摘要:

350 YUEN: DETERMINATION OF TRACES OF [Vol. 83 Determination of Traces of Manganese with Leucomalachite Green BY S. H. YUEN (Imnpertal Chemical Industrzes Ltd., jealott's Hill Research Station, Bracknell, Berks ) A sensitive absorptiometric method for determining traces of manganese, leucomalachite green being used as reagent, has been developed. Manganous ions are oxidised t o permanganic acid a t room temperature by potassium periodate in a sodium acetate - acetic acid buffer solution, and a solution of leucomalachite green is added ; this reagent is oxidised by permanganic acid to form a blue-green colour, which is measured a t about 620 mp. The effects of various conditions and of 36 foreign ions are examined. When applied to plant materials, the recommended procedure gives results in good agreement with the absorptiometric permanganate method, and with a precision of approximately 1 5 per cent.FOR many years the absorptiometric measurement of permanganate ionsi has been applied to the determination of manganese in plant materials and soil extracts. The method is simple and relatively free from interference, but, for small sample weights or very low manganese contents, it is often not sufficiently sensitive. An extremely sensitive test for detecting traces of manganese, using $9'-tetramethyldiaminodiphenylmethane (tetrabase or methane base) as reagent, has been widely used.2~3~4~~ Harry697 first described a quantitative determination with this reagent, ammonium persulphate being used as the oxidant. More recently, Nicholas and Fisher,8lQ who used potassium periodate as oxidant, measured the blue colour formed initially with tetrabast:.This colour, however, is too fugitive for satis- factory measurement; further, at the low 1 emperature (below 10" C) specified, mist forms on the walls of the optical cell and causes errors in absorptiometry. Cornfield and Pollardlo made use of the green colour that forms 10 minutes after the addition of the reagents, but, both in the presence of manganese and in blank tests, this colour continues to change, and, although it is less transient than the initial blue colour, it cannot be accurately measured. The use of a related compound, $$'-t etramethyldiaminotriphenylmethane, also known as leucomalachite green, for determining traces of manganese was reported by Schwarz, Fischer and Hagemannll and by Gates and E:llis.l2 Schwarz, Fischer and Hagemann developed the colour in a lactic acid solution.Gates and Ellis used potassium periodate in a mixture of nitric and phosphoric acids at 100" C to oxidise manganous ions, and then produced a yellow colour with leucomalachite green a t 80" C. I found that the latter procedure gave excessive blank values and that the effects of variations in temperature or time were too critical to permit adequate reproducibility. It was found, however, that leucomalachite green produced a blue-green colour with traces of permanganate at room temperature in a sodium acetate- acetic acid buffer solution, and this appeared to be suitable for the quantitative determination of manganese.Under suitable conditions the blank value was negligible and the colour intensity was constant. An investigation into the various factors influencing colour formation and the effects of various foreign ions resulted in the development of a satisfactory method for traces of manganese, which gives good agreement with the absorptiometric permanganate method and is about 200 times more sensitive. EXPE RIMENTAL ABSORPTION CHARACTERISTICS OF THE COLOUR- In preliminary experiments, the coloulr formed by oxidising leucomalachite green with manganous ions and potassium periodate was found to be similar to that formed by direct oxidation with an equivalent amount of potassium permanganate, and also to that of malachite green. Ferric or iodide ions in the presence of periodic acid also produced a similar, but much less intense colour.The identity of the various colours was confirmed by comparing absorption spectra measured at from 400 t o 700mp with a Unicam SP350 spectrophotometer. In each instance a maximum absorption at about 620 mp and a lesser peak at about 420 mp were found. The reactions were carried out at 19" C, at pH 3-35 obtained with a sodium acetate - acetic acid buffer solution.June, 19581 MANGANESE WITH LEGCOMALACHITE GREEN 351 Potassium periodate alone did not produce a colour a t this temperature, but, when heated to 100" C, a colour developed that had a maximum absorption at 590 mp. It was evident that the colour formed in the presence of manganese is due to permanganic acid and not directly to the periodic acid.The absorption spectrum showed that a red filter (Chance OR2) is suitable for measurement of the colour. CONDITIOXS REQUIRED FOR THE DETERMIXATION- It was found that factors influencing colour formation included acidity, amounts of potassium periodate and leucomalachite green present, temperature and the times allowed for preliminary oxidation of manganese and for colour development. These all need to be standardised ; the optimum conditions were established as follows. EfSect of acidity--4 satisfactory colour was produced only in a narrow pH range, namely from 3 to 4. Below pH 3, the development of colour was slow and incomplete, and above pH 4 leucomalachite green was precipitated. A buffer solution was therefore necessary to keep the pH within this range.Sodium acetate - acetic acid and sodium borate - acetic acid buffer solutions were found to be equally suitable. The effects of variations in pH within the range of 3 to 4 and of two different concentrations of sodium acetate, 0.25 and 0.5 g of the hydrated salt per 50 ml, were investigated. The colours formed by 0, 0.4 and 0.8 pg of manganese, as manganous sulphate, were developed in 50-ml calibrated flasks by adding the required amounts of sodium acetate and acetic acid, 3 ml of 0.2 per cent. potassium periodate solution and 0.5 ml of a 0.1 per cent. solution of leucomalachite green in 0.1 N hydrochloric acid. The solutions were diluted to 50 ml and their optical densities measured 1 hour later in 2-cm cells with a Hilger Biochem absorptiometer, a Chance OR2 red filter being used.The results are shown in Fig. 1. PH o'Fig. 1. Effect of acidity: curve .1, 0.4 pg of manganese with 0.25 g of sodium acetate added; curve B, 0.4 p g of mangaiiese with 0.5 g of sodium acetate added; curve C, 0.2 pg of manganese with 0.25 g of sodium acetate added; curve D, 0.2 p g of manganese with 0.5g of sodium acetate added; curve E, no manganese with 0.25g of sodium Fig. 2. Effect of pH on calibration acetate added; curve F, no manganese with curve: curve A, pH 3.55; curve B, pH 3.35; 0.5 g of sodium acetate added curve C, pH 3.25; curve D, pH 3.10 The sensitivity decreased rapidly as the acidity increased. '0 0.2 0 4 0.6 0.8 1.0 Manganese, p g An intense colour appeared rapidly at pH 4, but an appreciable blank value was found at pH 3.5 and above.The higher concentration of sodium acetate had a suppressing effect, so 0.25 g was used in all subsequent work. The effect of acidity was further examined at four different pH values. Buffer solutions were prepared by dissolving 50 g of sodium acetate and 200, 300, 400 and 500 ml of glacial acetic acid in water and diluting the solutions to 1 litre to give pH values of 3.55, 3.35, 3.25 and 3.10, respectively. Solutions containing from 0 to 1 pg of manganese were treated by adding 5 ml of buffer solution, 5 ml of 0.2 per cent. potassium periodate solution and 1 ml of 0.1 per cent. leucomalachite green solution, and then diluting to 50 ml. The absorptiometer readings are shown in Fig. 2.362 YUEN: DETERMINATION OF TRACES OF [Vol. 83 The optical densities increased as the acidity decreased; at pH 3-55 the sensitivity was about twice that at pH 3.10.None of the four curves is linear, but those at lower pH values are more nearly so. A sodium acetate - acetic acid buffer solution at pH 3-25, which gives a final pH of 3.35, was selected, as it gives a reasonably low blank value with adequate sensitivity. In these conditions, the further addition of acid equivalent to 0.25 ml of 0.1 it' hydrochloric acid had no appreciable effect on the optical density, but increased amounts proportionately reduced it. Hence, the amount of excess acid contained in extracts of plant ash should be the minimum required for complete dissolution. Effect of various amounts of potassium periodate-At first, saturated potassium periodate solution was used.I t was soon found thi2t the concentration of periodate in the saturated solution varied with temperature and length of storage. A 0.2 per cent. solution was therefore prepared instead. This reagent is stable for about 5 days. The effect of various amounts of periodate on the colour intensity was examined by adding from 1 to 10 ml of 0.2 per cent. periodate solution per test. The optical densities for 0, 0.4 and 0.8p.g of manganese are shown in Table I. TABLE 1 EFFECT OF VARIOUS AMOUNTS OF 0.2 PER CENT. POTASSIUM PERIODATE SOLUTION Amount of 0.2 per cent. Absorptiometer Absorptiometer Absorptiometer potassium periodate reading in absence of reading in presence of reading in presence of solution added, ml manganese 0.4 pg of manganese 0.8 pg of manganese 1.0 0.006 0.088 0.170 2.0 0.012 0.178 0.320 3.0 0,014 0.270 0.492 4.0 0.020 0.332 0.560 5.0 0.030 0.390 0-640 7.0 0.038 0.450 0.700 10.0 0.044 0.486 0.720 The optical density due to manganese increased with the amount of periodate added; so did that of the blank solution.The amount was standardised at 5ml of 0.2 per cent. potassium periodate solution, since this ga.ve nearly the maximum sensitivity, while larger amounts gave increasingly wide deviations from a linear concentration - optical density relationship. When smaller amounts were used, the colours formed were less stable and continued to increase after 1 hour. Effect of various amounts of leuconacdachite green-Leucomalachite green is readily soluble in acetone, chloroform or ethanol, but these solvents are unsuitable for use in the test.A 0.1 per cent. solution in 0.1 N hydrochloric acid, which is stable for about 1 week, was finally adopted. The effect of various amounts from 0.25 to 3 ml of this solution was examined. The results are shown in Table 11. TABLE I1 EFFECT OF VARIOUS AMOUKTS OF 0.1 PER CENT. LEUCOMALACHITE GREEN SOLUTION Amount of 0.1 per cent. Absorptiometer Absorptiometer Absorptiometer leucomalachite green reading in absence of reading in presence of reading in presence of solution added, ml manganese 0.4 pg of manganese 0.8 eg of manganese 0.25 0.012 0.288 0.488 0.50 0.014 0.350 0.574 1.0 0.018 0.366 0.630 2.0 0.016 0.340 0.553 3.0 0.012 0.230 0.460 Since the optical density increased as the volume of 0.1 per cent. leucoinalachite green solution was increased up to 1 ml and larger ;amounts reduced the intensity (probably because of the hydrochloric acid contained in the reagent), the addition of 1 ml per test was adopted.Effect of temperature-Both the colour intensity due to manganese and that of the blank solution increased somewhat with temperature up to 25" C, and above 30" C the results were erratic. Effect of variation in the times of reaction-The effects of variation in the times allowed for oxidation of manganous ions to permanganate and for the reaction of this with leuco- malachite green were examined. One hour was allowed for the second reaction when the A temperature of 25" C was therefore adopted as standard.June, 19581 MANGANESE WITH LEUCOMALACHITE GREEN 353 time for the first was varied; when the time for the second reaction was varied, 20 minutes were allowed for the first.The results are shown in Fig. 3. .E ob i o i o 6;o 80 160 110 Time, minutes Fig. 3. Efect of time. Time of standing before adding leucomalachite green: curve A, 0.8 pg of manganese; curve C, 0.4 pg of man- ganese. Time of standing after adding leuco- malachite green: curve B, 0.8 pg of manganese; curve D, 0.4 p g of manganese; curve E, no man- ganese When a longer time was allowed for the preliminary oxidation, the colour subsequently developed was more intense and more consistent. It appears that a 30-minute standing period before the addition of leucomalachite green is sufficient to complete the preliminary oxidation of up to 1 pg of manganese. After leucomalachite green had been added, the colour developed rapidly for 30 minutes, and then more slowly. Standing for 1 hour gave maximum stable coloration without unduly increasing the blank value.Periods of 30 minutes for preliminary oxidation and 1 hour for colour development were adopted as the standard procedure. I KTE RFERENCE- and anions as potassium or sodium salts. present without causing interference are shown in Table 111. In this investigation various cations were added in the form of chloride, sulphate or nitrate, The maximum amounts of 36 ions that could be TABLE 111 EFFECT OF FOREIGN IONS hIaximum Effect Maximum Effect amount that on colour amount that on colour does not cause intensity of does not cause intensity of Ion interference, excess of ion Ion interference, excess of ion mg mg Aluminium .. 0.1 Increase Chloride . . 15 Decrease Ammonium .. 5 Decrease Chromate . . 0.06 Increase Calcium. . .. 10 Decrease Citrate . , . . 0.1 Decrease Cobalt . . .. 0.1 Increase Cyanide. . . . 0.2 . Increase Copper . . . . 0.2 Decrease Dichromate , . 0.05 Increase Ferric iron . . 0.002 Increase Fluoride . . 0.5 Decrease Ferrous iron . . 0.01 Decrease Iodate , . .. 2 Decrease Lead . . .. 0.1 Decrease Iodide . . * . 0.001 Increase Magnesium . . 8 Decrease Molybdate . . 0.5 Increase Nickel . . . . 0.1 Increase Nitrate . . . . 2 Increase Potassium . . 10 Decrease Nitrite . . .. 0.002 Decrease Sodium . . . . 12 Decrease Orthophosphate 10 Decrease Zinc . . . . 0.3 Decrease Oxalate . . 0.05 Increase Acetate .. 15 Decrease Perchlorate , .8 Increase Increase Arsenate .. 1 Increase Silicate , . . . 1.5 Bicarbonate , , 5 Increase Sulphate . . 12 Decrease Borate . . . . 10 Increase Sulphite .. 0.01 Decrease Bromide . . 0.5 Increase Thiosulphate . . 0-01 Increase Among the cations, calcium, magnesium, potassium and sodium in fairly large amounts Both ferric and ferrous ions seriously interfered; as little as 2 pg of ferric had no effect.354 YUEN: DETERMINATION OF TRACES OF [Vol. 83 ion caused a significant error. Other metals, such as cobalt, copper, lead, nickel and zinc, affected the test only when present in amounts greater than 100 times the manganese content. Among the anions, chromate, dichrcrmate, iodide, oxalate and thiosulphate increased the optical density; reducing agents, such as nitrite and sulphite, decreased it. Iodide is unlikely to occur in interfering amounts in plant material, but the iron present in some plants or arising from contamination may be sufficient to interfere with the deter- mination.Attempts to minimise the effect of ferric ions by adding a small amount of phosphate, fluoride, oxalate or citrate to the test before addition of the reagents were without success. Removal of iron as ferric chloride by means of a diisopropyl ether extraction13 was found to be both simple and effective. Incidentally, the tetrabase method was found to be sensitive to interference by ferric iron to the same extent as the proposed method, and the statement made by Cornfield and Pollardlo that addition of a small amount of phosphate minimised the effect in their method was not confirmed. METHOD REAGENTS- Bufer solution-Dissolve 50 g of analytical-reagent grade hydrated sodium acetate in water containing 400 ml of analytical-reagent grade glacial acetic acid and dilute to 1 litre.Potassium periodate solution, 0.2 per cent.-Dissolve the required amount of analytical- reagent grade potassium periodate in water, with heating to 80" C and stirring. Leucolnalackite green solution, 0.1 per cent.-Dissolve 0.1 g of leucomalachite green (pp'-tetramethyldiaminotriphenylmethane) in water containing 1 ml of concentrated hydro- chloric acid at 80" C. The leucomalachite green obtained from Hopkin & Williams Ltd. (slightly brownish colour) was found to be suitable for this method; the solution is slightly pink.A solution prepared from a supply of the product of the British Drug Houses Ltd. (pale green) was green in colour and gave a high blank value. Standavd manganese stock solution-D issolve 0.203 g of analytical-reagent grade man- ganous sulphate, IvlnS0,.4H,O, in water containing 2 ml of concentrated sulphuric acid and dilute to 1 litre. Standard manganese working solution--Dilute 2ml of stock solution to 1 litre with water. PREPARATION OF THE EXTRACT- Weigh 0.1 g of plant material into a small silica basin and ignite in a muffle furnace at 550" C. Cool the basin and add a few millilitres of water and 1 ml of concentrated hydro- chloric acid. Evaporate to dryness on a sand-bath and ignite the residue over a small flame. Add 3 ml of approximately 7 S hydrochloric acid to the ash and stir with a thin glass rod until iron oxides have dissolved, warming 1:he mixture if necessary.Transfer the solution to a small separating funnel and rinse the basin with 2 ml of 7 N hydrochloric acid. Add 5 ml of diisopropyl ether and shake for about 1 minute. When the layers have separated, run the hydrochloric acid layer into the original basin and evaporate to dryness on a sand-bath. To the residue, add 15 ml of water acidified with 0.5 ml of approximately 0.1 S hydrochloric acid, heat to the boiling-point, and then filter through a Whatman NO. 40 filter-paper into a 100-ml calibrated flask. PROCEDCRE- By pipette, place an aliquot of the extract containing up to 1 pg of manganese in a 50-ml calibrated flask and dilute it to 35 ml with water.Add 5 ml of buffer solution and 5 ml of 0-2 per cent. potassium periodate solution with shaking. Place the flask in a thermo- statically controlled water bath at 25" C for 30 minutes and then add 1 ml of 0.1 per cent. leucomalachite green solution. Dilute to the mark, mix well and replace the flask in the water bath. One hour after mixing, measure the optical density against water absorptio- metrically in a 2-cm cell, using a red (Chance 10R2) filter. At the same time carry out a blank experiment and prepare a calibration curve for 0 to 1 pg of manganese by treating aliquots from 0 to 10ml of the standard working solution by the same procedure. From this curve, read the manganese concentration of the sample solution and express the result as p.p.m. in dry matter.Cool the solution and dilute it to 100 ml. This solution contains 50 p.p.m. of manganese. This solution contains 0.1 pg of manganese per ml. When cool, dilute to the mark with water.June, 19581 MANGANESE WITH LEUCOMALACHITE GREEN 355 ACCURACY OF THE PROPOSED METHOD The manganese contents of 17 plant materials were determined by both the proposed method and the absorptiometric permanganate method. With the former, determinations with and without diisopropyl ether treatment were made. The results are shown in Table IV. TABLE IV DETERMINATION OF MANGANESE IN PLANT MATERIAL Manganese found by Manganese found by the proposed method- the permanganate r A > method, p.p.m. in in presence of iron, in absence of iron, Sample dry matter p.p.m. in dry matter p.p.m. in dry matter Potato leaf .. . . 76 85 75 Potato leaf . . . . 266 260 254 Potato tuber . . . . 6 9 6 Mustard . . . . .. 16 27 18 Fodder beet leaf . . 14 22 16 Clover leaf . . . . 54 60 58 Pea leaf . . . . . . 16 15 16 Pea leaf . . . . . . 67 69 70 Lucerne . . . . . . 21 28 30 French bean leaf . . 177 170 I65 Broad bean leaf . . . . 115 125 117 Buckwheat . . . . 38 52 40 Oat grain . . . . 22 24 23 Oat straw . . . . 17 17 19 Grass . . . . . . 73 Barley , . . . . . 12 Wheat . . . . . . 172 72 13 185 -. 77 14 165 There was satisfactory agreement between the two methods over a wide range of plants. The removal of iron did not always give a closer agreement. On the basis of duplicate determinations in the course of routine analysis of a number of samples of plant material by the proposed method, a precision of about + 5 per cent.is indicated. DISCUSSION The mechanism of the formation of malachite green by oxidation of the leuco-compound is uncertain, but it is probably similar for a number of oxidising agents of high potential, including permanganate.14 In the present work, ferric and iodide ions in the presence of periodate were found to produce the reaction ; it has also been used to determine goldi5 and to detect peroxide.l6 The reaction is therefore not specific for manganese; however, only iron need be considered as a likely source of interference in plant materials and this is readily removed by extraction with diisopropyl ether. The sensitivity of the proposed method is similar to that of the tetrabase r n e t h o d ~ , ~ ~ ~ ~ and about 200 times greater than that obtained by absorptiometric measurement of the permanganate ion.Its advantage over the latter is its applicability to small sample weights or to materials that have very low manganese contents. It is superior to tetrabase methods in the formation of a stable colour, giving low blank values and a satisfactory calibration curve for the range 0 to 1 pg. The colour is readily extracted with chloroform, if necessary with a much smaller volume, thereby increasing the sensitivity attainable. In addition to plant material, ammonium acetate extracts of a number of soil samples have been satisfactorily analysed by the proposed method. It seems likely that it could be usefully applied to water analysis. I thank Mr. J. H. Dunn for valuable criticism and help in the preparation of this paper. 1. 2. 3. 4. REFERENCES Willard, H. H., and Greathouse, L. H., J . Amer. Chem. Soc., 1917, 39, 2366. Tillmans, J., and Mildner, H., J . Gasbel, 57, 496, 523 and 544; Chewz. Abstr., 1914, 8, 3085. Wenger, P., Duckert, R., and Busset, 51.-L., Helv. Chim. Acta, 1941, 24, 1143; Chem. Abstr., 1942, Mellan, I., “Organic Reagents in Inorganic Analysis,” The Blakiston Company, Philadelphia, 36, 3450. 1941, p. 452.356 5. 6. 7. a. 9. 10. 11. 12. 13. 14. 15. 16. REES-EVANS, RYAN AND WELLS : INORGANIC [Vol. 83 Vogel, A. I., “A Textbook of Qualitative Chemical Analysis,” Third Edition, Longmans, Green & Co. Ltd., London, 1948, p. 209. Harry, R. G., Chem. G. Ind., 1931, 50, 796. -, J . SOC. Chem. Ind., 1931, 50, 434. Nicholas, D. J. D., Nature, 1946, 157, 696. Nicholas, D. J. D., and Fisher, D. J., Ann. Rep. Agric. Hort. Res. Sta., Bristol, 1950, 115 to 120. Cornfield, A. H., and Pollard, A. G., J . Sci. Food Agric., 1950, 1, 107. Schwarz, G., Fischer, 0.. and Hagemann, B., Deutsch. Molkerei-Ztg., 1943, 64, 143; Chem. Abstr., Gates, E. M., and Ellis, G. H., J . Biol. Chem., 1947, 168, 537. Dodson, R. W., Forney, G. J., and Swift, E. H., J . Amer. Chem. SOC., 1936, 58, 2573. Swain, C. G., and Hedberg, K., Ibid., 1950, 72, 3373. Kul’berg, L. M., Zavod. Lab., 1936, 5, 170; Chem. Abstr., 1936, 30, 4782. Glavind, J., and Hartmann, S., Ada Chem. Scand., 1949, 3, 1021; Chem. Abstr., 1950, 44, 4056. 1944,38, 4708. Received J d y 6CA, 1967

ISSN:0003-2654    

DOI:10.1039/AN9588300350

出版商:RSC    

年代:1958

数据来源: RSC

摘要:

356 REES-EVANS, RYAN AND WELLS : INORGANIC [Vol. 83 Inorganic Chromatography on Cellulose Part XVII.* The Separation of the Non-volatile Platinum Metals BY D. B. REES-EVANS, W. RYAN AND R. A. WELLS (Chemical Research Laborntovy, Teddington, Middlesex) Rapid and simple methods are described for the separation of macro amounts of platinum, palladium, iridium and rhodium. Quantitative separations of widely different amounts of these metals are made possible by the use of cellulose powder as adsorbent and an acidified ketone as solvent, and by control of the valency level of the complex chlorides. The separations are based on the widely different mobilities of the complex chlorides of the four non-volatile platinum metals in a suitable solvent system and in contact with the cellulose powder.THE work described in this paper is an extension to the macro scale of earlier work1 on the chromatographic separation of microgram amounts of the platinum metals. With mixtures of platinum metals, an upper limit of 100 mg of any one metal has been only rarely exceeded, although this does not imply that larger a.mounts cannot be separated under the correct conditions. The chromatographic separations take place on a column of cellulose supported in a glass tube of internal diameter 1.8 0.1 cm, the column height being adjusted, within limits, to the amounts of the metals present and to the separation being attempted. By using a suitable solvent mixture for elution, the platinum metal chloride complexes display widely different mobilities on cellulose columns. The oxidising or reducing nature of the experi- mental conditions is of importance when the separations involve iridium.If oxidising conditions are maintained throughout, iridium is rapidly eluted together with platinum. Reducing conditions, such as are obtained with stannous chloride, immobilise the iridium and permit its separation from the mobile platinum and palladium. If the experimental conditions are insufficiently strongly oxidising or reducing, the chromatographic behaviour of iridium is unsatisfactory. I t is noteworthy that both the cellulose and the solvent used, hexone (isobutyl methyl ketone), are liable partly to reduce iridium. SOLVENTS- In the separation procedures described, two solvent mixtures are used, i.e., an acid solvent and an oxidising solvent.The former is hexone containing 3 per cent. v/v of concentrated hydrochloric acid and the latter is, essentially, the acid solvent with the addition of chlorine dioxide. The oxidising solvent is prepared by adding 100 ml of hexone containing 4 per cent. v/v of concentrated hydrochloric acid to a mixture of 4 g of sodium chlorate and 12 g of cellulose powder; the cellulose adsorbs the water formed by the reaction. The solvent (generator solvent) is decanted from the cellulose and then mixed with an equal volume of hexone containing 4 per cent. v/v of concentrated hydrochloric acid. Under these conditions, even after allowance has been made for the acid consumed in the reaction with the sodium * Part XVI of this series appeared in J . Appl. Cham., 1954, 4, 539.June, 19581 CHROMATOGRAPHY OK CELLULOSE.PART XVII 357 chlorate, the hydrochloric acid content of the mixed oxidising solvent cannot fall below 2 per cent. v/v. The elution behaviour of the complex chlorides of the non-volatile platinum metals with such solvent systems can be summarised by considering the steps involved in the separation of a four-component mixture of platinum, iridium, palladium and rhodium. Under oxidising conditions, platinum, iridium and zinc (a necessary addition, see later) move together in a fairly wide diffused band and are collected as one fraction. The slower-moving palladium is then collected as a second fraction. Rhodium is almost immobile with both oxidising and reducing solvents, but it can be readily eluted with water.Platinum and iridium must be separated on a second column, the iridium being reduced with stannous chloride and the platinum eluted with the acid solvent. The iridium is finally recovered by elution with water. EXPERIMENTaL For all separations the platinum metals were initially present as solutions of their complex chlorides in a limited amount of hydrochloric acid. Hexone was added to form a solvent solution for column separation. The preparation of such sample solutions frequently involves the evaporation to dryness of hydrochloric acid solutions of the platinum metals. This often causes the formation of insoluble platinum compounds, and there is evidence that iridium is similarly affected. The addition of sodium chloride prevents this partial insolu- bility, but tends to have a deleterious effect on the subsequent chromatographic separation.Mercuric chloride has been successfully used with mixtures, such as iridium and rhodium, that are to be separated under oxidising conditions, and it possesses the advantage of being volatile at the temperature of ignition of the platinum metals. Mercuric chloride cannot be used under reducing conditions as it is reduced to metallic mercury, which coats the walls of the vessels and tends to clog the column. Zinc chloride has proved to be generally satis- factory for preventing the partial insolubility of platinum and iridium. It is precipitated in alkaline solutions, but the addition of hydrazine dihydrochloride to precipitate the platinum re-acidifies the solution, and the zinc re-dissolves as chloride.The separation of the platinum metals from hydrolytic precipitates, e g . , iron, nickel, zinc and copper, in faintly alkaline solution is carried out by the addition of sodium nitrite, which forms nitrite complexes with the platinum metals.2 To be fully effective, the weight of base metals present should be less than the weight of platinum metals. The palladium - nitrite complex is unstable above pH 10, which makes its separation by this procedure difficult, As palladium is always obtained free from base metals in the chromatographic separation, it can be directly precipitated from acid solution with dimethylglyoxime. PREPARATION OF THE CELLULOSE COLUMNS- The glass extraction tube was of the standard Chemical Research Laboratory design.$ Burstall and Wells have described the preparation of the cellulose column; the only modifications necessary are described below.Short columns (15 cm) were found to be adequate for two-component systems in which one component was immobile or nearly immobile, use of such short columns being economical in terms of both time and solvent consumption. Short columns were used to separate platinum, palladium or iridium from rhodium, and also platinum from iridium. Longer columns (30 cm) were needed to separate other two-component systems and when separating three or four-component systems. Fine grade Whatman ashless cellulose powder was used throughout, and the columns were equilibrated by passing about 5 ml of acid solvent per cm of cellulose-column height through them before commencing a separation.When oxidising conditions were required, a minimum of 50ml of oxidising solvent was used to equilibrate the column before the solution of the platinum metals was transferred to it. SEPARATION OF BASE METALS- The presence of base metals lengthens the assay of any individual platinum metal and their prior separation is desirable. The nitrite procedure2 for the separation of base metals has been found to be satisfactory when the total weight of platinum metals exceeds that of the base metals, but the accuracy of the separation is lowered when the reverse circumstances apply. Adsorption of the platinum metals takes place on the hydrolytic precipitates of the358 REES-EVANS, RYAN AND WELLS INORGANIC [Vol. 83 base metals formed when the solution 01 the nitrite complexes of the platinum metals is made alkaline, and, if this precipitate is bulky, double precipitation fails to recover the platinum metals quantitatively.In this respect the final determination of platinum is more difficult than that of palladium, iridium or rhodium. A number of metals, including ferric iron, copper and zinc, are found in the platinum fraction when oxidising conditions are used. Nickel is immobile at the top of the column and is eluted, together with rhodium, by water. Mercuric chloride is mobile and moves with the iridium when the metal is separated from rhodium under oxidising conditions. When reducing conditions apply, stannous chloride is found in the platinum metal fraction together with the chlorides of copper and zinc.Ferrous chloride and reduced iridium remain almost immobile at the top of the column. In the separation of tin from platinum, the bulk of the tin was removed as stannic chloride by distillation from hydrochloric acid. This procedure was followed by a nitrite separation to ensure that no tin accompanied platinum when the latter was reduced with hydrazine dihydrochloride. METHOD REAGEKTS- Oxidising solvent-Prepare as described on p. 356. Acid solvent-Prepare as described on p. 356. Chlorine. Stannous chloride. PROCEDURE FOR THE SEPARATION OF PLATINUM, PALLADIUM, IRIDIUM AND RHODIUM- Bring the platinum metals, present as their acid chlorides, to their fully oxidised states and add sufficient zinc chloride to ensure that the weight of zinc is at least equal to the combined weights of the platinum and iridium.If nitric acid is present, evaporate the sample to dryness on a steam-bath and then repeatedly evaporate the residue with small portions of concentrated hydrochloric acid. Add a few millilitres of dilute hydrochloric acid and saturate the solution with chlorine. Evaporate again, avoiding any overheating of the residue, and remove excess of acid and condensate from the walls of the beaker at as low a temperature as possible. Dissolve the residue in 1.0 ml of concentrated hydrochloric acid saturated with chlorine, add 10ml of generator solvent and then 10ml of hexone. Replace each withdrawal of solvent by the addition to the generator of the same volume of hexone containing 4 per cent. v/v of concsntrated hydrochloric acid and stir the mixture.The initial acidity of the sample solution is rather high, but it decreases as the elution proceeds. I t is important to limit the initial volume of the sample solution and the solvent washings of the sample beaker. If the total volume of sample solution is too great, chromatographic separation commences before the last of the platinum-bearing solution comes into contact with the cellulose. If a few drops of aqueous phase separate from the solvent solution of the sample, this aqueous phase should be retained in the beaker by careful decantation of the solvent and the solvent washings. Some of the rhodium present in the sample will remain in the aqueous phase and is indicated by a characteristic rose-red colour. The remaining rhodium is immobile at the top of the column and is indicated by a narrow red band.Allow the level of the sample solution to fall to the top of the cellulose column, and then rinse the beaker with successive 3 l o 4-ml portions of oxidising solvent, any aqueous phase being retained in the beaker. The volume of washing solvent must be kept to a minimum. Elute the mixture by addition of oxidising solvent, care being taken to ensure that the level of the solvent does not fall below the top of the cellulose column. A wide diffuse band of iridium moves rapidly down the column, its brown colour masking the presence of some platinum that moves with it, sometimes a little ahead, but not detached from the iridium band. The bulk of the platinum follows the iridium as an adjacent merging band.As elution proceeds, the deep yellow platinum band moves completely away from the fairly compact orange-brown palladium band. Palladium moves away from the immobile rhodium, but at a much slower rate than does the wide platinum - iridium band. Platinum and iridium are collected as a single fraction of about 200 to 250 ml in a 500-ml round-bottomed flask fitted with a ground-glass joint to facilitate subsequent distillation. Zinc and any iron and copper present will move with the platinum and iridium, but any nickel present will be retained at the top of the column together with the rhodium. To avoidJune, 19581 CHROMATOGRAPHY ON CELLULOSE. PART XVII 359 any loss of iridium from the platinum - iridium fraction, the palladium band is allowed to approach the bottom of the column before the palladium fraction is collected.Fully oxidising conditions must be maintained to prevent any retention of iridium in the palladium band. The palladium fraction is collected in a similar 500-ml round-bottomed flask, and the addition of solvent to the column is discontinued as soon as all the palladium has been eluted; the palladium fraction is usually about 100 to 150 ml. Transfer any rhodium remaining in the sample beaker to the column with a jet of dis- tilled water and acidify the aqueous solution with a few drops of dilute hydrochloric acid. Since both rhodium and nickel move rapidly down the column, the major proportion of the oxidising solvent remaining in the column should be run either into the palladium fraction or to waste, otherwise the aqueous rhodium fraction will be accompanied by an inconveniently large volume of solvent.Collect the rhodium fraction, about 100 ml, in a 250-ml beaker and evaporate the layer of solvent under an infra-red lamp. Separation of platinum from iridium-Add water to the flask containing the first fraction and fit a Claisen head and condenser. Pass a current of air through a fine capillary reaching almost to the bottom of the flask and distil the solvent and a limited amount of water. Transfer the remaining aqueous solution to a 150-ml beaker and remove any carbonaceous deposit (resulting from breakdown of the solvent) from the walls of the flask with hot aqua regia. Add the acid solution to the beaker and evaporate the mixture to dryness, finishing on a steam-bath. Repeatedly evaporate the residue with small portions of concentrated hydrochloric acid to ensure the complete removal of nitrate in conditions that avoid over- heating the residue.Dissolve the residue in 1.0 ml of concentrated hydrochloric acid, add 200 mg of stannous chloride and dissolve the solid by gentle agitation. Add 20 ml of hexone t o the solution, which is highly coloured owing to the formation of an orange-red tin - platinum complex. This complex is unstable in solvent solution, but, if it does not break down in the cold with the production of the characteristic yellow colour of chloroplatinic acid, heat the contents of the beaker for a few seconds on a steam-bath to accelerate the decomposition. Equilibrate a 15-cm cellulose column with the acid solvent and decant the reduced platinum - iridium mixture on to it, avoiding the transfer of any of the aqueous phase.Rinse the beaker repeatedly with 3 to 4-ml portions of solvent and transfer the washings to the column, still retaining the aqueous phase. Collect the platinum fraction in a 500-ml round- bottomed flask and remove the solvent by distillation as described above, transferring the aqueous solution and acid washings of the flask to a 400-ml beaker. Elute the reduced iridium from the top of the column by addition of the faintly acidified water used to rinse the sample beaker in the same way as that described for the elution of rhodium. Collect the iridium solution in a 400-ml beaker and evaporate the solvent.At this stage, the four platinum metals are quantitatively separated and the assay methods selected will depend on the identity and concentration of any base metals present. Palladium is free from such contamination and can be precipitated as the dimethylgloxime complex. Evaporate the aqueous platinum solution to small volume, add concentrated hydrochloric acid and then remove tin by distillation as stannic chloride, assisted by the addition of bromine to the hydrochloric acid solution. After evaporation of the chloride solution to dryness, determine platinum by the standard method2 of precipitation of the metal by reduction with hydrazine dihydrochloride; determination of platinum by this method has proved to be satisfactory in the presence of copper and zinc.The remaining platinum metals can also be determined by standard procedures,2 iridium by bromate hydrolysis and rhodium b y precipitation with hydrazine dihydrochloride. PROCEDURES FOR THE SEPARATION OF TWO-COMPONENT MIXTURES- The description of the method for the separation of the four non-volatile platinum metals will clarify the chromatographic behaviour of these metals in the solvent systems recom- mended, and the general requirements for the separation of any pair will be fairly clear. Some two-component systems are described, however, with the intention of amplifying certain details. Separation of iridiztm from rhodium-Add 250 mg of mercuric chloride to the hydrochloric acid solution of the metal chlorides contained in a 150-ml beaker and evaporate to dryness on a steam-bath.Add a few millilitres of dilute hydrochloric acid, saturate the solution with chlorine and again evaporate to dryness. Dissolve the residue in 1-Om1 of concentrated360 REES-EVANS, RYAN AND WELLS: INORGANIC p o l . 83 hydrochloric acid saturated with chlorine, add 10 ml of generator solvent and then 10 ml of hexone. Use a 15-cm cellulose column equilibrated with solvent and pass through a minimum of 50 ml of oxidising solvent before decanting the sample solution on to the column. Retain any aqueous phase and rinse the beaker with successive 3 to 4-ml portions of oxidising solvent. Maintain oxidising conditions throughout the separation to prevent any retention of trace amounts of iridium. Collect the iridium fraction in a 500-ml round-bottomed flask and elute the rhodium with water.Seflaration of rhodium from other platinum metals-Use the oxidising solvent even in the absence of iridium, since it is desirable to repress the formation of any slower-moving reduction product of platinum. Use a 15-cm column for the separation of rhodium from platinum or palladium, but a 30-cm column for the separation of all three metals. Separation of palladium from iridiunz-Use a 13-cm column and maintain oxidising conditions throughout to keep iridium moving away from the less mobile palladium. A separation based on the retention of reduced iridium will fail because of the marked effect of reduction on the chromatographic behaviour of palladium. Separation of platinunt from iridium--This separation has already been described, but it is noteworthy that, under the reducing conditions that apply in this instance, the chromato- graphic behaviour of platinum differs from its behaviour under fully oxidising conditions.When the platinum - iridium mixture has been reduced by stannous chloride and the solvent solution transferred to the column, the pla.tinum invariably divides into two forms. A wide mobile yellow band accounts for most of it, and a slow-moving dull pink band represents the remainder. The two bands may separate, but both should be collected in one fraction. Sufficient solvent should be used to remove all trace of the less mobile product before com- mencing to elute iridium with water. Water elutes iridium very rapidly. If a relatively large amount of platinurn is being separated from a trace of iridium, it may be impossible to detect the presence of iridium at the top of the column by visual inspection.An operator unfamiliar with the characteristic colours of the various platinum metal complexes could mistake the slow-moving platinum band for iridium and collect it as such in a separate fraction. Iridium would then be rejected with the column packing, part of the platinum would be reported as iridium and the va.lue found for platinum would be less than the true one. If sufficient iridium is present to be plainly seen at the top of the column, its colour is a dull greyish green; it cannot be confused with the dull pink of the slow-moving platinum band. Separation of platinum from fialladizcm-Use a 30-cm column and oxidising conditions throughout the separation.RESULTS OF CHROMATOGRAPHIC SEPARATIONS Table I shows the assay results of separations in which both the components and the concentrations were varied. I t is considered that such discrepancies as exist were due to errors in assay rather than to any defect in the separation procedures. Gravimetric methods by standard procedures2 were used in the majority of the determinations, palladium being weighed as the dimethylglyoxime complex, and the other platinum metals as the metallic residue obtained by ignition in a stream of hydrogen. Worcester porcelain crucibles 4 cm in diameter were used, with Rose-type quartz covers with a central hole for the delivery of hydrogen from a quartz delivery tube. After ignition in hydrogen, leaching with hot dilute hydrochloric acid to remove occluded sodium and re-ignition in hydrogen, the residues were cooled in a stream of carbon dioxide and weighed on a semi-micro balance.Micro methods were used in the determination of amounts of platinum metal too low to be accurately assessed by a gravimetric procedure. The high platinum recovery in experiment 3 is due to the occlusion of iron in the reduced metal. In early experiments, iridium was reduced in the sample solution by passing sulphur dioxide through it, and a trap, consisting of a mixture of cellulose powder and solid ferrous sulphate, was placed at the top of the column in an attempt to maintain reducing conditions. Some oxidation of iron occurred with the result that iron appeared in the solvent fraction containing platinum and in the aqueous iridium fraction.This, in itself, is not a serious objection, but it was found that reduction of iridium by sulphur dioxide in the sample solution was frequently difficult or incomplete, hence it was decided to reduce the iridium with stannous chloride and to dispense with the trap.June, l ! M j CHROMATOGRAPHY ON CELLULOSE. PART XVII 361 TABLE I KECOVERY OF THE NOS-VOLATILE PLATISUM METALS BY THE PROPOSED METHOD Palladium found, mg - - __ 50.10 48.17 - - - 5.4 57.55 18.91 1.02 0.53 0 4 2 Iridium present, mg .- __ 1.02 4.70 4.70 52.23 5.06 49.12 19.83 -~ - - __ at least 0.24 Iridium found, n% __ .- 0.98 4.90 4.62 52.39 5-12 49.66 19.61 - - - - 0.42, 0 4 4 Rhodium present, mg 0.87 107.8 - 5 4 12 57.07 3.71 __ 114.1 5.7 5 i .6 22.53 - __ - In experiments 1 and 3, rhodium was determined photometrically and iridium by a micro-titration method described e1sewhere.l Base metals were included in the sample for separation in experiment 11. These were ;ttlded as chlorides equivalent to 100 mg of zinc and 20 mg each of copper, nickel and iron. The results show that the presence of limited amounts of these metals has no influence on the chromatographic behaviour of the platinum metals. Experiment 12 is typical of the analysis of a platinum-base alloy containing a small amount of palladium; it small amount of copper was also included. 13xperiments 1 3 and 14 show the results of analysis of platinum-base alloys supplied by thc Assay Office, Goldsmith’s Hall. The figures shown in the “present” columns of Table I are ttie values submitted by the A%ssay Office and are regarded as reliable.In experi- ment 13, phtinum \vas alloyed with small amounts of copper, palladium and gold, and in cxpcriment 14, the platinum was associated with copper, palladium, gold, iridium and rtiotlium. COSCLVSIOXS The proposed procedures provide rapid and simple methods for the quantitative separa- t ion of the non-\datile platinum metals. The apparatus required is limited and simple i n tlcsign, and, by following the methods described, the manipulator should experience no tlifticulty i n scprating the characteristically coloured bands of the platinum metals. 1 he cliromatographic analyses of platinum-base alloys agree well with those submitted by o t h r laboratories (the Assay Office and the blond Nickel Company), who use different rncthods of separation and assay. Separations on a larger scale have not been undertaken, I)ut no tlifhculties are foreseen in dealing with much greater amounts of the platinum metals. \Ve tliank the Deputy Warden, Assay Office, Goldsmith’s Hall, for samples of platinun- 1)ase alloys antl for his permission to publish the results of analyses. Frequent references li;tvc 1)cen mxle to “The Analysis of &Iinerals and Ores of the Rarer Elements.”2 The section tlrvotctf to the platinum metals has been of value in problems connected with nssny. lliis paper is published by permission of the Director of the Chemical Kesearcli 1,abor;rtory. I . REFERENCES I , 2. 3. Kciiilwr, S . I;., and \Vclls, R. .\., .4mzlyst, 19 Scilrwller, \\’. I<., antl Pow-ell, A. K., “The Ana i3urstall. 1;. H., ant1 \f’eIk, I<. .I., .lrzdyst, 1951, 76, 3!)6. rals and Ores of the Rarer IClcnients.” C . (;riltin & Cu. Lttl., London, 1!140. Receivcd F d w u v y 1 .It//, 1 !I57

ISSN:0003-2654    

DOI:10.1039/AN9588300356

出版商:RSC    

年代:1958

数据来源: RSC

摘要:

362 DIXON: SPECTROCHEMICAL METHODS OF ANALYSIS AS APPLIED TO [Vol. 83 Spectrochemical Methods of Analysis as Applied to Mineral Matter Associated with Coal BY K. DIXON (Coal Survey Laboratory, Government Buildings, Chalfont Drive, Nottingham) An account is given of the application of spectrochemical solution tech- niques to the analysis of mineral matter associated with coal. Solutions are prepared by fusing the sample and then extracting the melt, or by wet oxida- tion. Calibration for the major constituents, silicon, aluminium, iron, calcium, magnesium, sodium and potassium, and the minor constituents, titanium, manganese, copper, strontium and barium, is carried out with synthetic standards prepared, whenever possible, from Specpure materials. Sodium and potassium are determined with a flame photometer.The tech- niques are appreciably faster than other published methods, and have a comparable precision. They are applicable to gas producer and cyclone burner slags, as well as to coal ash. THE intensive use of coal in modem industrial appliances for combustion and gasification is bringing to light many new problems in the behaviour of the mineral matter associated with coal and coal seams, and a better knowledge of the composition and behaviour of this mineral matter is urgently needed. Interest centres chiefly on the major rather than the trace constituents, and the current paucity of information is a measure of the difficulty and length of the classical methods of chemical analysis, which have, until recently, been regarded as standard methods.Within the last few years, three paper~l9~9~ have been published that describe rapid colorimetric and composite schemes of analysis ; the methods proposed are some four or five times as rapid as the classical methods4 and are of acceptable precision, the coefficient of variation of the determination of a major constituent being approximately 2 per cent. Spectrochemical methods are potentially several times more rapid than the recently published methods and the aim of this investigation has been to devise a suitable spectrochemical procedure and to assess its precision, speed and range of applicability. EXPERIMENTAL PREPARATION OF SOLUTIONS- Solu- tions of coal or ash can be prepared by wet oxidation or by fusion and then extraction of the melt with water or acid.Wet oxidation is particularly useful when coal and dirt bands that have high contents of organic matter are to be analysed without recourse to ashing. In such instances, the silica must be determined gravimetrically and any residue after its removal as tetrafluoride must be fused, extracted and the extract added to the original filtrate. The most convenient approach, however, is to ash the sample according to the British Standard procedures and then to proceed by means of a fusion technique-information about the ash content of a sample is usually required. A separate solution, prepared by digestion with a mixture of hydrofluoric and sulphuric acids to ensure complete breakdown of silicate materials, i s then used for the determination of the alkali metals.Clay minerals and slags are treated in the form in which they are received, carbon dioxide, moisture at 105' C, water of constitution, sulphur, carbon and phosphorus being determined chemically to complete the analysis. When an exceptionally high concentration of an element is likely to occur, sample weights of less than 0.1 g are usually taken. Chromium was chosen as the internal standard to be added to each solution. This element gives a useful number of spectral lines in the wavelength range 2500 to 4600 A and occurs only in trace amounts in coal ash and associated clay minerals. EXCITATION- Three methods are most frequently used to introduce solutions into a discharge gap, namely, the Feldman porous cup, the rotating graphite wheel and a method in which the Samples may be received as coal, coal ash, or extracts in acid or aqueous media.June, 19581 MINERAL MATTER ASSOCIATED WITH COAL 363 solution is absorbed by a porous graphite electrode before sparking, see Fig.1. The last- named was chosen because of its ease of preparation, the small amount of high-purity electrode required and its availability for the possible future determination of trace elements by means of an arcing technique. t inch Feldman porous cup 0.040 inch n Graphite counter electrode v $-inch dia. inch Graphite counter n bl N.C.C. eraDhite counter i l e i t r o d e U 1 ;..-L. U f-inch dia. electrode 7 1 1 1 L . ( 1 “,OG+O inch f to f - w h e e l f l y o t inch dia. 0,1285 0 - f inch graphite cup =0.125 inch jM4B eraDhite Dedestal Graphite Porous N.C.C.- - - - - - - - - - - - - - Silica boat containing solution \- $-inch dia. (0) (bj (4 Fig. 1. Types of electrodes used for the analysis of solutions The small sections of high-purity graphite electrodes were packed into a shallow fused silica dish and made porous by heating at 800” C for 15 minutes. Spectra were produced with an uncontrolled condensed-spark discharge. Such a source unit gives one or more heavily damped discharges across the analytical gap per half cycle of the supply frequency. The peak current of an individual discharge is about 100 amperes. The operating conditions finally chosen were as follows- Spectrograph- Spark gap- Voltage across the ga9, r.m.s.- Added inductance- Capacitance- Upper electrode- Lower electrode- Slit Width- Source to slit distance- Exposuve- Plate- Hilger large quartz.2.5 mm. 15,000 volts. 0.015 mH. 0.005 pF. N.C.C. graphite electrode, having a 40’ cone with a 1 mm flat. N.C.C. graphite electrode, having a crater 0.04 inch deep and 0.20 inch 0.020 mm. 38 cm; image formed at collimating lens. 2 superimposures, each of 50 seconds. Ilford special rapid. in diameter. By making two superimposures, the use of salt solutions of concentrations greater than 4 per cent., which would crystallise on the top of the lower electrode on sparking, was avoided, and a smaller sample could be used. This is an important consideration when coal is analysed in the form of air-borne dust. PHOTOGRAPHIC CONSIDERATIONS- Control of the photographic processes is maintained by plate calibration using the iron- line group method, see Table I.The slope of the emulsion-characteristic curve, gamma, is obtained from a plot of the Seidel function of density against the relative intensity of each iron line. Tables were prepared to convert densitometer scale units directly to relative intensity for values of gamma within the range 1.05 to 1.20 in steps of 0.03. TABLE I RELATIVE INTENSITIES OF THE IRON-LINE GROVP Wavelength, A . . 2783.7 2793.9 2799.3 2812.5 2819.3 2823.3 2827.4 2828.6 2831.6 2833.5 Relative intensity . . 8.30 1.39 1.00 0.15 0.19 0.69 0.43 1.09 4.73 2.57364 [Vol. 83 All spectrum-line readings are corrected for background, a single measurement in the vicinity of the analytical line being applied to both this and the control line.After correction, calibration curves are linear over at least a hundred-fold concentration range. For line pairs of wavelength greater than 3400~, plate calibration is by means of a rotating-sector spectrogram of an iron - carbon arc. A rhodium-sputtered filter having two steps of 100 and 10 per cent. transmission can be used in front of the spectrograph slit to increase the ranges covered for certain elements, e g . , strontium and barium. STANDARDISATION- Calibration was carried out with synthetic standards prepared from the following Specpure and AnalaR materials : aluminium, iron and magnesium metals, calcium and sodium carbonates, potassium chloride, manganese sulphate and potassium titanyl oxalate. The hydrated sulphates of aluminium, iron and manganese are stable, and can be used as alternatives to the pure metals. Silica was added in the form of a freshly prepared and chemically standardised solution of sodium silicate.When this is not convenient, chemically standardised samples such as coal ash and diabase W1 can be used. If it is preferred to use pure metals rather than salts in the fusion method, iron and magnesium can be dissolved by prolonged heating with nitric acid and aluminium by fusion with sodium hydroxide; in the wet-oxidation method these metals can be dissolved in diluted hydrochloric acid (1 + l), after which sulphuric acid, sp.gr. 1.84, is added and the solution is evaporated to fumes to remove the hydrochloric acid. Standards for calibration were prepared from two solutions, both of which contained 20 ml of internal-standard solution per 25 ml, acid and fusion mixture when appropriate, and one of which contained also the maximum concentrations of the elements to be deter- mined.The two solutions were mixed to give a series of six standards with a dilution factor of two. Calibration curves were plotted from the mean of quintuplicate spectrograms for each standard on each of four plates, thereby permitting the precision of the technique at different concentration levels to be found. The chosen line pairs, together with the concentration ranges covered and limits of detection, are given in Tables IIA and IIB. DIXON: SPECTROCHEMICAL METHODS OF ANALYSIS AS APPLIED TO Element Silicon . . Aluminium Iron . . Calcium . . Magnesium Titanium . . Manganese . . Copper .. Strontium . . Barium . . . . * . . . . . . . . . . . .. . . . . TABLE IIA SPECTROPHOTOMETRIC DATA Excitation Concentration Line pair, potential, range covered, A eV % Si I 2881.6 - 5.1 1.0 to 60.0 Cr I 3021.6 - A1 I 3082.2- 4.0 1.0 to 40.0 Cr I 3021.6 - Fe I1 2755.7 - 13.3 1.0 to 50.0 Cr I1 2757.7 12.6 Ca I1 3179.4- 13.1 0.25 to 20.0 Cr I1 3180.7 13.1 Mg I1 2802.7 - 12.0 0.02 to 2.0 Mg I1 2798.0 - - Cr I1 2757.7 12.6 Cr I1 2757.7 12.6 0.50 to 10.0 Ti I1 3349.0 - 711.1 0.20 to 2.0 Cr I1 3180.7 1.3.1 Mn 2933.1 - J.2.8 0.10 to 2.0 Cr 2935.1 - c u I 3247.5- 3.8 0.10 to 5.0 Cr I 3021.6 - Sr I1 4077.7~ - 8.7 0.05 to 0.50 Cr I 4289.7~ 2.9 Ba I1 4554.0 - 7.9 0.05 to 5.0 Cr I 4289.7 2.9 Sensitivity, % <0.50 0.20 0.10 0.10 0.002 0.05 0.05 0.05 0.02 0.02 F = filtered portion of the line.3 65 June, 19581 MINERAL MATTER ASSOCIATED WITH COAL TABLE IIB FLAME-PHOTOMETRIC DATA Element Wavelength of line or band, Concentration range, A p.p.m.Sodium . . . . 5890 0.0 to 4.0 Potassium . , .. 7665 0.0 to 6.0 Calibration curves for the fusion technique have slightly lower slopes, 42" to 44", than have those for the wet-oxidation technique, 43" to 45". This can be attributed to the different acid radicle and the preponderance of sodium ions associated with the fusion technique. The elements with low excitation potentials give the lowest slopes, whichever method is used. METHOD PREPARATION OF SOLUTIOXS- Internal-standard solution-This solution is prepared from chromium trioxide that has been dried at 105" C for 1 hour, and contains 0,3846 g per 200 ml of 12 per cent.v/v nitric acid. Solution of coal ash-Approximately 0.1 g of ash, ground to pass through a 240-mesh B.S. sieve, is heated to 775" C for 30 minutes to remove moisture and carbon dioxide; it is then cooled in a desiccator, weighed and transferred to a platinum crucible of capacity 30 ml. The sample is covered with 1 0.05 g of fusion mixture (5 parts of anhydrous sodium car- bonate and 3 parts of pre-fused sodium borate) and gently heated for 1 to 2 minutes, then for a further 2 to 3 minutes at a dull red heat (450" to 550" C) and finally in the flame of a Meker burner at full heat (900" to 1000" C) for 5 minutes to ensure complete fusion. When cool, 20 ml of internal-standard solution are added and then 5 drops of 90 to 100-volume hydrogen peroxide to retain manganese and titanium in solution.The crucible lid is washed with 5 ml of 12 per cent. v/v nitric acid, the washings being added to the solution. The cold solution is then stirred mechanically for 10 to 15 minutes to complete the extraction. Note that care is needed in the extraction to prevent the precipitation of silica, which may occur if the temperature is above 50" C or if the stirring is inadequate. Even in the presence of a citric acid - sodium citrate buffer solution,6 temperatures above 70" to 80" C are not recommended. The solutions have a final volume of 25 2 1 ml; greater precision is not required in a ratio method. They should preferably be used within 3 hours of preparation, as when they are set aside for longer than 6 hours barium may separate as barium sulphate and there is some risk of silica also being precipitated. Solution for alkali determination-When the above-described procedure is used to prepare a solution of ash, alkalis are determined on a separate 0.1-g sample.This is weighed into a platinum crucible of capacity 30 ml, 6 ml of 50 per cent. v/v sulphuric acid and 5 ml of 40 per cent. w/w hydrofluoric acid are added, and the mixture is heated until fumes of sulphur trioxide are evolved and held at that stage for 15 minutes. When cool, the solution is quantitatively transferred to a 100-ml calibrated flask, diluted to the mark and set aside for 1 hour. Wet oxidation of coal-A 1-g sample of air-dried coal (2 g if the ash content is 5 per cent.) is ground to pass through a 72-mesh sieve and then weighed into a 250-ml tall beaker.Fifteen millilitres of nitric acid, sp.gr. 1.42, are added and the beaker is strongly heated for 15 minutes, more nitric acid being added when required to prevent baking of the sample. This preliminary treatment destroys much of the organic matter, and also reduces the frothing that occurs when sulphuric acid is added. When cool, 3 ml of sulphuric acid, sp.gr. 1.84, and 2 ml of nitric acid, sp.gr. 1.42, are added and the solution is then heated until fumes of sulphur trioxide are evolved, this procedure being repeated after each addition of nitric acid. After removal of the organic matter, indicated by the pale straw colour of the solution, 5 ml of 72 per cent. perchloric acid are added, and subsequently removed by evaporation, to leave the silica in a form that can be readily removed by filtration.When cool, the walls of the beaker are washed with 20 to 30 ml of hot water and the solution is again evaporated to fumes, A further addition of 20 to 30 ml of water is made in washing the walls of the beaker, the solution is brought to a gentle boil and then filtered hot through an 11-cm Whatman No. 542 filter-paper. If the residual silica is slightly coloured due to absorption of iron, the oxidation should be repeated on a fresh sample, as the iron cannot be recovered by washing.?TABLE IV COMPARISON OF THE PROPOSED FUSION METHOD WITH CLASSICAL CHEMICAL METHODS Analyses by chemical methods4 were carried out at the British Coal Utilisation Research Association.In all instances Sample No. 2461* 2860t 2869* 3034* 3341: 3343* 3345* 3346: 3360* 3446t 3450t Diabase W1 3344: Sample No. 2461* 2860t 2869* 3034* 3341: 3343* 3344: 3345* 33461 3360* 3446t 3450t Diabase W1 SiO, found by- chemical fusion method, method, 21.5 21-0 47.4 48.2 36.7 36.5 44-0 44.2 380 37.2 48.2 47.9 47.5 48.3 48.9 48-2 46.1 46-5 54.4 54-7 51.5 51.7 44.9 44.1 52.6 53.2 % % MgO found by- chemical fusioii method, method, 3.9 3.8 3.5 3.2 5.0 4-9 3-2 3.0 2.3 2-0 1.3 1.1 1-3 1-2 1.6 1-4 1.6 1.5 1.8 1.5 2-3 2 - 2 2.5 2.3 6-6 6-4 & % % the alkalis were determined by flame-photometric methods Difference, % - 0-5 + 0.8 - 0.2 + 0.2 - 0.8 - 0.3 + 0.8 - 0.7 + 0.4 t 0 . 3 + 0.2 -0.8 -t- 0.6 Diff ereiicc, -0.1 - 0-3 -0.1 - 0.2 - 0.3 - 0.2 -0.1 -0.2 -0.1 - 0.3 -0.1 - 0.2 - 0-2 % R1,0, found by- chemcal fusion method, method, 18.8 19-8 25.8 24.5 20-2 20.2 22-6 23-4 25-3 24-7 29.6 29-0 29.0 29.0 23.2 24.0 24.0 24-7 24.7 23-8 28.2 28.6 27-1 26.9 15.2 14.6 - % % TiO, found by- C G Z T f u s i o d method, method, 0.5 0.5 0.7 0.9 0.6 0.6 0.7 0.8 1.2 1.2 0-9 0.9 1.1 1.1 0.9 0.9 1.1 1.1 1.1 1.0 1.0 0.9 0.8 0.8 1.1 1-1 % % Difference, % + 1.0 - 1.3 0.0 + 0.8 - 0-6 - 0.6 0.0 + 0-8 +0.7 - 0.9 + 0.4 - 0.2 - 0.6 Difference, O/ / o 0-0 -t 0.2 0-0 +0-1 0.0 0.0 0.0 0.0 0.0 -0.1 -0.1 0.0 0.0 Fe,O, found by-- chemical fusion method, method, 38-9 39-3 8.6 9.0 17.0 17.8 11.6 11-1 12.1 11.9 5.8 5.6 5.9 5.8 6.6 6.1 11.2 11.2 10.1 9-9 6.0 5.8 16.3 15.9 10.9 11-0 - % % Na,O found by- chemical fusion method, method, 1.1 1.0 0.8 0.7 0-8 0.4 3-0 3-0 4.3 4.2 1.0 1-0 1.6 1-6 1.7 1.6 2-8 2-7 0.6 0.5 0.3 0.3 0-5 0.4 2.1 2.0 & % % * Coal ash prepared a t 775" C.t Slag. : Flue dust. Difference, % + 0.4 + 0-4 + 0.8 -0.5 - 0.2 - 0.2 -0.1 - 0.5 0.0 - 0.2 -0.2 -- 0.4 +0.1 Difference, -0.1 -0.1 - 0.4 0.0 -0.1 0.0 0.0 -0.1 -0.1 -0.1 0.0 -0.1 -0.1 % CaO found by- szlz-GG method, method, % % 6.7 6.8 9.8 10.0 11.5 12.1 7.3 7.8 10.4 10.9 6.7 7.2 5-3 5.4 10.5 10.8 5.9 6.0 2-4 2.3 7.9 7.7 3-4 3.2 10.9 10.9 K O found by- chemcal fusion method, method, & % % 0.5 0.5 3-2 3. I 2.8 2.9 3.4 3.5 0.6 0.8 3-8 3.7 4.4 4.4 3.5 3.4 3.8 3.8 3.4 3-3 2.8 2.9 3-1 3.2 0.7 0.7 Difference, Y O S0.1 + 0.2 + 0.6 + 0.5 + 0.5 + 0.5 +o-1 + 0.3 + 0.1 + 0-1 + 0.2 + 0.2 0.0 Difference, % 0.0 - 0.1 + 0.1 10.1 + 0.2 -0.1 0-0 - 0.1 0.0 -0.1 + 0-1 + 0.1 0.0 w Q, a'4 G s F W en 00 TABLE V COMPARISON OF THE PROPOSED FUSION AND WET-OXIDATION METHODS WITH A CHEMICAL METHOD u The samples used were coal ash, prepared at 380" C.The chemical analyses were carried out by a colorimetric semi-micro method3 with use of the same solvent acids as in the wet-oxidation method CaO found by- SiO, found by- A1,0, found by- Fe,O, found by-- A -7 f > r L-_-7 r - - 7 Sample No. 3622 4761 4731 1844.3 18444 47H50 48H50 49E-I50 2521150 54/275 45/4 chemical method, 36.6 31.0 35.0 36.7 29-1 20.7 21.9 19.7 24.2 17.7 24.7 % Sample No. 3622 4761 4731 18443 18444 47H50 48H50 49H50 252H50 54/275 45/4 fusion chemical method, method, % % 37.3 28.3 31-7 24.3 35.1 26.6 37-7 28-4 20.4 23.3 20.3 16.7 21-8 13-4 20.8 16.6 23.8 19.9 18.2 9.5 25-1 26.4 MgO found by- wet- Oxidation method, 29-2 23.4 26.1 28.5 23.2 16.3 12.7 15.8 19-3 9.6 264 % chemical method, 1.90 2-76 2.26 1.92 3.60 1.83 3.00 3.95 4.45 0.59 2.54 % wet- oxidation method, 1.96 2.45 1-92 1.86 3.15 1.80 2-77 3.60 4.40 0.56 2.08 Yo fusion method, 1.91 2.50 1.80 1.82 3.50 1-84 2.97 4.10 4.65 0.55 2-21 % fusion method, 29.2 23.8 27-1 28.6 23-5 16.8 12.8 16.5 20.4 10.0 27.0 % chemical method, % 4.75 6.3 5.50 4.95 6.6 43.0 19.9 23.8 9.2 43.6 9.8 wet- oxidation method, % 6.00 6-5 5.65 5.00 6.8 42.9 19.3 22.7 9.2 42.9 9.5 TiO, found by-- c---h-- 7 wet- chemical oxidation fusion method, method, method, 0.81 0.79 0.80 0.76 0.77 0.75 0.78 0.75 0.74 0.80 0.80 0.83 0.68 0-67 0.66 0.47 0.46 0.45 0-53 0.52 0.5 1 0.52 0.48 0.51 0.99 0.97 1.03 0.32 0.30 0.33 0.63 0.54 0.54 % % % fusion method, % 4.80 6.3 5.50 5.00 7- 1 44.1 20.5 23.7 9-4 44.5 9.6 chemical method, 8.8 11-7 9-7 8- 1 13-2 3.95 144 11.4 11.3 12.2 % 5-10 wet- oxidation method, % 9.5 11.8 9-4 8.5 13.3 4.45 14-4 11.8 11.5 12-0 4.80 Mn,O, iound by- 7 chemical method, 0.21 0.34 0.24 0.24 0.44 0.21 0-27 0.41 0.52 0.05 0.14 O/ / O A wet- oxidation method, 025 0.32 0.21 0.23 0.41 0.20 0-26 0.33 0.48 < 0.15 0.17 % -7 fusion method, 0.18 0-33 0.23 0.22 0.44 0-20 0.24 0.28 0.49 G0.15 0.16 % fusion method, 5 % z 9.2 kx R $ 12.8 9.8 8.6 E 14.4 5 4.05 ci M kx 14.9 12.1 4-60 g 13.7 2 11.9 % % 8 n 0 $TABLE VI THE EFFECT OF ASHING ON THE MAJOR CONSTITUENTS The ashing was carried out by the British Standard procedure? The chemical analyses were carried out by a colorimetric semi-micro method3 and the spectrochemical analyses by the wet-oxidation method, the same solvent acids being used El ?4 in both methods.The results for coal samples are calculated on the basis of ash Sample No. 5355 5356 5357 5358 5359 SiO, Original found by form of chemical sample method, 42-0 { :? 41-5 29-7 { :tl 30.0 28.7 { :tl 28.9 36-2 { z1 35.8 22.2 21-7 % Al,O, found by- & spectro- chemical chemical method, method, 30.7 29.3 30.0 20.4 19.7 - 21.7 24.7 24.2 __ 24.0 24-9 23-7 - 24.3 1s.a l9.0 - 18-9 % % - MgO found by- - Original form of chemical Sample No. sample method, % 3.19 0.81 5-10 4-2 3-55 5355 { zl? - 5356 { - 5357 { :i1 - 5368 {::I - 5359 { ztl - spectro- chemical method, % 3-20 3-21 0.78 0.84 5-25 5-10 4.3 4.2 3.65 3-35 Difference, - 1.4 - 0-7 - 0.7 + 1.3 - 0.5 -0.7 - 1-2 - 0.6 ~ 0.3 + 0.6 % Fe,O, found by- - spectro- chemical chemical method, method, 6-9 7-6 - 7-3 43.6 43.6 - 44.7 22-3 22.8 - 21.6 21.0 20.4 __ 20.7 31-0 32-5 - 31.2 % % Difference, % +0.01 + 0.02 - 0.03 + 0-03 +0*15 0.00 +0-10 0.00 +o-10 - 0.20 TiO, found by- & spectro- chemical chemical method, method, % % 0-98 0.98 - 0.96 0.57 0-55 - 0.56 0-69 0.67 - 0.65 1-18 1.17 - 1-15 0-68 0.68 - 0.69 Difference, % + 0 7 + 0 4 0.0 + 1-1 + 0.5 - 0-7 0.6 - 0.3 + 0.2 I 1 C T 1-0 z ..CaO found by- - spectro- 8 chemical chemical 4 % % % c) 11.5 11.1 -04 8 11.8 + 0.3 z 1.81 1-80 -001 s - 1.91 +0.10 g - 14.2 - 0.3 E 7.9 7-6 - 0-3 2 method, method, Difference, W 0 __ 14.5 14.3 - 0.2 - 7.9 0.0 0 t( 25-2 15.6 +04 0’ 15.5 + 0.3 0 - hl Difference, 0.00 - 0-02 - 0.02 - 0.01 - 0-02 - 0.04 -0.01 - 0.03 0.00 +0.01 % Mn,O, found by-- -7 chemical method, % 0.37 0.06 0.28 0.19 0.46 - - - - - spectro- chemical method, % .- 0.41 < 0.15 0-26 0.18 0.45 - - - - - - 0-02 - -0.01 -0.01 -June, 19581 MINERAL MATTER ASSOCIATED WITH COAL 369 The residue is washed with 150ml of hot water, and, after the addition of 20ml of internal-standard solution per 25ml of final volume, the filtrate is evaporated to about 15 to 20ml before being transferred to a 25-ml calibrated flask and diluted to the mark.For samples containing more than 10.0 per cent. of calcium, the final volume may be 50 ml, as the solubility of calcium sulphate is only about 5 mg per 10 ml. With this technique the solution is diluted to a known volume to permit a 2-ml aliquot to be taken for the determination of alkalis.Silica is determined by a standard gravimetric procedure. Residues after removal of the silica as tetrafluoride seldom exceed 1 mg, and when they do, they are fused and the extract added to the original filtrate. In these circum- stances, another portion of the sample must be dissolved in the same way as for the preparation of the solution for alkali determination. Acid or aqueous extracts of coal-To such extracts, 3 ml of sulphuric acid, sp.gr. 1.84, are added and the solution is heated to fumes. When cool, 20 ml of internal-standard solution are added and the solution is diluted to 25ml. All soluble constituents can be determined on this solution, the same calibration curves as for the wet-oxidation technique being used.EXCITATION PROCEDURE- A 0.05-ml portion of the prepared solution is added to a porous electrode, see Fig. 1 (c). After 40 to 60 seconds in which to dry, the electrode is sparked for 50 seconds against a high- purity graphite counter-electrode. The electrode is then allowed to cool for 10 seconds, after which a second drop of solution is added to it, and a second exposure is superimposed on the first. A duplicate spectrogram is prepared from the same solution, the mean of the two forming a single determination. Duplicate spectrograms for twelve samples can be prepared in 1 hour, PHOTOMETRIC PROCEDURE FOR ALKALIS- A 2-ml aliquot of the wet-oxidation solution, or 5 ml of the solution prepared by hydro- fluoric - sulphuric acid digestion, is diluted and used for the determination of sodium and potassium by standard flame-photometric procedures; an E.E.L.flame photometer (Evans Electroselenium Ltd.) was used. The two solutions have approximately the same normality with respect to sulphuric acid, so that the same operating conditions can be used for both. The determinations were found to be unaffected by changes in normality with respect to sulphuric acid between 0.05 and 0.5 N . The maximum concentrations of sodium and potassium measured were 4 and 6 p.p.m., respectively, and at these levels no mutual inter- ference was encountered in the sulphuric acid medium; the only direct interference encountered was from calcium, as noted by Jackson and Smith.* This effect was small, as both sulphate and aluminium ions depress the calcium emission, and could be ignored for calcium concentrations below 10 per cent.The use of an interference filter for sodium would reduce this interference to a still lower level. ACCURACY- Accuracy represents the degree of confidence that can be placed in a particular method of analysis and is determined by two factors, reproducibility (or precision) and bias. The former gives a measure of the random errors inherent in the method and the latter a measure of any systematic inaccuracies that can be eliminated by altering some aspect of the method. Precision-A measure of the precision of the excitation and reading procedures is given in column 3 of Table 111. These data were obtained from duplicate analyses on the same plate by using a single solution and can be used to assess the suitability of the chosen line pairs.The precision of the complete method was determined from replicate analyses of one sample sparked in quadruplicate on six plates taken over a period of 6 weeks. The coefficient of variation for major constituents is not as low as that from standard chemical methods of analysis, but it is, however, of the same order as that found by other physical methods. The lack of precision is largely related to the use of much smaller sample weights and to errors that arise from plate-emulsion defects. Bias-Bias is assessed by considering the difference between the observed, in this instance spectrochemical, and the "true" concentration, as determined by an independent chemical method, and should be considered in relation to the precision of either method DISCUSSION OF THE METHOD370 DISOK: SPECTROCHEMICAL METHODS OF ANALYSIS AS APPLIED TO TABLE 111 PRECISION OF THE PROPOSED NETHODS [Vol.83 Coefficient of variation for single determinations Fusion method Amount of (--- A 7 compound Variations Plate Wet- Compound in the within the calibration Complete oxidation determined sample, plate, data, method, method, 0 YQ YO % /O O/ /Q SiO, 31.0 A1.0, 24.0 F&Oi 6.55 CaO 11.80 MgO 2.45 TiO, 0.77 Mn,b, 0.32 Na,O 1.96 K,O 0.62 2.0 1.7 1.9 1.9 1.7 2.3 2.3 2.5 2.3 2.0 1.8 1.9 3.5 9.6 2.8 2.5 2.2 2.4 2.4 2.8 3.1 2.6 2.4 - 3.4 2.5 3.3 3.3 3.1 5.1 2.8 2.1 before a decision is taken on its significance. The three most probable causes of bias associated with spectrochemical solution methods are (a) failure to obtain complete solution of the sample, (b) the use of unsuitable salts for calibration purposes, e.g., hydrated salts that readily absorb moisture, or complex salts the composition of which may not be as specified, and (c) the possibility of inter-element effects.Table IV gives a comparison between the re:;ults of the fusion and the classical methods of analysis for coal ash, flue dusts and slags of the type encountered in gas producers. Small deviations are apparent for calcium, magnesium and manganese. The lower chemical figures for calcium could arise from its incomplete precipitation as oxalateg and loss as sulphate on digestion of the sample. Any calcium not removed as oxalate would be determined along with magnesium to give high results for this element.This is confirmed by the results shown in Tables IV and V and is most apparent with large sample weights. A smaller sample was used for the analysis of the mineral matter in coal and these effects are therefore not evident in the results shown in Table VI. No comparison is available for copper, strontium and barium. The comparison with the colorimetric semi-micro method, shown in Table V, is equally as good as that with the classical method, and no bias for manganese is evident. The results of the chemical analyses for manganese are not of high accuracy and have not been subject to much consideration, as this element is of minor importance in the analysis of ash. The deviations for two samples were originally large and the chemical figures were subsequently amended, which indicates that the observed bias of -0.1 per cent.is probably of chemical origin. The results given in Table VI verify the view that no loss of the major constituents occurs, within the limits of experimental error, when coal is ashed at 775" C. This has been confirmed by H. R. Brown, Officer-in-Charge, Coal Research Section, C.S.I.R.O., in a private communication. No chemical data for the alkalis were available, but the work of Logie and Raynorlo indicates that no loss occurs on heating up to 1000" C. CoNCLUsloss Spectrochemical solution methods are readily applicable to the analysis of the mineral matter in coal, coal ash and a wide variety of associated minerals. The mineral matter in coal itself can be analysed without prior ashing by using the wet- oxidation procedure to destroy the coal substance and to dissolve all the compounds except silica.These compounds, including the alkalis, are determined on one solution and silica is determined separately by a standard grakimetric procedure. Dissolution of ash is effected by means of an alkali-fusion technique, all compounds except the alkalis being determined in the solution of the melt. A second solution must be prepared by digestion with hydrofluoric acid for the determination of the alkalis. The precision is of the same order as that for other physical methods, i.e., $2 to 3 per cent., and the proposed methods are some three or four times as rapid. This work formed part of the programme of the Coal Survey Branch of the Scientific Department of the National Coal Board. I am indebted to Mr. A. Dawe, Chief Coal SurveyJune, 19581 MINERAL MATTER ASSOCIATED WITH COAL 37 1 Officer, Nottingham, under whose guidance the work was carried out, to Mr. J. M. Bromley and Mr. C. Fletcher, who did much of the experimental work, to Mr. D. Flint of the British Coal Utilisation Research Association, who supplied the analytical data obtained by classical chemical methods and to Mr. H. R. Brown for his private communication on the loss of major constituents when coal is ashed. I wish to thank the National Coal Board for permission to publish this paper. 1. 3. 4. 5. 6. 7. 8. 9. 10. 9 Y. REFERENCES Banerjee, K. N., and Collis, B.A., Fuel, 1955, 34, 571. Shapiro, L., and Brannock, W. W., U.S. Geological Suvvey Bulletin No. 1036c, 1956. Pringle, W. J . S., Fuel, 1957, 36, 257. D.S.I.R., Fuel Research Survey Papev No. 50;, H.M. Stationery Office, London, 1949. “The Analysis and Testing of Coal and Coke, Hartlief, G., and Kornfeld, H., Awh. Eissenhiittenw., 1955, 26, 329. Nishida, H., Japan Analyst, 1956, 5, 389. Jackson, P. J., and Smith, A. C., J . Appl. Chern., 1956, 6, 547. Toribara, T. Y,, Dewey, P. A., and Warner, H., Anal. Chern., 1957, 29, 540. Logie, D., and Raynor, J. E., J . Inst. Fuel, 1953, 26, 146. British Standard 1016 : 1942. Received July 29th, 1957

ISSN:0003-2654    

DOI:10.1039/AN9588300362

出版商:RSC    

年代:1958

数据来源: RSC

摘要:

June, 19581 MINERAL MATTER ASSOCIATED WITH COAL 37 1 THE DETERMINATION OF STARCH IN SAUSAGE MEAT AND CALCULATION OF THE NITROGEN CORRECTION FOR CEREAL CONTENT THE method published by the Society for Analytical Chemistry for estimating meat content1 relies on the Stubbs and More calculation, i.e., an assessment of the amount of filler by means of a "difference" figure, followed by deduction of the filler nitrogen from the total nitrogen, after which the corrected nitrogen figure is related to lean meat by means of a factor. The method described for the determination of starch as a check on the amount of carbohydrate in the filler is rather cumbersome, and the proposed method for determining the cereal-nitrogen correction makes use of a procedure that has been applied to different types of wheat flour2 and which is easier and more rapid.METHOD REAGENTS- Acidi$ed calcium chloride solution-Dissolve 620 g of calcium chloride hexahydrate crystals in 180 ml of distilled water and filter the solution until clear. Add a solution containing 18 g of sodium acetate trihydrate in 50 ml water to the clear filtrate and adjust the mixture to pH 2.3 by the addition of analytical-reagent grade glacial acetic acid. Adjust the specific gravity of the solution to 1.30 at 20" C. Cavrez's solution I-Dissolve 21.9 g of zinc acetate dihydrate and 3.0 ml of glacial acetic acid in 100 ml of water. Cavvez's solution 11-Dissolve 10.6 g of potassium ferrocyanide in 100 ml of water. PREPARATION OF THE SAMPLE- a rn0rtar.l Prepare the sample by passing it through a mincing machine and then mixing rapidly in PROCEDURE- Add 50 ml of acidified calcium chloride solution and heat in an autoclave for 10 minutes a t a pressure of 15 lb.per sq. inch. Cool the mixture by immersion in cold water and transfer it to a 100-ml calibrated flask by washing with acidified calcium chloride solution until the volume is approximately 90 ml. Add 2.0 ml of Carrez's solution I, shake the mixture well and add 2-0 ml of Carrez's solution 11. Shake the mixture again and then dilute to 100 ml at 20" C with acidified calcium chloride solution. Filter the dispersion on a Whatman No. 541 filter-paper. Discard the first 15 to 20 ml of filtrate and obtain the polarimeter reading at 20" C in a 20-cm tube on the subsequent runnings. If P is the reading at 20" C in a 20-cm tube and [a]ioo is 200, then the amount of starch present is given by- Mix about 5 g of sample with 10 ml of water in a 400-ml tall beaker.The filtrate should be perfectly clear. P x 104 per cent. 400 x weight taken372 NOTES [Vol. 83 If the reading is made with a saccharimeter and the reading is V , then the amount of starch present is given bv- - V x 0.3462 x lo4 per cent. 400 x weight taken Determine the amount of nitrogen associated with this amount of starch in the filler by multiplying the starch content by 0.025. Carry out a Kjeldahl determination on the sample to find the total nitrogen content, correct the result for the filler nitrogen and relate the corrected figure to the lean meat content by means of the usual factors, Le., for beef, 100j3.4 and for pork, 100/3.6.If the total meat content is desired, carry out a f a t determination. DISCUSSION Three typical results on pork sausage are given in Table I. The results are slightly lower than those obtained by using the Stubbs and Mc’re calculation. This is probably because the Stubbs and More method assumes a rather lower nitrogen content in the estimated filler than would be usual in present-day sausage rusks.a TABLE I DETERMINATION OF LEAN MEAT IN PORK SAUSAGES Starch found, % . . . . .. . * * . . . 9.78 10.74 9.33 Nitrogen due to filler, yo . . . . . . . . . . 0.24 0.27 0.23 Total nitrogen, Yo . . . . . . * . .. . . 1.59 1.45 1.67 Nitrogen due to lean meat, % , . . . .. . . 1.35 1.18 1.44 Lean meat found,* % . . . . . . . . .. .. 37.5 32.8 40.0 Lean meat found by Stubbs and More’s method, yo . . 38.3 33.6 4043 *The factor used was 100j3.6. If the proposed method were adopted for checking in manufacture, the true ratio of starch (determined polarimetrically) to nitrogen in the filler could be used instead. The values suggested here are based on the figures found for the composition of many bulked samples representative of the types of flour availahle in the United Kingd~m,~ from which it can be seen that the mean starch to protein ratio is 7 to 1 Since flour protein is given by the percentage of nitrogen found multiplied by a factor of 5.7, the nitrogen correction corresponding to the amount of starch found is given by the percentage of starch found multiplied by a factor of 0.025. Normally, a polarisation technique for the determination of starch in products such as sausage would be open to objection. Errors could be made (clue to (a) possible optically active constituents in the meat itself, ( b ) the degree of modification of the starch when the flour is made into rusks, and (c) the presence of other organic matter in the itour.The examination of both beef and pork by the proposed method has failed to show the presence of any optically active matter. Experi- ments on rusks by methods evolved for wheat flour2 indicate that, on average, the modification of the starch during rusk manufacture necessitates only a small correction of its specific rotatory power from [a]F = 203 t o [CC];~” = 200. The final objection, (c), has been dealt with by Clen- denning6 and for practical purposes is negligible.The method is not intended to replace the other methods for determining meat content, but, because of its ease and speed of application, it could be used as a check on the Society for Analytical Chemistry’s method and the Stubbs and More calculation, and would certainly provide a simple routine method. It would, of course, be unsuitable without modification if fillers such as soya or dried-milk powder were present. We are indebted to Dr. S. M. Herschdoerfer for samples of rusks used in sausage making, and to the Government Chemist for permission to publish this Note. REFEREXES Analytical Methods Committee, “The Analysis of Meat Products,” Analyst, 1952, 77, 543. f Fraser, J . R., Brandon-Bravo, M., and Holmes, D.C., J . Sci. Food Agric., 1966, 7, 577. Kent-Jones, D. W., and Amos, A. J., “Modern Cereal Chemistry,” Fifth Edition, Northern Pub- Fraser, J. R., and Holmes, D. C., J . Sci. Food Agric., 1956, 7, 589. Clendenning, K. A., Canad. J . Res. B., 1946, 23, 245. J. R. FRASER CLEMENT’S INN PASSAGE D. C. HOLMES 1. 2. 3. 4. 5. lishing Co. Ltd., Liverpool, 1967, p. 383. DEPARTMENT OF THE GOVERNMENT CHEMIST STRAND, LONDON, W.C.2 Received October 10th. 1957June, 19581 NOTES 373 FLAME-PHOTOMETRIC DETERMINATION OF SILVER IN BLISTER COPPER THIS investigation was undertaken in order to reduce the time taken in the routine assay of metallurgical products. Rathje' has shown that silver can conveniently and accurately be determined by flame photometry; i t was decided, therefore, to combine the precipitation of silver as halidea with the flame analysis.METHOD APPARATUS- A Zeiss PMQll quartz spectrophotometer with flame attachment was used. The oxy- hydrogen flame, at pressures of 200 mm of hydrogen and 0.3 kg per sq. cm of oxygen, was found to be most satisfactory. REAGENTS- All reagents should be of recognised analytical grade. Sodium chloride solution-A 20 per cent. aqueous solution. Nitric acid, 20 per cent. v/v. Nitric acid, 5 per cent. v/v-A 5 per cent. aqueous solution containing 1 per cent. of sodium Nitric acid, diluted (1 + 1). Nitric acid, concentrated. Ammonia soZution-Ammonia solution, sp.gr. 0.880, diluted (1 + 3) with water. Silver nitrate. Bromine. chloride. PROCEDURE- Weigh an amount of clean blister (2 assay tons*) to give a final solution containing about 50 p.p.m.of silver and dissolve it in sufficient concentrated nitric acid, about 225 ml, adding the acid in 30-ml portions. When the vigorous reaction has subsided, add 1 ml of bromine t o oxidise the sulphur and set aside for 10 minutes. Add 50 ml of diluted nitric acid (1 + 1) and heat to dissolve any remaining copper. Boil the solution to remove bromine and nitrous fumes and then add 100 ml of water and 5 ml of sodium chloride solution. Boil the solution for a further 10 minutes and then set it aside in a dark place overnight. Filter the cold solution through a Whatman No. 40 filter-paper and wash out the beaker thoroughly. Wash the filter-paper alternately with cold water and cold 5 per cent. nitric acid wash solution.Place a clean 250-ml beaker under the funnel and pour 10 ml of ammonia solution over the precipitate and filter-paper. Allow to drain and pour a further 10 ml of ammonia solution through the filter-paper. Follow this with 20 ml of hot 20 per cent. nitric acid and then two further 10-ml portions of ammonia solution. Finally, wash the filter-paper three times with small portions of water. Make the filtrate up accurately to 100 ml with water and use this solution for the flame-photometric determination. When the copper has been removed, wash well with water. TABLE I RECOVERY OF SILVER FROM BLISTER COPPER The blister used in each test was known to give 25 p.p.m. of silver per assay ton in the final solution Amount of sample Amount of silver Amount of silver Amount of added silver taken, assay tons added, p.p.m.found, p.p.m. recovered, p.p.m. 2 Xi1 51 1 2 10 59 9 2 50 101 51 1 50 75 50 Set the instrument to give full-scale deflection with a reference solution containing 100 p.p.m. of silver, 40 ml of ammonia solution and 20 ml of 20 per cent. nitric acid in a volume of 100 ml. Prepare a calibration graph from readings taken on known concentrations of silver in ammonia solution. * An assay ton is the weight of sample that must be taken so that 1 mg is equivalent to 1 oz Troy per ton, i.e., in this work, 1 assay ton is 29.17 g, representing 1 ton of 2000 lb.374 SOTES [Vol. 83 RESULTS The most suitable wavelength was found to be 338-3mp. The flame background at this wavelength was very small and, in practice, could be neglected.If accurate results are required, the background radiation can be measured at 335 and 341 mp and allowed for in the usual way. Results of some recovery experiments are given in Table I. The proposed method was com- pared with a method that involves cupellation and then weighing the silver bead produced. The Tesults of three comparative analyses were as follows- Weight of sample taken, assay tons . . .. 2 2 2 Silver found by proposed method, p.p.m. . . .. 48 53 49 Silver found by cupellation method, p.p.ni. .. 46 54 48 The proposed method reduces working time to about 2 hours and obviates the necessity The calibration graph was found t o be slightly curved and to approach a limiting No interference was noted from sodium or magnesium, as these were removed during filtration.for cupellation. value as the silver concentration increased. None of the common metals interfered. thank the Roan Antelope Copper Mines Ltd. for permission to publish this Note. REFERENCES 1. 2. ROAN ANTELOPE COPPER MINES LTD. Rathje, A. O., Anal. Chem., 1955, 27, 1583. Higgs, D. G., Analyst, 1944, 69, 270. CHEMISTRY LABORATORY LUANSHYA N. RHODESIA N. McN. GALLOWAY Received November 29th, 1957 THE SIGNIFICANCE OF THE POINTAGE TITRATION IN THE CONTROL OF ZINC PHOSPHATING BATHS THE complete analysis of phosphating solutions can be carried out by such procedures as those of Bush, Higgs and B0x.l However, in industrial operations, sufficient control is obtained by using the well known pointage titration. The pointage is usually defined as the titre of 0.1 N sodium hydroxide solution required to neutralise a 10-ml aliquot of the bath solution, phenol- phthalein being used as indicator.The titration is variously stated to measure the total ortho- phosphoric acid2 or simply the total acid,3 as distinct from the free orthophosphoric acid, which is titrated when methyl orange is used as indicator. The reactions that occur and the way in which the total phosphate content can be related to the pointage are subjects seldom discussed. Cupr and Hemala4 studied the comparatively simple example of the titration of an acid solution of zinc orthophosphate with sodium hydroxide and considered that the precipitate was a mixture of zinc secondary and tertiary phosphates. Rathje5 precipitated zinc phosphate by mixing solutions of a zinc salt, potassium dihydrlogen phosphate and sodium hydroxide, and deduced that only in a certain range of concentrations is a precipitate of constant composition, 3Zn,(P04),.Zn(OH2), formed.In practice, complications arise owing to the fact that any heavy metals present in excess of the equivalent combined phosphate can also be expected to react during the pointage titration. EXPERIMENTAL Some data, which help to explain the reactions that occur, have been obtained during operation of a commercial nitrate-accelerated zinc phosphate bath at an operating strength of 70 to SO points and a temperature of about 185' F. COMPOSITION OF PRECIPITATE- diffraction pattern was similar to that of hopeite, Zn,(PO,) ,.4H20. indicated the presence of 51.2 per cent.of zinc and 47.9 per cent. of phosphate, as PO,. formula Zn,(PO,), requires 50-8 per cent. of zinc and 49.2 per cent. of phosphate. suggest that the precipitate is mainly zinc orthoph.osphate, Zn,(PO,) ,.4H,O. The precipitate present at the end-point of the pointage titration was examined. The X-ray Analysis of ignited residues The The resultsJune, 19581 NOTES 375 CALCULATION OF POINTAGE- zinc dihydrogen orthophosphate in the bath to the titre at the phenolphthalein end-point. these are termed titre 1 and titre 2, respectively, the following reactions can be postulated- It was now possible to consider the separate contributions of free orthophosphoric acid and If H,PO, + 2NaOH -----f Na,HPO, + 2H,O . . . . . . . . titre 1 3Zn(H,PO,), + 8NaOH -+ Zn,(PO,), + 4Na,HPO, + 8H,O .. titre 2 Pointage titres were calculated from these reactions for comparison with the observed values. The pointage was found after taking a 5-ml aliquot from a cooled sample and diluting it to about 50ml with distilled water. Titre 1, the free orthophosphoric acid content, was determined by direct titration, methyl orange being used as indicator. Titre 2 was found after the total phosphate present had been determined by Gabrielson's ion-exchange technique,6 in which the cooled solution is passed through a cation-exchange column. The results are shown in Table I, expressed as the amounts of 0.1 AT sodium hydroxide solution equivalent t o 10 ml of phosphate solution. The results show that, although the pointage titration is a measure of the total phosphate content for the synthetic solution, it measures some constituent in addition to the phosphate content in the nitrate-accelerated bath.If alkali metals had been present equivalent to the nitrate content, the pointage titration would have been unaffected. It appeared, therefore, that heavy metals were present in excess of the zinc required for titre 2. DETAILED STUDY OF SELECTED BATH- The major constituents, expressed in g per litre, were as follows: free acid, as H,PO,, 16.1; combined phosphate, as H,PO,, 19.5; nitrate, 51.2; zinc, 21.3; sodium, 7.1. Curves for potentiometric titrations with 0.1 N sodium hydroxide solution showed only one sharp inflexion a t pH 4.28. After passage through a cation- exchange column, a further potentiometric titration curve showed points of inflexion a t pH 4.65 and 9.06, which correspond to the first and second neutralisation stages of orthophosphoric acid.TABLE I The first bath in Table I was then studied in more detail. CALCULATION OF POINTAGE Zinc dihydrogen acid (titre l), (titre 2 ) , (titre 1 + titre 2 ) , Free orthophosphoric orthophosphate Total acidity ml ml ml 32.8 26.8 59.6 30.8 29.5 60.3 27.6 26.1 53.7 27.4 23.2 50-6 35.8 54* 41.2 Pointage found, ml 82.0 85.0 78.8 69.0 41*3* * Results from a synthetic solution prepared by dissolving zinc oxide in orthophosphoric acid. A large excess of zinc is present over that equivalent to combined phosphate, so that to titres 1 and 2 there must be added a third titre, to account for this excess of zinc, which probably reacts in the following way- 3Zn2+ + 2Na,HPO, + 2NaOH -+ Zn,(PO,), + 6Na- + 2H,O .. titre 3 Titre 3 was calculated from the analytical data and the calculated pointage (the sum of titres 1, 2 and 3) was compared with the observed value for the first bath in Table I. Titre 1 was 32.8 ml, titres 2 and 3 were 26.8 and 16.3 ml, respectively, thus giving a calculated pointage of 75.9 ml. The agreement between this figure and the observed pointage, 82.0 ml, is much better. The remaining discrepancy is probably accounted for by co-precipitation effects and the start of dissolution of the precipitate. COKCLUSIONS The results indicate, as one might expect, that the pointage titration for zinc phosphate baths (and probably also for iron and manganese phosphate baths) is a measure of the total orthophosphate, according to the usual definitions, only if other heavy-metal salts, such as zinc nitrate, are absent.I n this instance, the combined phosphate (titre 2) can be approximately deduced from a separate free orthophosphoric acid titration. When zinc is present in excess of376 NOTES [Vol. 83 combined phosphate, as in the commercial bath st.udied, such a simple calculation is impossible and the pointage titration is a measure of all the major active bath constituents, i.e., free ortho- phosphoric acid, zinc dihydrogen orthophosphate and the accelerator. I thank Dr. T. Emmerson, Director of Research, G.K.N. Group Research Laboratory, for permission to publish this Note and Mr. D. Wilkinson for carrying out the X-ray examination. REFEREN CES Bush, G.H., Higgs, D. G., and Box, F. W., Analyst, 1955, 80, 885. Machu, W., “Die Phosphatierung,” Verlag Chemie, G.m.b.H., Weinheim, 1950, p. 286. Gilbert, L. O., “Phosphating Materials and Processes,” Rock Island Arsenal Lab. Tech. Rep. Cupr, V., and Hemala, M., Publications de la Facult4 des Sciences de 1’Univarsite’ Masaryk (Brno), Rathje, W., Ber., 1941, 74, 357. Gabrielson, G., Metal Finishing, 1953, 51, 76. 1. 2. 3. 4. 5. 6. No. 54-2906 (PB.111726), 1954, p. 61. No. 307, 1948. G.K.N. GROUP RESEARCH LABORATORY LANESFIELD, WOLVERHAMPTON BIRMINGHAM NEW ROAD F. J. ARMSON Received September 2nd, 1957 A POTENTIAL SOURCE OF SAMPLING ERROR IN THE DETERMINATION OF NITROGEN IN TITANIUM ALLOYS DURING the course of experimental work on the apparent segregation of nitrogen in large billets of titanium alloys, erratic results caused us to suspect that serious contamination by nitrogen could be caused during preparation of samples by milling or drilling in the manner customary for the sampling of metals.Solid specimens, millings and drillings taken at several cutting speeds, were prepared from two different types of titanium alloy to test the effects of drill speed, size and wear on contamination of titanium samples by atmospheric nitrogen caused by heating at the cutting edge of the tool. When drillings were compared with lump samples, the latter were pickled in dilute sulphuric acid (1 + 4) to remove worked layers and the small particles of cutting tool that were welded into the machined surfaces. Nitrogen was determined by a method used for steel, namely, Kjeldahl digestion with sulphuric acid and sodium sulphate and then steam- distillation from sodium hydroxide solution and titration of the ammoniacal distillate with 0.01 N hydrochloric acid.A solution of 3.5g of the sample in hydrochloric acid was used for the determination. The results on samples from an alloy containing 2.6 per cent. of chromium and 1.5 per cent. of iron are summarised in Table I. The results of experiments on drilling samples at several cutting speeds from an alloy containing 4 per cent. of manganese and 4 per cent. of aluminium are shown in Table 11. TABLE I Both alloys were fairly hard, their Birinell numbers being between 320 and 370. EFFECT OF THE METHOD OF SAMPLE FREPARATION ON THE APPARENT OF CHROMIUM AND 1-5 PER CENT.OF IRON NITROGEN CONTENT OF A TITANIUM AL,LOY CONTAINING 2.6 PER CENT. Method sampling tool of Sampling Billet slice Drilling # inch H Hacksaw (11 inches x 11 inches) } Lumps Drilling Q inch H Milling 1 inch MC Billet slice (9 inches x 9 inches) Lumps Hacksaw Bar } lvJvJgp: 1 inch MC (2 inch x $inch) Hacksaw Speed of tool, r.p.m. 1000 1000 130 130 - - Cutting speed, feet per minute 98 98 34 34 - - - Number of samples 10 12 4 7 4 3 5 Range of nitrogen results, 0.046 to 0.058 0.038 to 0-040 0-053 to 0.058 0.049 to 0.050 0.046 to 0.048 0.069 to 0.070 0.037 to 0-039 % Mean, 0.0515 0.039 0.0545 0.0495 0-0465 0.069 0.0385 % H = high-speed drill. MC = carbon-steel milling cutter. With all drills it was found that with each successive hole drilled the apparent nitrogen content increased.Results seriously in error could be obtained before the drill became too blunt to use. The %-inch carbide-tipped drill retained a sharp cutting edge for longer than a similar high-speed steel drill, but even the former was starting to show wear after drilling six holes each 4 inch deep.June, 19581 NOTES 377 TABLE I1 EFFECT OF THE METHOD OF SAMPLE PREPARATION ON THE APPARENT NITROGEN CONTENT OF A TITANIUM ALLOY CONTAINING 4 PER CEXT. OF MANGANESE AND 4 PER CENT. OF ALUMINIUM Samples were taken from a billet 12 inches in diameter Nethod of sampling Drilling Drilling Drilling Drilling D r i 11 in g Lumps Sampling tool 8 inch H 8 inch H or C 8 inch H or C $ inch H or C 1 inch D Hacksaw Speed of tool, r.p.m. 1000 440 150 80 25 - Cutting speed, feet per minute 98 43 14.7 7.9 6.5 Number of samples 13 6 6 6 2 12 Range of nitrogen results, ”/o 0.035 to 0.097 0.037 to 0.042 0.035 to 0.038 0.035 to 0.037 0.037 0.030 to 0.038 Mean, 0-0565 0.0385 0.037 0.0355 0.037 0-034 % H = high-speed steel drill.C = carbide-tipped drill. D = spearpoint carbon-steel drill. The results show clearly that it is essential to drill at low speeds and with a relatively coarse feed, a freshly sharpened drill being used if contamination of the sample by nitrogen is to be avoided. Since this work was carried out, a general warning on errors in sampling titanium has been published,l and low cutting speeds with a feed rate of 0-005 to 0.006 inch per revolution have been recommended for drilling titanium alloys.1? However, the magnitude of the possible error has not been generally known, and this Note is published to call it to the attention of chemists engaged in the analysis of titanium alloys. We thank Mr.E. W. Colbeck, Metallurgical Director of Hadfields Limited, for permission to publish this Note. REFERENCES 1. 2. “The Analysis of Titanium and its Al:oys,” Imperial Chemical Industries Ltd., London, 19.56, “The Machining of the Rarer Netals,” Murex Ltd., Rainham, Essex, 1952, p. 11. pp. 6 and 7. HADFIELDS LIMITED EAST HECLA WORKS SHEFFIELD, 9 J. D. HOBSON D. SWINBURN Received May 7th, 1957 A RAPID SPECTROSCOPIC METHOD FOR THE DETERMINATION OF WATER IN GLYCEROL UNTIL recently, spectroscopic work in the infra-red region has been centred on the fundamental 2 to 15-11 region.The near infra-red region, 0.7 to 2.0 p , is a region that is only recently beginning to be exploited. This region has a number of advantages over the fundamental region, e.g., path lengths are measurable with a micrometer, and liquids can be contained in simple robust glass ce1ls.l A particular advantage of this region is that, when overlapping of absorption frequencies occurs in the fundamental region, a useful separation often exists between the combination frequencies. This occurs with the fully bonded hydroxyl-group bands of water and alcohols.2 These bands are hardly separable a t 3 p, but the combination bands near 2 p are clearly separated by over 300cm-l. A band a t 1.93 p , attributable to absorption due to water, is clearly separated from the band arising from the hydroxyl groups of the glycerol at 2.1 p.Typical spectra of two glycerols containing 1.82 and 9.85 per cent. of water are shown in Figs. 1 and 2. These spectra were measured with a Cary recording spectrophotometer, model 14, and a glass cell of path length 1 rnm was used. The double-beam arrangement of the Cary instru- ment, with a glycerol containing a trace of water in the reference beam, was used to record difference spectra, which are also shown. Calibration of glycerols containing known amounts of water ranging from 1 to 20 per cent. with the optical density of the 1-9-11 band gave an excellent linear relationship. The procedure for determining the water content of an unknown glycerol is merely to fill a l-mm glass cell with the glycerol, scan the 1-9-p region, determine the optical density of the 1-9-p band and deduce the water content from the calibration curve. With the limited number of determinations so far made, the standard deviation corresponds to +0.06 per cent.of water over the range 1 to 20 per cent. on pure glycerols. This is the basis of the present method.378 NOTES, [Vol. 83 A number of typical results by this method 011 glycerols containing only water as impurity are shown in Table I and compared with results obtained by means of a Karl Fischer apparatus. Wavelength, I.I Fig. 1. Spectra of glycerol containing 1.82 per cent. of water: curve A, difference spectrum against water; curve B, near infra-red spectrum I =M 0.2 0 1.8 2.0 2.2 B ! ! ! ! i ! ! ! j ! ! ! 1.0 (2.0) 0.8 ( I .8) 0.6 (1.6) 0.4 ( I .4) 0.2 (I .2) 0 (1.0) 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4 2.6 Wavelength, IJ Fig.2. Spectra of glycerol containing 9.85 per cent. of water: curve A, difference spectrum against water; curve B, near infra-red spectrum TABLE I COMPARISON OF RESULTS Water found Water found by spectroscopic method, by Karl Fischer method, % % Pure glycerol . . .. 0.45 0.60 0.84 10.29 19.1 Crude glycerol 6.89 7.24 7.63 13.65 0.42 0.60 0.78 10.40 19-41 6.82 7.14 7.55 13.71June, 19581 NOTES 379 Some crude discoloured glycerols were also examined and their water contents deter- mined by both methods; the results by the spectroscopic method were consistently higher (average error + 0.8 per cent.). This is considered to be due to scattered radiation from the dispersed salt and iron present in the samples.A constant correction factor improved the over-all agreement between the two methods. A simpler instrument than the more versatile Cary spectrophotometer used in the present investigation could be devised to carry out this determination. We thank T. H. Newlove and G. G. Jones for the moisture determinations with the Karl Fischer apparatus, and the Directors of Unilever Limited for permission to publish this Note. REFERENCES 1. 2. PORT SUNLIGHT, CHESHIRE Kaye, W., Spectrochim. Acta, 1954, 6 , 257; 1955, 7 , 181. Miller, R. G. J., and Willis, H. A., J . AppZ. Chem., 1956, 6, 385. UNILEVER LIMITED D. CHAPMAN J. F. NACEY Received July 23rd, 1957 FINENESS OF FERTILISER SAMPLES FOR a great many years we have found it possible to obtain concordant duplicates in fertiliser analyses by grinding to pass a 22-mesh sieve (British Standard 410: 1943), which is a little finer than the sieve recommended in the 1955 Regulations, i.e., “apertures about one-millimetre square.” Recently we have found that, with the more concentrated fertilisers, such as those having N : P,O,: K,O in the ratio of 10: 10: 16, it is impossible to weigh out duplicate 2-g portions that will give the same analysis.Inquiry showed that, for technical reasons connected with granulation, a new crystal form of ammonium sulphate is now favoured, namely, short thick prisms instead of needles. Examination with a low-power microscope indicates that this, together with the high proportion of ammonium sulphate, accounts for the separation and lack of uniformity (10 per cent.of nitrogen corresponds to about 47 per cent. of ammonium sulphate). We find that grinding to pass a 30-mesh sieve (British Standard 410: 1943) makes it easy to obtain close duplicates, and microscopic examination of the more finely ground material shows that the crystals are largely fractured and that the tendency to separate is obviated. I t is hoped that the publication of this Note will save other analysts from some rather puzzling results. WATERFALL & O’BRIEN 4 QUEEN SQUARE BRISTOL, 1 H. S. HOWES Received December 19th, 1957 A MICRODIFFUSION METHOD FOR THE DETERMINATION OF HYDROGEN SULPHIDE HYDROGEN sulphide either in biological materials or that liberated in enzymic reactions has frequently been determined by blowing the gas into a solution of a heavy-metal salt with a stream of nitrogen or carbon d i o ~ i d e , l ~ ~ 9 3 and then determining the heavy-metal sulphide either by titrations or colorimetrically.1~2 9 3 Such methods suffer from the disadvantage that, when protein solutions are used, serious foaming is liable to occur.The application of the well known principles of microdiffusion4 to the determination of amounts of the order of 1 p g of hydrogen sulphide is described. The method is simple and the high sensi- tivity combined with the high rate of diffusion have permitted it to be used for more detailed studies of the kinetics of certain enzymic reactions in which hydrogen sulphide is liberated than could have been attempted by other methods.It seems possible that, with minor modifications, the method might be of fairly general use for determining small amounts of hydrogen sulphide. A diffusion technique has been applied previously5 ,B to the determination of enzymically produced hydrogen sulphide, but the method was relatively insensitive and details of the rate of diffusion were not given. The method described by Conway on p. 244 of his book4 also lacks high sensitivity. In the proposed method, an apparatus (see Fig. 1) similar to that introduced by Brown et aZ.7 for the determination of ammonia is used. A filter-paper, D, moistened with 0.01 ml of zinc acetate reagent solution (see p. 381) is held near the centre of a glass bottle of capacity about 50 ml, which380 YOTES [Vol. 83 is rotated about its axis on horizontal rollers at about 120 r.p.m.; the hydrogen sulphide diffuses from the solution in the bottle on to the paper and is then eluted and determined colorimetrically. A groove, A, in one side of the bottle provides two separate compartments, B and C, whenthe bottle is held in the position shown in Fig.1 and the solutions to be used are placed in these before the stopper and paper are inserted. Sometimes it is necessary to work under anaerobic conditions, and, when this is so, a stream of nitrogen saturated with water is passed through the assembled apparatus by way of the poly(viny1 chloride) and polythene tubes, H and E, for 5 minutes. The bottle is then closed by doubling back the poly(viny1 chloride) tubes, H, and securing them w-ith the rubber bands, G (see lower tube H, Fig.1); the nitrogen supply is then disconnected and the bottle is tilted to mix the contents of B and C and placed on the rollers for a suitable length of time. A = B = c = D = Groove on one side of E = Polythene tubes, bore Up per corn part ment F = Soft rubber stopper Lower compartment G and I = Rubber bands cut from Filter-paper, diameter 2-3 rubber tubing cm, folded and held H = Poly(viny1 chloride) tubes between polythene tubes bottle 1.5 mm Fig. 1. Section through bottle To measure the hydrogen sulphide taken up, the paper is removed with forceps and dropped into a test-tube containing 3.5 ml of amine reagent solution (see p. 381) and the tube is stoppered at once. The stopper is then removed momentarily, 0.5 ml of ammonium ferric sulphate reagent solution is added and the stopper is then replaced and the contents of the tube are mixed.The blue colour of the solution is measured with a spectrophotometer a t 675mp, preferably 1 to 10 minutes after the addition of the oxidising agent, as slight fading sometimes takes place on standing. Time on roners, minutes Fig. 2. Diffusion of 1.4 pg of hydrogen sulphide from phosphate buffer solution, pH 7, at about 24" C on t o zinc acetate papers. Values are from several different experiments ; the average recovery when diffusion was complete was 80 per cent. of that from experiments in which sodium sulphide solution had been added directly to the colorimetric reagents Hydrogen sulphide. yg - Fig. 3. Calibration curve for hydrogen sulphide prepared by dif- fusion from sodium sulphide solution and phosphate buffer solution for 8 minutesJune, 1958; NOTES 381 Results with 2 ml of phosphate buffer solution a t pH 7 in C and sodium sulphide solution in B are shown in Figs.2 and 3. Fig. 2 shows that under the conditions used two-thirds of the hydrogen sulphide diffused in the first minute on the rollers and that diffusion was virtually complete after 5 to 10 minutes. Fig. 3 shows a typical calibration curve with different amounts of sodium sulphide solution and a rolling time of 8 minutes; the slight deviations from linearity shown appeared to be fairly reproducibile. In order to speed the diffusion in other work on total hydrogen sulphide determinations, it might well be advantageous to replace the phosphate buffer solution by a more acid medium; the pK value of hydrogen sulphide is about 7 .For determinations of enzymically produced hydrogen sulphide, the enzyme sample is placed in C and the substrate in B, and the tests are carried out in an atmosphere of nitrogen with the rollers inside a constant-temperature box. I t may generally be assumed in such experiments, in view of the high diffusion rate shown in Fig. 2, that the rate of uptake of hydrogen sulphide by the paper is virtually equal to the rate of its production by the enzyme system; the method is therefore well suited for rapid studies of the kinetics of such enzymic reactions. The reagents required are as follows (compare Johnson and Xishita*)- Zinc acetate reagent solution-A solution of 10 g of zinc acetate dihydrate and 2.5 g of sodium acetate trihydrate in 50 ml of water. A i n i n e veugent solution-A 0.01 per cent. solution of dimethyl-p-phenylenediamine sulphate (Hopkin & Williams Ltd.) in 1.5 per cent. vjv sulphuric acid. (This solution should not be stored for more than 1 week.) A m m o n i u m fevvic sulphnte yeagent solution-A 2 per cent. solution of ammonium ferric sulphate in 1.3 per cent. v/v sulphuric acid. I thank Professors F. Bergel and A. Haddow for their interest in this work, n-hich was supported by grants from The British Empire Cancer Campaign, the Jane Coffin Childs Memorial Fund, the Anna Fuller Fund and the National Cancer Institute of the National Institutes of Health, U.S. Public Health Service. REFEREKCES Almy, L. H., J . Bmev. Chem. Soc., 1925, 47, 1381. Desnuelle, P., and Fromageot, C., EizzyynoZogia, 1939, 6, 80. Ohigashi, I<., Tsunetoshi, 1., Uchida, M., and Ichihara, K., J . Biochent., Tokyo, 1962, 39, 211 Conway, E. J., “WIicrodiffusion Analysis and Volumetric Error,” Fourth Edition, Crosby Lockaood Kallio, R. E., and Porter, J. R., J . Buct., 1950, 60, 607. Smpthe,.C. V.> iz Colowick, S. P., and Kaplan, K. A, Editors, “Methods of Enzymology,” -1cademic Brown, R. H., Duda, G. D., Korkes, S., and Handler, I?., Arch. Biochem., 1957, 66, 301. Johnson, C. RI., and Nishita, H., A d . Chem., 1952, 24, 736. 1. 2. 3. 4. 5. 6. 7. 8. & Son Ltd., London, 1957. Press Inc., Kew York, 1955, p. 315. CHESTER BEATTY RESEARCH INSTITUTE INSTITUTE OF CANCER RESEARCH: ROYAL CANCER HOSPITAL FULHAM ROAD R. C. BRAY LONDON, S.W.3 Received September 25th, 1967

ISSN:0003-2654    

DOI:10.1039/AN9588300371

出版商:RSC    

年代:1958

数据来源: RSC

摘要:

June, 1958: NOTES 381 Book Reviews STABILITY COXSTANTS OF METAL-ION COMPLEXES, WITH SOLUBILITY PRODUCTS OF IXORGANIC SUBSTANCES. Compiled by JANXIK BJERRUM, GEROLD SCHWARZENBACH and LARS GUSNAR SILLEN under the Auspices of The International Union of Pure and Applied Chemistry. Compiled from publications earlier than January lst, 1956, with collaboration by G. AXDEREGG and S. E. RASntESSEX. Pp. xvi + 105. London: The Chemical Society, 1957. Price 60s. (Special Publication No. 6). The ability to adjust the position of ionic equilibria in solution forms the basis of many analytical procedures. It is certainly true for qualitative and quantitative separations, for precipitation and solvent extraction, for masking and sequestration and in the production of coloured complexes for absorptiometry.But nowhere is the effect so apparent as in the use of aminopolycarboxylic acids (complexones), and Schwarzenbach has shown in his monograph on Complexometric Titrations (Methuen, 1957) how a knowledge of the relevant dissociation constants of proton - ligand and metal - ligand complexes permits a correct choice of conditions in, e.g., the determination of hardness in water with EDTA and a metal indicator. Similar data are of great value for making an intelligent choice of masking agent and ascertaining the optimum pH for its use. PART I : ORGANIC LIGAXDS.382 BOOK REVIEWS [Vol. 83 For some years a group of scientists has been collecting data on the stability of complexes and the present volume is a result of international co-operation directed by the Commission on Equilibrium Data of the I.U.P.A.C.The literature survey has been completed up to the end of 1955 and includes data for 464 organic ligands, which are listed in order of their empirical formula according to Beilstein’s system. For each ligand, L, are quoted the stability constant for every published metal complex, the pK values for the conjugate acids HL,H,L, etc., the methods used in their determination, the ionic strength and temperature of measurement and the literature references. To indicate the coverage, there are 11 stability constants itemised for the first entry (CH,O,, formic acid) and 30 for the last entry (C:32H32012N2, phthaleincomplexone), and no less than 148 under C,,H1,0,N2 (ethylenediaminetetra-xetic acid). There are excellent indexes of metals and of ligands, a concise introduction that is an effective and brilliant survey of methods of measuring and computing stability constants and a most valu- able 5 pages headed “How to use the tables.” Fublication by the Chemical Society as one of their Special Publications has enabled the price of this unique compilation to be kept reasonably low.Together with the companion volume dealing with Inorganic Ligands (which will shortly be published) this is certain to become an indispensa.ble reference book for every up-to-date analyst. H. IRVING ORGANIC ELECTRODE PROCESSES. By MILTON J. ALLEN. Pp. xiv + 174. London: Chapman This book deals essentially with the instrumental methods that might be used by the investi- gator interested in the electrolysis of organic compounds and further with the products arising from their cathodic reduction or anodic oxidation.The publication gives useful information about the nature and yield of the products obtained in the electrolysis as well as the mechanism of the processes. Particular attention is paid to methods of electrolysis under controlled conditions. The literature has been covered very thoroughly, its is shown by the list of 464 references. At first sight it might be thought that this book will appeal mostly to organic chemists seeking new ideas on preparative methods; polarographers, however, interested in cathodic reduction and anodic oxidation processes, will find the publication worth studying. Moreover, one has the feeling after reading this work that analytical chemists have not paid as much attention to the electrolysis of organic compounds as they might profitably have done.& Hall Ltd. 1958. Price 32s. J. HASLAM METHODS OF BIOCHEMICAL ANALYSIS. Volume V. Edited by DAVID GLICK. Pp. x + 502. K. B. Augustinson of Stockholm contributes a chapter on assay methods for cholinesterases (ChE), which catalyse the hydrolysis of choline esters, acetylcholine being the most important. These enzymes are numerous and may also split non-choline esters and exhibit very different specificity patterns, but only those esterases that are inhibited by M eserine should be designated ChE. Within the family of ChE enzymes there are two main groups-acetylcholin- esterases (I) and a more heterogeneous group (11), including the unspecific ChE or pseudo ChE. The complicated problems that arise from this situation are surveyed clearly and in adequate detail.An article on biological standards in biochemical analysis by J. H. Humphrey, D. A. Long and W. L. M. Perry emanates from the Xational Institute for Medical Research, London. “Biological standards are necessary only for impure substances and their rawon d’etre is evanescent, in that their useful life is limited to the time taken to elucidate, by biochemical analysis, their exact chemical composition and to prepare b y synthetic methods their active constituents, Can the standards themselves aid in shortening this period, can they commit felo de se?” With this as the main theme the authors say “drug control authorities speedily discovered that there was no way of eliminating animal variation and that the only way to reduce its effect on activity measurements was to carry out comparisons of the sample under test with a standard preparation, or reference substance, which was examined concurrently at every stage of the investigation.” After describing the arrangements for intern(itiona1 co-operation, the authors proceed to discuss antibiotic standards, hormone standards, enzyme standards, miscellaneous pharmacological standards and finally immunological standards.I t is interesting that the vitamin standards have served their turn and that, with the possible exception of cholecalciferol, they are not likely to be renewed when supplies are used up. New York and London: Interscience Publishers Inc.1957. Price $9.50; 75s. This is an interesting and authoritative chapter.June, 19581 BOOK REVIEWS 383 The next chapter, by W. J. P. Neish of Sheffield (formerly of Edinburgh), surveys the different a-keto acids known to occur in plants and animals, and outlines the development of analytical methods. Preliminary chromatographic separations of a-keto acids can be carried out and the disadvantages of non-specific methods can thus be overcome. Micro-methods for determining pyruvic and a-ketoglutaric acids are described, together with some apparently specific methods for determining these two acids and also oxalacetic acid, separately or in mixtures. B. E. Saltzman and R. G. Keenan of the U S . Public Health Service deal with the micro- determination of cobalt in biological materials, a problem of interest because of the role of cobalt in animal nutrition, in enzyme systems and in haemopoiesis via the cobalamines.Various colori- metric methods are described and a detailed procedure is given based on extraction with l-nitroso- 2-naphthol followed by spectrophotometric determination as the nitroso-R-salt complex (range 1 to 25 p g and sensitivity 0.1 pg). New ground will be broken for many readers in the chapter by B. A. Loveridge and A. A. Smales of A.E.R.E., Hanvell, on activation analysis and its application in biochemistry. “Many elements may be determined in the sub-microgram region by measuring the radioactivity induced when the sample is exposed to bombardment by nuclear particles. Neutrons are frequently used, but successful analyses have also been completed, using charged particles-protons, deuterons and helium nuclei-to create radioactive isotopes.” At first it was necessary to rely upon the particle-accelerating machines, but nowadays it is the “high neutron flux of the atomic reactor which has made radioactivation analysis a formidable tool for analytical purposes.” The intensity of the characteristic radiation emitted after bombard- ment is linearly related to the weight of the stable isotope activated.A weighed sample and a reference standard (with a known proportion of the element under analysis) are subjected to nuclear bombardment side by side. A chemical separation is then needed to remove all other induced radioactivity and an inert carrier for the element under study may be used.It is quite clear that the method has very great possibilities; it is highly specific, reasonably accurate and reproducible, very sensitive, and, once the irradiation has been done, it is immune to contamination errors. R. E. Thiers of Harvard contributes a chapter on Contamination in trace-metal analysis and its control. “Contamination . . . is the presence of an error . , . due to the entrance (or exit) of unwanted, unexpected and usually unknown quantities of the substance being measured.” This leads Thiers to the concepts of positive con- tamination, negative contamination and fiseudo-contamination. To designate unduly low results due to unexpected losses as negative contamination is a somewhat curious use of words; in the main it refers in practice to losses incurred during dry-ashing.Incidentally, pre-ashing (with a hyphen) seems a little less disconcerting than pveashing. This chapter must at one point or another be of interest to every analyst. It deals with handling samples, purifying reagents (including water), with the selection and cleaning of apparatus and with many ways of circum- venting the consequences of contamination. W. A. Bauld and R. M. Greenway of Montreal present an authoritative account of the chemical determination of oestrogens in human urine. The subject is discussed fully and the final procedures (a) based on adsorption chromatography of the methyl ethers and colorimetric determinations (J. B. Brown) and ( b ) based on column partition chromatography and colorimetric determination (Bauld) are set out in detail.Dr. Bauld devoted much time to this problem during his stay in Marrian’s laboratory at Edinburgh and it is pleasant to read how successfully it has been solved. H. Rosenkrantz of the Worcester Foundation, Mass., contributes a chapter on the infra-red analysis of vitamins, hormones and co-enzymes. He used throughout a spectrometer (Perkin - Elmer 12C) in which energy intensities between 3640 and 2500 cm-l were sacrificed so as to obtain a relatively constant transmission line at lower frequencies. The samples were used as potassium bromide pressed prisms, a method of examination that eliminates solvents. More than half the absorption curves so obtained were checked against, and found to agree with, earlier records by other procedures for preparing the absorbing sample. The method of presenting curves is, unfortunately, on a scale that is linear neither to wavelength nor frequency. The commentary in the text is, however, very valuable in that special features of the spectra are noted; in any case many of the curves can, when necessary, be enlarged and studied carefully, Rosenkrantz makes his point about the advantages of dispersing samples in potassium bromide, but does not overstate his case. In his conclusion he rightly says “interpretation of infra-red spectra comes with much experience and does not necessarily follow hard and fast rules.” This is a long and comprehensive article. Emission-spectrographic techniques are also described. He defines the “trace level” as 100 p.p.m. or less.384 PUBLICATIONS RECEIVED Surveying this volume as a whole, it is impossible for anyone who has himself battled with troublesome analytical problems not to feel respect and admiration for the authors and for their work. R. A. MORTON

ISSN:0003-2654    

DOI:10.1039/AN9588300381

出版商:RSC    

年代:1958

数据来源: RSC

摘要:

384 PUBLICATIONS RECEIVED RECOMMENDED METHODS ITOR THE ANALYSIS OF TRADE EFFILUENTS REPRINTS of the Recommended Methods prepared by the Joint A.B.C.M. - S.A.C. Committee on Methods for the Analysis of Trade Effluents are available from the Secretary, The Society for Analytical Chemistry, 14 Belgrave Square, :London, S.W.l; price to members 1s. 6d., or to non-members 2s. 6d. each. Remittances made out to the Society for Analytical Chemistry must accompany orders, and these reprints are not available through Trade Agents. Reprints Nos. 1 to 10 were listed in the November, 1957, issue; reprints Nos. 11 to 13 were listed in the February, 1968, issue. Reprints Nos. 3 and 8 are now out of print. The following reprints, which complete the series, are now available- Reprint No. 14. Reprint No. 15. Determination of Phosphorus and Acid-soluble Sulphate (January, 1958). Determination of Residual Chlorine, Cyanides and Thiocyanate, Fluoride, Formal- dehyde and Sulphite and Thiosulphate (April, 1858).

ISSN:0003-2654    

DOI:10.1039/AN958830384b

出版商:RSC    

年代:1958

数据来源: RSC