluoromethyl- benzoic acid; such samples were covered with magnesium o ~ i d e . ~ When a flux or reagent was used, the sample was weighed in a long narrow boat, which was then placed in a quartz sleeve before insertion into the platinum spiral.Initially, tests were carried out in which the sample was introduced into the hot zone by means of a short form of the holder E, which was placed immediately behind the cap of the tube, so eliminating sleeve G. However, this simple method of moving the sample holder into the hot zone was not always satisfactory, as the holder tended to rotate when moved forward by means of the magnet, and sample or added reagent was tipped out into the combustion tube. It is hoped that this simple combustion procedure will eventually be able to cope with all types of compounds difficult to analyse by other established methods, and work is proceeding with this aim in mind.The method of combustion provides a means of analysing the “difficult” organo-metallic compounds now being prepared, because suitable reagents can be sought to assist decomposition; the use of such reagents, coupled with a high temperature of combustion, permits quantitative oxidation. The possibility of increasing the rate of flow of oxygen to 100 ml or more per minute is being considered, as such an increase would cut the sweeping-out period by half. Other aspects being investigated are the determination of the water and carbon dioxide mano- metrically, so that the method could be extended to the determination of compounds containing carbon-14. The method of combustion also provides a simple way of decomposing samples for the determination of the halogens and sulphur, particularly when combustion is found to be difficult by the oxygen flask procedure.Work on these methods is in progress, and it is hoped that a full account of the investigations will be published later. REFERENCES Belcher, R., and Ingram, G., Anal. Chim. Acta, 1950, 4, 118. Korbl, J., Coll. Czech. Chem. Comm., 1955, 20, 948. Schoniger, W., Mikrochinz. Acta, 1955, 123; 1956, 869. Malissa, H., in “Proceedings of the International Symposium on Microchemistry, 1968,” Pergamon Belcher, R., Fildes, J , E., and Nutten, A. J., Anal. Chim. Acta, 1955, 13, 431. 1 . 2. 3. 4. 5. 6. Throckmorton W. H., and Hutton, G. H., Anal. Chem., 1952, 24, 2003. Press, Oxford, London, New York and Paris, 1960, p. 97. COURTAULDS LIMITED, RESEARCH LABORATORY MAIDENHEAD, BERKS.G. INGRAM Received January 5th, 1961 THE DETERMINATION O F NITRITE IN WATER MOST methods described for the determination of nitrite depend on diazotisation. The diazonium compound formed is usually determined colorimetrically, either by direct measurement1 or after being coupled with a suitable reagent. The latter principle is employed in the Griess - Ilosvay method, various modifications2 of which are commonly used in water analysis. This method, although reliable, suffers from a number of disadvantages : the colour formed is rather unstable, the reagent, I-naphthylamine, deteriorates when stored, and samples may require excessive dilution on account of the small range over which results are reproducible. The method described in this Note is largely free from these disadvantages. Garner, Baumstark, Muhrer and Pfander determined nitrate by a method: later improved by Hill, Pivnick, Engelhard and B ~ g a r d , ~ in which a bacterial culture was used to reduce the nitrate to nitrite; this ion was then determined as in the Griess - Ilosvay method, except that N-( 1-naphthy1)ethylenediamine was used instead of I-naphthylamine.Sulphanilic acid and N-( 1- naphthy1)ethylenediamine had previously been used for the indirect determination of nitrogen dioxide by Saltzman,S who found that these reagents were more satisfactory than the Gness - Ilosvay reagents or the combination of sulphanilamide and N-( 1-naphthy1)ethylenediamine introduced by Shinn.6 By modification of the conditions used by Garner and his co-workers, we have developed a convenient method for determining nitrite in water.The reagents are added separately, instead of together as in the earlier procedures, and the acidity of the solution is established with potassiumJune, 19611 NOTES 416 hydrogen sulphate, which has a greater buffering capacity than the acetic acid normally used. Beer's law is obeyed up to 2 pg of nitrite-N, and amounts up to 8 pg ca.n be determined with use of a calibration graph. (The official Griess - Ilosvay procedure2 formerly used in this laboratory does not give good reproducibility for amounts of nitrite-N greater than 1 pg.) The colour pro- duced is more intense by 13 per cent., and blank values are less than half of those obtained in the official procedure.The N-( 1-naphthyl) eth ylenediamine hydrochloride solution can be stored in the dark for at least 2 months. TABLE I COMPARISON OF RESULTS BY PROPOSED AND OFFICIAL Sample Percolating-filter effluent A Percolating-filter effluent B I Thames water from Tilbury (salinity 13.4 parts per thousand) -*I Percolating-filter effluent C . . i Digester effluent containing phenols r I Sea water from Worthing . . f . Nitrite added (as N), p.p.m. 0.0 4.0 6.0 8.0 0.0 0.1 0.2 0.3 0.0 0.2 0.4 0-6 0.0 0.2 0.4 0.6 0.0 0.1 0.15 0.2 0.3 0.0 0.1 0.15 0.0 0.4 0.8 1.2 0.0 0.4 0.8 1-2 0.0 0.05 0.1 0.15 0.0 0.05 0.1 0.15 Total nitrite (as N) found by- r proposed method, p.p.m. 3.0 7.0 9.1 0.13 0-23 0.33 0.43 0.15 0.36 0.56 0.77 0.15" 0.34" 0*53* 0.72" 0.042 0.142 0-194 0.241 0.042" 0.140" 0.183" 0.79 1-21 1.58 2.00 0.73" 1.11" 1-48" 1.88" 0-008 0-055 0.108 0.157 0.008" 0.060" 0.112" 0.162" - - official method, p.p.m.3.0 7-0 11.1 - 0-14 0.23 0.33 0-42 0.17 0.37 0.5 7 0.76 - - - - 0.040 0-135 0.234 0.339 - - - - 0-82 1-24 1.61 2.02 - - - - 0.004 *0*051 0.101 0.144 - - - - METHODS Bias (as N) of- - proposed method, p.p.m. 0.0 +0.1 - - - 0.0 0.0 0.0 + 0.01 + 0.01 + 0.02 -0.01 - 0-02 - 0.03 0.0 + 0.002 - 0.001 - - - - - - 0.002 - 0.009 + 0.02 - 0.01 + 0.01 - 0-02 - 0.05 - 0.05 - - - - 0.003 0.0 - 0.001 + 0.002 + 0.004 + 0-004 - official method, p.p.m. 0.0 + o w 1 -0.01 - 0.01 - 0.02 0.0 0.0 - 0.01 - - - - - - - - - - 0.003 - 0.006 - 0.00 1 - - I - - + 0.02 -0.01 0.0 - - - _- - - 0.003 - 0.003 - 0.010 - - - - * Large sample taken; optical density measured in 1-cm cells.NO quantitative statements on the stability of the colour produced in the Griess - Ilosvay method have been found in the literature, although there are reports of precipitation or intensi- fication of the colour. In a 24-hour test of the stability of the colour produced by the official procedure2 in this laboratory, the colour derived from a standard solution of pure nitrite decreased by 2 per cent. and those from sewage effluents showed increases ranging up to 6 per cent. Under similar conditions, the colours obtained in the proposed method remained within 2.5 per cent. of their initial intensities for 2 days. When typical samples were analysed by both methods, there was good agreement between the two sets of results, some of which are shown in Table I.The samples, which included sea416 NOTES [Vol. 86 water and water from the Thames Estuary, were then fortified with known amounts of standard nitrite solution and again analysed. The increments observed normally agreed with the calculated values, but certain samples initially extremely low in nitrite (less than 0.1 p.p.m. of nitrite-N) showed a bias towards low recovery by both procedures. For samples from the Stevenage Brook, this effect is associated with warm and dry weather conditions. -4 similar bias was observed with effluents from the biological treatment of farm wastes containing phenolic disinfectants ; for these effluents, however, the bias could be removed by taking much smaller samples for the determination. Although most of the optical-density measurements were made at 550 m p with a spectro- photometer, an identical calibration graph was obtained when a Spekker absorptiometer and a Spectrum Green No.604 filter were used. METHOD REAGENTS- sulphanilic acid in water, and dilute to 1 litre. PROCEDURE- Clarify the sample, if necessary, by filtration or centrifugation, measure a suitable volume (not more than 30 ml) into a 50-ml calibrated flask, and adjust the temperature to between 20" and 30" C. (For effluents containing phenols, use the smallest practicable sample; otherwise, take a sample containing between 0.2 and 1.0 pg of nitrite-N unless the concentration is high, when up to 8 pg may be tolerated.) Dilute to a t least 10 ml, add 2.5 ml of sulphanilic acid solution, and set aside for 10 minutes.Add 2.5 ml of N-( 1-naphthy1)ethylenediamine hydrochloride solution, set aside for 20 minutes, dilute to the mark, and measure the optical density at 550 mp against a reagent blank solution; use 4-cm cells unless the amount of nitrite-N present exceeds 1.5 pg, when 1-cm cells are satisfactory. For the most accurate work, or with turbid or highly coloured samples, prepare an additional blank solution by carrying out the procedure described above, but omit the N-( 1-naphthy1)ethylenediamine hydrochloride solution. Measure the optical density against distilled water, and subtract the value from the optical density found in the test. This correction is usually negligible. Ascertain the nitrite content of the sample by reference to a calibration graph plotted from the results obtained for standard solutions of nitrite; we have found that solutions of analytical- reagent grade sodium nitrite (previously dried at 110" C) in freshly distilled water are suitable for this purpose and are stable for a t least 1 month. The purity of the sodium nitrite was checked by titration against a standard solution of potassium permanganate.This Note is published by permission of the Department of Scientific and Industrial Research. Sul$hanilic acid solution-Dissolve 27.2 g of potassium hydrogen sulphate and 3.46 g of N-( 1-Waphthy1)ethylenediamine hydrochloride solution, 0.04 per cent.-Store in the dark. REFERENCES 1. Bark, L. S., and Catterall, R., Mikrochim. Acta, 1960, 563. 2. 3. 4. 5. 6. Ministry of Housing and Local Government, "Methods of Chemical Analysis as applied t o Sewage Garner, G.B., Baumstark, J. S., Muhrer, M. E., and Pfander, W. H., Anal. Chem., 1956, 28, 1589. Hill, R. M., Pivnick, H., Engelhard, W. E., and Bogard, M., J . d g r i c . Food Chem., 1959, 7, 261. Saltzman, B. E., Anal. Chem., 1954, 26, 1949. Shinn, M. B., Ind. Eng. Chem., Anal. Ed., 1941, 13, 33. and Sewage Effluents," Second Edition, H.M. Stationery Office, London, 1956, p. 15. DEPARTMENT OF SCIENTIFIC AND INDUSTRIAL RESEARCH WATER POLLUTIOX RESEARCH LABORATORY STEVENAGE, HKRTS. H. A. C . MONTGOMERY JOAN F. DYMOCK Received February 3rd. 1961 INTERFERENCE BY CHLORIDE ION IN THE DETERVINATION OF SHAW and McFarlane developed a convenient colorimetric method for determining tryptophan in proteins.1 The method is an adaptation of the Hopkins - Cole reaction2 and has in general proved satisfactory in this laboratory, where i t has been extensively used for analysing proteins and in studies on the preservation of tryptophan by sulphur dioxide during acid hydrolysis of proteins? TRYPTOPHAN BY SHAW AND McFARLANE'S METHODJune, 19611 NOTES 417 In work on the rates of hydrolysis of whole casein and the various fractions of the casein complex, by proteolytic enzymes and by acids, 4 g of Warner casein (containing 14.1 per cent.of nitrogen4) and 80 ml of 3 N hydrochloric acid were boiled under reflux. Samples (3 ml) of the mixture (casein hydrolysate No. 1) were taken a t various time intervals and were analysed for tryptophan by Shaw and McFarlane’s methodl and for total nitrogen by a micro Kjeldahl method.6 A sample of the casein used was also analysed and found to contain 0.10 g of tryptophan per g of nitrogen.This experiment was then repeated with 3 pa sulphuric acid instead of hydrochloric acid ; the results were- Time of heating, hours . . 0-25 0.5 0.76 1.0 1.25 1-5 2.0 2.25 2.50 2.75 3.0 Tryptophan found in HC1 hydrolysate, g per g of nitrogen.. . . . . 0.128 0.124 0.122 0.118 0.110 0.112 0.106 0.099 0.094 0.090 0.084 Tryptophan found in H,SO, hydrolysate, g per g of nitrogen . . 0.099 0.098 - 0.097 - 0.092 0.084 - 0.074 - 0.086 These figures show- that the tryptophan content of the hydrochloric acid hydrolysate was (a) higher than that of the sulphuric acid hydrolysate and (b) did not attain that of the unhydrolysed casein (0.10 g of tryptophan per g of nitrogen) until the casein - acid mixture had been heated for more than 1.5 hours.This indicated that the hydrochloric acid had interfered with the determination of tryptophan. Accordingly, a mixture of 1.0 g of casein and 20 ml of a 20 per cent. solution of sodium hydroxide was heated for 20 minutes on a steam-bath. (This is the first step in the determination of tryptophan in proteins by Shaw and McFarlane’s method.) The volume of the resulting solution was adjusted to 25 ml, and various amounts of sodium chloride were added to 1-ml portions of the solution, each containing 0.04 g of casein. Tryptophan was determined in these samples by Shaw and McFarlane’s method; the results were- NaCl added, millimoles . . .. . . 0.0 0.1 0.3 0.5 1.0 2.0 4-0 6.0 8.0 Tryptophan found, g per g of nitrogen .. 0.100 0.102 0.118 0.137 0.131 0.132 0.131 0.128 0.124 which suggested that the presence of chloride had increased the intensity of the Hopkins - Cole reaction. Lesuks has reported that chloride ions increase the intensity of the colour produced by the reaction of tryptophan with p-dimethylaminobenzaldehyde.’ Spies* was able to overcome inter- ference from chloride ions by treating the tryptophan-containing solutions with silver sulphate. Attempts were therefore made to remove chloride ion by adding silver sulphate to solutions pre- pared by treating casein with a 2.0 per cent. solution of sodium hydroxide according to Shaw and McFarlane’s procedure. However, removal of chloride ion in this manner was unsatisfactory in so far as it led to the precipitation of protein. A sample of casein hydrolysate No.1 prepared in the previous experiments was extracted with di-2-ethylhexylamine solution (1 part by volume of di-Bethylhexylamine plus 10 parts by volume of chloroform) ; this removed virtually all the chloride. The hydrolysate was extracted with freshly distilled chloroform to remove the di-2-ethylhexylamine and was then heated on a steam-bath to remove chloroform. The hydrolysate, which then contained no detectable chloride ion or chloroform, was analysed for tryptophan by Shaw and McFarlane’s method. It was observed that the blanks (extracted hydrolysate to which no glyoxylic acid had been added) were faintly pink in colour, so that it was impossible to obtain an accurate value for the tryptophan content of the hydrolysate.The pink colour appearing in the blanks had the same peak absorption as did that produced by the action of glyoxylic acid on tryptophan. These observations show that Shaw and McFarlane’s method is subject to interference from chloride ion and that this possible interference should be taken into account when the method is used. REFERENCES 1. 2. Winkler, S., Hopfie-Seyl. Z., 1934, 228, 50. 3. 4. 5. Shaw, J. L. D., and McFarlane, W. D., Canad. J . Res. B., 1938, 16, 361. Pedersen, J. W., and Baker, B. E., J . Sci. Food Agric., 1954, 5 , 549. Warner, R. C., J . Amer. Chem. SOC., 1944, 66, 1725. Lepper, H. A., Editor, “Official Methods of Analysis,” Seventh Edition, The Association of Official Agricultural Chemists, Washington, D.C., 1950, p.745.418 NOTES [Vol. 86 Lesuk, A,, U S . Pharmacopoeia Amino Acids Advisory Committee, “Report on Collaborative Study on Chemical Tests and Standards for Amino Acids,” Letter 135, November 26th, 1948. Spies, J. R., and Chambers, D. C., -4naZ. Chem., 1948, 20, 30. Spies, J. R., Ibid., 1950, 22, 1447. 6. 7. 8. MACDONALD COLLEGE OF MCGILL UNIVERSITY QUEBEC, CANADA G. 0. HEHNEBERRY B. E. BAKER Received November 28th, 1960 DETERMINATION OF COPPER IN PLATINUM JEWELLERY ALLOYS HAYWOOD and Sutcliffe’s method1 for determining copper in steel with use of biscyclohexanone oxalyldihydrazone has been extended, with certain essential modifications, to the determination of 0.05 to 6 per cent. of copper in platinum jewellery alloys.As only a small amount of sample is required, the method is particularly suitable for application t o delicate articles of jewellery, for which binary alloys of platinum and copper are commonly used. The proposed direct procedure, based on spectrophotometric measurement of the blue copper - reagent complex, is simple in comparison with older methods. O S S O l t 0.510 0 20 40 60 80 100 120 140 Time after addition of reagent, minutes Fig. 1. Variation of optical density of the copper - biscyclohexanone oxalyldihydrazone complex with time. (The values are for a copper - platinum alloy containing approximately 5.5 per cent. of copper) EXPERIMENTAL It was established that- In preliminary experiments, it was found that Haywood and Sutcliffe’s method required (i) The addition of ammonia solution must be deferred until the final-aliquot stage to avoid precipitation of ammonium chloroplatinate.(ii) Accurate adjustment of pH is necessary. This can be achieved by using neutral red indicator as recommended by Wetlesen and Gran.2.s (iii) The biscyclohexanone oxalyldihydrazone solution must be freshly prepared and added after the final adjustment of pH. (iv) The blue colour of the complex attains a maximum in about 7 minutes, after which there is slight progressive fading (see Fig. 1). Results are satisfactory, however, provided that the optical density is measured after a fixed interval from the time that the reagent solution is added; the recommended interval is 20 minutes. The relationship between copper content and optical density found by the proposed procedure is linear; an optical density of 1.000 (4-cm cell) is equivalent to 105 pg of copper in the final s o h tion.certain modifications for our purpose.June, 19613 NOTES 419 METHOD REAGENTS- Hydrochloric acid, s$.gr. 1. I -. Nitric acid, sp.gr. 1-42?. Citric acid solution-Dissolve 10 g of citric acid in 200 ml of water. Ammoniu solution-Dilute 100 ml of ammonia solution, sp.gr. 0.880, to 400 ml with water. NeutraE red ilzdicator solution-Dissolve 0.05 g of neutral red in 100 ml of water. Copper reagent solution-Dissolve 0.1 g of biscyclohexanone oxalyldihydrazone in a mixture of 10 ml each of industrial methylated spirit and hot water, dilute to 200 ml, and filter if necessary. PROCEDURE- Weigh 0.02 g of sample in the form of fine powder or drillings into a 150-ml beaker, add 7.6 ml of hydrochloric acid and 2.5 ml of nitric acid, cover the beaker, and heat on a hot-plate until solution is complete; add further portions of hydrochloric and nitric acids if necessary.(Platinum - iridium alloys may require modified treatment to effect solution.) W e n the sample has dissolved, remove the cover-glass, evaporate almost to dryness, add about 0.05 g of sodium chloride and a few drops of hydrochloric acid, and evaporate to dryness. Dissolve the residue in a little water, add 5 ml of hydrochloric acid, and dilute to 100 ml in a calibrated flask. Transfer a suitable aliquot of this solution (5ml for samples containing 2 to 6 per cent. of copper) to a TABLE I COPPER CONTENTS FOUND IN STANDARD ALLOYS Alloy No.Nominal Aliquot of alloy Copper copper content, solution taken, content found, % ml % 1 3.5 5 3-47 2 3.0 5 2-96 3 2.5 5 2-55 4 1.0 5 1.00 5 6 7 8 0.5 0.2 0.1 0.05 {l: 25 25 0.50 0.5 1 0.23 0.21 0.11 0-057 second 100-ml calibrated flask, add 5 ml of citric acid solution and 2 drops of neutral red indicator solution, and dilute to about 70 ml. Add ammonia solution until the indicator just turns yellow, and then add 1 drop in excess. Add 20ml of freshly prepared copper reagent solution, with thorough mixing, dilute to the mark, set aside for exactly 20 minutes, and then measure the optical density at 5950 A in 4-cm cells with a Uvispek spectrophotometer or similar instrument. Carry out a reagent blank determination concurrently with the sample determination, and apply any necessary correction. Determine the copper content of the sample from the corrected optical density by reference to a graph previously prepared from measurements made on solutions of known composition.(The solutions used in preparing the graph should have approximately the same content of platinum as encountered in the determination.) RESULTS A series of single deter- minations carried out on a copper - platinum alloy on successive days gave results of 5-55, 5.53, 6-53, 5.52 and 5.55 per cent. of copper. Haywood and Sutcliffe found that only iron, nickel, chromium and cobalt interfered with their rneth0d.l With platinum jewellery alloys, iron and nickel would have negligible effect on the results, whereas chromium and cobalt are unlikely to be present in amounts sufficient to cause any significant interference.The effects of some other elements occasionally found in small amounts were, however, studied ; none was found to interfere. To each of five samples of a copper - platinum alloy containing approximately 5.5 per cent. of copper was added a solution of the The reproducibility of results by the proposed method is good.420 NOTES [Vol. 86 element to be tested, the amount of element added being equivalent to a concentration of 6 per cent. in the alloy; the results were- Element added . . . . None Gold Palladium Rhodium Ruthenium Iridium Copper content found, % . . 5.55 5.53 5.55 5.53 5.57 5.54 The results obtained when the proposed procedure was applied t o eight standard copper - This Note is published by permission of the Wardens of the Worshipful Company of Gold- platinum alloys are shown in Table I.smiths. REFER.ENCES 1. 2. 3. Haywood, L. J. A,, and Sutcliffe, P., Analyst, 1956, 81, 651. Wetlesen, C.-U., and Gran, G., Svensk Papperstidn., 1952, 55, 212. Johnson, W. C., Editor, “Organic Reagents for Metals, and Other Reagent Monographs,” Fifth Edition, Hopkin & Williams Ltd., Chadwsell Heath, Essex, 1955, p. 31. ASSAY OFFICE GOLDSMITHS’ HALL LONDON, E.C.2 J. S. FORBES D. B. DALLADAY Received Ja?zuary 13t12, 1961 THE PHOTOMETRIC DETERMINATION O F SILICA IN ROCKS AND REFRACTORY MATERIALS INVESTIGATIONS by Stricklandl and Ringbom, Ahlers and Siitonen2 have demonstrated the utility of the yellow a-form of molybdosilicic acid for the photometric determination of silica in rocks and refractory materials. At lorn acidity, the a-form is stable; the 8-form is developed a t high acidity and slowly changes to the a-form.Both forms yield blue solutions when reduced by l-amino-2-naphthol-4-sulphonic acid3 or p-methylaminophenol ~ulphate,%~ and optical-density measurements are made at about 810 mp. In general, the sensitivity of this method is much too high for rocks containing 35 to 95 per cent. of silica. The great dilution involved may be a source of error, and the strongly acid solution results in initial development of the 8-form of the acid. Further, the time elapsing between development of the yellow molybdosilicic acid and its reduction is critical, and errors may also arise from deterioration of the reducing solution and the addition of reagents to mask iron.In the procedure described below, the sample is fused with sodium hydroxide,3!6 and the cooled melt, instead of being dissolved in a solution of disodium ethylenediaminetetra-acetate: is taken up in dilute sulphuric acid. This leads to much clearer solutions, particularly for rocks rich in iron. The pH of a suitable portion of the solution is then adjusted to between 3.0 and 3.1, as previously recommended,* before the yellow molybdosilicic acid is developed. The effect of iron is overcome by measuring the optical density of each sample solution against a blank prepared from that sample rather than against distilled water or a “fusion” blank. A correction is made for P20,, which is seldom present to an extent greater than 0.5 per cent.in most rocks. METHOD SPECTROPHOTOMETRY- All optical-density measurements were made in matched 10-mm glass cells with a Unicam SPSOO spectrophotometer. In accordance with the method described by Ringbom and &terholm,S the concentrations of the final solutions were adjusted so that the optical densities were between 1.0 and 1.6 at a sensitivity of 1 to 10. GLASSWARE- thoroughly rinsed with water just before use. should be stored filled with water. REAGENTS- Calibrated flasks should be set aside overnight filled with diluted sulphuric acid (1 + 1) and They should never be allowed to become dry and All reagents should be stored in polythene bottles. Sodium hydroxide pellats. Sulphzrric acid, 2.5 N.June, 19611 NOTES 421 Buffeev solution-Dissolve 200 g of chloroacetic acid in water, and dilute to 1 litre.Slowly add diluted ammonia solution (1 f l), with stirring, until the pH of the cold solution is between 3-15 and 3.20 (approximately 160 ml of the ammonia solution will be needed). Ammonium mdybdate solution-Prepare a solution containing 35.3 g of (NH,),Mo,0,,.4H20 per litre. PROCEDURE- Place 150-mg portions of standard and samples (weighed to the nearest 0.01 mg) in separate 30-ml silver crucibles, add 1-5 g of sodium hydroxide pellets to the contents of each crucible, and heat the covered crucibles in a muffle furnace for 8 to 10 minutes at 800" C. When cool, half fill each crucible with water, and warm gently on the top of a steam-bath until most of the cake has broken up. Transfer the contents of each crucible to a 500-ml calibrated flask containing about 250 ml of water and 20 ml of 2.5 N sulphuric acid, scrub the crucible and lid with water and then with 5 ml of 2.5 N sulphuric acid to remove any adhering material, and add the rinsings to the contents of the flask.Dilute the solution to the mark with water, and shake thoroughly. Transfer the solution to a dry polythene bottle, insert a stopper, and set aside to clear for 3 or 4 hour s or overnight. TABLE I SILICA CONTENTS FOUND BY PROPOSED METHOD For N.B.S. samples certified values are shown in columns 2 , 3 and 4; for rocks G-1 and W-1 the values in these columns are those recommended by various w ~ r k e r s ~ ~ * ~ ~ Total iron SiO, found present, as p20, SiO, by proposed Sample Fe,O,, present, present, method, % % % % Burnt refractory (W.B.S.No. 76) . . 2.38 0.069 64-68 64*5(1) 5 4.6 (5) 54-6(0) 54.5 (8) 59*0( 7) 59*0(5) Granite, G-1 . . . . .. .. 1-96 0.09 72-66 72.4( 7) 72.5(1) 7 2.5 (8) 72*5(5) Diabase, W-l . . .. . . . . 11-10 0.126 62-64 52-5(1) 52-5(9) 52-5(7) 52*6( 1) Plastic clay (N.B.S. No. 98) . . .. 2.05 0.08 59.1 1 69.1 (9) By pipette, place 25-ml portions of standard and sample solutions in separate 150-ml beakers, add 15ml each of buffer and ammonium molybdate solutions, and mix thoroughly. Heat the beakers in a bath of boiling water for 20 minutes, cool to room temperature, transfer each solution to a 100-ml calibrated flask, and dilute to the mark with rinsings from the beaker. Set these solutions, which must all have the same pH (3.0 to 3.1), aside for 12 hours or overnight.Prepare blank solutions for the standard and each sample in exactly the same way, but omit the addition of ammonium molybdate solution and subsequent heating. Measure the optical density of each solution against the appropriate blank solution. RESULTS The standard used throughout was National Bureau of Standards sample Xo. 99 (soda feld- spar), which contains 68.66 per cent. of SO,, 0.067 per cent. of Fe,O, and 0-142 per cent. of P,O,; this standard gave an optical density of 1.498. The lowest working concentration was about 45 per cent. of silica (optical density of 1.0) and the upper limit about 75 per cent. (optical density of 1-6). This range of concentrations covers a large proportion of common igneous rocks, and suitable adjustments to the final concentration can be calculated for samples having silica contents outside this range. The effect of phosphorus pentoxide was determined by adding this constituent to the standard up to a concentration of 3 per cent. I t was found that the presence of 1 per cent. of phosphorus422 NOTES vol. 86 pentoxide increased the optical density by 0.015 (equivalent to 0.75 per cent. of silica), and the results for silica in the samples were corrected accordingly. The results of several separate determinations of silica in two National Bureau of Standards refractory samples and two thoroughly investigated rocks, G-1 and W-1, are shown in Table I. 1. 2. 3. 4. 5. 6. 7. 8. 9. REFERENCES Strickland, J. D. H., J . Amer. Chem. Soc., 1952, 74, 862. Ringbom, A., Ahlers, P. E., and Siitonen, S., Anal. Chirn. Acta, 1959, 20, 78. Shapiro, L., and Brannock, W. W., “Rapid Analysis of Silicate Rocks,” U.S. Geological Survey Mullin, J. B., and Riley, J. P., Anal. Chim. Acta, 1955, 12, 162. Riley, J. P., Ibid., 1958, 19, 413. Ringbom, A., and Osterholm, K., Anal. Chem., 1953, 25, 1798. Fairbairn, H. W., Schlecht, W. G., Stevens, R. E., Dennen, W. H., Ahrens, L. H., and Chayes, F., “A Cooperative Investigation of Precision and Accuracy in Chemical, Spectrochemical and Modal Analysis of Silicate Rocks,” U.S. Geological Survey Bulletin 980, 1961. Fairbairn, H. W., Geochim. Cosnzochirn. Acta, 1953, 4, 143. Stevens, R. E., Fleischer, M., Niles, W. W., Chodos, A. A., Filby, R. H., Leininger, R. K., Ahrens, L. H., and Flanagan, F. J., “Second Report on a Cooperative Investigation of the Composition of Two Silicate Rocks,” U.S. Geological Survey Bulletin 1113, 1960. Bulletin 1036-C, 1956. DEPARTMENT OF GEOLOGY UNIVERSITY COLLEGE SWANSEA T. W. BLOXAM Received September 7th, 1960