摘要:

706 HODGSON AND GLOVER : THE POLAROGRAPHIC DETERMINATION [Vol. 76 The Polarographic Determination of Aluminium, Zinc and Tin in Water BY H. W. HODGSON AND J. H. GLOVER (Presented at the meetiftg of the ljolarographic Discussion Pan,el of the Physical Methods Group o n Friday, July 6th, 1951) A method is described for the polarographic determination of aluminium, zinc and tin. The factors studied, which influence the reduction wave of aluminium, are the pH and the composit:ion and concentration of the base solution. Conditions have been selected for the simultaneous estimation of zinc and aluminium. The polarographic reduction of aluminium is considered to be irreversible, with a half-wave potential of -1.63 volts against the saturated calomel electrode. DURING an investigation into the corrosion of soldered aluminium it became evident that a polarographic method should be developed €or the rapid determination of small concentrations of aluminium, zinc and tin.Stackelberg’ has reviewed published work and outlined the difficulties encountered in the quantitative polarography of aluminium. In alkaline solutions aluminates are not reducible at a dropping-mercury electrode and in acid solutions the hydrogen-ion wave interferes; as hydrolysis and precipitation occur in neutral solutions it is evident that careful adjustment of pH is necessary. Prajzler2 obtained a well-defined wave for aluminium in 0-05 N barium chloride solution, but Stackelberg states that the results are not precise owing to hydrolysis. Much of the work on the polar’ography of aluminium is connected with the analysis of light alloys; Gull3 has described the determination of aluminium, zinc, lead and manganese in magnesium alloys by use of a solution of the sample in hydrochloric acid as the base solution. Exact neutralisation to bromophenol blue was necessary and the solutions were left to stand for at least three hours to attain pH equilibrium before analysis.In a paper on the polaro- graphy of light metals and alloys, Stross4 stated that Gull’s procedure was satisfactory for alloys containing more than 1 per cent. of aluminium and indicated that a magnesium chloride base solution was only suitable for amounts of aluminium between 4 and 40 mg per 100 ml. Semerano and Ronchi5 criticised Gull’s method and proposed a modification in which the hydrochloric acid solution of the.alloy was neutralised with ammonia to pH 3.25, dimethyl yellow and methylene blue being used as a mixed indicator. ‘Heller and Zanko6 described a very similar procedure in which the solution was neutralised with lime water to pH 3.3.Koslova and Portnow’ examined the effect of pH and electrolyte concentration on the aluminium wave in calcium chloride base solution over the pH range 3.7 to 4.18, and recommended an optimum pH of 4.0 in 0.1 N calcium chloride solution. EXPERIMEL TAL Preliminary trials indicated that the only suitable base solutions for the reduction of aluminium were those containing either magiiesiurn or barium chloride. Magnesium chloride solution proved to be a suitable base solution for zinc and stannic tin, but well-defined waves for aluminium were only obtained at higher concentrations than the amounts under considera- tion.In the concentration range required, the aluminium wave obtained was not suitable for analytical use, as the top of the wave sloped steeply, particularly in the presence of tin, as shown in Fig. 1. In barium chloride solution, aluminium gave a wave which was more suitable for ana- lytical purposes. The half-wave potential was -- 1-63 volts %gainst the saturated calomel electrode. The addition of zinc ions to the solution resulted in a well-defined wave for zinc, which had a half-wave potential of - 1.1 volts against the same anode, and the two waves were sufficiently well separated to permit simultaneous determination of the two metals.The addition of tin interfered with the aluminium wa.ve and no conditions could be found that would allow all three metals to be determined on one polarogram.Dec., 19511 OF ALUMINIUM, ZINC AND TIN I N WATER 707 The removal of tin is easily accomplished by volatilisation after conversion to stannic bromide, and in the proposed method aluminium and zinc are estimated together after the removal of the tin. The tin is determined separately on a further aliquot of the sample. The pH of the base solution was found to have a considerable effect on the shape of the aluminium wave, and the following experiments were made to evaluate this factor. J i 3 - Voltage Fig. 1. Polarogram of zinc, tin and aluminium in 0.1 N magnesium chloride. Concentrations of zinc and aluminium 0.5 mg per 100 ml; tin 1.0 mg per 100 ml f 2 - Voltage Fig.2. Polarogram of zinc and aluminium in 0-05 N barium chloride a t pH 4.6. Curve A, concentrations of zinc and aluminium 2.0 mg per 100 ml; curve B, concentrations of zinc and aluminium 1.0 mg per 100 ml To a 5-ml aliquot of a standard aluminium solution 1 ml of N hydrochloric acid was added and the whole diluted to 10 ml with water. From a burette, 0.1 N barium hydroxide solution was added dropwise, with stirring, until the desired pH was attained. The solution was finally diluted to 20 ml and a final pH adjustment made with 0.02 N hydrochloric acid. The pH was determined electrometrically with the glass electrode. It was found necessary to leave the solutions to stand for at least one hour in order to attain pH equilibrium.After the removal of oxygen, a polarogram of the solution was taken over the range -0.8 to -2.0 volts. Two series of experiments were made with different amounts of aluminium and the results are shown in Table I. TABLE I EFFECT OF pH ON THE ALUMINIUM WAVE Wave height, Slope of top PH galvanometer units of wave Series 1- - - 4.0 . . . . .. . . . . 4.3 . . .. . . .. .. 7.2 0.5 4.7 . . .. .. .. .. 7.2 0.3 4.85 . . ,. . . .. .. 6.8 0.2 5.0 . . - . .. .. . . 4.6 0.1 Series 2- - - 4.05 . . .. .. . . .. 4.3 . . .. .. . . .. 4.9 0.5 4.7 . . . . .. .. .. 4.9 0.4 4*5* . . .. .. .. .. 4.9 0.5 4.8 . . .. .. .. * . 4.7 0-3 5.0 . . . . .. .. * . 3.8 0.1 * pH 4-6 determined by use of B.D.H. “4.6” Indicator; other pH values determined electrometrically.708 HODGSON AND GLOVER THE POLAROGRAPHIC DETERMINATION [Vol.76 At pH 4.0 no distinct wave was recorded. The slope of the top of the wave decreases with increasing pH; the values quoted in Table I represent the tangent of the angle that the linear portion of the wave makes with the horizontal. Above pH 4.7 the waves approach the ideal shape but the height diminishes. This is presumably due to precipitation of aluminium hydroxide which, according to Britton,s would occur at p€I 4.76 under these conditions. Ideal waves were obtained only when some aluminium hydroxide had been precipitated; it is thus neces- sary to work at a pH as near the precipitation point as is practicable. This is in agreement with the findings of Kozlova and Portnow,’ who used an optimum pH of 4.0 in calcium chloride solution. Britton* found that precipitation of aluminium hydroxide occurs at pH 4-14 in this medium.Attempts to buffer at pH 4.5 with phosphate, phthalate, citrate, tartrate and acetate solutions either precipitated the barium ions or resulted in the disappearance of the aluminium wave. Satis- factory results were only obtained by adjustment of an acid solution to pH 4-5 with barium hydroxide solution using B.D.H. “4.5” Indicator.* The effect of variations in the base solution concentration was next considered; the pH was adjusted to 4.5 in the manner described. Various amounts of hydrochloric acid were used to give different concentrations of barium chloride in the final solution. The results are shown in Table 11.The wave height remains constant at pH values between 4.3 and 4.7. These considerations indicated that the optimum condition was at pH 4.5. TABLE I1 EFFECT OF CONCENTRATION OF BASE SOLUTION ON THE ALUMINIUM WAVE Strength of barium chloride base solution Wave height, microamps. 0.02 M 5.6 0.05 M 6-5 0.1 M 5.2 0.2 M Approx. 5-0 In 0.2 M barium chloride the wave was so ill-defined that an accurate measurement of height could not be made. The optimum conditions were therefore fixed at pH 4.5 in 0.02 to 0.05 M barium chloride base solution. Calibrations were made under these conditions with standard solutions of zinc and aluminium, for both of which a linear relation between wave height and concentration was obtained. Details of the aluminium calibration are shown in Table 111, and typical polarograms of both metals are shown in Fig.2. TABLE I11 CALIBRATION OF THE POLAROGRAPH FOR ALUMINIUM Capillary constants- m = 2-29 mg per second. m2/3 = 2.09 mg2P set. -112 Aluminium, mg per 100 ml of final solution 0-5 1.0 1.5 2.0 2.5 t = 2.98 seconds. Temperature = 20” C. Wave height, microamps. 1-3 2.4 3.7 4.75 6.0 Attempts to elucidate the electrode reaction by a plot of log i/(& - i) against cathode potential indicated the reduction of aluminium to be irreversible, a linear relation was obtained for the top portion only of the wave; the slope of the line was 0.057 volt. The half-wave potential deduced from this relation was found to be - 1.63 volts against the saturated calomel electrode at 20” C. * Made by The British Drug Houses Ltd., Poole, Dorset.Dec., 19511 OF ALUMINIUM, ZINC AND TIN I N WATER 709 A trace of iron was known to be present in the solutions under consideration, so additions of ferric iron were made to standard aluminium solutions in order to determine if there was any interference; it was found that 5 mg of ferric iron per 100 ml could be present without affecting the aluminium wave over the range 0 to 5 mg per 100 ml.Lead, in concentrations up to 5 mg per 100 ml, was also found to be without effect. A blank determination on the reagents gave a well-defined aluminium wave when recorded on the highest galvanometer sensitivity. The height of the wave corresponded to 0.02 mg of aluminium per 100 ml. This represents the lowest concentration determinable by the method.The method used for the estimation of tin was that due to Godar and Alex- ander,g in which a polarogram is made of the tin in ammonium chloride base solution after a preliminary separation by precipitation on aluminium hydroxide. The tin is estimated in the stannic condition, thereby avoiding the difficulties encountered in many methods wherein it is necessary to maintain tin in the divalent state. The combined lead and tin waves are measured, ammonium citrate is added to suppress the wave due to tin, and the diffusion current due to lead is thus estimated. METHOD FOR THE DETERMINATION OF ZINC AND ALUMINIUM Under these conditions, lead is reduced at the same potential as tin. PROCEDURE- Transfer, by means of a pipette, 50 ml of sample solution, containing up to 50 p.p.m.of aluminium or zinc, into a 100-ml beaker, add three drops of bromine and 1 ml of concentrated hydrochloric acid, and evaporate to dryness. To the residue add 1 ml of concentrated hydro- chloric acid, 5 ml of water and one drop of bromine, and evaporate to dryness. Dissolve the residue with 1 ml of concentrated hydrochloric acid and 5 ml of water, and evaporate just to dryness. Add 10ml of water, one drop of B.D.H. “46” Indicator and adjust to the grey shade with 0.1 N barium hydroxide solution. The amount of barium hydroxide used must be at least 4m1, so that the final concentration is between 0.02 and 0.05M. If less than 4 ml is needed, make alternate dropwise additions of 0.1 N hydrochloric acid and 0.1 N barium hydroxide until the desired concentration is reached.Adjust the pH finally with 0.02 N hydrochloric acid, dilute to 20 ml and leave to stand for one hour, making pH adjustments when necessary by dropwise addition of 0.02 N hydrochloric acid. Remove oxygen from the solution by passing nitrogen for 15 minutes and record a polarogram over the range METHOD FOR THE DETERMINATION OF TIN -0.8 to -2.0 volts. REAGENTS- concentrated hydrochloric acid. AZuminium solution-Dissolve 0.5 g of pure aluminium in the minimum quantity of Hydrochloric acid-A 50 per cent. v/v solution. Ammonizm chloride-A 30 per cent. w/v aqueous solution. PROCEDURE- Transfer, by means of a pipette, 25 ml of sample solution, containing up to 50 p.p.m. of tin, into a 50-ml beaker. Add 5 ml of aluminium solution and one drop of methyl red indi- cator.Make just alkaline by addition of ammonium hydroxide, sp.gr. 0.880, add one drop in excess and filter. Dissolve the precipitate on the paper in 5 ml of hydrochloric acid reagent, by re-passing the filtrate through the filter-paper several times. Wash the filter-paper with 5ml of ammonium chloride reagent, adding the washings to the 5ml of hydrochloric acid solution. Finally add one drop of cresol red indicator and dilute to exactly 10ml with ammonium chloride reagent. Remove oxygen from the solution by passing nitrogen for 15 minutes, and record a polarogram over the range 0 to -0.9 volt. Measure the height of the second wave. STANDARD METAL SOLUTIONS- the minimum volume of hydrochloric acid. to a 500-ml volumetric flask and dilute to volume. Dilute to 500 ml with distilled water. PREPARATION OF CALIBRATION CURVES Weigh 0.5 g of the metal (zinc, tin or aluminium) into a 250-ml beaker and dissolve it in When solution is complete, boil, cool and transfer 1 ml of the solution = 1 mg of the metal.7 10 SHORT : THE DETERMINATION OF TUNGSTEN [Vol.76 CALIBRATION FOR ZINC AND ALUMINIUM- Prepare solutions of zinc and aluminium by diluting 10 ml of the standard metal solution to 100 ml. Take 1-ml, 2-ml, and 3-ml aliquots of the diluted solutions in three separate 50-ml beakers. Add 1 ml of standard tin solution to each and proceed as in the determination of zinc and aluminium. Plot the wave height against the zinc and aluminium content. CALIBRATION FOR TIN- Dilute 10 ml of standard tin solution to 100 11-11. Take 1-ml, 2-ml, and 3-ml aliquots of the solution and proceed as for the determination of tin. Plot the wave height against tin content. The authors thank the Directors of The British Oxygen Company Limited for permission to publish. REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 9. RESEARCH AND DEVELOPMENT DEPARTMENT THE BRITISH OXYGEN Co. LTD. Stackelberg, M. V., “Polarographische Arbeitsmethoden,” W. de Gruyter & Co., Berlin, 1960, Prajzler, J., Coll. Czech. Chem. Comm., 1931, 3, 407. Gull, H. C., J . SOC. Chem. Ind., 1937, 56, 177. Stross, W., Analyst, 1949, 74, 285. Semerano, G., and Ronchi, I., Atti R. Ist. Venei’o. Sci., 1941, 100, 11, 397; Chem. Abs., 1944, 38, Heller, B. A., and Zanko, A. M., Zavod. Lab., 1!939, 8, 1030; Chem. Abs., 1940, 34, 1583. Kozlova, A. A., and Portnow, M. A., Ibid., 1940, 9, 287; Chem. Abs., 1940, 34, 6783. Britton, H. T. S., “Hydrogen Ions,” Chapman ,& Hall Ltd., 1942, Vol. 11, p. 49. Godar, E. M., and Alexander, 0. R., Ind. Eng. Chem., Anal. Ed., 1946, 18, 681. pp. 116-118. 3923. ANALYTICAL LABORATORY LONDON, S.W.19 March, 1961

ISSN:0003-2654    

DOI:10.1039/AN9517600706

出版商:RSC    

年代:1951

数据来源: RSC

摘要:

7 10 SHORT : THE DETERMINATION OF TUNGSTEN [Vol. 76 The Determination of Tungsten and Molybdenum in Titanium‘!‘ BY H. G. SHORT A colorimetric method has been developed for the determination of small amounts (0.002 to 0.10 per cent.) of molybdenum and tungsten in titanium metal. Coloured complexes are formed with toluene dithiol, and these are extracted from aqueous solution with am.yl acetate for measurement of the colour intensity. Molybdenum reacts at low acidity and is extracted first; tungsten is then extracted from strongly acid solution. It has been shown that the thiocyanate method, based on that used for tungsten in steel, gives high results when used for titanium. A STANDARD absorptiometric method for the determination of tungsten in steel, making use of the reaction between reduced tungsten and thiocyanate ions, has recently been developed, and it appeared probably that a similar procedure could be applied to the determination of tungsten in titanium.When trials were made, it was found necessary to reduce the titanium before addition of the thiocyanate in order to obtain reproducible results. As a result, the small absorption due to the yellow tungsten complex was measured in the presence of a deep violet colour due to titanous chloride (Tic&). No suitable wavelength could be found for the measurement of the tungsten colour a t which titanous chloride did not also have a considerable absorption, and in fact a small test reading was always measured in the presence of a considerable back- ground value. It was at first assumed that the lowest background value obtained ona number of samples corresponded to tungsten-free titanium, and the calibration graph was based on this “blank” value.* Communication from the National Physical Laboratory.Dec., 19511 AND MOLYBDENUM IN TITANIUM 711 Various titanium samples, prepared at the National Physical Laboratory by arc melting, when tested by this thiocyanate method, showed tungsten contents ranging from 0.02 to 0.10 per cent., but suspicion was aroused when a sample from the U.S.A., which was stated to have been prepared by melting in graphite, was found to give a reading indicating 0.04 per cent. of tungsten. Another reagent that has been applied to the determination of small amounts of tungsten in steel is dithio1,l and it was therefore decided to examine the possibility of using it in an alternative method for tungsten in titanium.In the procedure for steel, stannous chloride was used as a reducing agent in the produc- tion of the greenish-blue coloured complex of reduced tungsten with dithiol. An examination of the standard reduction potentials for the following reactions shows, however, that stannous tin is not really a strong enough reducing agent for the purpose and that trivalent titanium can be expected to be better- W205 -i- H,O = 2W0, + 2H' + 2e' E, = -0.15 . . . . . . * * (1) 2W0, -+ H20 = W20, + 2H' + 2e' E, = 0.0 . . . . . . * * (2) + 2e' E, = -0-15 . . . . . . * - (3) Ti"' + H,O = TiO" + 2H' + e' E, = -0.1 (approx.) . . - * (4) - I Sn"" Sn" From equations (1) and (2) it is apparent that reduction of tungsten is favoured by high acidity.Reduced molybdenum gives with dithiol a green coloured complex ; this reduction (of molybdenumV1 to molybdenum") is also favoured by a high acidity, but the reduction occurs much more readily than that of tungsten- 2H20 + MoO,' = H,MoO, + 2H' + e' E , = -0.4 (approx.) Advantage is taken of this in the steel methodl to extract the molybdenum-dithioi complex at low acidity (with hydroxylamine as a reducing agent) and then to add a stronger reducing agent, stannous chloride, and extract the tungsten at a higher acidity. Molybdenum has not been found in appreciable amounts in any of the titanium samples so far examined, but a dithiol method for tungsten would have to take account of its presence, and the possibility of separations of both molybdenum and tungsten from titanium on the above lines was therefore investigated.EXPERIMENTAL SIMULTANEOUS DETERMINATION OF TUNGSTEN AND MOLYBDESUM WITH A SPECTRO- The possibility of extracting tungsten and molybdenum together as complexes with dithiol and determining both by absorptiometric measurements at two suitable wavelengths was considered. The absorption curves of the two complexes were found to be fairly close, the points bf greatest separation being as follows- PHOTOMETER- Optical density (100 pg of metal per 25 ml) 6000 A 6900 A Tungsten . . .. . . .. 0.342 0.280 Molybdenum . . . . .. .. 0.240 0.87 The inaccuracy introduced by the measurement of the optical densities at these two wavelengths and calculation by proportion is about & 3 per cent.at the 100-pg level, +_ 6 per cent. at 50 pg, and greater for smaller quantities. This might be adequate for trace analysis, but a more serious difficulty was the fact that no conditions could be found under which it was possible to extract both tungsten and molyb- denum quantitatively at the same time. A strong reducing agent, which is necessary for the reduction of tungsten, reduces some molybdenumV to molybdenum11r, and the latter does not give a coloured dithiol complex. FORMATION OF TUNGSTEN - DITHIOL COMPLEX- The structure of the complex and the valency of tungsten in the molecule are unknown. A reducing agent is necessary for the formation of the greenish-blue complex, and in previous work stannous chloride has been used, but as considered above, it was probable that a stronger reducing agent was required.This was shown to be so by measuring the colour produced by 100 pg of tungsten and dithiol under similar conditions of heating, acidity, and so on, (a) with712 SHORT: THE DETERMINATION OF TUNGSTEN [Vol. 76 5 ml of 20 per cent. stannous chloride solution as a reducing agent and (b) with titanous chloride containing 0-2 g of titanium (see Table I). From the erratic and low results for (a) compared with those for ( b ) , it appeared that the formation of the tungsten complex was incomplete. TABLE I REDUCTION OF 1OOpg OF TUNGSTEN Colour produced, Reduced with Spekker units (a) SnCl, . . . . .. . . 0.290, 0.316, 0.278 ( b ) TiC1, . . . . . . . . 0.423, 0.430, 0425 In the present work titanous chloride could be produced in situ by solution of the titanium metal sample in acid, and a calibration curve was constructed by dissolving 0.2-g portions in concentrated hydrochloric acid and adding known amounts of tungsten.The solution was then diluted to 80 per cent. v/v of hydrochloric: acid and warmed. Ten millilitres of 0.5 per cent. dithiol in amyl acetate were added and the hot solution shaken for 15 to 20 minutes. The amyl acetate layer was washed as described below under Procedure, and its absorption measured on the absorptiomFter with Ilford 607 filters. With up to 1OOpg of tungsten a linear relationship was obtained (see Table 11). TABLE: I1 Spekker absorptiometer with Ilford 607 filters and l-cni cell ABSORPTION OF TUNGSTEN - DITHIOL COMPLEX Tungsten, pg per 25 ml .. .. .. 30 60 100 Drum difference reading . . .. .. 0.133 0.255 0.423 In the absence of molybdenum the determination of tungsten in titanium is therefore The sample is dissolved in hydrochloric acid and the dithiol complex formed simple. immediately and extracted. FORMATION OF MOLYBDENUM - DITHIOL COMPLEX- As with tungsten, the structure of the complex and the valency state of the molybdenum are unknown. For the formation of the molybdenum - dithiol complex it was decided to work in a more dilute acid solution and to use conditions less favourable to the simultaneous reaction of the reagent with tungsten. (1) The colour produced by a given amount of molybdenum was very much less in the presence of titanium'= than with other, less powerful, reducing agents.The prob- able explanation is that some molybdenum had been reduced to molybdenumn1, which did not form a green dithiol complex. (2) Working conditions of dilute acid (50 per cent. v/v hydrochloric acid or less) and room temperature were imposed by the fact that otherwise, i.e., in strong acid or hot solutions, some tungsten would react and accompany the molybdenum. In dilute acid a fairly strong reducing agent was necessary to give maximum colour production from a given amount of molybdenum. Using stannous chloride in 50 per cent. v/v hydrochloric acid at room temperature, no reaction with tungsten occured on shaking for twenty minutes. As these conditions offered some hope of a satisfactory separation, a calibration curve for molybdenum was prepared by shaking known amounts in 50 per cent.v/v hydrochloric acid with 10 ml of 0.5 per cent. dithiol in amyl acetate for 15 minutes at room temperature. The amyl acetate layer was then washed as described below under Procedure, and the absorption of the green solution measured. The molybdenum complex has an absorption maximum at 6800 to 6900 A, and the cali- bration points recorded in Table I11 were measured on a Spekker absorptiometer with Ilford 608 filters. Owing to the broad transmission of these filters at the red end of the spectrum the readings depart from linearity. Measured on a spectrophotometer a t 6900 A the readings fall much more closely on a straight line, but for ordinary working purposes a Spekker absorp- tiometer can be used satisfactorily.The following facts were established by preliminary tests-Dec., 19511 AND MOLYBDENUM I N TITANIUM TABLE I11 713 ABSORPTION OF MOLYBDENUM - DITHIOL COMPLEX Molybdenum, Spekker difference : Spectrophotometer : pg per 25 ml 608 filters and 1-cm cells 27 0.235 0.26 54 0.438 0.485 90 0.675 0.785 density a t 6900 A SEPARATION OF TUNGSTEN AND MOLYBDENUM WITH DITHIOL- It was shown (see (1) in previous section) that molybdenum does not form the green dithiol compound quantitatively in the presence of titanous chloride. As molybdenum reacts with dithiol more readily than does tungsten, it must be extracted first for quantitative separation of the two elements, and therefore the titanium, after solution in acid, must be oxidised. This oxidation was done by drop-wise addition of potassium permanganate to the warm acid solution until the purple colour of titanium'" was just discharged.From the resulting solution, diluted to 50 per cent. v/v of acid, the molybdenum was quantitatively extracted by shaking with dithiol solution in the presence of stannous chloride. A second extraction from the same solution with fresh reagent gave no further molybdenum colour. If the acidity of the resulting aqueous layer were increased to 80 per cent. v/v and the solution warmed, some titanium'II was produced and the liquid became violet in colour. It can be shown from standard oxidation potentials that, for equivalent amounts of titaniumxv and tin" as reactants, approximately one tenth of the titanium is reduced to titaniumm at equilibrium in N acid.In the present example the initial concentration of tinU is much greater than that of titaniumm, and the acidity is high, both factors being favourable to the production of titanium". On applying the usual conditions for the production of the tungsten - dithiol complex, the same amount of colour was produced from 100 pg of tungsten a s in the presence of 0.2 g of titanium, dissolved in acid and not oxidised. In other words, after extraction of the molybdenum in 50 per cent. acid, the tungsten could be extracted from stronger acid solution on warming. A second extraction for tungsten from the same solution with fresh reagent gave no further tungsten colour. This scheme of separation was then tried out on known amounts of molybdenum and tungsten added to 0.2-g portions of titanium, with the results shown in Table IV.TABLE I V SEPARATION OF MOLYBDENUM AND TUNGSTEN FROM TITANIUM (100 pg = 0.5 per cent. on an 0.2-g sample) Molybdenum Experiment Taken, PQ 50 75 100 25 90 45 67.5 Found, Pg 49 75 96 19 90 40 62 Tungsten -7 Takcn , Found, Pg Pg nil nil ' 25 26 100 100 75 76 50 47 50 45 25 24 INTERFERENCE OF OTHER ELEMENTS IN DITHIOL TEST- The following elements also form coloured compounds with dithiol: silver, mercury, arsenic, copper, cadmium, tin, lead and bismuth. Of these, only mercury and arsenic react in 50 per cent. hydrochloric acid solution. The mercury compound, which is pale yellow in colour, is insoluble in the amyl acetate used to extract the tungsten and molybdenum com- plexes. In any event, the occurrence of appreciable amounts of mercury in titanium metal is most unlikely.Large amounts of arsenic give a pale yellow colour in the amyl acetate extract, but as 20 mg of arsenic give a colour equivalent to 15 pg of tungsten, the interference of this element can be neglected.7 14 SHORT [Vol. 76 RESULTS ON SAMPLES- The graphite-melted titanium, mentioned in the introduction (p. 711), which had shown 0.04 per cent. of tungsten by the thiocyanate method, when re-tested by the dithiol method showed less than 0.01 per cent. The tungsten content of two other samples, which had shown 0.03 and 0.10 per cent., likewise showed reduced values of 0.017 and 0.022 per cent. respectively. METHOD FOR THE DETERMINATION OF MOLYBDENUM AND TUNGSTEN IN TITANIUM METAL REAGENTS- Stannous chloride, 20 per cent.w/v-Dissolve 20 g of SnC1,.2H20 in 50 ml of concentrated hydrochloric acid and warm until the solution is clear. Toluene dithiol, 0.5 per cent. w/v-Dissolve 1 g of solid reagent in 200 ml of amyl acetate. The solution remains fit for use for about one week. PROCEDURE- Add saturated potassium permanganate solution drop by drop until the purple colour of titanium”I is just discharged. Cool and transfer the solution to a measuring cylinder, washing out the beaker with small amounts of concentrated hydrochloric acid. Separation and determination of mol,~?,bdenum-Pour the solution into a 100-ml conical flask and add an equal volume of water, then 5 m,l of 20 per cent. stannous chloride and 10 ml of dithiol reagent. Transfer to a separating funnel and wash out the flask with a small amount of amyl acetate.Run the aqueous layer back into the same conical flask. Wash the solvent with a few millilitres of 50 per cent. v/v hydrochloric acid, add the washings to the liquid in the flask and reserve them for the determination of tungsten. Make the volume of the amyl acetate layer in the funnel up to about 20 ml and wash twice with 15 ml of 80 per cent. v/v hydrochloric acid. Run the solvent layer through a dry Whatman No. 41 filter-paper into a stoppered graduated cylinder. Wash out the funnel once or twice with 1 to 2 ml of amyl acetate, running this through the same filter. Dilute to 25 ml with amyl acetate, mix well and measure the absorption on the Spekker absorptiometer with Ilford 608 filters and l-cm cells, or on the spectrophotometer at 6900 A.Read off the amount of molybdenum from a calibration graph prepared under similar conditions. Determination of tuungsten-To the aqueous layer from the separation of molybdenum add 20 ml of concentrated hydrochloric acid. Warm the solution until the violet colour of titaniumnT appears. Add 10 ml of dithiol reagent and stand on a hot-plate for 20 minutes with very frequent shaking. Cool and transfer to a separating funnel. Discard the aqueous layer. Make the volume of the solvent layer up to about 20 ml with amyl acetate and wash twice with 15 ml of 80 per cent. v/v hydrochloric: acid. Run the solvent layer through a dry Whatman No. 41 filter-paper into a stoppered graduated cylinder. Wash out the funnel once or twice with 1 to 2 11-11 of amyl acetate, running this through the same filter. Dilute to 25 ml with amyl acetate, ;mix well and measure the absorption on the Spekker absorptiometer with Ilford 607 filters and l-cm cells, or on the spectrophotometer at 6000 A. Read off the amount of tungsten from a calibration graph prepared under similar conditions. Dilute to 100 ml with water. Dissolve 0.2 g of titanium metal in 10mI of concentrated hydrochloric acid. Shake well for 10 to 15 minutes. Reject the acid washings. Reject the acid washings. The work described above has been carried out as part of the research programme of the National Physical Laboratory, and this paper is published by permission of the Director of the Laboratory. REFERENCE 3 . Bagshawe, B., Analyst, 1947, 72, 189. May, 1951

ISSN:0003-2654    

DOI:10.1039/AN9517600710

出版商:RSC    

年代:1951

数据来源: RSC

摘要:

Dec., 19511 MAYER AND BRADSHAW 715 Photometric Determination of Small Amounts of Iron in Magnesium and Magnesium Alloys by Thioglycollic Acid BY A. MAYER AND G. BRADSHAW A method is described for the determination of 0.0005 to 0.015 per cent. of iron in commercial magnesium-base alloys. It is based on the formation in ammoniacal solution of the pink ferrous iron - thioglycollic acid complex, the absorption of which is measured on a Spekker absorptiometer. The colour is stable for 18 hours, is unaffected by small changes in temperature or by excess of the reagent, and develops immediately in the absence of manganese; in the presence of large amounts of manganese about 20 minutes are required for the stable colour to form. Results by this method are easily reproducible. Investigations of the effect of many elements have shown that the method could be applied to aluminium-base or zinc-base alloys, and it has been used in our laboratory for the analysis of zirconium-base materials.AN investigation has been made to find a method of determining 0.0005 to 0.015 per cent. of iron in magnesium-base alloys. Although several colorimetric reagents have been suggested for the determination of traces of iron, only a few were sufficiently sensitive and selective for the present purposes. These are thiocyanate,l y 2 1 : lO-phenanthroline,l32J 2 : 2’-dipyridyl,4,6 thioglycollic acid,s p7 s8 y 9 7-iodo-8-hydroxyquinoline-5-sulphonic acidlo (“Ferron”) and 1 : 2- dihydroxybenzene-3 : 5-disulphonic acid.ll ,12 The last reagent was tried but, although the green colour developed in acid solution is sufficientlyintense for photometric work, it is too sensitive to small changes of pH and is therefore unsuited to routine work.The red complex formed in alkaline solution is less sensitive to pH changes, but unfortunately copper interferes by forming a yellow complex and would have to be removed; for several reasons this was considered undesirable. 2 : 2’-Dipyridyl and 1 : 10-phenanthroline are very expensive reagents ; copper causes interference with 2 : 2’-dipyridyl, and zinc with both. “Ferron” was tried, but was found to be insufficiently sensitive and, in addition, gave deep yellow coloured reaction products with some of the major constituents of magnesium alloys. Thiocyanate had been in use in our laboratory for some time as a reagent for the determination of iron by a direct procedure, iron thiocyanate not being extracted, but it was found to be unreliable for routine determinations. This was partly because many factors have to be controlled accurately and partly because the laboratory has to deal with several types of alloys with varying major constituents, for each of which separate calibrations would be required, which makes the procedure rather cumbersome.The method involving the use of 1 : 10-phenanthroline had also been in use in the laboratory for some time, but could not be applied to alloys containing more than about 1 per cent. of zinc. The procedure is, however, extremely reliable in the absence of zinc and it was thought that a simple separation might make it generally applicable.During work on this procedure it became apparent that iron was introduced, not only from reagents (which can be allowed for by a blank determination), but also in random quantities from atmospheric dust particles introduced during filtration or ignition and from filter-papers with variable soluble-iron content. This last source was particularly unexpected because the filter-papers were washed with dilute hot hydrochloric acid before use, but were found to liberate more iron on treatment with hydrochloric acid after having been in contact with ammoniacal ammonium chloride solution. From these observations it was clear that a method was required that involved as few preliminary steps as possible, preferably one in which filtration was unnecessary and which used as few reagents as possible.Statements in the literature indicated that thioglycollic acid (mercapto-acetic acid) might be suitable; it reacts with ferrous iron in alkaline solution to form a pink complex. Ferric iron is reduced to the ferrous state by the reagent, so that the total iron content can be determined. It appeared that zinc reduced the intensity of the iron - thioglycollic acid complex and that manganese gave a transient colour. We confirmed Swank and Mellon’s statements that zinc did not interfere when sufficient of the reagent was716 MAYER AND BRADSHAW PHOTOMETRIC DETERMINATION [Vol. 76 present to combine with it. Manganese gives a transient yellow-green colour with thiogly- collic acid but does not interfere, even when present in large amounts, if the reagent is present before the solution is made alkaline.EXPERIMENTAL REAGENTS- Standard iron solution-This was made by dissolving pure iron in hydrochloric acid to give a solution containing 0.01 mg per ml. Magnesium base soZution-This was made by dissolving 100 g of pure magnesium in hydrochloric acid, adding 0.05 g of aluminium, oxidising iron by the addition of a few drops of hydrogen peroxide, and precipitating iron and aluminium by adding ammonia and filtering. The solution was diluted to 2 litres and contained 1 g of magnesium in 20 ml. When tested with o-phenanthroline, less than 0-0003 g of iron was detected in 100 g of magnesium. ThiogZycoZZic acid reagent-This was made by dissolving 100 ml of thioglycollic acid in 500 ml of diluted ammonium hydroxide (1 + 1) and diluting to 1 litre.APPARATUS- Measurement of absorption was made with a Spekker photo-electric absorptiometer with a tungsten filament lamp. The instrument was fitted with a long right arm so that a 20-cm cell could be used. PR ELIMINARY EXPERIMENTS- Choice of coZour $Zters-The iron - thioglycollic acid complex gave the greatest absorption values at 530 mp with Ilford No. 604 green filters and these were used throughout in conjunction with Hilger heat filters H 503. Table I records absorptions measured with a series of filters. CHOICE OF COLOUR FILTER TABLE I Hilger H503 filter used with Ilford filter No. 60 1 602 603 604 605 606 607 608 Transmission range, mP 390 to 470 450 to 490 480 to 520 500 to 540 540 to 570 560 to 610 580 to 680 620 to 800 Peak transmission, mC1 430 470 495 520 550 580 600 700 Absorption by iron - thioglycollic acid complex.Drum difference reading 0.50 0.53 0.62 0.69 0.67 0.54 0-34 0.11 The transmission curve was plotted with a Hjlger “Uvispec” spectrophotometer and is shown in Fig. 1. Maximum absorption is at 530 rnp, which agrees with that shown by the curves recorded by Swank and Mellon6 and SandelL2 Range-In order to cover adequately the required range of 0.0005 to 0.015 per cent. of iron, it was found necessary to use a 20-cm cell; to fill this cell, a final volume of 500 ml of solution was required. From preliminary tests it was found that a 4-g sample of the alloy dissolved and diluted to 500 ml would cover the range 0 to 0.008 per cent.of iron when the absorptiometer was adjusted to a water - water setting of 1.00. One drum division was then equivalent to approximately 0.0001 per cent. of iron. This range was considered adequate, and for greater amounts of iron it was proposed to use a proportionately smaller weight of sample. Amount of thiogZycoZlic acid-By varying the amount of thioglycollic acid for a given zinc content and conversely by varying the zinc content for a given amount of thioglycollic acid, it was found that just over two molecules of the acid were required per atom of zinc before the iron colour developed fully. From this the amount of 10 per cent. v/v thioglycollic acid reagent was fixed at 10 ml; this is sufficient for 4-15 samples containing up to 6 per cent. of zinc. E’ert of pH on colour development-To obtain full colour development a pH greater than 7-5 is required, but addition of a large excess (20 nil) of ammonium hydroxide, sp.gr.0.880, causes no change.Dec., 19511 OF SMALL AMOUNTS OF IRON I N MAGNESIUM 717 E$ect of temperature-The temperature of the final solution before addition of the thio- It was found that increasing the temperature from 15" glycollic acid reagent was varied. to 30" C reduced the intensity of the colour by 2 per cent. DEVELOPMENT OF THE COLOUR WITH TIME- When the thioglycollic acid is added after the solution has been made ammoniacal, the colour develops slowly. However, when the thioglycollic acid is present before the pH is raised, the colour develops immediately, even in the presence of large amounts of other ions and, in diffused daylight, remains stable for at least 18 hours.Mehlig and Shepherd14 report Wavelength. mp Fig. 1. Transmission curve of iron thiogIycollate solution containing 0.25 mg of iron per 100 ml, measured in 1-cm cell. fading after 2 hours, while other workers report fading after 30 minutes and state that the colour can be restored by shaking (air-oxidation). Our solutions were kept in 20-cm absorp- tion cells, which expose a large surface area, and this may account for the stability of the colour. An exception to this is when more than 20 rng (0.5 per cent.) of manganese are present. When 75 mg of manganese are present, stability of the colour is only attained after about 5 minutes, while in the presence of 120 mg of manganese about 15 minutes are required.Most commercial magnesium alloys contain less than 0.5 per cent. of manganese, an exception being alloy AM 503, which contains 1.5 to 2.5 per cent. EFFECT OF ALLOYING ELEMENTS- NormaZ alloying elements-None of the elements normally found in magnesium alloys interferes. Tests were carried out on 4 g of magnesium to which additions of the chlorides of the elements had been made, 10 ml of thioglycollic acid being used in each experiment. The colour was measured after 5 to 10 minutes. The maximum alloying element contents that were tested corresponded to 10 per cent. of aluminium, 6.2 per cent. of zinc, 4 per cent. of rare earths, 1 per cent. of zirconium, 2.5 per cent. of manganese and 0.3 per cent. of copper. Typical results are shown in Table 11.TABLE I1 EFFECT OF NORMAL ALLOYING ELEMENTS Alloying elements present A 7 Zinc, Mischmetall, Zirconium, Aluminium Manganese, Copper, mg mg mg mg mg mg 250 200 50 - 200 20 20 200 - 130 - - 200 40 - - I L - 100 - - - - 400 100 - - - - Iron added mg 0.140 0.250 0.080 0.120 0.240 Iron found, mg 0-140 0-248 0-079 0.1 19 0.240 Special alloying elements-The results recorded in Table I11 show that large amounts of In the presence of thallium Chromium lithium, calcium, strontium, indium and thallium do not interfere. a precipitate of thallous salts forms but redissolves on addition of more ammonia.718 MAYER AND BRADSHAW: PHOTOMETRIC DETERMINATION [Vol. 76 does not interfere if allowance is made for the green colour of chromic ions by measuring the blank. Cadmium uses up the reagent; up to 0.2 g can be tolerated, but larger amounts cause serious errors unless more thioglycollic acid is added.Stannic tin also uses up the reagent and its effect can be overcome by adding more thioglycollic acid. Silver, even in small amounts, interferes by formation of a white precipitate with thioglycollic acid and must be entirely absent. Large amounts of bismuth cause the formation of turbidity; they should be removed. Large amounts of lead are a cause of erratic results; further, if fluoride is added, as stipulated in the final procedure, to dissolve zirconium, lead fluoride is precipitated. It is therefore desirable to remove lead. If large amounts of silicon are present the solution of colloidal silicic acid tends to adsorb the iron complex and leads to low results; this can be overcome by the addition of fluoride.Removal of silver, lead and bismzdh-Major quantities of silver can be removed by addition of chloride but it is necessary to remove the last traces of silver. This can be carried out conveniently by hydrogen sulphide. Lead and bismuth can be removed by a similar procedure (see note 11, p. 722). When silver and tin are present together in major amounts, as,in the alloy specified by D.T.D. 350 A (up to 6 per cent. of tin and 1 per cent. of silver) treatment with hydrogen sulphide leads to losses of iron by adsorption on the tin sulphide. This can be overcome by inhibiting the precipitation of tin sulphide with oxalic acid (see note 11, p. 722). TABLE I11 EFFECT OF SPECIAL ALLOYING ELEMENTS 4 g of magnesium and 0-300 mg of iron taken in all experiments Alloying element added 200 mg of calcium 200 mg of strontium 200 mg of lithium 200 mg of indium 200 mg of thallium 200 mg of lead 200 mg of lead 200 mg of bismuth 200 mg of silver 193 mg of cadmium 579 mg of cadmium 820 mg of cadmium 926 mg of cadmium 820 mg of cadmium 12 mg of silicon 47 mg of silicon 12 mg of silicon 47 mg of silicon 47 mg of silicon 17 mg of chromium 250 mg of stannic tin 250 mg of stannic tin 400 mg of stannic tin 400mg of stannic tin + 110 mg of silver 400 mg of stannic tin + 110 mg of silver 400mg of stannic tin + 110 mg of silver Iron found, mg 0.304, 0.301 0.302, 0.302 0-304, 0.304 0.302, 0.302 0.296, 0.298 0.298, 0.357, 0.294 0.302, 0.290, 0.298 0.296, 0.294 0.305, 0.298 0.294 0.284 0.226 0.064 0-296 0.280, 0.287 0.255, 0-263 0.294, 0.297 0.290, 0.295 0.287, 0-287 0.297, 0.303 0.230, 0.270 0.295, 0.297 0.303, 0-305 0.210, 0.080 0.307, 0.285 0.288, 0.297 Remarks pH greater than 9 Direct method; fluoride absent After sulphide separation; fluoride present After sulphide separation After sulphide separation 40 ml of thioglycollic acid reagent used 1 ml of potassium fluoride (200 g per litre) solution 1 ml of potassium fluoride (200 g per litre) solution 5 ml of potassium fluoride (200 g per litre) solution added added added 20 ml of thioglycollic acid reagent 40 ml of thioglycollic acid reagent Treated with hydrogen sulphide 6 g of oxalic acid added and 40 ml of thioglycollic acid rcagent used ; treated with hydrogen sulphide 10 g of oxalic acid added and 40 ml of thioglycollic acid reagent used ; treated with hydrogen sulphide Efect of large amounts of nickel-On adding thioglycollic acid to a solution prepared from spectrographically pure nickel, a colour similar to thak given by iron was observed; 20 mg of nickel gave the same absorption as 0-32mg of iron.Normally, magnesium alloys do not contain more than 0.01 per cent. of nickel, which would lead to a positive error equivalent to 0.0002 per cent. of iron and can be neglected.Dec., 19511 7 19 It was found that nickel could be satisfactorily removed with dimethylglyoxime and the iron determined on the filtrate after destroying the excess of dimethylglyoxime. For details see note I, p. 721. To a magnesium alloy containing 0.0063 per cent.of iron, the equivalent of 1.4 per cent. of nickel was added; after separation of nickel, 0.0066 per cent. of iron was found by this met hod. Other cations-From the literature it appears that even small amounts of cobalt produce a yellow to red colour, uranyl ions an intense orange colour, gold an intense amber colour and mercury’ a black precipitate. None of these elements are likely to be present in magnesium alloys. Less than 100 mg of arsenicII’ or antimonyIII, or less than 30 mg of mercuryII cause no interference.6 Sodium and potassium in extremely large amounts are said to bleach the colour slightly. Efect of aniorzs-Sulphate, chloride, bromide, fluoride, nitrate, sulphite, oxalate and citrate were found not to interfere. Swank and Mellon6 report that iodide, orthophosphate, chlorate, thiocyanate, acetate and tartrate in concentrations of 5 g per litre have no effect on the colour.Berg and Reimersl3 found that borate and tartrate decreased the sensitivity of the reaction, although other inve~tigators~1~~~ report no interference. We have found that small amounts of tartrate cause no interference, but that the colour intensity of a solution containing 16 g of ammonium tartrate per litre is reduced to about 60 per cent. of that obtained when 16 g of ammonium citrate per litre is present. Tartrate in moderate amounts can be tolerated if the calibration graph is prepared from solutions containing similar amounts of tartrate. Swank and Mellon6 further state that 2.5 g of B,O, (as tetraborate ion) per litre, 2 g of P20, (as pyrophosphate) per litre, 20 mg of molybdenum (as molybdate) per litre, 20 mg of tungsten (as tungstate) per litre, and 500 mg of arsenic (as arsenate) per litre can be tolerated, but that nitrite and cyanide interfere seriously.OF SMALL AMOUNTS OF IRON I N MAGNESIUM APPLICATION TO MAGNESIUM ALLOYS- It was found that some magnesium alloys contained small amounts of insoluble matter, probably slag or flux inclusions, which caused appreciable absorption when the solution was measured in a 20-cm cell. This absorption varied from sample to sample and therefore could TABLE IV EFFECT OF FLUORIDE ION ON IRON - THIOGLYCOLLIC ACID COMPLEX Potassium fluoride (200 g per litre) Iron added, solution, Iron found, mg ml mg 0.084 0.5 0.086 0.084 - 0.082 0.320 0-5 0.316 0.320 - 0.322 not be allowed for by a blank determination. It was also shown that this turbidity was slightly greater in ammoniacal solution than in acid solution, although it rarely exceeded an extinction of 0.08, i.e., 0.032 mg of iron.Filtration was undesirable as it invariably led to the random introduction of iron. The difficulty was overcome by adjusting the solution before the addition of thioglycollic acid to about 470 ml in a marked beaker, making it just ammoniacal and measuring the absorption. The solution was then returned to the beaker and acidified with hydrochloric acid; thioglycollic acid was added, the solution was adjusted to the correct pH with ammonia and diluted to 500 ml. The iron colour was measured and the absorption due to turbidity deducted.Since the absorption due to turbidity is small, the error introduced by its measurement when the solution has a volume of 470 ml, as compared with the final volume of 500 ml, is negligible. Magnesium - zirconium alloys-Alloys containing zirconium contain also some “insoluble zirconium,” which can be seen as a fine black sludge after the sample has been dissolved in acid. I t has been found that most of the iron is contained in this sludge and it is necessary, therefore, to decompose it, as generally the total iron content has to be determined. The “insoluble zirconium” dissolves very slowly in acid, but is attacked readily by fusion or by treatment with hydrofluoric acid.MAYER AND BRADSHAW : PHOTOMETRIC DETERMINATION [Vol. 76 720 As filtration, ignition and fusion of this residue seemed undesirable because of the danger of introducing comparatively large amounts of iron, it was decided to use hydrofluoric acid.It was shown first (see Table IV) that fluoride had no effect on the iron - thioglycollic acid colour. When large volumes of potassium fluoride solution were added to samples of rare-earth - zirconium - magnesium alloys, a precipitate of rare-earth fluoride formed, but as 0.5 ml of the 200 g per litre solution was sufficient to decompose all the “insoluble zirconium,” no difficulty was experienced when this technique was used on several hundred samples. It was proved that the decomposition of the insoluble zirconium was complete by carrying out the above procedure and comparing the results with those obtained by the following modifica- tion.After dissolution, “insoluble zirconium” was filtered on a small circle (1-8 cm in diameter) of filter-paper held in a sintered-glass crucible under suction. The filter-paper was ignited in a platinum crucible and the residue -was digested with concentrated hydrofluoric acid, fumed with sulphuric acid and extracted in a small amount of hydrochloric acid; this solution was combined with the main filtrate. ’The iron determination was completed in the usual manner. As shown in Table V, there was no significant difference between the two procedures and it was concluded that the potassium fluoride treatment was effective. TABLE IT TEST OF POTASSIUM FLUORIDE DECOMPOSITION OF “INSOLUBLE ZIRCONIUM” Normal method with 0.5 ml of KF.Insoluble matter filtered and digested with HF. Alloy Iron found, Iron found, YO YO 3% of rare earth + 0.6y0 of zirconium . . 0.0017, 0.0015 0.0015, 0.0017 6% of zinc + 0.6y0 of zirconium . . .. 0.0016, 0.0017 0.0016, 0.0014 CALIBRATION GRAPH- same whether the base was pure magnesium or alloy-magnesium. The calibration graph has remained unaltered for two years and was found to be the RECOMMENDED METHOD Principle-Dissolution in acid, formation of the iron - thioglycollic acid complex at pH greater than 7.5 and measurement of the absorption of the pink colour at 520 mp. APPARATUS- Spekker absorptiometer fitted for use with 20-cm cells and tungsten filament lamp. All glassware must be cleaned with hot hydrochloric acid and then rinsed with distilled Storage vessels for distilled water should be protected against contamination by dust.All reagents should be of recognised analytical quality. Hydrochloric acid, sp.gr. 1-18. Bromine water-Saturated. SztZPphur dioxide solution-A freshly prepared saturated solution. Ammonium citrate solzdion-Dissolve 250 g of citric acid in about 700 ml of water, neutralise to litmus paper with ammonia, filter, cool and dilute to 1 litre. ThiogEycoZlic acid reagent-Dilute 50 ml of the acid with about 300 nil of water, neutralise to litmus paper with ammonia and dilute to 500 ml. Ammonium hydroxide, sp.gr. 0.880. SAMPLE- With a hacksaw cut slices of the alloy weighing about 4g. Clean these by repeated dipping in diluted nitric acid (1 + 9), rinse in distilled water and dry with filter-paper before weighing.For samples containing more than 0.008 per cent. of iron use a proportionately smaller water. REAGENTS- Keep tightly stoppered. wpipht nf snmnle hiit nnt lpcc than 1 P Fnr the rnnpp 0.015 to 0.lfi mq- cent. of iron 11qeDec., 19511 OF SMALL AMOUNTS OF IRON I N MAGNESIUM 721 PROCEDURE- and 10 ml of hydrochloric acid per gram of sample. lution is complete add 10 ml of bromine water to dissolve copper and similar elements. boil the solution for about 5 minutes to remove most of the excess of bromine. If zirconium is present in the alloy, add from a pipette 0.5 ml of potassium fluoride solution, 200 g per litre, to dissolve “insoluble zirconium,” and boil for a few minutes. If more than 0.1 per cent. of silicon is present, add 1 ml of potassium fluoride solution.Cool to room temperature, add 10 ml of sulphur dioxide solution and 40 ml of ammonium citrate solution. Dilute to about 400 ml and add ammonium hydroxide from a dropping bottle until the solution is just alkaline to litmus paper. Dilute to the 470-ml mark with distilled water. Transfer the solution to a 20-cm cell and measure the absorption with Ilford No. 604 green filters and H 503 heat filters, and a water - water setting of 1.00. Return the solution to the beaker, acidify with hydrochloric acid, and add 10 ml of thioglycollic acid reagent. If the alloy contains more than 5 per cent. of zinc in conjunction with over 1 per cent. of manganese, or if 1 to 2 per cent. of cadmium is present, 20 ml of thioglycollic acid reagent should be added; with 15 per cent.of cadmium present 40 ml should be added. If the alloy contains major amounts of tin add a 10-ml excess of thioglycollic acid reagent for every 0.1 g of tin present. Again make just alkaline to litmus paper and dilute to 500ml in a graduated flask. Measure the absorption as before in a 20-cm cell.* The difference between the two readings is a measure of the iron content of the solution and has to be corrected for iron introduced by reagents. Carry out a blank determination with 5 ml of hydrochloric acid, but otherwise proceed exactly as in the above procedure; deduct the drum-difference reading obtained from that obtained with the sample. Should this value be exceeded, a careful check should be made and any reagent suspected of high iron content replaced immediately.Results are reproducible to within tO.0002 per cent. of iron when a 4-g sample is taken. Transfer the sample to a 600-ml beaker marked at 470m1, add about 100ml of water Cover with a clock glass and when disso- Then Normally the blank will not exceed an extinction of 0.07. The time required for a determination is 18 hours. Range-- A 4-g sample in a 20-cm cell covers the range 0 to 0.008% of iron. A 2-g sample in a 20-cm cell covers the range 0-008 to 0.016% of iron. A 1-g sample in a 4-cm cell covers the range 0.015 to 0.15% of iron. CALIBRATION- hydrochloric acid (1 + 1) and dilute to 1 litre. hydrochloric acid (1 + 9). Standard iron solution, 1 g per Zitre-Dissolve 1.00 g of pure iron in 100 ml of diluted DiZuted standard iron solution-Dilute 10 ml of the above solution to 1 litre with diluted 1 ml of solution EE 0.000010 g of iron, i.e., E 0.00025% of iron on a 4.0-g sample.To 4-g samples of magnesium of low iron content add diluted standard iron solution from To one sample add no iron solution and use Plot a graph of drum differences against milligrams of iron added. A typical a burette and carry out the given procedure. this as a blank. graph corresponded to the equation- Milligrams of iron in 500 ml = Drum difference x 0.390. NOTE I- If the sample contains more than 0.01 per cent. of nickel, dissolve the sample in hydro- chloric acid, oxidise with bromine water and add the ammonium citrate. Heat the solution to 80” C, add sufficient 1 per cent. aqueous solution of the sodium salt of dimethylglyoxime to precipitate all the nickel, make alkaline with diluted ammonium hydroxide (1 + 1) and leave on a steam bath for half an hour. Filter through a sintered-glass crucible and wash * If more than 0.5 per cent. of manganese is present, allow the solution to stand for 20 minutes before In the presence of only small amounts of manganese the colour develops immediately and is measuring.stable for 18 hours.722 MAYER AND BRADSHAW [Vol. 76 the precipitate with hot water. Acidify the filtrate and add 20ml of concentrated nitric acid. Bring the solution to the boil and continue boiling for half an hour. Cool the solution, add 10 ml of sulphurous acid and continue as in the normal procedure. NOTE II- Omit the addition of sulphur dioxide solution and, after the addition of the ammonium citrate solution, neutralise to litmus with ammonium hydroxide and then just acidify with hydro- chloric acid.Saturate the solution with hydrogen sulphide and filter through a small circle of Whatman No. 40 filter-paper held on a sintered-glass crucible under suction. Before use wash the crucible and paper thoroughly with hot dilute hydrochloric acid. Wash the pre- cipitate with cold citric acid solution, 10 g per litre, saturated with hydrogen sulphide. Remove hydrogen sulphide from the filtrate by boiling, add a few drops of bromine water, boil for a few minutes, then cool and add 10 ml of sulphur dioxide solution. Dilute to 400 ml, add ammonium hydroxide and continue as described in the procedure. If the sample con- tains more than 0-5 per cent.of tin in addition to silver, add 10 g of oxalic acid before neutral- ising and passing hydrogen sulphide. If the sample contains silver, lead or bismuth, remove these elements as follows. TYPICAL RESULTS The method has been applied to most commercial magnesium alloys and has given satisfactory results as judged by reproducibility between several operators and compared with TABLE VI SOME TYPICAL RESULTS Alloy 99.5% Mg 99.5% Mg AM503 M5A AZM A.8 A.8 A.8 AZG AZG ZRE.l ZKE. 1 z 5 z ZA ZA z5z Composition, yo r \ A1 Mn Zn P earths Zr A Rare - 1.0 0.5 0-5 0-6 3 3 2.6 2.6 5 5 Iron determined by r - r thioglycollic throline acid method, method, 0.044 0.042 0.019 0.019 0.014 0-015 0.0036 0.0039 0.0025 0.0022 0*010 0.01 1 0.0048 0.0048 0.024 0.026 0-0060, 0.0062 0.0006, 0.0005 0.0021, 0.0019 0~0010, 0.0009 0.0014, 0.0015 0.0037, 0.0036 0.0027, 0.0026 0.0014, 0*0017 o-phenan- Yo % the o-phenanthroline method on alloys containing up to 1 per cent. of zinc. A few typical results are shown in Table VI. The authors wish to express their thanks to the Chairman and Directors of Magnesium Elektron, Limited, for permission to publish this report. REFERENCES 1. “Methods of Analysis of Magnesium and of Magnesium Alloys,” F. A. Hughes & Co. Ltd., London, 2. Sandell, E. B., “Colorimetric Determination of Traces of Metals,” Interscience Publishers Inc. 3. Fortune, W. B., and Mellon, M. G., Ind. Eng. Chem., Anal. Ed., 1938, 10, 60. 4. Schulek, E., and Floderer, I., Bey. ungar. $harm. Ges., 1939, 15, 210; Chem. Abst., 1939, 33, 4901. 5. Moss, M. L., and Mellon, M. G., Ind. Eng. Chem., Anal. Ed., 1942, 14, 862. 6. Swank, H. W., and Mellon, M. G., Ibid., 1938, 10, 7. 7. Snell, F. D., and Snell, C. T., “Colorimetric Methods of Analysis,’’ D. Van Nostrand Inc., New 1946, pp. 14 and 15. New York, 1950, p. 362 el seq. York, 1949, Vol. 2, p. 319.Dec., 19511 LEYTON : AN IMPROVED FLAME PHOTOMETER 0. 9. 10. 11. 12. 13. 14. Chirnside, R. C., J . SOC. Glass Tech., 1938, 22, 41. Chirnside, R. C., and Pritchard, C. F., Ibid., 1939, 23, 26. Yoe, J. H., and Hall, R. T., J . Amer. Chem. SOC., 1937, 59, 872. Greenburg, R. H., Ind. Eng. Ckem., Anal. Ed., 1946, 18, 255. Yoe, J. H., and Jones, A. L., Ibid., 1944, 16, 111. Berg, K. B., and Reimers, F., Dansk Tzdsskr. Farm., 1950, 24, 316. Mehlig, J. P., and Shepherd, M. J., Chemist Analyst, 1946, 35, 8. CHEMICAL RESEARCH LABORATORY MAGNESIUM ELEKTRON LIMITED CLIFTON JUNCTION NR. MANCHESTER 723 February, 1961

ISSN:0003-2654    

DOI:10.1039/AN9517600715

出版商:RSC    

年代:1951

数据来源: RSC

摘要:

Dec., 19511 LEYTON : AN IMPROVED FLAME PHOTOMETER 723 An Improved Flame Photometer BY L. LEYTON The paper describes the design, construction and performance of a relatively simple flame photometer possessing a number of new features that have led to definite improvements in the reliability and performance of the instrument. The incorporation of a concentric type of atomiser and a new burner system allows reliable analyses to be made on less thaD 1 ml of test solution with a sensitivity of the order of 0.2 p.p.m. for potassium and 0.5 p.p.m. for calcium. Selectivity has been greatly increased by the introduction of interference filtgrs. THE increasing popularity of flame photometry for the quantitative determination of elements such as sodium, potassium and calcium is proof of its great advantages over more orthodox analytical procedures.The principles underlying this technique are now well known and need not be discussed here in any detail, Briefly, a solution containing the element or elements to be determined is converted into a spray by means of an atomiser. The spray is then injected into a flame (e.g., air-acetylene) under controlled conditions and the intensity of the spectral radiation measured by some light-measuring device (e.g., photocell and galvan- ometer). When more than one element is present in the solution, the respective radiations are separated by means of appropriate light filters. The instrument is calibrated with standard solutions analysed under the same conditions (for a review of the development of the tech- nique see Leytonl).A number of instruments are now commercially available, but the high cost of these instruments is sufficient to put them beyond the means of most laboratories. Some years of experience with flame photometry has now led to the design of a simple, rela- tively inexpensive model that has proved reliable and accurate and well suited to routine determinations. A number of features have been introduced, which, it is believed, represent a definite advance in photometer design. These features include (a) a concentric type of atomiser, (b) a simple burner tube, up which the test solution is sprayed directly into the flame, (c) a stainless-steel burner giving a flat, spade-shaped flame, ( d ) a “vernier” water manometer to allow more precise control of the air pressure, (e) sintered-glass filters in the air and acetylene streams to remove troublesome solid impurities and (f) special interference filters for particular elements.CONSTRUCTIONAL AND WORKING DETAILS THE ATOMISER- The atomiser, A, shown in Fig. 3, is the key component of the photometer in determining its efficiency and reliability, and in this respect the concentric type has proved far superior to the “right angled” types previously used. Being fixed it is not subject to sudden changes that may occur in the adjustable types; it is readily removed from the apparatus and easily cleaned or replaced. Air forced through the outer annulus, b, creates sufficient suction in the central feed tube, c, to lift the solution through a height of about a foot.and sometimes more, so that by using a relatively short feed tube, small variations in the depth to which the tube is immersed in the solution lead to insignificant variations in the amount of spray delivered to the flame. The spray is much more uniform than with other types of atomiser and the reduced losses involved in the conversion of liquid to spray allow the use of much smaller volumes of test solution. The model is illustrated in Figs. 1 and 2.724 LEYTON: AN IMPROVED FLAME PHOTOMETER (Vol. 76 -2 I Depending on the atomiser, reliable determinations have been made on as little as 0.5 ml of test solution. Whilst the dimensions of the orifices do not appear to be critical to enable the atomiser to function, they do influence the efficiency of its performance, the amount of test solution required and the volume of air used.Atomisers with relatively large orifices use an apprec- iable amount of air and test solution, although the sensitivity may be quite high because of the large amount of spray delivered to the flame. Experience with a number of atomisers I H - - O-- \ Atr SUPP’Y Mercury M o n o r n e t ~ ~ l or Pressure Gauge 13 inches 1 I -0 Fig. 3. A, concentric atomiser; D, burner tube; H, burner. For details see text X Y Fig. 4. A, “vernier” manometer; B, gas filter. For details see text had led to the development of models with very fine orifices. These require much higher air pressures, but the volume of air (and therefore the amount of acetylene) used is much less, while their efficiency is very high.The following dimensions and associated data are given for a typical atomiser that has proved very efficient in practice- Outer tube, external diameter a t tip . . .. .. .. 79 99 internal diameter a t tip . . .. . . .. Inner tube, external diameter a t tip . . .. .. .. 99 91 internal diameter a t tip . . .. . . .. Feed tube, length to tip of atomiser . . .. .. .. Air pressure (lb per sq. inch) . . .. .. .. .. Acetylene pressure (cm of water)* . . .. .. . . Rate of uptake of test solution (ml per minute) 97 79 external diameter . . .. .. .. .. )) 39 internal diameter . . .. . . .. .. . . .. 4.0 mm 0.96 mm 0.76 rnm 0.12 rnm 4.0 mm 1-8 mm 11-8 cm 20 19” 1-5 * This value applies when a 19-cm length of 1-mm capillary tubing is inserted in the acetylene lead. Without this extra resistance, the reading of the acetylene water manometer is of the order of 1 to 2 cm and is too low to monitor with sufficient accuracy.In practice it is impossible to produce two atomisers with identical performances and the efficiency of a particular model can only be tested in the photometer itself by observing the sensitivity achieved with different air and acetylene pressures for a given test solution.~ ~~ - - Fig. 2. Flamc photometer. Top view with lid removed Ci,ri/rrp : 13iirnc.r in asbcstos flanic chambcr I-efr : Potassinin unit with water ccll, intcrfrrcmce iiltcr ant1 photociectric cell I?iqhf ; Calcium unit with water cc.11, tlidymiiim filter, intcrfcrcncc filter ;m(l Ijrrrk : Sodium iiiiit with \vatcr ccll, condensing IPIIS, \ o l i i t i o i i ; ~ i i t l light l i l t c .i . 5 wlen ium ccll and sclcniriin ccllDec., 19511 LEYTON: AN IMPROVED FLAME PHOTOMETER 725 In the construction of the atomiser, both inner and outer tubes are drawn out concentri- cally so that the annular space between them tapers towards the tip. The tip is then ground back until the atomiser functions satisfactorily; it is then inserted into the burner tube and the air pressure adjusted to give a suitable rate of uptake of solution. This air pressure will fix the amount (pressure) of acetylene required for complete combustion in the flame. The sensitivity can be increased by increasing the air and corresponding acetylene pressures, but as this entai1s.a greater uptake of test solution, the final choice will be determined by the conditions of the investigation.BURNER TUBE AND BURNER- The burner tube, D, shown in Fig. 3, is blown from Pyrex glass and is provided with an acetylene inlet, e, and an outlet for condensed spray, f. The atomiser is fixed in the base of the burner tube by means of a short length of rubber tube, acting as a bung, so that its tip projects above the waste outlet. The efficiency of this unit has been greatly improved by introducing a small glass bulb, g, 5 mm in diameter, through the side of the burner tube and fixing it about 1 cm above the atomiser tip, to break up the larger droplets in the spray impinging upon it. The choice of this size of bulb was purely arbitrary and it is evident that both the size of the bulb and the distance between it and atomiser tip will affect the amount of spray finally reaching the flame and hence the sensitivity.The slight loss in sensitivity involved is more than compensated for by the increase in stability and more rapid clearing of the flame after a reading has been made. The dimensions of the burner tube do not appear to be critical. The burner, H, Fig. 3, was made from a length of stainless steel rod, 2 inch in diameter, one end of which was drilled to make a smooth fit over the end of the burner tube, the other end being turned and carefully squashed in a vice over a former to give an elongated jet about 1 mm wide and 3 cm long. It is essential that this jet be uniform, but the necessary adjust- ments are easily made by trial and error with a coal-gas flame. The spatulate flame so produced, besides reducing the zone of unburnt gas to a narrow band, gives a much greater sensitivity than does a conical flame. When a lens system is incorporated, the flat side of the flame presents a large luminous plane surface to the lens, while in the absence of a condensing lens system, for example when interference filters are used, the flame is arranged with its edge towards the optical system.The burner unit also has the advantage that the acetylene supply can be turned off without any of the annoying “blow back” that appears to be inherent in certain other designs. FLAME CHAMBER- Much of the design of the flame chamber (Figs. 1 and 2) can be left to the ingenuity of the designer. In the model illustrated it consists of a rectangular box of &inch asbestos sheet, measuring 2$ inches by 34 inches by 44 inches high, open at top and bottom.Two-inch holes are drilled in three of the side walls and a smaller hole, 1 inch in diameter, is made in the front wall for inspection and ignition. The lid, also of asbestos, carries a simple metal tubular chimney, 7; inches long and 2$ inches in internal diameter, open at the top. Over each of the side holes is clamped a 3-inch length of 2;-inch diameter metal tubing into which plane glass windows are cemented about 2 inches apart. These form cylindrical chambers for water, which provides cooling between the flame and the optical system. Access to the water chambers is provided by two small holes drilled in the wall at the top. MANOMETERS- Originally the manometers were similar to those described by Doming0 and Klyne,2 and consisted of a mercury U-tube manometer for the air supply and a similar water mano- meter for the acetylene supply.Small changes in air pressure have an appreciable effect on sensitivity, but are not easily detectable as changes in the height of the mercury column. An improvement in monitoring has been made by introducing into the air system a simple “vernier” water manometer, as shown in the diagram, Fig. 4A. With tap T open, the air pressure is adjusted to the correct working value, as shown by the mercury column; this produces the same pressure in bulbs X and Y. When tap T is closed, Y is maintained a t the original pressure, but X is now subject to any subsequent fluctuations. As the U-tube between X and Y is filled with water, small changes in air pressure can be detected with far greater precision than in the main mercury column.Under normal laboratory conditions726 LEYTON: AN IMPROVED FLAME PHOTOMETER [Vol. 76 i t is unlikely that changes in temperature will upset this monitoring system. Bulbs X and Y are merely safety devices to ensure no loss of water if the main air supply is turned off at the end of a run before tap T is opened. Though simple in construction, manometers of this kind are rather cumbersome, and it is possible to replace them with more compact dial pressure gauges. This has been done for the air supply because the introduction of atomisers with very fine orifices required the use of air pressures much higher than those that could be accommodated with the existing mercury column.No accuracy in monitoring has been lost, however, as the “vernier” manometer is still incorporated in the system. The original water manometer for measuring acetylene pressure has been retained. AIR AND ACETYLENE FILTERS- The air and acetylene filters, Fig. 413, consist of $inch diameter discs of No. 2 porosity sintered glass fused into glass bulbs and inserted in the air and acetylene streams in front of the atomiser. OPTICAL SYSTEMS- The introduction of interference filters has improved the selectivity far beyond the limits possible with the ordinary light filters used in earlier designs. These filters consist of several layers of metallic film separated by dielectric of such dimensions that, because of well known interference effects, they act essentially as monochromatic filters.Suitable filters are now available* with maximum transmissions at wavelengths equivalent to the spectral radiations of sodium, potassium, calcium, lithium and so OIL Compared with ordinary light filters, the narrower transmitted waveband results in rather lower sensitivities, but despite this, adequate sensitivity is still achieved. The recently developed all-dielectric interference filters allow a much higher transmission and should, therefore, prove more suitable. As interference filters are arranged to work with parallel beams of light, it appears that, in the absence of a point source of light, an optical condensing-lens system is of little use. Various experiments with lens systems have so far resulted in no appreciable gain in sensitivity, and there is always the tendency to increase the width of the transmitted waveband, with a consequent decrease in selectivity. The conventional lens system fou.nd in photometers of this kind has therefore been omitted, the light being transmitted directly from the flame, through the interference filter, to the detecting device.The triple light path allows three filters to be set permanently in the instrument and so avoids troublesome delays in changing from the measurement of one element to another. By means of a simple switch, any one of the three optical systems may be chosen. It is theoretically possible to determine three elements at the same time, but this is impracticable with the present simple design.In the existing model, two interference filters are used, one for potassium, with a maxi- mum transmission at 768mp, and the other for calcium, with a maximum transmission a t 616mp. The calcium filter transmits some sodium light and is therefore combined with a didymium filter (Wratten 77A) to reduce interference by sodium still further. The sensitivity and degree of selectivity achieved with these filters is shown by the following figures- Test solution Galvanometer reading Potassium filter . . .. .. . . K 20p.p.m. 90 mm Ca 500p.p.m. 1 mm Na 500 p.p.m. 1 mm Calcium filter . . .. .. . . Ca 20p.G.m. 48 mm K 500p.p.m. 8 mm Na 500p.p.m. 14 mm Although the potassium filter transmits insignificant amounts of calcium or sodium radia- tion, the calcium filter is not so selective.The relative transmission, however, is only of the order of 0.5 per cent. for potassium and 1.0 per cent. for sodium, so that the corresponding error in the calcium determination is usually negligible. Sometimes, however, as in the deter- mination of low concentrations of calcium in the: presence of high concentrations of sodium, this interference cannot be ignored. Fortunately both the potassium and sodium interfer- ences appear to be additive, and reliable calcium estimations can still be made in the presence * From Messrs. Barr and Stroud, Anniesland, Glasgow, or B. Lange, 14-18 Hermannstrasse, Berlin- Zehlendorff.Dec., 19511 LEYTON: AN IMPROVED FLAME PHOTOMETER 727 of large concentrations of these elements if their concentration is known and transmission through the calcium filter is allowed for.The third light path is temporarily fitted with ordinary light filters for sodium (a combina- tion of Li’ratten numbers 57 and 23A). These filters still pass an appreciable amount of potassium radiation, so that a liquid filter (1 cm of dilute copper sulphate) has been included to reduce this. A condensing lens system is introduced and the combination is arranged to work with the flat side of the flame.* For calcium and sodium, selenium barrier layer cells (Evans Electroselenium Ltd.) are used, and for potassium, a gas-filled caesium cell (Cinema Television Type GS 18) is used in conjunction with a 100-volt dry battery. Selenium cells (Megatron Ltd.) are now available which are sensitive to infra-red radiations and therefore to potassium radiation, but in the absence of comparative measure- ments no details can be given of their performance.The filters and corresponding detectors are mounted in small wooden boxes held in posi- tion against the ends of the water chambers, the optimum position having been determined experimentally before-hand (see Fig. 2). As it is essential to exclude from the detectors any light other than that from the flame, the flame chamber and associated optical systems are enclosed in a light-tight box (see Fig. 1) provided with lids for easy access to the components. The output terminals of the detectors are connected via a switch and variable shunt box to a sensitive galvanometer (Tinsley model SS 6.45, coil resistance 207 ohms, sensitivity 990 mm per microamp).I t is recommended that one terminal of the galvanometer be earthed in order to avoid the pick-up of stray currents in the leads. Experiments are now being made to convert one of the light paths so that the internal- standard technique can be used. The detecting devices differ little from those recommended elsewhere.2 THE TECHNIQUE OF FLAME PHOTOMETRY- The apparatus described above is still subject to the errors and defects inherent in the flame photometric procedure and due care has to be taken if the instrument is to give reliable results. It is advisable to allow the instrument to warm up for at least a quarter of an hour before taking readings, as there is always an initial drift in sensitivity, presumably during the stabilisation of the detecting cells.I t has been the custom, when dealing with a large number of samples, to analyse them in batches of five or six and to refer to standard solutions between each batch. This frequent reference to standards has proved advantageous because of the unfortunate tendency of the atomiser to become blocked with solid particles originating in the test solution or air supply. Much trouble of this kind can be avoided by using air filters and clean solutions (either filtered or centrifuged), but even so, occasional stoppages occur. When the atomiser ceases to function, the trouble is easily located and remedied, but sometimes the blockage is only partial, resulting in a decreased uptake of solution and therefore a decreased sensitivity. Only by continual reference to standards can this source of error be detected.The flame clears itself fairly rapidly but, particularly after concentrated solutions, there may be an appreciable “residual” effect, and it may be necessary to wash with distilled water for some time before the galvanometer spot returns to zero. It has often been recommended that two successive runs be made on each solution, the first reading being ignored. If sufficient test solution is available, more than one reading can be taken and this procedure has obvious advantages. A useful adjunct to the apparatus is a movable stage, such as the one illustrated in Fig. 1, whereby the solution can be held under the atomiser feed tube and so allow the full attention of the operator to be directed towards the galvanometer and controls.For calcium in the range 0 to 100 p.p.m. and for sodium in the range 0 to 20 p.p.m., or perhaps greater, there is a linear relationship between concentration and galvanometer reading. For potassium, the curve is logarithmic, the instrument being most sensitive at low concentrations. Over the range 0 to 100 p.p.m. of potassium a straight line is obtained when the graph is plotted on log - log paper. Washing the atomiser with distilled water between each test solution is essential. * Since the submission of this paper, a sodium interference filter has been incorporated in the apparatus; The sensitivity attained is about 0.02 p.p.m. transmission of potassium and calcium radiation is negligible. of sodium.728 LEYTON: AN IMPROVED FLAME PHOTOMETER [Vol.76 PERFORMANCE The instrument has undergone exhaustive tests in a variety of analyses and some details are given below. DETERMINATION OF POTASSIUM, CALCIUM AND SODIUM I N PLANT TISSUES (CONIFER NEEDLES)- About 200 mg of the dried plant samples were weighed into silica crucibles and ashed in a muffle furnace a t 550 to 600” C for 7 hours, after which there was no further change in weight. The ash was extracted with 2 drops of concentrated hydrochloric acid and hot distilled water and, when cool, made up to 50ml. Potassium, calcium and sodium were determined directly in centrifuged samples of the extract. Preliminary information on the approximate proportions of the elements in the ash decided the composition of the standard solutions. These contained 0 to 100 p.p.m.of calcium, 0 to 100 p.p.m. of potassium and 0 to 40 p.p.m. of sodium as chlorides, with the equivalent of two drops of concentrated hydrochloric acid per 50 ml. Each standard solution contained potassium, calcium and sodium in the ratio of 5 : 5 : 1. DETERMINATION OF SODIUM AND POTASSIUM IN PLASMA, WHOLE BLOOD, URINE, GASTRIC ASPIRATIONS AND PERITONEAL FLUIDS- For sodium, analyses were normally made directly on 1 + 40 dilutions with distilled water. Protein in red cells was precipitated with a 10 per cent. solution of trichloro-acetic acid and readings were made on centrifuged sa.mples, the standards containing the same concentration of the precipitant. Because of large variations in the amount of sodium in plasma, potassium determinations were made on plasma diluted about 20 times with a diluent containing 3,300 p.p.m.of sodium. This proved sufficient to reduce the otherwise variable sodium interference to a substantially constant level. For urine and gastric aspirations, potassium was determined directly on samples diluted 20 times. For red cells, potassium was determined directly 011 protein-precipitated samples diluted about 60 times. The standard solutions contained the same high sodium concentrations. DETERMINATION O F CALCIUM I N TISSUES AND BODY FLUIDS OF XEXQPUS- For plasma, the proteins in 0.1 ml of sample were precipitated with trichloro-acetic acid and the supernatant fluid made up to 1 ml. For tissue sections, about 100 mg were ashed at 550” C in a platinum crucible and the ash was dissolved in 1 ml of dilute hydrochloric acid. Determinations on replicate samples and recovery tests showed a reproducibility for all elements of within &3 per cent. and recoveries of the order of 100 -j= 5 per cent. The figures vary somewhat with the particular material used, but in the majority of these estima- tions an ove1-all accuracy of better than +5 per cent. can usually be achieved. Not all the modifications included in the apparatus described above are original. Some of the improvements have been noted in apparatus designed and made in other laboratories and due acknowledgement is made to these sources, published or otherwise. The author is indebted to Mr. Vincent of the Biochemistry Department, Oxford, for his invaluable help in designing and making the numerous concentric atomisers tried out in the course of this work, and to Dr. Streeten and Dr. Foulkes for valuable criticism and for details of the biochemical procedures. REFERENCES 1. 2. Leyton, L., “Flame Photometry,” Ann. Rep. Chem. SOL, 1948, 45, 326. Domingo, W. R., and Klyne, W., “A Photo-electric Flame Photometer,” Biochem. J., 1949, 45, 400. DEPARTMENT OF FORESTRY OYFORD UNIVERSITY ERRATUM: August (1951) issue, p. 451. Eleventh and tenth lines from foot of page; the equations should read- X = 2.611 EM - EL - EH) = 2.611 (2E at 325 mp - E at 310 mp - E at 340 mp).

ISSN:0003-2654    

DOI:10.1039/AN9517600723

出版商:RSC    

年代:1951

数据来源: RSC

摘要:

Dec., 19511 NOTES 729 NOTE ON ELECTROSTATIC CHARGES ON GLASS AND OTHER VESSELS* THE Note by Armstrong’ in a recent number of The Analyst drew attention to a source of error which, although noted by earlier workers, might easily pass unsuspected when an object is weighed on an undamped balance and would more probably be overlooked when an aperiodic balance is used. The electrostatic effects referred to are quite distinct from the adsorption, convection and buoyancy effects that may disturb the weighings for some minutes after wiping any vessel, although they are often associated with them. High-silica glasses are particularly prOne t o acquire a charge, but porcelain crucibles may show the same effect; some do so more than other, apparently similar, crucibles supplied by the same maker.The difficulty usually arises when the relative humidity is low, e.g., about 40 per cent. Even soda-glass absorption tubes on a combustion train may then be troublesome. To discharge the apparatus many recommendations have been made, e.g., passing the apparatus through a non-luminous flame, breathing on the surface2 or humidifying the laboratory air.3 Ultra-violet light4 has been stated to reduce a charge to a negligible amount in 10 minutes (a change of 0.05 mg or less between weighings at 5-minute intervals on a 150-ml Pyrex flask was regarded as negligible). Also high-frequency discharge,s such as from a vacuum-line leak detector, has been stated to be effective in a matter of seconds. We suspect that precautions taken to deal with electrostatic charges are not always fully effective.Part of the charge can be dispersed fairly readily, but the residue may defy considerable efforts to remove it. For example, the National Physical Laboratory had to abandon the use of quartz (rock crystal) weights as reference standards because, even when they were shielded with metal boxes that were earthed immediately before each weighing, different results were found on different balances. Radio- active material and X-rays also proved unsuccessful in eliminating the charges. None of the methods of discharging apparatus that we have tried has proved fully reliable. I t seems better to prevent a charge being acquired. In microchemical work, this can be achieved by cooling the crucibles, after ignition, in the usual way in a metal block, but in a moist atmosphere such as can be produced by inserting under the glass cover a crucible filled with water or with a saturated solution of a salt that has a vapour pressure corresponding to a relative humidity of approximately 60 per cent.Thiers and Beamish6 suggest the use of calcium nitrate for this purpose, as it gives a relative humidity between 50 and 60 per cent., but reference can be made to O’Brien’ for alternatives. On a dry day the weighing tube of a combustion train should be covered with a damp filter-paper or a damp cloth during the run. This is removed when the tube is placed in the balance, and sufficient time is then allowed for equilibrium conditions to be established before the actual weighing takes place. Armstrong’s paper relates to Vitreosil crucibles.Where high-accuracy weighing is needed this can be a limiting factor. This Note is published by permission of the Director of the National Physical Laboratory- REFERENCES Armstrong, A. W., .4naZyst, 1951, 76, 551. “Standard Methods for Testing Petroleum and its Products,” 11th Edition, 1961, p. 604. Hayman, D. F., Ind. Eng. Chem., Anal. Ed., 1936, 8, 342. Rodden, C. J., Ibid., 1940, 12, 693. Thiers, R. E., and Beamish, F. E., Anal. Chem., 1947, 19, 434. O’Brien, F. E. M., J . Scz. Instrum., 1948, 25, 73. 1. 2. 3. 4. 6. Van Straten, F. W., and Ehret, W. F., Ibid., 1937, 9, 443. 6. 7. I. H. HADFIELD P. H. BIGG August, 1951 * Communication from the National Physical Laboratory.730 NOTES [Vol. 76 THE TOXICOLOGY OF ANTHISAN THREE recent fatalities in New Zealand involving young children have brought to light further information on the toxicology of anthisan [mepyrarnine maleate (B.P.Addendum, 1951), pyranis- amine maleate (U.S.A.), neoantergan maleate (Mei-ck)]. The children were boys aged 3& years (case I ) ; 2 years (case 11); 2 i years (case 111). The amounts of anthisan taken were approxi- mately- Case I : 0.5 to 0.7 g. Case I1 1.1 g. Case 111: 0.5 g. Case I was prescribed one 0-05-g anthisan tablet three times a day after meals for a nasal complaint for three days before the fatal overdose. DISTRIBUTION OF THE DRUG- comparison with the amounts taken. Stomach and contents-The amounts of anthisart base recovered in all cases were very small in Vomit-This was examined in cases I and 11.Intesfines-In case I1 the whole intestinal tract, weighing 670 g, was examined. Base equivalent to 0.094g of anthisan was recovered. Liuer-In case I, base equivalent to 0-069 g of anthisan per 1000 g of tissue was recovered. In case 11, the liver, spleen, kidneys and some blood from these organs were submitted for analysis in the same container and were therefore examined as one sample. Base equivalent to 0.021 g of anthisan per 1000 g of tissue was recovered. Heart, thymus, lungs and trachea-In case II, base equivalent to 0.014 g of anthisan per 1000 g of tissue was recovered. Bladder and urine-In case 11, a trace of the drug was detected, although the amount of urine present was less than 5 ml. The urinary excretion of anthisan was studied in an experiment in which an adult female subject was given two 0-1-g tablets, one a t 6 p.m., the other a t 10 p.m.Analysis and spectro- photometric examination of the urine collected periodically during the next day showed that up to 5 p.m. approximately 0.05 g of base, equivalent to 0.07 g of anthisan, had been eliminated. The urine passed at 6 a.m. was not available. The urine passed at 9 a.m. contained only a trace of the drug compared with the urine passed at 1 p.m. and 5 p.m. The three case histories follow closely that reported by Miller and Pedley,l with the exception that a green coloration of the alimentary tract or contents was not observed. Green staining from the tablet coating was found only about the mouths. It is of interest to note that in the case described by Miller and Pedley the drug could not be detected in any organs other than the stomach and contents.The clinical details will be published more fully in the New Zealand Medical Journal. Traces of the drug were found. METHOD- alcoholic extracts. water followed by filtration and extraction with 'chloroform. ammonia and extracted with chloroform. As anthisan was known to have been taken, the tests were used for confirmation. The conventional Stas - Otto process was followed, tartaric acid being used to acidify the The ammoniacal chloroform extracts were purified by solution in acidified The oily base was liberated by The base was tested by the methods given by Idson.2 I am indebted to the Director, Dominion Laboratory, for permission to publish this Note. REFERENCES 1.2. Miller, A. A., and Pedley, E., Brit. Med. J . , May 13th, 1950, p. 1115. Idson, B., Chem. Rev., 1950, 47, 383. DOMINION LABORATORY WELLINGTON NEW ZEALAND A. P. OLIVER June, 1951Dec., 19511 NOTES 731 THE DETERMINATION OF SMALL QUANTITIES OF COPPER I N IRON IN a paper’ published in 1937 the determination of small quantities of copper in a plain carbon steel and “Bearco” by internal electrolysis has been described. In the experiments previously described the steel was dissolved in approximately 10 per cent. v/v sulphuric acid (z.e., 10 ml of 96 per cent. sulphuric acid made up to 100 ml with water) in an atmosphere of carbon dioxide under a Contat - Gockel valve. The amount of acid used was equivalent to that theoretically required to dissolve the steel, plus 3 ml of 96 per cent.sulphuric acid in excess. REFERENCE 1. 329 HIGH HOLBORN LONDON, W.C. 1 Fife, J. G., and Torrance, S., Analyst, 1937, 62, 29. J. G. FIFE September, 195 1 THE DETERMINATION OF SMALL QUANTITIES OF RHODIUM I N THE PRESENCE OF LEAD AND ZINC IN the fire assay of rhodium, Allen and Beamishl recommend dissolving the lead button in nitric acid and precipitating the small quantity of rhodium that dissolves in this acid by means of zinc. After the residue has been dissolved in concentrated sulphuric acid and the solution diluted and filtered, the rhodium is finally precipitated with thiobarbituric acid; the complex is ignited and weighed as rhodium metal. As the quantity of rhodium involved is usually very small, it seems that a colorimetric determination would provide a more convenient finish to the assay. Accordingly the possibility of using 2-mercapto-4 : 5-dimethylthiazole for determining rhodium in solutions TABLE I DETERMINATION O F RHODIUM WITH 2-MERCAPTO-4 : 5-DIMETHYLTHIAZOLE AFTER TREATMENT O F SULPHURIC ACID SOLUTIONS WITH SODIUM CHLORIDE Rhodium Experiment taken, Pg Per ml 1 1.33 2 3.33 3 1-33 4 3.33 5 1.33 6 1.33 7 3.33 8 1.33 9 3.33 Rhodium found, r-g Per ml - - 1.35 3-37 1.35 1-33 3-32 1.30 3.32 Optical density 0.187 0.418 0.173 0-428 0-172 0.166 0.418 0.162 0-416 Remarks Rhodium chloride solution containing 10 g As expt.1 Rhodium sulphate solution evaporated with As expt. 3 As expt. 3, but with 100mg of lead and As expt. 3, but with 100 mg of lead present As expt. 3, but with 100 mg of zinc present Rhodium precipitated from lead nitrate solution by zinc, dissolved in sulphuric acid, and procedure applied As expt.8 of sodium chloride sodium chloride 100mg of zinc present of this kind in the presence of small quantities of lead and zinc was investigated. In an earlier paper2 the author has described the colorimetric determination of rhodium in hydrochloric acid solutions with this reagent and shown that small amounts of zinc and lead do not interfere; further tests have confirmed this, provided that the amount of lead present does not exceed the solubility of lead chloride (about 1 g per 100 ml). The dimethylthiazole procedure cannot be applied directly to rhodium solutions containing sulphuric add, but excellent results are obtained if sodium chloride is added and the sulphuric acid solution evaporated until crystals of sodium chloride separate; this procedure converts the rhodium into chloride, and the sodium sulphate formed a t the same time has no effect on the resu Its. PROCEDURE- To the solution containing the small quantity of rhodium dissolved in 5 ml of concentrated sulphuric acid add 50 ml of a 20 per cent.sodium chloride solution and evaporate until crystals of sodium chloride separate. Dilute to approximately 50 ml, add 10 ml of concentrated hydro- chloric acid, and heat to boiling. Add 2 ml of a 0.5 per cent. 2-mercapto-4 : 5-dimethylthiazole732 APPARATUS [Vol. 76 solution for each microgram of rhodium per millilitre expected and boil gently for one hour. Cool, dilute accurately to 100 ml and measure the optical density on an absorptiometer. The results shown in Table I were obtained with a Spekker abs,orptiometer, l.O-cm cells and an Ilford No. 601 filter. The results show that small amounts of rhodium can be determined colorimetrically by 2-mercapto-4 : 5-dimethylthiazole in sulphuric acid solutions that have been evaporated in the presence of sodium chloride and that milligram quantities of lead and of zinc do not interfere. Appreciation and thanks are expressed to Mr. L. S. Theobald, in whose laboratory this work was carried out, for his encouragement, help and patience. The author also desires to record his appreciation to Lord Beaverbrook f o r providing the scholarship, through the Beaverbrook Overseas Scholarship Fund, that enabled this work to be done. REFERENCES 1 . 2. IMPERIAL COLLEGE OF SCIENCE Allen, W. F., and Beamish, F. E., Anal. Chc.tn., 1950, 22, 451. Ryan, D. E., AnaZyst, 1950, 75, 557. DEPARTMENT OF CHEMISTRY LONDON, S.W.7 D. E. RYAN First submitted February, 1951 Amended June, 1951

ISSN:0003-2654    

DOI:10.1039/AN9517600729

出版商:RSC    

年代:1951

数据来源: RSC

摘要:

732 APPARATUS [Vol. 76 Apparatus DETERMINATION OF JELLY STRENGTH OF GLUES AND GELATINS BY THE “BOUCHER” JELLY TESTER THE determination of jelly strength is an important factor in the classification of gelatins and glues. Various types of apparatus have been recommended for this purpose; for example, the B.A.R. tester described by Campbell,’ Lockwood and Hayes’s apparatusZ and the Bloom gelo- meter.3A The Lockwood and Hayes instrument measures the percentage sag of an unsupported gel and the Bloom gelometer is based on the principle of resistance of gels to a force applied at the surface. We required an instrument capable of testing gels of all strengths with equal accuracy at a uniform concentration. During 1932 an apparatus was designed in our laboratories and was constructed in April, 1933; it successfully met these requirements.I t was found to be a very sensitive instrument giving consistent results and a.t the same time being easy to operate. In this instrument a 5-mm depression in the surface of the sample gel is produced by a plunger 13 mm in diameter. The force required to produce this depression is applied by a quantity of water, the volume of which is afterwards measured; each millilitre is interpreted as one “jellogram.” In the B.A.R. tester, a torque is applied to a blade embedded in the gel. OPERATION OF THE INSTRUMENT- The instrument is illustrated in Fig. 1. The tt:st bottle containing the jelly is placed on the table, I, and raised until the plunger, M, just comes in contact with the surface of the jelly and the zero-indicating line is seen t o be in the centre of the microscopic zero finder, L.The needle valve, D, is then opened and water is allowed to run from the container, C, into the water receptacle, B. I t will be seen from the diagram that the weight of the water in this receptacle is conveyed through the beam, A, t o the plunger, M, depressing it into the sample jelly. When the plunger reaches a precise depression of 5 mm the electric points at N make contact and immediately stop the water supply to the water receptacle. This is an extra wide-mouthed glass bottle of the following specification- A standard container has been adopted for the preparation of the jelly. .. .. . . 160ml Capacity . - .. .. Diameter of body, inside . . .. .. .. 66 mm Diameter of body, outside .. .. .. .. 60 mm Height . . .. .. .. .. ,. .. 82 mm I t takes a rubber stopper with a diameter of 42/46 mm. size throughout when comparative tests are to be made. It is essential t o keep t o the standardDec., 19511 APPARATUS 733 To meet the particular needs of the user, any degree of concentration, temperature or pH In the same way any convenient can be adopted for the purpose of comparative tests of gelatins. shape and size of the plunger, i.e., conical, hemispherical or spherical, can be employed. THE STANDARD BOUCHER TEST- To make the standard Boucher test, weigh 5 g of gelatin and transfer without loss to the test bottle. Add 100 ml of distilled water, preferably by means of an automatic pipette, and close the bottle with a rubber stopper.Then place the sample in a water-bath or other suitable compartment a t a temperature of 10" to 15" C and allow it to soak until the gelatin is completely swollen. The soaking time may vary according to the nature of the sample to be tested, but generally it is at least 1 hour for powdered gelatin, 3 hours for coarse and flake gelatin and a t least 6 hours for kibbled and sheet gelatin; the sheet gelatin must be cut into small pieces before- hand. I - I Fig. 1. Diagram of instrument After the sample has been soaked, place the test bottle in a thermostatically-controlled bath of hot water maintained at a temperature not exceeding 65OC for 10 to 15 minutes. In these conditions the temperature inside the test bottle will reach between 60" C and 62" C. Then dissolve the sample by gentle movement and by inverting the bottle several times to obtain uniform solution.Violent agitation of the hot solution must be avoided, as this is detrimental and can cause foaming. Remove the dissolved sample from the bath of hot water and allow it to cool, at a temperature of 15" to 25" C, for about half an hour. Then place it in a thermostatically-controlled chill bath at a constant temperature of 10" i 0.1" C for a period of 16 to 18 hours. The test must be performed within 2 minutes after removal of the sample from the chill bath, as the change in temperature in that time is negligible and has no significant influence on the result. The full nominal range of jelly strength expressed in "jellograms" is from 4 to 500, and the highest result obtained so far is 396.734 APP AF:AT U S pol.76 COMPARATIVE TESTS- comparative tests have been carried out and the results are shown in Fig. 2 (a). and Fig. 2 (b) shows comparative results at this concentration. To establish a relation between our apparatus and the Bloom gelometer, a large number of When Bloom tests give results less than lOOg, a 12-5 per cent. concentration is prepared, 360- 320 280 i . g240- (3200- 160- 120 - - - Fig. 2. Comparisons between Bloom strength and Boucher standard. (a) Bloom at 6-66 per cent. concentration; (b) Bloom a t 12.6 per cent. concentration Any given Bloom result may be calculated from the graphs in Fig. 2 or by means of the following formulae- 2 3 7 3 Bloom at 6.66 per cent. concentration = Boucher x - + 18 Bloom at 12.5 per cent. concentr.ation = Boucher x - + 46 Several years of experience has proved that the method described is advantageous in securing reproducibility and accuracy of results. It also allows a long and continuous series of tests to be made with simplicity and speed of operation on all gelatins and glues a t a uniform concentration. REFERENCES 1. 2. 3. 4. Campbell, L. E., J. SOC. Chem. Ind., 1938, 57, 413~. Lockwood, H. C., and Hayes, R. S., Ibid., 1931, 50, 145~. Richardson, W. D., Chem. Met. Eng., 1923, .28, 551. Bloom, 0. T., U.S. Patent 1,540,979, dated June 9th, 1925. RESEARCH LABORATORY UNION GLUE & GELATINE Co. LTD. GARRETT STREET, GOLDEN LANE LONDON, E.C.1 W. S. KOPROWSKI First submitted June, 1960 Amended June, 1961

ISSN:0003-2654    

DOI:10.1039/AN9517600732

出版商:RSC    

年代:1951

数据来源: RSC

摘要:

Dec., 19513 MINISTRY OF FOOD 735 Ministry of Food CIRCULARS* CIRCULAR MF 19/51 Approved Oxidising and Preservative Agents This circular (price 2d.), dated August 31st, 1951, refers to Circular M F 11/50 (Analyst, 1950, 75, 504) and to Circulars M F 17/50 and MF 5/51 (Analyst, 1951, 76, 321), and gives the name of a further product whose use for the cleansing of milk tankers, vessels or appliances has been approved by the Minister of Agriculture and the Minister of Food, as follows: “BLUE RAM,” AGRICULTURAL GRADE. CIRCULAR MF 21/51 Use of the word “Butter” in the Description of Confectionery This circular (price 2d.), dated November 6th, 1951, draws attention to a Code of Practice that has been agreed between the Ministry of Food and representatives of the Chocolate and Sugar Confectionery trade on the use of the word “Butter,” or a synonym, in the description of chocolate and sugar confectionery, as follows- (1) Where the word “butter” or a synonym is used without qualification in the description of sugar confectionery or chocolate products, e.g., “butter drops,’, “butterscotch,” “butta mint,” the butter-fat content of the products shall be not less than 4 per cent.except where ingredients are used with which butter-fat will not mix, e.g., nuts, or where the product consists of two distinct parts, e.g., a boiled sugar or toffee casing and a chocolate centre; in such cases the butter-fat content shall be not less than 4 per cent. of the part of the product with which the butter-fat is mixed. (2) Where butter-fat is present but in an amount less than the minimum provided for in para. (1) and it is desired to use the word “butter” or a synonym in the description of the goods, the word shall be qualified by the term “flavoured.” CURRENT STATUTORY INSTRUMENTS AND STATUTORY RULES AND ORDERS RELATING TO FOOD The Index of Current Statutory Instruments and Statutory Rules and Orders, Sectional List No. 33, has (See been revised to September 30th, 1951, and may be obtained from H.M. Stationery Ofice at cost of postage. Analyst, 1951, 76, 499.) * Obtainable from H.M. Stationery Office. Italics indicate changed wording.

ISSN:0003-2654    

DOI:10.1039/AN951760735b

出版商:RSC    

年代:1951

数据来源: RSC

摘要:

736 BOOK REV![EWS [Vol. 76 British Standards Institution DRAFT SPECIF [CATIONS A FEW copies of the following draft specifications, issued for comment only, are available to interested members of the Society, and may be obtained on applicittion t o the Secretary, Society of Public Analysts and Other Analytical Chemists, 7-8 Idol Lane, London, E.C.3. Draft Specifications prepared by Technical Commit tee ISE/ 18-Sampling and Analysis of Iron and Steel. CN(1SE) 5442-Draft B.S. Method for the Sampling of Ferro-Alloys. CN(1SE) 5443-Draft B.S. for the Absor ptiometric Determination of Small Amounts of Chromium in Iron and Steel. Draft Specifications prepared by Technical Committee FCC/4--Solvents and Allied Products. CN(FCC1 5174-Draft B.S. for Methyl isoButyl Ketone. CN(FCC) 525GDraft B.S. for Methyl Ethyl Ketone.

ISSN:0003-2654    

DOI:10.1039/AN951760736a

出版商:RSC    

年代:1951

数据来源: RSC

摘要:

Dec., 19511 PUBLICATIONS RECEIVED 739 Publications Received KWALITATIEVE CHEMISCHE ANALYSE. By C. J. VAN NIEUWENBURG and J. W. L. VAN LIGTEN. Third Edition. Pp. 321. Amsterdam: D. B. Centen’s Uitgevers-Maatschappij N.V. 1951. DIE CHEMISCHE ANALYSE. Edited by WILHELM BOTTGER. Band XXXIII. NEUERE MASS- ANALYTISCHE METHODEN. By E. BRENNECKE, K. FAJANS, N. H. FURMAN, R. LANG and H. STAMM. Third Edition. Pp. xx + 347. Stuttgart: Ferdinand Enke, Verlag. 1951. Price DM 46 (paper); DM 49 (cloth boards). CALENDAR OF THE PHARMACEUTICAL SOCIETY OF GREAT BRITAIN, 1951-52. Pp. iv + 308. London: The Pharmaceutical Press. 1951. Price 12s. 6d. GERMAN BOOKS ON CHEMICAL AND COGNATE SUBJECTS, PUBLISHED 1939-50. Second Edition. By A. E. CUMMINS and S. VINCE. London: Lange, Maxwell & Springer Ltd.1951. Free of charge on request. Thefivst edition, 1946, was published by The Chemical Council, London, by whose permission the present publishers have produced a revised and extended edition. Pp. ii + 102. Priced and dated list of titles arranged by subjects.740 PUBLICATIONS RECEIVED [Vol. 76 TEXTBOOK OF PHARMACOGNOSY. By T. E. WALLIS, DSc., F.R.I.C., Ph.C., F.L.S. Second Edition. Pp. xi + 556. London: J. & A. Churchill Ltd. 1951. Price 35s. CHEMISTRY RESEARCH, 1950. Pp. vi + 104. London: H.M. Stationery Office. 1951. Price 3s. 6d. Report of the Chemistry Research Board with the Report of the Director of the Chemical Research Laboratory for the year 1950. AIDS TO PHARMACEUTICAL CHEMISTRY. By W. A. MUTEHAM, M.P.S. Pp. viii + 348. London: Baillikre, Tindall & Cox.1951. Price 6s. 6d. OFFICIAL METHODS OF ANALYSIS OF THE SOCIETY OF LEATHER TRADES’ CHEMISTS. Second Edition. Pp. vi + 200. Croydon: Society of Leather Trades’ Chemists. 1951. Price 17s. 6d. PHARMACOPOEA INTERNATIONALIS. Volume I. Pp. xviii + 406. Geneva: World Health Organisa- tion. London: H.M. Stationery Office; The Pharmaceutical Press. New York: Columbia University Press. 1951. Price 35s.; $5.00; Sw.fr. 20. Available also in French and Spanish editions. T h i s i s Bulletin of the World Health Organisation, Supplement 2. TEXTILE LABORATORY MANUAL. By WALTER GARDNER, MSc., F.R.I.C., F.T.I. Second Edition. Pp. x + 574. London: The National Trade Press Ltd. 1951. Price 30s. PHYSIOLOGY AND PHARMACOLOGY FOR PHARMACEUTICAL STUDENTS. By HAROLD HAYDEN BARBER, B.Sc., Ph.D., F.R.I.C.Third Edition. Pp. x + 622. London: BailliBre, Tindall & Cox. 1951. Price 25s. MICROCHEMISTRY GROUP THE Honorary Secretary of the Group, Mr. D. F. Phillips, has changed his address to- 101 South Promenade, St. Annes-on-Sea, Lytham St. Annes, Lancashire. JOINT MEETING OF THE STIRLINGSHIRE AND DISTRICT SECTIONS OF THE ROYAL INSTITUTE OF CHEMISTRY AND THE SOCIETY OF CHEMICAL INDUSTRY AN official invitation has been extended to members of the Society of Public Analysts and Other Analytical Chemists to attend the Joint Meeting of the Stirlingshire and District Sections of the Royal Institute of Chemistry and the Society of Chemical Industry to be held in the Golden Lion Hotel, Stirling, on Thursday, January 17th, 1952, at 7.30 p.m. At this meeting, Dr. C. L. Wilson of Queen‘s University, Belfast, will be speaking on “Ultramicro Analysis-Past, Present and Future.” ANNUAL GENERAL MEETING Change of Place, Date and Time THE Annual General Meeting of the Society will be held in the Meeting Room of the Chemical Society, Burlington House, on Friday, March 7th, 1952, at 4.30 p.m., followed by the Bernard Dyer Memorial Lecture at 5 p.m., and not as announced on the Programme of Meetings for the Session, 1951-52.

ISSN:0003-2654    

DOI:10.1039/AN9517600739

出版商:RSC    

年代:1951

数据来源: RSC