ISSN:0003-2654    

DOI:10.1039/AN95782FX013

出版商:RSC    

年代:1957

数据来源: RSC

ISSN:0003-2654    

DOI:10.1039/AN95782BX015

出版商:RSC    

年代:1957

数据来源: RSC

ISSN:0003-2654    

DOI:10.1039/AN95782FP049

出版商:RSC    

年代:1957

数据来源: RSC

ISSN:0003-2654    

DOI:10.1039/AN95782BP055

出版商:RSC    

年代:1957

数据来源: RSC

摘要:

APRIL, 1957 THE ANALYST Vol. 82, No. 973 PROCEEDINGS OF THE SOCIETY FOR ANALYTICAL CHEMISTRY JOINT MEETING A JOINT Meeting of the Society with the Fine Chemicals Group of the Society of Chemical Industry was held at 7 p.m. on Friday, March ZZnd, 1957, in the Chemistry Lecture Theatre, King’s College, London, W.C.2. The Chair was taken by the President of the Society for Analytical Chemistry, Dr. J. H. Hamence, M.Sc., F.R.I.C. The following paper was presented and discussed : “Organic Reagents in Inorganic Analysis: Some Recent Developments,” by H. M. N. H. Irving, M.A., D.Phil., F.R.I.C., L.R.A.M. ORDINARY MEETING AN Ordinary Meeting of the Society, organised by the Physical Methods Group, was held at 7 p.m. on Wednesday, April 3rd, 1957, in the meeting room of the Chemical Society, Burlington House, London, W.l.The Chair was taken by the President, Dr. J. H. Hamence, M.Sc., F. R. I .C. The following papers were presented and discussed : “The Spectrometry of Fluorescence,” by E. J. Bowen, M.A., DSc., F.R.S. ; “Some Experiments with Spectrofluorimeters and-Filter Fluorimeters,” by C. A. Parker, B.Sc., Ph.D., F.R.I.C. ; “Spectrofluorimetry,” by Professor R, T. Williams, Ph.D., D.Sc.; “A Direct-reading Fluorimeter,” by L. Brealey, B.Sc., and R. E. Ross, A.M.Brit.1.R.E. NEW MEMBERS ORDINARY MEMBERS Shelagh Maureen Burns, B.Pharm. (Nott.), M.P.S.; Ethel Neil, B.Sc. (Dunelm.) ; Patricia Deloraine Parr-Richard, B.Sc. (Lond.), A.R.C.S. JUNIOR MEMBER Muriel Cessford Gray, B.Sc. (Edin.). DEATH Douglas Arnold Yoxall. SCOTTISH SECTION AN Ordinary Meeting of the Section was held at 7.15 p.m.on Friday, February 22nd, 1957, in the Central Hotel, Glasgow. The Chair was taken by the Vice-chairman of the Section, Mr. A. N. Harrow, A.H.-W.C., F.R.I.C. The following paper was presented and discussed: “Some Recent Developments in Analytical Chemistry,” by R. Belcher, Ph.D., DSc., F.Inst.F., F.R.I.C. AN Ordinary Meeting of the Section was held at 7 p.m. on Wednesday, March 13th, 1957, in the George Hotel, George Street, Edinburgh. The Chair was taken by the Chairman of the Section, Dr. Magnus Pyke, F.R.I.C., F.R.S.E. The following papers were presented and discussed: “Some Aspects of the Estimation of Uronic Acid in Carbohydrate Materials,” by D. M. W. Anderson, B.Sc., Ph.D., -4.R.I.C.; “The Routine Semi-micro Determination of Molecular Weights,” by J.Brooks, M.A., A.R.I.C., and A. F. Williams, B.Sc., F.R.I.C. 217 WE record with regret the death of218 PROCEE.DINGS [Vol. 82 WESTERN SECTION AN Ordinary Meeting of the Section was held a t 6.30 p.m. on Friday, February 22nd, 1957, at the College of Technology, Ashley Down, Bristol. The Chair was taken by the Chairman of the Section, Mr. P. J. C. Haywood, B.Sc., F.R.I.C. A lecture on “The Oxygen Demand of Trade EfAuents with Respect to River Pollution” was given by C. J. Regan, B.Sc., F.R.I.C. A JOINT Meeting of the Section with the South Wales Section of the Royal Institute of Chemistry was held at 6.30 p.m. on Friday, March 15th, 1957, in the Chemistry Lecture Theatre, University College, Swansea. The Chair was taken by the Chairman of the Section, Mr.P. J. C. Haywood, B.Sc., F.R.I.C. A lecture on “Some Recent Developments in Metallurgical Analysis” was given by G. W. C. Milner, M.Sc., A.Inst.P., F.R.I.C. MIDLANDS SECTION AND PHYSICAL METHODS GROUP A JOINT Meeting of the Midlands Section and Physical Methods Group was held at 7 p.m. on Tuesday, February 12th, 1957, in the English Theatre, The University, Edmund Street, Birmingham, 3. The Chair was taken by the Chairman of the Physical Methods Group, Dr. J. E. Page, F.R.I.C. The Meeting took the form of a discussion on “High-frequency Titrations” and the subject was introduced as follows: “Instrumentation,” by J. Allen, A.R.1.C:; “Applications,” by E. S. Lane, B.Sc,, Ph.D., F.R.I.C. MIDLANDS SECTION A JOINT Meeting of the Section with the Birmingham and Midlands Branch of the Royal Institute of Chemistry was held at 7 p.m. on Tuesday, March 5th, 1957, in the Main Chemistry Theatre, The University, Edgbaston, Birmingham, 15.The Chair was taken by the Chairman of the Midlands Section, Dr. R. Belcher, F.Inst.F., F.R.I.C. A lecture on “Thermo-gravimetric Analysis” was given by Professor C. Duval (Paris). MICROCHEMISTRY GROUP THE Thirteenth Annual General Meeting of the Group was held at 6.45 pm. on Friday, January 25th, 1957, in the meeting room of the Chemical Society, Burlington House, London, W.1. The Chairman of the Group, Dr. G. F. Hodsman, A.Inst.P., presided. The following Officers and Committee Members were elected for the forthcoming year :-Chairman- Mr. D. F. Phillips.Hon. Secretary-Mr. D. W. Wilson, Department of Chemistry, Sir John Cass College, Jewry Street, Aldgate, London, E.C.3. Hon. Treasurer-Mr. G. Ingram. Members of Committee-Mrs. D. Butterworth and Messrs. R. A. Chalmers, R. Goulden, G. F. Hodsman, W. I. Stephen and C. L. Wilson. In order to adjust the number of Committee Members retiring each year, two Members, Messrs. Hodsman and Stephen, were elected to serve €or one year only. Dr. L. H. N. Cooper and Mr. H. Childs were re-appointed as Hon. Auditors. The Annual General Meeting was followed by an Ordinary Meeting of the Society, organised by the Group. THE ninth London Discussion Meeting of the Microchemistry Group was held at 6.30 p.m. on Wednesday, February 27th, 1957, in the restaurant room of “The Feathers,” Tudor Street, London, E.C.4. The Chair was taken by Dr. G. F. Hodsman, A.1nst.P. A discussion on “The Micro-determination of Halogens” was opened by F. Oliver and R. Goulden, A.R.I.C. Vice-Chairman-Mr. F. Holmes. BIOLOGICAL METHODS GROUP AN Ordinary Meeting of the Group was held at 6.30 p.m. on Wednesday, March 6th, 1957, in the restaurant room of “The Feathers,” Tudor Street, London, E.C.4. The Chair was taken by the Chairman of the Group, Dr. S. K. Kon, F.R.I.C. The Meeting took the form of a discussion on “The Experimental Assessment of Tran- quillisers,” which was opened by A. Spinks, B.A., B.Sc., Ph.D., D.I.C.

ISSN:0003-2654    

DOI:10.1039/AN9578200217

出版商:RSC    

年代:1957

数据来源: RSC

摘要:

April, 19571 GAGE 219 The Potentiometric Titration of Weak Acids and Bases in Dilute Aqueous Solution* BY J. C. GAGE The limitations in the application of the Henderson equation to the titration of weak acids and bases is discussed. It is shown that the graphical method of finding the concentration, c, and the (Brransted) dissociation constant, K , may be given a slightly enlarged range of usefulness by means of the exact form of the Henderson equation. Of more general application is the transformation of the titration curve to a linear form; this permits the determination of c and K when one or both end-point inflexions are obscured, provided that c is not less than K or K,/K. Examples are given from the titration of chlorophenols at a concentration of 10-4 M, singly and in admixture, and of certain alkaloids.THE usual procedure for the quantitative and qualitative determination of weak acids and bases by means of potentiometric titration is based on the Henderson equation, which is presented in equation (1) in the form for the titration of a weak acid with a strong alkali; c is the concentration of the acid, m the concentration of alkali, and pK is the negative logarithm of the (Brmsted) dissociation constant, K. This equation is an approximation and involves the assumption that the hydrogen-ion concentration, h, and the hydroxyl-ion concentration, Kw/h, are negligible in comparison with m; in the exact equation m must be replaced by (m + h-Kw/h). According to equation (l), the titration curve should be asymptotic to m = 0, m = c, but at high and low pH values the assumption that h and K,/h are negligible is no longer valid and the ends of the curve are flattened, the asymptotes being replaced by points of inflexion.In the application of equation (1) to the potentiometric titration of weak acids, it is assumed that these points of inflexion correspond to the end-points of the titration, and the pH of the point half-way between them gives the pK value of the acid; a precise mathematical study of the potentiometric-titration curve, which is at present being prepared for publication, reveals that none of these assumptions is entirely justified and that the systematic error in making them increases with increasing dilution. Equation (1) does not suggest that the scope and precision of potentiometric titration should be in any way affected by the value of pK, or by the dilution provided that the precision in determining the ratio m/c is not thereby affected. The experimental difficulty in the determination of pK and c of weak acids and bases at low concentration is due to the deviation of the titration curve from that expected from equation (1).The flattening of the curve becomes more pronounced as the hydrogen and hydroxyl-ion concentrations approach m ; as the dilution increases or as the pK vzlue departs from 7, one or both end-point inflexions become progressively less well defined, resulting in a progressive loss of precision until a point is reached at which the inflexion is PO longer visible. This is demonstrated in the titration curves for $-chlorophenol and 2 : 4-dichlorophenol shown in Fig.1, curves A and B, at a concentration in the region of M ; the weak inflexion on curve B appears as a shallow peak on the differential curve of d(pH)/dm against m, while no inflexion is apparent in curve A, nor any peak in the corresponding differential curve. If two acids are present in the solution titrated, it may be impossible to measure graphically their separate con- centrations, as is shown in Fig. 6 (p. 226) for a mixture of 2:4:5-trichlorophenol and 2 : 3 : 4 : 6-tetrachlorophenol, each approximately M , although the pK values of these two phenols differ by 0.76 unit. The theory of potentiometric titration of acids and bases, singly and in admixture, has been studied by Auerbach and Smolczykl and has been further extensively developed * Presented a t the XVth International Congress on Pure and Applied Chemistry (Analytical Chemistry), Lisbon, September 8th to 16th, 1956.220 GAGE : THE POTENTIOMETRIC TITRATION OF WEAK ACIDS [Vol.82 by Ricci,2 who includes a mathematical treatment of the feasibility of a titration. These authors do not, however, consider what information can be extracted from a titration curve when the graphical method is impossible or can be shown to lead to unacceptably large errors. I 0 1 1 - Fig. 1. Potentiometric-titration curves: curve A, pH - m for p-chlorophenol; curve B, pH - nz for 2 : 4-dichlorophenol ; curve C, pH - x for p-chlorophenol; curve D, pH - x for 2 : 4-dichlorophenol The precision of the potentiometric titration of weak acids and bases, particularly of those with a low aqueous solubility, may be increased by substituting a suitable polar organic solvent, such as acetic acid or pyridine, for water. Although this procedure may permit an accurate quantitative determination, it is of limited value when the substance to be investigated is only available in aqueous solution.Moreover, it is not possible to calculate from a titration in a non-aqueous solvent the dissociation constant in aqueous solution, The importance of a knowledge of the dissociation of acids and bases under physiological conditions, in the study of the relation between chemical structure and pharmacological action, has been emphasised by many investigators, particularly by Albert .3 During a comparative study in these laboratories of the biological action of a series of chlorinated phenols, it became evident that a knowledge of their dissociation constants would be desirable.Although figures for all of the substituted phenols under investigation had been published, the methods used varied widely, and an attempt was made to develop a standard procedure that could be applied to the whole series. Potentiometric titration is an attractive method, as it is simple to perform and should be applicable to substances of unknown purity, but the wide range of acidities of the series of chlorinated phenols, and the low solubility of some of its members, necessitated a re-investigation of the mathematical treatment of the results of pot en t iomet ric tit rat ion. THEORETICAL TREATMENT TITRATION OF SINGLE ACIDS- The general equation for the titration of a weak acid by a strong alkali is given by (2). This may more conveniently be expressed in the form (3), where m + h - K,/h = x.Equation (3) may be written as (4), which has the same form as (l), but it is an exact and h c-m-h + K,/h K - -h-K,/h .. * - (2) .. . . _ - h C-x . . * - (3) .. .. .. .. K= xApril, 19571 AND BASES IN DILUTE AQUEOUS SOLUTION 221 not an approximate equation, and, if the pH values from a potentiometric titration are plotted against x instead of against m, the curve obtained is asymptotic to x = 0, x = c. In the titration of a single acid the value of x cannot exceed c, and can only be made to approach this value when m is in excess of c and it is, therefore, necessary to add more than one equivalent of alkali to the acid in order to obtain the end-point on the pH - x curve.Similarly when m = 0, x = h-Kw/hJ and in order to obtain the lower end of the curve it is necessary to titrate the weak acid with a strong acid. This is shown in Fig. 1; in curve B positive values of m correspond to the concentration of sodium hydroxide and negative values to the concentration of hydrochloric acid. It is evident from curve D, which has been derived from curve B, that the end-point inflexion on the pH - x curve is much more clearly defined than on the pH - m curve. TITRATION OF MIXTURES OF ACIDS- If more than one acid is present, the inflexions at the end-points of the pH - x curve may not be clearly defined, interfering with the accurate determination of c and pk'.The general equation for the titration of a mixture of weak acids by a strong alkali is given in equation (5), in which c,, c2, c3 .... and K,, K,, K3, .... are the concentrations and dissociation constants of the component acids. Equation (5) does not permit a formal solution unless certain simplifications are possible. If K , + h, and K,, K , ....< h, all the terms on the right-hand side of ( 5 ) , except for the first, vanish, leaving an equation that may be converted to (3). As the presence of the other acids may prevent the accurate graphical evaluation of c1 and pK, from the titration curve, equation (3) may be converted to the linear form (6), from which it is evident that the values of 1/K, and c1 may be found by plotting xlz against x.Fig. 2 shows such a plot of xh against x derived from the values of pH and x for 2 : 4 : 5-trichlorophenol presented in Fig. 3, curve A. "I I 2 Fig. 2. trichlorophenol xh - x curve for 2 : 4 : 5- If h h K,, and K,, K3 ...> h, equation (5) reduces to (7), which may be written in the If now xh is plotted against x , a straight line is not obtained and this has been form (8).222 GAGE : THE POTENTIOMETRIC TITRATION OF WEAK ACIDS [Vol. 82 t I 0.5 1.0 I r x ~ ~ 5 -.- I 2 P 0.5 1-0 x ~ 1 0 3 Fig. 3. pH - x curves : curve A, 2 : 4 : 5-trichlorophenol; curve 2 : 3 : 4 :~&tetrachlorophenol; curve C, quinine; curve D, brucine + c2 + ca ...... . . .. . . x = C1K1 K , + h . . .. . . *O B, observed with most of the chlorophenols studied, indicating their lack of purity.Equation (8) cannot be directly converted to a linear form, but after differentiation (9) can be re- arranged to give (lo), in which x = d-dh/dx. It can be shown that the curve obtained by plotting z against h. from the potentiometric titration of a mixture of two acids is ____- . . . . . . --=-- dx ClKl dh ( K , + h)2 zdc,K, = Kl $- h . . . . . . .. . . (10) compounded from the intersection of the lines of each acid separately and may permit the separate determinations of the two concentrations and dissociation constants, provided that each component is present as a sufficiently large proportion of the total, and the constants are sufficiently widely separated. The several dissociation constants of a polybasic acid or polyacid base may be determined in a similar manner.TITRATION OF BASES- The mathematical treatment of the titration of a weak base by a strong acid is similar to that outlined above. The dissociation of the base is regarded as that of the cation, BH+ + B + H+, and no distinction is made between the constants of acids and bases. In the following equations x’ = [A-] + Kw/h--h, where [A-] is the concentration of strong acid used in the titration; alternatively, if it is preferred to add an equivalent or excess of strong acid to the weak base, followed by back-titration with strong alkali, x’ may be replaced by [A-]--x, where x has the previously assigned definition. For the titration of a single weak base, equation (4) is transformed to (ll), while for a mixture of bases, when Kl fi h and K,, K,, .....> h, equation (6) becomes (12).The differ- ential equation (10) is unchanged and applies to both acids and bases.April, 19571 AND BASES IN DILUTE AQUEOUS SOLUTION 223 . . .. . . (11) xf K, + x' = c1 . . . . .. .. EXPERIMENTAL Into the titration cell, A, dips a glass electrode, B, and the saturated-calomel half-cell, C, is connected through the hair capillary, D. The junction of the saturated potassium chloride solution at D was freshly made before each titration by allowing the solution to enter A from the reservoir E; this was achieved by rotating E until the ridge F coincided with the hole G in the cone of the standard taper joint, as shown in the diagram. During the titration, E was rotated to isolate the reservoir. The potentiometric titrations were performed in the apparatus shown in Fig.4. Fig. 4. Titration assembly: A, titration cell; B, glass electrode; C, saturated-calomel half-cell ; D, hair capillary; E, reservoir; F, ridge; G, hole; H, funnel; I, tube; J, three-way tap; K, glass syringe Liquid or washing water may be introduced into the cell through the funnel, H, and the cell is emptied by applying suction to tube I. This tube carries a three-way tap, J, the second arm of which is used to introduce a stream of pure nitrogen into the cell, A, to mix the contents during titration and to protect against the ingress of atmospheric carbon dioxide. The titration cell was immersed in a water bath at 25" C, and the nitrogen, before entering the cell, was bubbled through water maintained at the same temperature. The titrant was 0.1 N sodium hydroxide for the chlorophenols and N sodium hydroxide for the alkaloids; it was contained in the glass syringe, K , to which is fused a thick-walled glass capillary tip.The solution was expelled from the syringe by means of an Agla micro- meter head, not shown in the diagram. This arrangement permits accurate additions of 0-002 ml; 0.1 N sodium hydroxide added in this manner to 40 ml of solution is equivalent to a concentration increment of 5 x All were commercially available materials, and with the exception of pentachlorophenol, which was recrystallised before use, were not subjected to purification. M ethanolic solution was prepared from each of the chlorophenols, and 0-1-ml portions of these solutions were added to 40ml of water to give approximately 10-4 M concentrations for titration.For the alkaloids brucine and quinine M . The substances titrated are listed in Table I. A 4 x2 24 GAGE : THE POTENTIOMETRIC TITRATION OF WEAK ACIDS [Vol. 82 4 x 10-3 M aqueous solutions were prepared with two equivalents of mineral acid, and 10-ml portions were diluted to 40 ml with water to give 10" M concentrations. TABLE I CALCULATED pK VALUES COMPARED WITH PREVIOUSLY PUBLISHED FIGURES (ALL MEASUREMENTS AT 25" C UNLESS OTHERWISE STATED) Substance Calculated pKS Value from literature o-Chlorophenol . . .. .. .. 8.65 8.4ga, 7-44b m-Chlorophenol . . .. .. .. 9.12 8 ~ 8 5 ~ , 9.070 p-Chlorophenol . . ,. . . , . 9.37 9 ~ 1 8 ~ 2 : 4-Dichlorophenol . . .. . . 7-86 7 ~ 7 4 ~ ~ 7-5lb, 7.W 2 : 6-Dichlorophenol .. .. .. 6-91 6.Sa 2 : 3 : 6-Trichlorophenol . . .. .. 5.98 6 ~ 1 3 ~ 2 : 4 : 5-Trichlorophenol . . .. .. 7.07 7.0d 2 : 4 : 6-Trichlorophenol . . 9 . .. 6-22 6*41b, 6-Xd 3 : 4 : 5-Trichlorophenol . . .. .. 7-83 8.3ljb 2 : 3 : 4 : 6-Tetrachlorophenol . . .. 5.46 5 ~ 1 4 ~ , 5 ~ 3 ~ Pentachlorophenol . . .. .. 5-00 4.Sd, 526b Quinine pK1* . . .. .. .. 4.13 4.15(20" C)" PKZS - - I . .. .. 8-52 8.3(2O0 C)e Brucine pK, . . . . .. .. 8.28 7-96f * K, and K , are the dissociation constants of the cations BH2++ + BH+ + H+ and BH* + B + H+. t These pK values have been derived by potentiometric titration of samples of unknown purity. a. Murray and G~rdon.~ b. Tiessen~.~ c. Hodgson and Smith.6 d. Blackman, Parke and G a r t ~ n . ~ e.Christophers.* f. Kolthoff.a The measurements of pH value were made with a Cambridge pH meter. This was standardised with phthalate and borate buffers in the usual way, and then as a check 25 ml of 8 x 10-4 M potassium hydrogen phthalate were titrated with 0.1 N sodium hydroxide; if the mid-point of the titration deviated from pH 5.42, the pK value calculated from the accurate determinations of Hamer and Acree,lo an appropriate correction was applied to the subsequent measurements, TABLE I1 CALCULATION OF c AND pK Duplicate determinations are shown on each substance 7- inspection from pH-m Substance curves 2 : 4-Dichlorophenol . . . . 1.20 1.10 2 : 6-Dichlorophenol . . . . 1.08 1.12 2 : 4 : 5-Trichlorophenol . . . * 1.10 2 : 4 : 6-Trichlorophenol . . . . 1.10 3 : 4 : 5-Trichlorophenol .. , . 1.00 1.10 1.12 1-00 10% by-- inspection from p H - x curves 1.19 1.12 1.08 1.15 1-12 1.14 1-12 1.14 1-05 1.02 -l means of equation (10) 1.13 1.03 1-16 1.18 1.11 1-12 0-86 0.88 1.01 1.04 7 inspection from curves 7-85 7.85 6-93 6-87 7-07 7.02 6.24 6.19 7.92 7.94 pH-m PK by- n inspection from curves 7.86 7.83 6.9 1 6-91 7.10 7.07 6-23 6.20 7.93 7.94 p H - x -- means of equation (10) 7.93 7.94 6.84 6.87 7.05 7.07 6-24 6.2 1 7.82 7.84 RESULTS Table I1 shows values of pK and c derived by inspection of the pH - m curves, by inspection of the pH - x curves and by means of equation (10). Only the values for two dichlorophenols and three trichlorophenols are included in this Table ; with the remainder of the chlorophenols studied, one end-point was unobtainable from the pH - m curve, and was subject to considerable uncertainty on the pH - x curve.April, 19571 AND BASES IN DILUTE AQUEOUS SOLUTION 225 From the pH - x curve for 2 : 4 : 5-trichlorophenol, shown in Fig.3, curve A, has been derived the plot of xh against x shown in Fig. 2. From the parameters of this straight line, c and pK have been calculated to be 1.13 x lo4 M and 7.09 by equation (6); these values agree well with those in Table 11. In the derivation of c and K by means of equation (lo), the value of z2 may be obtained graphically from a plot of h. against x, but more simply and with no loss of accuracy by deriving Ah and Ax from the differences between alternate values of these variables. This is illustrated in Table 111, which shows the stages in the calculation of z from some of the measured values of pH and m for 2 : 4 : 5-trichlorophenol.Four examples of h - x curves are given in Fig. 5; curves A and B are derived from the pH - x curves A and B in Fig. 3 for 2 : 4 : 5-trichlorophenol and 2 : 3 : 4 : 6-tetrachlorophenol, and curves C and D represent the two titration steps of the pH - x curve for quinine in Fig. 3, curve C, from which the two dissociation constants may be calculated. The values of pK, calculated by means of equation (lo), for all of the substances examined are listed in Table I, together with pK values taken from the literature. ZXI o2 ZXlO Fig. 5. h - z curves: curve A, 2: 4: 5-trichlorophenol; curve B, 2 : 3 : 4 : 6-tetrachlorophenol; curves C and D, quinine Fig.6 shows the pH - x curve for a mixture of 2 : 4 : 5-trichlorophenol and 2 : 3 : 4 : 6- tetrachlorophenol. each Fig. 7, curve B, is the plot of h against z for this mixture; the lower end of this curve has been plotted with extended co-ordinates in Fig. 7, curve A. M ,226 GAGE : THE POTENTIOMETRIC TITRATION OF WEAK ACIDS [Vol. 82 TABLE I11 CALCULATION OF 2 = 4- nh/nx FROM pH AND VZ FOR 2 4 : 5-TRICHLOROPHENOL __- PH 1 0 5 ~ 108h 105x .-108Ah 106Ax z = lOa2/-Oh/Ax 6-75 3.5 18.2 3-52 6.83 4.0 14.8 4.01 5.9 9.9 7.7 6.9 1 4.5 12.3 4-5 1 4.6 10.0 6.8 6-99 5.0 10.2 5-0 1 3-8 9-9 6-2 7-14 6.0 7.25 6.00 2.35 9.9 4.85 7-21 6.5 6.15 6-40 2.1 9.8 4-65 7.29 7.0 5.15 6.98 2.0 9.9 4.6 7.38 7.5 4.15 7.48 The close resemblance between curves A and B of Figs. 5 and 7 will be apparent; the pK values calculated from Fig.7, curves A and B, differ by less than 2 per cent. from those derived from Fig. 5, curves A and B, while the concentrations are less accurate, being within 15 per cent. of the expected value. I01 4- 1:5 2:o 0 4 5 Fig. 6. pH-x curve for a mixture of 2 : 4 : 5-trichlorophenol and 2 : 3 : 4 : 6-tetrachloro- phenol DISCUSSION 017 RESULTS Potentiometric titration has advantages over other physical procedures for the deter- mination of concentrations and dissociation constants of weak acids and bases, in that it may yield the required results from impure solutions about which no other information is available. The conventional graphical method based on the Henderson equation (1) may be used if the concentration of hydrogen and hydroxyl ions is negligible in comparison with the concentration of strong acid or base used in the titration, but if this condition is not fulfilled a serious error may be introduced into the assessments of the end-points of the titration and the point of half neutralisation.The close agreement in Table I1 between t:he values of c and pK derived by the usual method from the titration curve, and those from the exact pH - x curve, exists only because substances with more extreme values of pK are excluded from this Table by the impossibility of directly defining by inspection both end-points of their curves. A comparison between the pH - m and pH - x curves for @-chlorophenol in Fig. 1, curves A and C, indicates that in the region of half neutralisation the former curve has a pH value about 0.2 units lower than the latter.Although the error implicit in the Henderson equation should be avoided by means of the exact form (4), it has been found in practice that this transformation of the titrationApril, 19571 AND BASES I N DILUTE AQUEOUS SOLUTION 227 curve does not greatly extend the graphical method of solution ; when h or Kw/h approaches or exceeds m in value, the evaluation of x is subject to considerable errors arising from the magnification of minor errors in the standardisation of the pH meter, or of minor deviations from linearity of response. This uncertainty in the location of the end-point may be avoided by converting the general equation to a linear form, and this has been effected by equations (6) and (10); the latter equation is of more general application and several examples of its use are presented here.As (10) is a differential equation, the experimental plot cannot be expected to yield the required parameters with as much accuracy as would the Henderson equation when this is admissable. Nevertheless, the graphs shown in Fig. 5 demonstrate that the scatter of the points is not so large as to involve an unacceptably large error in defining the slope and position of the straight line and, therefore, in the calculated values of c and K. zxi o2 Fig. 7. h - x curves derived from Fig. 6 The reasonably close agreement shown in Table I1 between the duplicate values of c and pK determined by equation (lo), and a comparison'of these values with those obtained by inspection of the pH - x curves, suggests that no serious error will be involved in those values obtained by equation (10) for which the inspection method is not available, provided that this method is not applied outside the limits prescribed below.The low value of c for 2 : 4 : 6-trichlorophenol obtained by equation (10) suggests that this substance contains a titratable impurity, probably another chlorinated phenol, which cannot be distinguished from the main component by inspection of the titration curve. The h - x curve gives an indication of such an impurity, but does not permit its pK value to be ascertained. In general, the derivation of K from the h - x curve, which is obtained directly from the intercept on the h axis, is subject to a smaller error than is the calculation of c, which necessitates the squaring of the intercept on the x axis.It has been shown by Ricci2 and others that two end-point inflexions in the titration of a weak acid occur only when c is not less than 27K or 27Kw/K. This consideration does not necessarily invalidate the method of interpreting titration curves described above ; nevertheless, when K is greater than c in magnitude, it is necessary for h or Kw/h to be in excess of m in order to obtain the useful part of the pH - x curve. This is attended by a considerable error, and in practice it has been found that the treatment is impracticable if the value of c is less than K or K,/K. A demonstration of this effect is to be found in the pH - x curves for quinine and brucine in Fig. 3, curves C and D.The pK, value of quinine, which has been calculated from Fig. 5, curve C, to be 4.13, is of a magnitude that permits the calculation to be made at a concentration of lo3 M.. The lower part of the brucine curve in Fig. 3, curve D, shows x almost invariant with pH, and no information on the pK, value can be made, apart from the statement that it must be of the order of 2 or less.228 SMYTHE : INTEGRATED-CURRENT SOURCE FOR [Vol. 82 As the application of equation (10) demands a measurement of ah, it is essential that the pH meter assembly used should be properly calibrated and also adequately stable and sensitive. The titration curves in Fig. 1 and 3 show that the useful pH range of the titration is less than 2 pH units; as it is desirable to take at least 20 measurements within thisrange, the pH meter should be capable of accurately measuring a pH difference in the region of 0.05 unit. Careful and patient technical assistance was provided in this investigation by Miss Sylvia Morrissey . REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. Auberbach, F., and Smolczyk, E., 2. physik. Chem., 1924, 110, 65. Ricci, J. E,, “Hydrogen Ion Concentration,” Princeton University Press, New Jersey, 1952. Albert, A., “Selective Toxicity,” Methuen & Co. Ltd., London, 1951. Murray, J. W., and Gordon, N. E., J . Anzer. Chem. SOC., 1935, 57, 110. Tiessens, G. J., Rec. Trav. Chim. Pays-Bas, 1929, 48, 1066. Hodgson, H. W., and Smith, R., J . Chem. Soc., 1939, 263. Blackman, G. E., Parke, M. H., and Garton, G., Arch. Biochem. Biophys., 1955, 54, 45. Christophers, S. R., Ann. Trop. Med., 1937, 31, 43. Kolthoff, I. M., Biochem. Z., 1925, 162, 289. Hamer, W. J., and Acree, S. F., J . Res. Naf. Bur. Stand., 1945, 35, 38. IMPERIAL CHEMICAL INDUSTRIES LIMITED THE FRYTHE, WELWYN, HERTS. INDUSTRIAL HYGIENE RESEARCH LABORATORIES September 27th, 1956

ISSN:0003-2654    

DOI:10.1039/AN9578200219

出版商:RSC    

年代:1957

数据来源: RSC

摘要:

228 SMYTHE : INTEGRATED-CURRENT SOURCE FOR [Vol. 82 Integrated-current Source for Automatic Coulometric Titrations* BY L. E. SMYTHE A simple and inexpensive integrated-current source for automatic coulometric titrations has been developed. The unit is suitable for use with a commercial automatic pH titrimeter and may be used with wide variations in mains voltage. “Fast” or “slow” coulometric titrations may be carried out with a choice of six current ranges (5 to 100 mA) for titrations up to or near to the end-point, and a “slow” 1-mA range for use when approaching the end-point. The “fast” and “slow” outputs from the automatic pH titrimeter provide for automatic changing from fast to slow coulometric titrations and stopping a t the selected end-point. The utility of the equipment is demon- strated by the determination of chromium in dichromate solutions by coulo- metric titration with electro-generated ferrous ions, use being made of potentio- metric end-point detection.The unit should prove suitable for most other coulometric titrations and could be adapted for amperometric and spectro- photometric end-point detection. COULOMETRIC titrations now form a well established section of electro-analysis and the rapidity of development may be gauged from the publication of some eighty papers in the past 2 years. The principles of coulometric analysis and details of earlier and more recent methods are given by Linganel and Delahay,2 and recent apparatus and methods are reviewed by De Ford.3 The coulometric method is of great importance in the field of automatic analysis, principally because of the ease of interpretation and utilisation of electrical and time-based phenomena with which it is concerned. The electronic control equipment may be set up some distance from the coulometric cell, and automatic coulometric titrations are thus well suited for the determination of constituents of certain active solutions.An example is the coulometric titration of chloride and of chromium in homogeneous reactor-type solutions .4 A disadvantage of coulometric titration procedures for every-day laboratory use has been the often elaborate and expensive equipment6ye required for close control of currents * Presented at the XVth International Congress on Pure and Applied Chemistry (Analytical Chemistry), Lisbon, September 8th to 16th, 1956.April, 19571 AUTOMATIC COULOMETRIC TITRATIONS 229 of the order 1 to 100 mA to within Ifr 0.01 per cent.Because of the limitations of the constant- current method, in which the number of coulombs is calculated from the product of current and time, some attention has been devoted to the development of a precise direct-reading coulometer. Parsons, Seaman and Amick’ have described a current source in which a low-inertia integrating motor fitted with a counter is used in conjunction with a 70-ohm series resistor immersed in transformer oil. The current source was useful for currents of 150 to 250 mA, and current levels of 12.5 to 13.7 mA were also studied. This circuit does not provide great flexibility as regards choice of current range, and for some current ranges the 24-volt motor is operated at a reduced input voltage owing to the resistor circuitry.There is a slight deviation from linearity of the motor-speed - voltage curve, which becomes greater than 0.5 per cent. when the motor is operated at an input voltage below 5 volts. The integrated-current source described in this paper obviates the difficulties mentioned above, is versatile and inexpensive and is suitable for operation from a 230 to 240-volt 50 cycles per second a.c. supply. EXPERIMENTAL DETAILS OF APPARATUS- The integrated-current source for automatic coulometrk titrations incorporating two low-inertia integrating motors provides the following facilities- Choice of six current ranges of 5, 10,20, 50, 75 and 100 mA for automatic or manual titrations up to or near to the end-point.A “slow” l-mA current range for automatic or manual coulometric titrations ; used when titrating microgram amounts of the constituent, or when approaching the end-point. Use of the “fast” and “slow” outputs from a commercial automatic pK titrimeter provides for automatic changing from “fast” to “slow” coulometric titrations and stopping at the selected end-point. Selected switching of 100 watt wire-wound resistors operates the integrating motors at & 10 per cent. of their specified input voltage for all current ranges. A simple d.c. power unit incorporated in the apparatus coupled with the efficient integrating property of the motor makes the equipment suitable for use with variation in the a.c. mains voltage (nominal 230 volts) of 20 per cent.over periods up to 10 minutes. Under these conditions the standard deviation for the factor on the 10-mA range was & 0-03 per cent. Fig. 1 shows the essential circuitry of the integrated-current-source for automatic coulometric titrations. Fig. 2 shows the complete equipment for automatic coulometric titrations, comprising an integrated-current source, a coulometric cell for the electro-genera- tion of titrant and the automatic pH titrimeter. Magnetically operated stirring is employed with the titration cell. The six switch positions on the integrated-current source, corresponding to the above- mentioned current ranges, were calibrated in terms of milli-equivalents per count of the “fast” motor. The l-mA range was also calibrated in terms of milli-equivalents per count of the “slow” motor.The calibrations were carried out by using for current measurement an oil-filled standard l-ohm resistor (H. Tinsley, type 1659,3-ampere rating, R = 1.00008 ohm at 20” C) and a high-precision potentiometer (Gambrell, type 12244/1). Counts were recorded (& 0.01 count) for periods of 600 seconds (timed with a calibrated electric permanent magnetic operated stop clock registering to 1/100th second over 12 minutes (Synchromatic Time Recording Co., model 310/0407). Current was measured at 0, 300 and 600 seconds, and averaged. The switch factors were then calculated from the expression- current (amperes) x time (seconds) 96.492” x counts = milli-equivalents per count. The cell used for coulometric titrations with electro-generated ferrous ions is essentially that of Cooke and Furman,S of capacity 50 ml and a scaled down 10-ml version.Provision was made for either tungsten - platinum or saturated-calomel electrode - platinum potentio- metric end-point detection. Although provision was made for titrations in an inert atmosphere * 96,492 coulombs = 1 chemical equivalent (Craig and Hoffman8).230 SMYTHE INTEGRATED-CURRENT SOURCE FOR [Vol. 82April, 19571 AUTOMATIC COVLOMETKIC TITRATIONS 23 1 (with sub-milligram amounts of the constituent to be determined), it was found that carrying out blank determinations obviated this need. The blank determinations also made allowance for traces of impurities in reagents and any slight positive or negative bias in end-point detection.The automatic pH titrimeter (model 24) was manufactured by Electronic Instruments IAd., Richmond, Surrey, and was capable of end-point settings within the ranges pH 3 to 11 and + 400 to - 800 mV. The titrimeter also provided (i) two separate input sockets, so that while one titration is being performed another can be set up, and (ii) “fast” to “slow” change over within the range 0 to 300 mV (as selected) in advance of the end-point. The titrimeter balance point is sensitive to 3 mV and the change in e.m.f. at the end-point, for the determinations studied, was usually 250 to 350 mV during a fraction of a count. The shape of the e.m.f. - count curve9 for end-point detection should be determined for the particular concentration range of constituent by manual incremental titration with the potentiometric electrode system coupled to the terminals of a direct-reading pH meter. -4lternatively, the course of such a titration may be followed by noting the balance point on the titrimeter itself after various increments (counts). Typical titrimeter settings for the titration of 1 mg of chromium in ;t cell volume of 10 ml were: cnd-point setting, - 100 mV; function switch set to MV RISING; “fast - slow change-over” setting, 80 mV; tungsten lead connected to glass-electrode terminal ; platinum lead corinected to reference-electrode terminal. The integrated-current supply was operated with the APPROX.MA switch on 5 mi2 and the control switch on AUTOMATIC. For the determination of sub-milligram amounts the whole titration should be carried out on the 1-mA range.A%lJTOMATIC COULOMETKIC TITRATIONS--- Various trials were carried out in order to demonstrate the utility of the equipment for automatic coulometric titrations. The determination of chromium in standard potassium dichromate solution, by titration with electro-generated ferrous ions, was chosen because potassium dichromate is a reliable primary standard and the volumetric titration with ferrous solutions is long-established. The determination of chromium in steels and cerium in the presence of uranium and lanthanum with electro-generated ferrous ions was tried, m d the accuracy of determination was comparable with that for chromium in dichromate. DETERMIXATION OF CHROMIUM I N STANDARD POTASSIUM DICHROMATE SOLUTIOSS -- For this determination the following solutions were prepared- Fevric alum solution-Prepared by dissolving 290 g of analytical-reagent grade ammonium ferric sulphate, (NH,),S0,.Fe,(S04),.24H,0 in 200 ml of distilled water containing 20 ml of concentrated sulphuric acid, sp.gr.1-84, in a 1-litre beaker. Then 90 ml of concentrated sulphuric acid were added and the solution was diluted to approximately 1300 ml; 30 ml of 100-volume hydrogen peroxide were added and the solution was warmed on a hot-plate at 50” to 70” C for (This procedure ensured that the solution was free from ferrous ions and was tnore satisfactory than that described by Cooke and F~irman.~) When the solution had cooled, it was passed through a sintered filter and diluted to 1 litre.Stmzdavd dichromate solutions-Three 0.1000 AT standard potassium dichromate solutions were prepared for cross-checking in the trials of the integrated-current supply. Andytical- reagent grade potassium dichromate was recrystallised and, after a preliminary drying, the crystals were ground to a fine powder iii an agate mortar and dried at 140” to 1,550” C to constant weight. The required weight was then used to prepare the dichromate solutions. Dilution of the 0.1 IV solutions was used for the preparation of 0.01 N and more dilute solutions. Add 2 ml of 18 Y sulphuric acid and then 5 to 30 ml of the ferric alum solution. Use about 5 ml of the ferric alum solution for sub-milligram amounts of chromium (1 or 5-mA current ranges), 10 ml for current ranges of 10 to 50 mA and 20 to 30 ml for current ranges of 70 to 100 mA, depending on the chromium content of the solution.Dilute the solution sufficiently to cover the electrodes after inserting a polythene-covered stirrer. To carry out the titration, set the titrimeter as previously described, record the “slow” and “fast” motor readings and turn on the titrimeter AUTOMATIC switch. On completing the titration, again record the motor hour or until the evolution of oxygen ceased. The procedure is as follows- By pipette place the solution to be aiialysed into the coulometric titration cell.232 SMYTHE : INTEGRATED-CURRENT SOURCE FOR [Vol. 82 readings. The titrimeter and integrated-current supply should be switched on 10 minutes before the titration is commenced and set on the instrument the required current range.This permits the resistors to attain normal working temperature. The results are shown in Table I. TABLE I DETERMINATION O F CHROMIUM IN POTASSIUM DICHROMATE Chromium found No. of Relative mean Cell volume, Chromium taken, (average), determinations error, 40 17-34 11.32 4 - 0.2 10 3.47 3.46 7 - 0.3 10 0.1734 0.1726 4 - 0.5 ml mg mg % 10 0.01734 0.0183 5 + 5.0 CONCLUSIONS The results show that good accuracy can be attained with the integrated-current source for automatic coulometric titrations, in the milligram and sub-milligram range down to a cell volume of 10ml. The methods offer advantages in the automatic determination of elements in active solution^.^ From a consideration of this work and other work with micro- coulometric equipment,lOJ1 ,l2 the accuracy of the method (with concentrations below 0.1 mg of the element in a cell volume of 10 ml) should be capable of further improvement, by using suitable micro-cells with provision for de-gassing.The equipment described should be suitable for the majority of coulometric titrations. The assistance of J. C. Boag with circuit design is gratefully acknowledged. APPENDIX LIST OF COMPONENTS USED I N THE CONSTRUCTION OF THE INTEGRATED-CURRENT SOURCE FOR AUTOMATIC COULOMETRIC TITRATIONS (Fig. 1) = 10-ohm, l-5-watt, wire-wound vitreous-enamelled resistance. = 33,000-ohm, 10-watt, wire-wound vitreous-enamelled resistance. = 150,000-ohm, l-watt, composition non-insulated resistance. = 24,000-ohm, l-watt, composition non-insulated resistance.= 4700-ohm, 1 OO-watt, wire-wound vitreous-enamelled resistance. = 2200-ohm, 100-watt, wire-wound vitreous-enamelled resistance. R,* Rl, = 1200-ohm, 100-watt, wire-wound vitreous-enamelled resistance. = 470-ohm, 100-watt, wire-wound vitreous-enamelled resistance. = 330-ohm, 100-watt, wire-wound vitreous-enamelled resistance. Rll = 220-ohm, 100-watt, wire-wound vitreous-enamelled resistance. Rl, = 1500-ohm, 1 OO-watt, wire-wound vitreous-enamelled resistance. R14 = 3300-ohm, 1 OO-watt, wire-wound vitreous-enamelled resistance. Rl, = 1 2,000-ohm1 100-watt, wire-wound vitreous-enamelled resistance. Rl, = 15,000-ohm, 100-watt, wire-wound vitreous-enamelled resistance. Rl, = 20,000-ohm, 1 OO-watt, wire-wound vitreous-enamelled resistance. = S-pF, 600-V, fixed paper - foil capacitor.= lOOO-ohm, 20-watt, variable wire-wound linear resistance. = 500-ohm coil, K3000-type1 magnetic relay (4 contacts, 2 of which are platinum; RLA RLB = 500-ohm coil, K300-type, magnetic relay (4 contacts, 2 of which are platinum). MR1, MR2 = Metal rectifier (S.T.C. code MDA25-20-1GZ). TR1 = Power transformer (Gardner's Radio type to AERA specification No. 182). M1 = 150-mA1 2+inch, flush-type moving-coil, d.c. ammeter (Sangamo Weston Ltd.) X1, X2 = 24-V, d.c., low-inertia integrating-type motor (Electro Methods type 913). PLA = Three-pole, position 5, fixed plug. PLB = Six-pole, position 0, fixed plug. PLC = Tow-pole, position 0, fixed plug. SWA SWB = 3-amp., double-pole change-over toggle switch. swc = Six-pole, three-way, non-shorting, rotary wafer switch.SWD Rl R, R,, R4 R6 Rl3 R,, Rl, Rll Rm 2V1 3 are used). = 3-amp., double-pole change-over toggle switch (used as double-pole ON - OFF). = Four-pole, six-way, shorting, rotary wafer switch. onlyFig. 2. Equipment for automatic coulometric titrationsApril, 19571 AUTOMATIC COULOMETRIC TITRATIONS 233 REFERENCES 1. Lingane, J., “Electro-analytical Chemistry,” Interscience Publishers Inc. , New York, 1953, 2. Delahay, P., “New Instrumental Methods in Electrochemistry,” Interscience Publishers Inc., 3. DeFord, D., Anal. Chem., 1956, 28, 660. 4. Schults, W. D., Thomason, P. F., U.S. Atomic Energy Commission Report ORNL-1846, Oak Ridge 5. Lingane, J., Anal. Chem., 1954, 26, 1021. 6. Gerhardt, G. E., Lawrence, H. C., and Parsons, J. S., Ibid., 1955, 27, 1752. 7. Parsons, J. S., Seaman, W., and Amick, R. M., Ibid., 1955, 27, 1754. 8. Craig, D. N., and Hoffman, J. I., National Bureau of Standards Circular 524, Washington, 1953, 9. Cooke, W. D., and Furman, N. H., Anal. Chem., 1950,22, 896. 10. Yamado, T., and Kondo, S., Bull. Nagoya Inst. Technol., 1952, 4, 232. 11. Yamado, T., Japan Analyst, 1954, 3, 215. 12. Schreiber, R., and Cooke, W. D., Anal. Chem., 1955,27, 1475. Chapters 17 and 18. New York, 1964, Chapters 5, 11, 14 and 19. National Laboratory, Tennesse, November, 25th, 1955, unclassified. pp. 13 to 20. ANALYTICAL CHEMISTRY GROUP ATOMIC ENERGY RESEARCH ESTABLISHMENT HARWELL, NR. DIDCOT, BERKS. September 20th, 1956

ISSN:0003-2654    

DOI:10.1039/AN9578200228

出版商:RSC    

年代:1957

数据来源: RSC

摘要:

April, 19571 AUTOMATIC COULOMETRIC TITRATIONS 233 A Method for the Determination of Acetic Anhydride in Mixtures with Acetic Acid* BY T. ELLERINGTON AND J. J. NICHOLLS The titration of amines in acetic acid with perchbric acid, with either a potentiometric or visual end-point, is applied to the determination of acetic anhydride in acetic anhydride - acetic acid mixtures. The end-points and accuracy are good. MANY methods for the determination of acetic anhydride in mixtures of acetic anhydride and acetic acid are modifications of the British Standard meth0d.l Such methods are somewhat tedious and suffer from the disadvantages that a small titration error leads to a comparatively large error in the final result. A more recent method involves reaction of the anhydride with an excess of rnorpholine and titration of the excess of reagent with a solution of hydrochloric acid in methanol? In the method described in this paper the acetic anhydride is dowed to react with excess of standard aniline solution in glacial acetic acid, and the excess of aniline is titrated with perchloric acid in acetic acid.3~~ The end-point may be determined either potentiometrically with use of a glass indicator electrode and a saturated-calomel reference electrode in con- junction with a high-impedance millivoltmeter or visually by using a suitable indicator (see p.235). A closed potentiometric titration cell incorporating agitation with dry nitrogen and fitted with a trap for acetic acid vapour is shown in Fig. 1. All the work to be described involves reactions in non-aqueous solution, and hence introduction of water into the titration cell is to be avoided as far as possible.It should also be noted that a glass electrode immersed in glacial acetic acid may assume a positive potential with respect to a calomel electrode, and that, if pH readings are taken, the values obtained are in arbitrary units, since pH has no significance in non-aqueous solution. EXPERIMENTAL PREPARATION AND STANDARDISATION OF REAGENTS- Sodizcm carbonate in glacial acetic acid, 0.1 N-Prepared by drying analytical-reagent grade anhydrous sodium carbonate at 270" & 10" C to constant weight, dissolving 2.6501 g of the solid in glacial acetic acid and making up to 500ml in a calibrated flask. PerchZoric acid in glacial acetic acid, 0.1 N-Anhydrous perchloric acid is difficult to obtain, and the acid is usually available as an approximately 70 per cent.solution having a specific gravity of 1.7. The reagent is prepared by diluting 9.5 ml of the 70 per cent. acid Lisbon, September 8th to 16th, 1956. * Presented at the XVth International Congress on Pure and Applied Chemistry (Analytical Chemistry),234 ELLERINGTON AND NICHOLLS: A METHOD FOR THE DETERMINATION OF [Vol. 82 to 1 litre with glacial acetic acid. This solution is standardised against the 0.1 N sodium carbonate in glacial acetic acid, the standard actually being sodium acetate, which functions as a base of comparable strength to aniline (see Fig. 2) under the conditions of the experiment. Potassium hydrogen phthalate can also be used as a standard.The reagent solution is not stable and requires re-standardising at least every 3 days. Millivoltmeter +ve terminal Trap for acetic acid vapour Fig. 1. Potentiometric-titration apparatus, with a trap for acetic acid vapour Volume of titrant Fig. 2. Standardisation curves: curve A, perchloric acid against sodium carbonate ; curve B, aniline against perchloric acid Aniline in glacial acetic acid, 0-1 N-Prepared by diluting 9.1 ml of redistilled aniline This solution is standardised against the 0.1 N perchloric to 1 litre with glacial acetic acid. acid by potentiometric titration ; it is unstable and requires standardising every day. DETERMINATION OF ACETIC ANHYDRIDE- In the method to be described acetic anhydride is allowed to react with excess of aniline until quantitative reaction has occurred, and the remaining aniline is then titrated with perchloric acid.April, 19571 ACETIC ANHYDRIDE IN MIXTURES WITH ACETIC ACID 235 The development of the method was effected in three main stages, namely- (a) a series of assays of pure acetic anhydride to establish the best technique, (b) a series of experiments to determine the time required for the reaction to go to completion, and (c) a series of determinations of acetic anhydride in mixtures of anhydride and acid of known composition.Development of technique-In the development of the technique of introducing the reactants to each other, three methods were tried, the third being successful. In the first, the anhydride was introduced into the aniline solution, contained in a stoppered flask, by means of a Lunge - Ray pipette.There was a tendency for the tip of the pipette to catch on the ground neck of the flask during this procedure and deposit a small amount of anhydride, which did not enter into reaction. Low results were, therefore, sometimes obtained. In the second method, the stoppered flask containing the standard aniline solution was weighed before and after introduction of the anhydride. There was a tendency for acetic acid vapours to “creep” between the ground-glass surfaces, so that, when the stopper was removed to allow the introduction of the sample, some acid evaporated, so giving a low apparent weight of sample, which led to high results. In the third, successful method, the anhydride was weighed directly into a clean dry calibrated flask and diluted to the mark with glacial acetic acid.A suitable aliquot was then put by pipette into the standard aniline solution. This method of introducing the reactants to each other consistently gave the expected results. Determination of time required for reaction-A 1-1350-g amount of pure acetic anhydride (99.9 per cent. by the British Standard method1) was weighed into a clean dry 100-ml calibrated flask and diluted to the mark with glacial acetic acid. Into each of four stoppered clean dry 150-ml flasks were put by pipette 50.0 ml of standardised 0.1 N aniline solution and 20.0 ml of the anhydride solution. The last mixture to be prepared was titrated as soon as possible after the first mixing, even so 3 minutes had elapsed, and the others at longer intervals.The recoveries of acetic anhydride in these four experiments, which may be taken as the percentages of reaction having occurred, were plotted against the time required, as shown in Fig. 3. Determination of acetic anhydride in mixtures of known composition-Four mixtures of acetic anhydride and acetic acid, together with a “pure” anhydride, were assayed by the method described. The anhydride gave a result of 96.9 per cent. w/w by reaction with aniline and titration with perchloric acid. The four mixtures were made up accurately by weight. The acetic anhydride contents of the mixtures, based on both chemical and potentio- metric assays of the pure anhydride used, are shown in Table I, together with the results by potentiometric titration and with use of a visual indicator.METHOD Into a clean and thoroughly dry 100-ml calibrated flask weigh accurately sufficient sample to contain approximately 1 g of acetic anhydride and dilute to the mark with glacial acetic acid. By pipette put into a 150-ml stoppered flask 50.0ml of standardised aniline in acetic acid solution. Add by pipette 20.0 ml of the prepared solution of the sample in acetic acid, set aside for a minimum of 40 minutes, and transfer the contents of the flask quantitatively to the titration vessel with the aid of glacial acetic acid; titrate the excess of aniline with standard perchloric acid in glacial acetic acid solution. Acetic anhydride, per cent. = (B - T ) x 0-01021 x F x 100, 0.2 x w where B = volume in ml of standard perchloric acid solution equivalent to 50.0 ml of T = titre in ml of perchloric acid solution after reaction, F = factor of 0.1 N perchloric acid solution, and W = weight of sample taken.aniline in acetic acid solution, USE OF AN INDICATOR FOR A VISUAL END-POINT- Wagner, Brown and Peters3 refer ,to crystal violet as being a suitable indicator for This indicator has been tried and, titrations, giving a good potentiometric-titration curve.236 ELLERINGTON AND NICHOLLS: A METHOD FOR THE DETERMINATION OF p o l . 82 once the operator had become accustomed to the rather subtle colour change at the end-point, the results agreed with those found by plotting the titration curve. When added to the mine in acetic acid, the indicator was violet and on titration with perchloric acid underwent the following colour changes- violet -+ blue-violet +- blue -+ green-blue + blue-green --+ “pure” green + yellow-green +- green-yellow -+ yellow.TABLE I DETERMINATION OF ACETIC ANHYDRIDE IN MIXTURES Acetic anhydride present in mixture, calculated- A I from a chemical from a potentiometric analysis, determination, % % 96.9 96-2 73.7 73.2 48.8 48.5 26.1 24.9 1.12 1.11 Acetic anhydride in mixture, determined with a- A r -l potentiometric visual end-point, end-point, 96-2 96.2 73.2 72.8 48.8 48.5 26.3 25.3 1.18 1-18 % % The colour indicative of the correct end-point varies with the substance being titrated, and it is advisable to mark the colour changes on an actual titration curve for any given titration to determine the correct visual end-point. The “pure” green was the correct visual end-point colour for aniline. Fig.3. Rate of reaction of acetic anhydride with excess of aniline a t 20°C The colour changes of the indicator were observed in various aqueous solutions by adjusting the pH by suitable addition of acid or alkali, in the hope of finding a solution that, on being added, would give the “pure” green colour of the indicator. Such a solution could be used for matching when the visual method of detecting the end-point is used. It was found that the colour of the indicator depended not only on the pH of the solution but also on its composition. Finally, however, an acetic acid - acetate buffer solution was selected, the pH being adjusted by means of 10.0 N hydrochloric acid. The colours of the indicator and the pH of these solutions are shown in Fig.4. A suitable indicator solution is prepared by dissolving 0.1 g of crystal violet in 100 ml of glacial acetic acid; 0.1 ml of indicator solution is used for every 50 ml of solution. Preparation of coZoz4r standards-Three colour standards were developed to provide matching to the indicator colour (a) just before the end-point, (b) at the end-point, and (c) just after the end-point. These are prepared by adding indicator solution (0.1 ml per 50 ml) to the appropriate buffer solution immediately before required. The buffer solutions are made by mixing 50-Om1 of N sodium acetate solution, 10-Om1 of 99 to 100 per cent. acetic acid and various amounts of 10.0 N hydrochloric acid, and then diluting to 250 mlApril, 19571 ACETIC ANHYDRIDE I N MIXTURES WITH ACETIC ACID 237 with water. For: (a) buf€er solution of pH 0-7, blue-green with indicator added, 84ml of 10.0N hydrochloric acid are required; (b) buffer solution of pH 0.6, “pure” green with indicator added, 10.5 ml of 10.0 N hydrochloric acid are required; (c) buffer solution of pH 0.5, yellow-green with indicator added, 12.5 ml of 10.0 N hydrochloric acid are required. I \ --& ‘ Fig. 4. Colour changes of crystal violet in acetic acid - acetate buffer solutions; sharp changes are indicated by full lines and more gradual changes by brok :n lines We are indebted to British Industrial Solvents (a Division of the Distillers Company Limited), in whose laboratories this work was carried out, for permission to publish this paper. REFERENCES 1. British Standard 2068: 1953. 2. 3. 4. Johnson, J. B., and Fauk, G. L., Anal. Chem., 1955, 27, 1464. Wagner, C. R., Brown, R. H., and Peters, E. D., J . Amer. Chem. SOC., 1947, 69, 2609. Becket, A. H., and Tinley, E. H., “Titration in Non-aqueous Solutions,” The British Drug Houses Ltd., London, 1955. BRITISH INDUSTRIAL SOLVENTS HULL SALT END HEDON August 2nd, 1966

ISSN:0003-2654    

DOI:10.1039/AN9578200233

出版商:RSC    

年代:1957

数据来源: RSC

摘要:

238 JOHNSON AND POLHILL: THE USE OF AN ANION-EXCHANGE RESIN [VOl. 82 The Use of an Anionlexchange Resin in the Determination of Traces of Lead in Food BY E. I. JOHNSON AND R. D. A. POLHILL Microgram amounts of lead are separated from most other ions by absorption from N hydrochloric acid solution on a column of the chloride form of an anion-exchange resin. The lead is recovered by elution with 0.01 N hydrochloric acid. This principle is used in a method for the determination of lead in foods. RECENTLY Kraus and Nelson,lJ Miller and HunteI3 and Jentzsch4,S have published informa- tion about the behaviour of various metals in hydrochloric acid solutions with the chloride form of strongly basic anion-exchange resins of the cross-linked polystyrene quaternary ammonium type. In particular, the paper by Kraus and Nelson2 on lead and bismuth, together with our experience of the method of Rush and Yoe6 for zinc, suggested that these resins could be used for the quantitative separation of microgram amounts of lead from acid solution.Miller and Hunte? found that Amberlite IRA-400 was as satisfactory as Dowex 1 (used by Kraus and Nelson and by Rush and Yoe) for the separation of zinc. Amberlite IRA-400 was used for the present investigation. It was found, as reported for Dowex 1, that a column of Amberlite IRA-400 would absorb microgram quantities of lead from N hydrochloric acid solution while allowing alkali metals, alkaline-earth metals, iron, and copper to pass through. It was also found that small amounts of phosphate and sulphate ions passed completely through the column when introduced in N hydrochloric acid solution. Examination of the distribution curves given by Nelson and Kraus and the elution constants by Jentzsch suggested that very dilute hydrochloric acid could be used to elute the lead.It was found that 0.01 N hydrochloric acid quickly eluted the whole of the lead present. It was found possible to adjust the con- ditions of elution so that the normality of the eluate containing the lead was very close to the normality of a 1 per cent. nitric acid solution. The lead in the eluate could then be easily and conveniently determined by the mono-colour method of Snyder.? Further work was directed to the development of a method for the determination of lead in foods and similar materials.The possibility of interferences with such a method for lead caused us to investigate the behaviour of zinc, bismuth, cadmium, tin and thallium under the conditions of the method. Table I shows the effect in terms of apparent lead of various amounts of these metals in N hydrochloric acid solution, placed on the column and subsequently treated as in a determination of lead. TABLE I EFFECT OF VARIOUS METALS ON LEAD DETERMINATION Metal Amount added, Apparent lead found, Pg Pg Tin.. .. .. .. 2000 Nil Zinc . . .. .. 500 Nil Cadmium .. .. .. 200 0.9 Bismuth . . ,. .. 230 0-5 Thallium . . .. .. 300 0-7 Tin and thallium pass through the resin with the N hydrochloric acid. Bismuth is retained and not eluted. Zinc is retained and eluted with 0.01 N acid. Cadmium is retained and eluted much more slowly than zinc.The conditions of the dithizone procedure sub- stantially suppress interference from zinc and cadmium if present in amounts similar to those in Table I. EXPERIMENTAL Amberlite IRA-400 anion-exchange resin (analytical grade) purchased in the hydroxyl form was treated by the method of Miller and H ~ n t e r . ~ Six grams were reduced by being ground in a mortar until the resin all passed through a No. 60 B.S. sieve. Fines were removed by a No. 100 B.S. sieve and the fraction not passing that sieve was used to make the column.April, 19571 IN THE DETERMINATION OF TRACES OF LEAD IN FOOD 239 The resin was soaked overnight in 2 N hydrochloric acid. Fines were removed by decantation and the remainder was transferred to the ion-exchange tube, which consisted of a glass tube 15 cm long and 8 mm in internal ’diameter.The bottom was fitted with a tap to control the flow rate and the top was sealed to a piece of wider tubing to give a reservoir of capacity about 30 ml. A 5-mm plug of cotton-wool was placed in the bottom of the tube and the resin washed in with N hydrochloric acid. After the resin had settled, the top of the resin column was held in place with a 2-mm plug of cotton-wool. The column was thoroughly washed with N hydrochloric acid, then with 0.01 N hydrochloric acid and finally with 10 ml of N hydro- chloric acid. The resin column had a length of about 8 cm. Substances were passed into the column in the form of solutions in 5 ml of N hydro- chloric acid, at a flow rate of 1 ml per minute.The column was then washed with 30 ml of N hydrochloric acid at a flow rate of 2 ml per minute. Elution with 0.01 N hydrochloric acid was done at a flow rate of 1 ml per minute. The column was allowed to drain under gravity between these operations, but no pressure was applied. It was found that on elution with 0.01 N hydrochloric acid the first 2-5 to 3 ml of eluate were undiluted N hydrochloric acid. The next 10 ml of eluate, which contained all the lead, had an acidity of about 0-2 N . Solutions containing elements likely to be present in the ash of foods were prepared in 5 ml of N hydrochloric acid solution and their behaviour on the resin column when sub- jected to the washing and elution procedure described above was studied. These solutions contained up to 1.5 mg of copper and iron, up to 0.25 g of calcium, up to 0.15 g of magnesium and up to 0.15 g of phosphorus as phosphate.All of the calcium and magnesium and the bulk of the iron, copper and phosphate were found in the first 15 ml of N hydrochloric acid wash. The remainder of the phosphate was contained in the next 5 ml. The last portion of the N hydrochloric acid wash contained only a trace of iron and copper. It was found that the retention of lead by the column was reduced by the presence of large amounts of calcium and magnesium phosphates, but that amounts up to 0.1 g of calcium, magnesium or phosphorus as phosphate did not effect retention or recovery of lead. Sulphate behaved similarly to phosphate. Small amounts such as result from the ashing of 5 g of food have no effect on retention of lead by the column, but large amounts such as would remain from the wet or sulphated ashing of a food temporarily destroy the capacity of the column to retain lead and most other metallic ions.The passing of 50 ml of N sulphuric acid through the column was in fact found to be a convenient method of clearing it from accumulations of ions normally retained and not eluted, e.g., bismuth and cadmium. The subsequent passage of 30ml of N hydrochloric acid restored the column to the chloride form. Amounts of lead up to 40 pg in 2 ml of N hydrochloric acid were run on to the column, which was then washed with 30 ml of N hydrochloric acid. All the lead was recovered in the first 10 ml of eluate with 0.01 N hydrochloric acid, after rejection of the 2.5 ml of un- diluted N hydrochloric acid that preceded it.Second and third 10-ml portions of eluate were found to be free from lead. We are of the opinion that this resin column made and used in the way we have described is a useful and convenient device for freeing a hydrochloric acid solution containing lead in microgram quantities from substances that would otherwise interfere with the determination of the lead. It has certain limitations besides those we have already indicated. We did not find it possible to use it as a simple means of concentrating quantitatively very dilute solutions of lead. Washing the column with volumes of N hydrochloric acid greatly in excess of the 30 ml normally used caused the lead to move down and ultimately off the column despite the maintenance of the concentration of the acid.This has been taken into consideration in the design of the following method for the determination of lead in foods. The method has the merit of requiring fewer and less reagents than any other of which we are aware. As a consequence of this, the method gives lower blanks than other methods in which reagents of the same “lead-free” standard are used. The lowness of this blank is not due to loss of lead in the N hydrochloric acid wash. This possibility was carefully investigated and no lead was found in the 35 ml of acid emerging from a column that contained 40 pg of lead. The method used would have detected 0.2 pg of lead. METHOD REAGENTS- Amberlite IRA-400 anion-exchange resin.Hydrochloric acid, 2 N, N and 0.01 N.240 JOHNSON AND POLHILL: THE USE OF AN ANION-EXCHANGE RESIN [VOL 82 Chloro f o w . Dithimne stock solzltion-A 0.1 per cent. w/v solution of dithizone in chloroform. Dithizone working solution-Shake 6 ml of the stock dithizone solution with 10 ml of 0-5 N ammonium hydroxide solution and reject the chloroform layer. Solution A-Mix 340 ml of ammonium hydroxide, sp.gr. 0-880, 75 ml of 2 per cent. w/v sodium sulphite, Na,SO,, solution, 30ml of 10 per cent. w/v potassium cyanide solution and 605 ml of water. Standard lead solution-(a) Dissolve 1-60 g of lead nitrate in water, add 10 rnl of concen- trated nitric acid and dilute to 1 litre. (b) Dilute 1 volume of (a) to 100 volumes with water. Prepare dilution (b) freshly as required. 1 ml = 10 pg of lead.Magnesium nitrate solution-A 10 per cent. w/v solution of magnesium nitrate, Mg(NOJ2.6H20, in water. Further details about the preparation in a lead-free state, storage and standards required of these reagents are given el~ewhere.~~~ APPARATUS- been described under "Experimental". An ion-exchange tube-The design and filling of this tube with resin have already PROCEDURE FOR DESTROYING ORGANIC MATTER AND SEGREGATION OF LEAD- Ash a suitable quantity of sample, containing not more than 5 g of dry matter, in a silica or platinum dish at a temperature not exceeding 500" C. If the material is otherwise difficult to ash, up to 5ml of magnesium nitrate solution may be added as an ashing-aid. Dissolve the ash in 4 ml of 2 N hydrochloric acid, cover the dish with a watch-glass and heat it on a steam-bath for 10 minutes.Filter through a cotton-wool plug held in the stem of a small conical funnel into a 10-ml measuring cylinder, washing the dish and filter with a few millilitres of water. Volumes must be kept low at this stage, the total of filtrate and washings not exceeding 8 ml, and the exact volume should be noted. To the resin column prepared as previously described and wet with N hydrochloric acid, add a 5-ml aliquot of the ash solution. The 5 ml should not contain more than 4-0 pg of lead. Adjust the flow rate to 1 ml per minute. When all the 5 ml of solution is below the level of the top of the resin, add 25 ml of N hydrochloric acid and allow it to pass through the column at the rate of 2 ml per minute.Elute the lead from the column with 0.01 N hydrochloric acid at a flow rate of 1 ml per minute, rejecting the first 2-5 ml, which should be undiluted N hydrochloric acid free from lead, and collecting the next 10ml. PROCEDURE FOR DETERMINING LEAD- To the eluate in a 100-ml separating funnel add 30 ml of solution A, exactly 10 ml of chloroform and 0.5 ml of the dithizone working solution; shake vigorously for 1 minute and allow to settle. Insert a plug of cotton-wool into the dry stem of the funnel and, after rejecting the first runnings, fill a 1-cm spectrophotometer cell with the chloroform solution. Measure the optical density against chloroform at 520 mp. Prepare a blank solution under the same conditions as the test, omitting only the sample, and determine the optical density.To prepare a calibration graph measure 0, 1.0, 2.0, 3.0 and 4.0 ml of standard lead solution into separating funnels containing 2 ml of N hydrochloric acid and add water to give a total volume of 10ml in each funnel. Proceed as described above. As the method is sensitive, all the precautions usual in this type of work must be o b ~ e r v e d . ~ ~ ~ Correct the observed result for the reduction due to the taking of part only of the ash solution for the final determination. Run off a little of the chloroform layer. RESULTS Typical results for lead found in various samples are shown in Table 11. Blanks were usually found to be less than 1 pg of lead and were constant for one set The resin column may be re-used repeatedly after regeneration with N hydro- Accumulations of ions not readily eluted with hydrochloric acid can be removed of reagents. chloric acid.April, 19571 IN THE DETERMINATION OF TRACES OF LEAD IN FOOD 241 by running 50 ml of N sulphuric acid through the column, followed by 30 ml of N hydro- chloric acid to regenerate the chloride form.TABLE I1 DETERM N OF LEAD Sample 15 pgof lead .. .. 40 pg of lead . . .. 5gof cocoa .. .. 5 g o f syrup .. .. 5 g of curry powder 6.5 g of cocoa + 30 pg lead . . (calculated 5.4 p.p.m.) usual Lead by laL _ _ - _ _ _ methods proposed method .. - 15.5 pg .. 0.84 p.p.m. 0.82, 0.84 p.p.m. .. 5-6 p.p.m. 5.3 p.p.m. .. 4.3 p.p.m. 4.3 p.p.m. 6 . - 5.4 p.p.m. .. - 39.9 pg We express our thanks to the Government Chemist for permission to publish this paper. REFERENCES 1. 2. 3. 4. 5. 6. 7. Snyder, L. J., Ibid., 1947, 19, 684. 8. 9. CLEMENT’S INN PASSAGE Nelson, F., and Kraus, K. A., J . Amer. Chem. SOL, 1954, 76, 5916. Kraus, K. A., and Nelson, F., Ibid., 1954, 76, 984. Miller, C. C., and Hunter, J. A., Analyst, 1954, 79, 483. Jentzsch, D., and Pawlik, I., 2. anal. Chem., 1955, 146, 88. Jentzsch, D., Ibid., 1956, 150, 241. Rush, R. M., and Yoe, J. H., Anal. Chem., 1954,26, 1345. Analytical Methods Committee, “The Determination of Lead in Foodstuffs,” Analyst, 1954,79,397. Johnson, E. I., and Polhill, R. D. A., Ibid., 1955, 80, 364. DEPARTMENT OF THE GOVERNMENT CHEMIST STRAND, W.C.2 November 5th, 1966

ISSN:0003-2654    

DOI:10.1039/AN9578200238

出版商:RSC    

年代:1957

数据来源: RSC

摘要:

April, 19571 IN THE DETERMINATION OF TRACES OF LEAD IN FOOD 241 Flame-photometric Determination of Magnesium in Plant Material A Study of the Emission of Magnesium in a Highly Reducing Oxygen - Acetylene Flame BY KARIN E. KNUTSON The sensitivity obtained with the 2852 A line emitted by the magnesium atom was considerably greater with a highly reducing carburising flame than with a neutral one. Examination of the sources of error showed that phos- phate and sulphate ions gave rise to negative interference, which could be largely eliminated by the presence of an excess of calcium ions. The accuracy of the method was checked by analysis of prepared samples of known mag- nesium content and by comparative chemical and flame-spectrophotometric analysis of plant material. In the study of the precision the standard deviation of a single determination was less than 1 per cent.The “detection limit” was 0.06 p.p.m. THE chemical determination of magnesium has always been an involved and time-consuming procedure when applied to samples of complex composition, such as plant and soil materials. Accordingly, methods for the flame-photometric and spectrophotometric determination of magnesium have been evolved. One difficulty of these methods lies in the flames used; for instance, acetylene - air, propane - oxygen - air and acetylene - oxygen mixtures them- selves produce a high emission at the utilisable wavelengths: 2851, 3721, 3811 and 3850 A. Hence, there is insufficient contrast between the magnesium emission and the background radiation, and it is difficult to determine small quantities of the element.1~2~3~4~6~6 Good results have, however, been achieved, although mostly with relatively high magnesium242 KNUTSON : FLAME-PHOTOMETRIC DETERMINATION OF [Vol.82 concentration~.7,8~~,lo There are many reports in the literature on the disturbing effect of other ions, such as those occurring in plant sample^.^ 9 4 9 8 y 9 >11 J2 Consequently, in the work described in the paper, attempts have been made to increase the sensitivy of the method. Attention has also been given to the significance of the presence of other ions. EXPERIMENTAL APPARATUS- Preliminary experiments with a Beckman DU spectrophotometer with attachment for a propane - oxygen - air flame showed that its sensitivity was unsatisfactory.Moreover, it is inconvenient to work with a compensation instrument in flame-photometric studies. For these experiments, therefore, the equipment consisted of a monochromator with a quartz prism, a burner housing with an atomiser burner for an oxygen - acetylene mixture and regulators (Beckman DU flame spectrophotometer with attachment13). In addition, a photo- multiplier tube, RCA 1P.28, was used with a power supply of 860 volts from dry batteries. There should be means of measuring the current strength or, if a high resistance is introduced, the voltage drop in the anode circuit of the photomultiplier tube. However, as it is necessary to operate the recorder from a low-impedance source, a cathode follower was inserted in the circuit. The circuit arrangement is shown in Fig. 1.The output signal level was adjusted by changing the resistance (resistors R, to R5) across which the input voltage was developed (when a current flows in the phototube). A setting of 0.2 was used for all measurements. The dark-current could be compensated for by means of a zero adjustment control. R,, Rz = 0.9-megohm resistance R3 = 2-7-megohm resistance R4 = 4.5-megohm resistance = 9-megohm resistance - - 500-oh m resistance R5 R6 (ZERO ADJUSTMENT CONTROL) = 470,000-ohm resistance - 300-ohm resistance R, R, R9 = 50,000-ohm resistance R,, = 25,000-ohm resistance R,,, R12, R,, = 10,000-ohm resistance Cl, c2 = 8-pF condensers Vl = 5692 valve v2 v3 = OC3 (VR 105) valve = 5Y3GT valve (anode to cathode plate voltage per plate (r.m.s.), 300 volts) F1 = O.8-amp.fuse PI = 0.2-amp. 6.5-volt bulb Fig. 1. Circuit diagram of cathode follower for Speedomax recorder, type G The voltage was measured by means of a Speedomax automatically recording potentio- meter, type G, model S, obtained from Leeds & Northrup Co., with a 2-2-second full-scale pen travel. The scale was calibrated in millivolts and the full-scale deflection was 10 mV (240 mm). A valuable advantage of the automatically recording potentiometer is that any error, such as that incurred by a tendency for the atomising efficiency to decrease, is im- mediately observed. In many measurements, however, the instrument was used without inserting the recording device. PROPERTIES OF THE FLAME AND THEIR INFLUENCE ON THE MAGNESIUM EMISSION- The type of flame is dependent on the volumetric proportions of the oxygen and acetylene supplied to the burner.In a neutral flame the acetylene content is about 48 per cent. and in a carburising flame about 55 per cent.l* The neutral flame contains carbon monoxideApril, 19571 MAGNESIUM I N PLANT MATERIAL 243 and hydrogen, and the more strongly reducing carburising flame also contains carbon, which, on account of the high temperature, emits an intensive yellowish white light. With the more strongly reducing carburising flame the ratio of the intensities of the 2852 A line emitted by the neutral magnesium atom and the background radiation is appreciably increased (see Table I). The higher the carburising zone the higher the potentiometer reading for magnesium. The background reading is kept constant by reduction of the slit width.The background radiation increases with the proportion of acetylene. TABLE I EMISSION FROM THE NEUTRAL MAGNESIUM ATOM AT 2852A FOR VARIOUS PARTS OF THE FLAME AND HEIGHTS O F THE CARBURISING ZONE For all determinations the reading for background radiation was adjusted to 4mV by means of the monochromator slit. This value has been subtracted from the figures given in the “Potentiometer reading for 10 p.p.m. of magnesium” column. Adjustment was made for dark-current Height of Height of Potentiometer Burner Oxygen Acetylene carburising exposed part reading rating, pressure, pressure, zone above of flame above for 10 p.p.m. of lb per Slit width, lb per lb per burner tip, burner tip, magnesium, sq. inch mm sq.inch sq. inch nim mm mV 19 0.036 9 1.8 0 to 10 11 to 43 1-3 19 0.025 9 3 0 to 25 11 to 43 2.4 19 0.023 9 5 0 to 45 11 to 43 3.8 19 0.037 9 1.8 0 t o 10 11 to 25 2.2 19 0.029 9 3 0 t o 25 11 to 25 3.2 19 0.030 9 5 0 to 45 11 t o 25 4.4 19 0.049 9 1.8 0 to 10 19 to 34 1-6 19 0.032 9 3 0 to 25 19 to 34 3.6 19 0.030 9 5 0 to 45 19 to 34 5.1 19 0-044 9 1-8 0 to 10 23 to 38 1.9 19 0.030 9 3 0 to 25 23 to 38 2.9 19 0.028 9 5 0 to 45 23 to 38 4.1 10 0.020 7 5 0 to 15 11 to 43 2.1 10 0.045 7 2 0 11 to 25 0.9 10 0.023 7 5 0 to 15 11 to 25 2.6 10 0.02 1 7 7 0 to 20 11 to 25 3.5 10 0-021 5 5 0 to 20 11 t o 25 2.7 14 0.019 5.5 5 0 to 30 11 to 43 2-4 14 0.026 5.5 5 0 to 30 11 to 25 3-6 14 0.03 1 5.5 5 0 to 30 23 to 38 4.6 TABLE I1 EMISSION FROM THE MAGNESIUM ATOM AT 2852 A AND FROM MAGNESIUM OXIDE AT The figures given in the “Potentiometer reading for background” column have been subtracted from the observed potentiometer readings to give the figures in the “Potentiometer reading for 100 p.p.m.of magnesium” column. Adjustment was made for dark-current. For all determinations a burner with a rating of 19 lb per sq. inch was used and the height of the exposed part of the flame was 11 to 25 mm above the burner tip 3721, 3811 AND 3 8 5 O A I N FLAMES OF VARIOUS REDUCING STRENGTHS Height of Potentiometer, Wave- pressure, pressure, zone above reading for for 100 p.p.m. of Iength, Slit width, lb per lb per burner tip, background, magnesium, Oxygen Acetylene carburising Potentiometer reading A mm sq. inch sq. inch mm mV mV 2852 0.034 19 3 0 1.5 3.1 2852 0.020 9 5 0 to 45 1-5 7.4 3721 0.029 19 3 0 1-5 0.5 3721 0.015 9 5 0 to 45 1.5 0.4 3721 0-051 19 3 0 6.0 2.1 3721 0.027 9 5 0 to 45 6.0 0.9 381 1 0.052 19 3 0 6.0 1.3 3811 0.025 9 5 0 to 45 6.0 0-5 3850 0.052 19 3 0 6.0 2.0 3850 0.024 9 5 0 to 45 6.0 0-8244 KNUTSON : FLAME-PHOTOMETRIC DETERMINATION OF [Vol.82 The sensitivity for magnesium at the 2 8 5 2 ~ line emitted by the neutral magnesium atom increases considerably in the carburising flame in spite of the simultaneous increase in the background radiation. This may be due to displacement of the equilibrium MgO + Mg in the flame owing to its stronger reducing effect, the concentration of magnesium atoms therefore being increased. This explanation is supported by the results of tests performed at 3721, 3811 and 3850 A, when the emission from magnesium oxide resulted in higher sensitivity with the more weakly reducing flame than with the carburising flame (see Table 11).The background radiation is extended over a wide wavelength range, but at 2852 A a sharp minimum was observed, as can be seen in Fig. 2. A slit width of about 0.03 mm, corresponding to a theoretical half-intensity band of about 1 A, was normally used with this instrument. The signal-to-background ratios were compared for slit widths of 0.03 and 0.02 mm, the latter with the use of a more sensitive photomultiplier tube (EM1 6255). How- ever, only a slight increase in the ratio was produced with the narrower slit width. 0 28 45 2850 2; Wavelength, A i5 2860 Fig. 2. Emission readings in the region of 2852 A: Approxi- Slit width, A, 10 p.p.m.of magnesium; B, background. mate readings for wavelength correction. 0.03 mm The flame light was masked so that only a 10-mm wide horizontal slit was left open in the wall of the burner housing. After careful trials a suitable region of the flame was chosen 11 to 25 mm above the burner tip. A suitable stable emission was produced in this position, but higher in the flame the stability was reduced. The interfering action of certain other ions was less marked in a carburising than in a neutral flame (see also p. 250 and Table IX). The consumption of oxygen should be reduced so that a carburising flame of suitable height can readily be produced. With each burner there are data on the pressure required to give an atomising rate of 1.5 to 2 ml per minute.It is, however, more convenient to reduce the pressure to give a consumption of about 1 ml per minute. The smaller the volume of acetylene required to produce the carburking flame, the less likelihood is there of soot being deposited on the burner tip. The flame will also be of a more suitable height and will fluctuate less. The ejector action in some burners is greater than others. Burners having the greatest ejector action were found to be the most suitable, as the oxygen consumption could be kept lower.April, 19571 MAGNESIUM IN PLANT MATERIAL 245 A high sensitivity to magnesium was achieved with all the atomiser burners tested, provided that the following conditions were observed- (a) the rate of atomisation was adjusted to about 1 ml per minute by regulating the oxygen pressure, (b) the acetylene pressure was set at 5 lb per sq.inch and the gaseous mixture ignited and (c) the acetylene supply was regulated by means of the fuel needle valve in the right side of the pressure-control unit, so that a yellowish white zone about 45 mm high was produced in the flame. For the flame to give a stable emission the gas pressure must be kept constant. This is effected by maintaining a certain critical pressure ratio.ls If the absolute gas pressure indicated on the fuel gauges of the gas cylinder and the Beckman unit are, respectively, PI and pz, the ratio for acetylene should be 1.8 or over and for oxygen 1.9 or over. In the instructions for the Beckman instruments,ls a pressure of 0.7 atmospheres above atmospheric is recommended for the acetylene cylinder.If, for example, 5 lb per sq. inch are required for the Beckman instrument the over-pressure shown on the gauge of the acetylene cylinder should be at least 1-4 atmospheres so as to give the critical pressure ratio. Com- parative values of the rate of acetylene flow at the burner tip are obtained not only from the pressures read on the fuel regulator. The flow rate is dependent also on the choke area of the needle valve in the pressure-control unit and on the design of the burner. PREPARATION OF STANDARD AND CONTROL SOLUTIONS- Hydrochloric acid (0.02 M) was the solvent in all the control tests except when aqueous solutions and acid solutions of different molarity were used for comparison.As control solutions, chlorides and acids of analytical-reagent grade were used except for the following. Magnesium chloride-Weigh magnesium in the form of strips that have previously been scraped bright. Check the magnesium concentration by gravimetric analysis. Calcium chloride-Weigh Iceland spar or calcium carbonate of analytical-reagent grade that has been twice precipitated as the oxalate and heated to 500" C to form the carbonate. Convert to the chloride. STANDARD CURVES- The magnesium emission for the 2852 A line at relatively high concentrations is weakened by the self-reversal of the atoms. This means that the gradient of the standard curve in linear co-ordinates is weakened at these concentrations. On the other hand, the signal- to-background ratio increases so that a higher precision could be achieved with more con- cent rat ed solutions.Convert the metal to the chloride. Concentration of magnesium, p.p.m. Fig. 3. Standard curves for magnesium (linear co-ordinates) : curve A, 0 to 1000 p.p.m.; curve B, 0 to 125 p.p.m.; curve C, 0 to 12-5 p.p.m.246 KNUTSON : FLAME-PHOTOMETRIC DETERMINATION OF [Vol. 82 As the samples of plant material available for magnesium determination are often very small, a sensitive method was desirable, and low magnesium concentrations were consequently examined. Fig. 3 shows the standard curve in linear co-ordinates, and Fig. 4 in logarithmic co-ordinates. The curve for low magnesium concentrations is practically linear in both systems. These two curves give very similar values for the magnesium content if the points on the curve lie as close together as 1.00, 2-50, 5-00, 7.50 and 10.00 p.p.m. of magnesium.Concentration of magnesium, p.p.m Fig. 4. Standard curves for magnesium (logar- ithmic co-ordinates): curve A, 50 to 1000 p.p.m.; curve B, 10 to 125 p.p.m.; curve C, 1 to 12.5 p.p.m. PROCEDURE FOR DETERMINING MAGNESIUM- Preliminary treatment of the plant sample-Weigh out about 1 g of the finely powdered sample and put it in a long-necked Kjeldahl flask. Digest with nitric acid and perchloric acid.17 Remove the silica by filtration, then drive off the excess of perchloric acid by evaporation over a hot-plate and then in an oven at 200" C. In this laboratory, the sample dissolved in dilute hydrochloric acid is also repeatedly evaporated to dryness on a steam- bath to transform all the phosphate present to orthophosphate; this is desirable for certain determinations carried out on other aliquots of the same sample.18J9 Dissolve the residue from the evaporation in 0.02 M hydrochloric acid and make up to 100 ml with the acid.Not more than about 15 ml are generally required for the magnesium determination. Filter the solution through Pyrex-glass-wool before atomising, If, in a sample, the weight ratio of calcium to phosphorus (as the phosphate) is lower than 2 to 1, it is safer either to remove the phosphate and sulphate ions (see p. 251), or to increase the calcium content of the sample by the addition of calcium chloride solution. The calcium compound of analytical-reagent grade should be checked for the presence of magnesium and, if necessary, purified further.PROCEDURE FOR TAKING MEASUREMENTS- Connect the cathode follower to a 220-volt a.c. power supply 30 minutes before taking the readings. Connect the photomultiplier tube to the auxiliary power supply and the potentiometer to a 110-volt a.c. supply. Set the wavelength at 2 8 5 2 ~ and the selector switch of the cathode follower circuit to 0.2. Turn on the oxygen and clean the palladium tube by introducing from below a wire of diameter 0.2 mm. Adjust the pressures to suitable values so that a constant supply of oxygen and acetylene is obtained and ignite the mixture. Adjust the acetylene pressure to give a carburising flame, according to the instructions above. Balance the dark-current by means of the zero adjustment on the cathode follower.Compensate also for the current due to the background radiation if a very low magnesium concentration is to be determined. Atomise a more concentrated solution con- taining, e.g., 50 p.p.m. of magnesium. Make a fine adjustment of the wavelength and, ifApril, 19571 MAGNESIUM I N PLANT MATERIAL 247 necessary, of the entrance mirror on the monochromator, until a maximum reading is produced on the potentiometer scale. Select the slit width to give a reading of about 10 mV for about 10 p.p.m. of magnesium (a slit of about 0.03 mm is required). The difference between the reading for 10 p.p.m. of magnesium and for the solvent alone is, with a good burner, about 5 mV, which corresponds to a scale deflection of about 120 mm.The reading for the flame alone is higher than that for the solvent and for low concentrations of magnesium, presumably owing to cooling and consequent lower emission of the flame on atomisation. Make a series of readings by atomising 0-02 M hydrochloric acid, these serving as a check of the stability of the reading on the potentiometer scale. The readings are taken symmetrically with respect to the standard solutions, until 6 to 10 determinations of the test solution have been made. The measurement technique is described by Ehrlin-Tamm.20 The potentiometer needle settles slowly unless the atomising is interrupted a few times at the beginning of each new sample by raising and lowering the vessel containing the solution. When the salt solutions have been atomised, traces of deposit on the burner tip may result in a fall in the reading.This may be avoided if the palladium tube is cleaned with, for instance, M hydrochloric acid instead of 0-02 M or distilled water after each supply of salt solution. INTERFERING FACTORS Temperature of the solution-The temperature of the solution influences the atomisation and therefore the emission. For a temperature difference of 6" C a deviation of about 3 per cent. was observed. Care should be taken, therefore, to ensure that the solutions are at room temperature before atomising and that the vessels containing the solutions are closed before and between tests. Acid concentrations of the solutions-The tests, results of which are given in Table 111, were carried out with 10.00 p.p.m. of magnesium in 0.02 M hydrochloric acid as a standard.It is evident from the Table that it is important to have the same concentration of acid in both the standard and unknown solutions and that 0.02 M concentrations of hydrochloric, nitric and perchloric acids have about the same effect on the emission. TABLE I11 THE INFLUENCE OF ACIDITY ON THE DETERMINATION OF MAGNESIUM IN MAGNESIUM CHLORIDE SOLUTIONS Each solution contained 10 p.p.m. of magnesium Concentration of Substance added substance added, M 0.00 1 - Distilled water . . . . Hydrochloric acid 0.10 Perchloric acid . . . . 0.02 Nitric acid . . .. 0.02 . . / E:, Magnesium found, p.p.m. 10.30 10.30 10.29 10.00 9.77 9.92 9.92 Error, + 3-0 + 3.0 + 2.9 % & 0.0 - 2.3 - 0.8 - 0.8 Various anions and cations-The presence of phosphate or sulphate ions resulted in negative errors (see Fig.5). The phosphate error in the determination of 10 p.p.m. of magnesium was fairly constant between 50 and 200 p.p.m. of phosphorus (as the phosphate) and was about -11 per cent. Sulphate gave an error that increased continuously and was about -8 per cent. for 50p.p.m. of sulphur (as the sulphate) and about -13 per cent. for 200 p.p.m. Phosphorus and sulphur compounds present in plant material are changed to the phos- phate and sulphate in the preparation of the samples for ana1ysis.l' The proportions of these ions are often high enough to involve appreciable errors in the analysis. These errors are largely eliminated if sufficient calcium is present (see Tables VI, VII, IX and XI).The effects of some cations were studied as can be seen in Tables IV and V. The first tests showed a positive error in the presence of large quantities of pure calcium salts. Analytical-reagent grade calcium chloride was used only for the samples that were prepared on the basis of the plant analyses, when a correction for the traces of magnesium248 KNUTSON : FLAME-PHOTOMETRIC DETERMINATION OF [Vol. 82 was also applied (see Tables IV and XI). In all other interference tests the calcium salt was freed from magnesium by precipitation twice as the oxalate. I A _-_--- '. '. '*. \ 1 20 Concentration of interfering ion, p.p.m. Fig. 5. Error in the determination of 10.00 p.p.m. of magnesium in the presence of interfering ions: curve A, sulphate sulphur; curve. B, phosphate phosphorus; curve C, aluminium TABLE IV DETERMINATION OF MAGNESIUM JN THE PRESENCE OF CALCIUM CHLORIDE All solutions were prepared with 0.02 M hydrochloric acid Calcium salt taken Calcium for conversion to chloride present, p.p.m.AnalaR calcium carbonate . . 1000 500 250 Iceland spar .. .. .. 2000 1000 200 100 Magnesium found (none added), p.p.m. 0-29 0-14 0.00 0-32 0.10 0.07 - Magnesium found (10 Uncorrected p.p.m. added) , error, 10.60 + 6.0 10.25 + 2.5 10.08 + 0.8 10.29 + 2.9 10.04 + 0.4 p.p.m. % - - - - Corrected* error, % + 3.1 + 1.1 + 0.8 * Corrected with respect to magnesium content before addition of magnesium sample. The Iceland spar was the purest substance tested. For this an error of less than 3 per cent. was obtained for 10 p.p.m. of magnesium in the presence of 1000 p.p.m.of calcium, no correction for the magnesium content before the addition of the sample being made. It is evident from this that in plant material interference due to calcium may be disregarded. For calcium, sodium, potassium, ammonium and iron ions only small errors were observed, as can be seen in Tables IV and V. If aluminium is present in a higher concentration than a few parts per million, it should be removed before carrying out the determination (see Fig. 5 ) . HutchinsonZ1 states that plant material contains as a rule no more than 0-002 per cent. of aluminium. At the Forest Research Institute of Sweden determinations were made of the aluminium content of spruce needles and birch leaves from an experimental area in Molna. The spruce needles contained about 0.01 per cent.and the birch leaves about 0-001 per cent. of aluminium,April, 19571 MAGNESIUM I N PLANT MATERIAL TABLE V DETERMINATION OF MAGNESIUM IN THE PRESENCE OF VARIOUS CHLORIDES ,411 solutions were prepared with 0.02 M hydrochloric acid and contained 10 p.p.m. of added magnesium 249 Substance added Sodium chloride . . Potassium chloride . . Ammonium chloride Aluminium chloride Ferric chloride . . Manganese chloride . . .. .. . I .. .. Concentration, p.p.m. 1000 of Na 100 of Na 10 of Na 1000 of K 100 of K 10 of K 1000 of NH, 100 of NH, 10 of NH, 200 of A1 100 of A1 10 of A1 1 of A1 100 of Fe 10 of Fe 260 of Mn 20 of Mn Magnesium found, p.p.m. 9.83 10.18 9.94 9.77 9-91 9-88 10.30 9.75 9.89 7.10 7-70 9-65 10.12 9.93 10.07 9.65 9.98 Error, % + 1-8 - 0.6 - 2.3 - 0.9 - 1.2 + 3.0 - 2.5 - 1-1 - 1.7 - 29 - 23 - 3.5 + 1.2 - 0.7 + 0.7 - 3.5 - 0.2 Manganese often occurs in plant material in concentrations similar to those of magnesium and it must be taken into account in most chemical methods of magnesium determination.In flame-spectrophotometric determination the presence of manganese does not appreciably affect the results of the analysis, this being evident from the results given in Tables V and XI. I t is clear that there was considerable negative interference from phosphate and sulphate ions and a negligible positive interference from calcium. In spite of this it was found, from the results in Table VI, that the error due to the phosphate in the presence of calcium was considerably reduced.With twice as much calcium as phosphorus (as the phosphate) the error was practically eliminated. TABLE VI ERROR IN DETERMINING 10 p.p.m. OF MAGNESIUM IN THE PRESENCE OF Analytical-reagent grade calcium chloride that had been further purified was used. The figures in parentheses are the number of series of results of which the mean was taken for the tabulated value. All solutions were prepared with 0.02 M hydrochloric acid PHOSPHATE AND CALCIUM IONS Phosphorus (as phosphate), p.p.m. 0 6.25 12-5 25 50 Error in determining 10 p.p.m. of magnesium in the presence of- 0 p.p:m. 6-25 p.p.m. 12.5 p.p.m. 25 p.p.m. 50 p.p.m. 100 p.p.m. 100p.p.m.* oi calcium, of calcium, of calicum, of calcium, of calcium, of calcium, of calcium, A I T, Y O Y O Yo % "4 I_ -- + 0*6(4) - - - % % - - - - _- - 1.0 - 0.3 - 4.3 - - 3-4 - 0.6 - 6.0 - 9.4(8) - .- - - 6.0 - 2*1(3) - 0.7(4) - - - - - - - 4.5 - 1.4 - - * Calcium chloride prepared from Iceland spar.The error due to the sulphate in the presence of calcium is given in Table VIP. The negative interference was almost completely removed even when the calcium content was equal to that of the sulphur (as the sulphate). In plant material calcium is often present in sufficient amounts to give these weight ratios. Tests were made with potassium and sodium instead of calcium, but no reduction of the interference by phosphate ions was effected.250 KNUTSON FLAME-PHOTOMETRIC DETERMINATION OF TABLE VII [Vol. 82 ERROR IN DETERMINING 10 p.p.m. OF MAGNESIUM IN THE PRESENCE OF Analytical-reagent grade calcium chloride that had been further purified was used. The figures in parentheses are the number of series of results of which the mean was taken for the tabulated value.All solutions were prepared with 0.02 M hydrochloric acid SULPHATE AND CALCIUM IONS Error in determining 10 p.p.m. of magnesium in the presence of- 0 p.p:m. 6-25 p.p.m. 12.5 p.p.m. 25 p.p.m. 50 p.p.m. 100 p.p.m. of calcium, of calcium, of calcium, of calcium, of calcium, of calcium, f A > % % % % % % - 0.8(2) + 4*3(2) + 0.7(2) - - 1.9(5) - + 1*6(2) + 1.2(2) - - - + 0-6(4) - - - - - - - - 4*7(5) - I + 0-9(2) + 0*4(2) - 7-5(5) - - - - 1.7(3) 0-0(4) Sulphur (as sulphate), p.p.m. 0 6.25 12-5 25 50 Comparative determinations were carried out with 10.00 p.p.m. of magnesium and the same sample diluted to 2.00 p.p.m.It can be seen from Table VIII that for samples con- taining phosphate, sulphate or aluminium ions, the relative errors were considerably reduced on dilution. In the presence of iron (see Table V) the error at different dilutions was negligible. TABLE VIII ERROR I N DETERMINING MAGNESIUM I N THE PRESENCE OF VARIOUS IONS AT DIFFERENT DILUTIONS OF THE MAGNESIUM SOLUTION Phosphorus (as phosphate) present, p.p.m. 10 2 250 50 50 10 - - 40 8 50 10 - - 250 50 Sulphur (as sulphate) present, p.p.m. Error in determining Magnesium present, p.p.m. 10.00 2-00 10.00 2-00 10.00 2.00 10.00 2.00 10.00 2.00 10.00 2-00 10.00 2.00 10.00 2.00 Calcium present, p.p.m. Iron present, p.p.m. Aluminium present, p.p.m. magnesium, % - 5.0 + 1.0 - 11 - 4.0 - 6.0 + 0-5 - 1.3 + 1-5 - 4.0 + 1.5 - 8.0 - 3.5 - 9-5 - 27 - 52 - 29 - 50 10 - 480 96 I 50 10 I 50 10 - 200 40 200 40 I I The interference by aluminium, phosphate and sulphate ions was very dependent on the type of flame used.The effect of calcium was only slight and appears to be independent of the type of flame used (see Table IX). Interference caused by aluminium, phosphate and sulphate ions decreased as the proportion of acetylene used increased. The presence of calcium will reduce the interference of phosphate and sulphate in strongly carburising, weakly carburising and neutral flames. REMOVAL OF THE INTERFERING SUBSTANCES- During the course of the study a closer examination was made of whether the phosphate and sulphate ions influenced the results in the analysis of plant material (see Tables VI and VII, and Fig.5). Attempts were therefore made to remove these ions by adsorption in an alumina column. The adsorption columns were prepared as described by Nydah1.22 The solvent for the samples was 0.02 M hydrochloric acid. Before use the column was rinsed with about 500 ml of 0.02 M hydrochloric acid to obtain an aluminium-free eluate. The eluate was testedfor aluminium by the procedure given by Gjems and Lyder~en,~~ but to the scale of 1 to 10.25 1 April, 19571 MAGNESIUM IN PLANT MATERIAL TABLE IX ERROR IN DETERMINING 10 l3.P.m. OF MAGNESIUM IN THE PRESENCE OF ALUMINIUM, CALCIUM, PHOSPHATE AND SULPHATE IONS IN DIFFERENT TYPES OF FLAME All solutions were prepared with 0.02 M hydrochloric acid Oxygen Acetylene pressure, pressure, lb per lb per sq.inch sq. inch 9 5 9 3-75 9 3 19 3 Oxygen Acetylene pressure, pressure, lbper lb per sq. inch sq. inch 9 4.5 9 3 19 3 The adsorption Type of flame carburising weak carburising weak carburising neutral Type of flame carburising weak carburising neutral Error in determining 10 p.p.m. of magnesium in presence of- 100 p.p.m. of 100 p.p.m. of 100 p.p.m. of phosphorus suIphur aluminium, calcium, (as phosphate), (as sulphate), A I \ 100 p.p.m. of % ’ % % % - 12 + 1.0 - 8 - 9 - 10 - 10 - 19 - 24 - - 9 - 11 - 47 + 0.5 - 21 - 27 - Error in determining 10 p.p.m. of magnesium in presence of- f A \ 60 p.p.rn. of 50 p.p.m. of phosphorus (as sulphur (as 50 p.p.m. of phosphate) and 50 p.p.m. of sulphate) and phosphorus (as 100 p.p.m. of sulphur (as 100 p.p.m.of phosphate), calcium, sulphate), calcium, % YO Yo YO - 7 - 1.5 - 4 + 0.5 - 9 - 3 - 8 - 1 - 19 - 6 - 16 - 2 tests were carried out in the following manner. The alumina column was washed uider pressure with a part of the solution to bye analysed. Another part of the same solution was then immediately passed through the alumina column. The magnesium determinations were performed on this part directly. It was readily established, by the method described by Feig1,24 that the phosphate ions were adsorbed. The sulphate ions presented some difficulty and could be checked only indirectly through the fall in the magnesium error. The error obtained after treatment in the alumina column was less than -3 per cent. of the magnesium value, compared with previous figures of -7 to -15 per cent. (see Table X).Hence this method of removing the phosphate and sulphate ions seems to be efficient. Tests were also made with Dowex 2, but the adsorption of phosphate ions from a solution 0.02 M with respect to hydrochloric acid was negligible. TABLE X DETERMINATION OF 10 p.p.m. OF MAGNESIUM IN SOLUTIONS IN THE PRESENCE OF PHOSPHATE AND SULPHATE IONS BEFORE AND AFTER ADSORPTION IN ALUMINA COLUMNS Phosphorus Sulphur (as phosphate) (as sulphate) present, present, p.p.m. p.p.m. - - - 50 100 - - 40 - 100 - 200 Magnesium found before adsorption, p.p.m. - 8-95 8.90 9.30 9.07 8.50 Magnesium found after Error, adsorption, % p.p.m. - 10.01 - 10.6 9-87 - 11.0 9.74 - 7.0 9.96 - 9.3 9.80 - 15.0 9-71 Error, YO + 0.1 - 1.3 - 2.6 - 0.4 - 2.0 - 2.9 Number of determina- tions 3 3 2 13 11 6 Standard deviation of a single determina- tion, % f- 2.2 2 1.6 & 0.4 + 1.2 1.0 - THE ACCURACY AND PRECISION OF THE METHOD Prepared samples-To check the method solutions were prepared having the same com- The error in the magnesium deter- Correction was made for the traces of magnesium present position as those given by the analyses of plant material.minations is given in Table XI. in the calcium salt.252 KNUTSON : FLAME-PHOTOMETRIC DETERMINATION OF TABLE XI FLAME-PHOTOMETRIC DETERMINATION OF MAGNESIUM I N SOLUTIONS PREPARED Solution I I I I I1 I11 I11 I11 IV IV IV V V VI VI VII VII VIII VIII I X IX X The ON THE BASIS OF PLANT ANALYSES All solutions were prepared with 0.02 M hydrochloric Concen- tration 1 0.8 0.6 0.3 1 1 0.2 0.1 1 0.8 0.3 1 0.33 1 1 1 0.67 1 0.4 1 0.4 1 Metal added in p.p.m.- Na K Ca A1 Mn PO,-P SO,-S Mg 10 100 100 - 50 35 - 25-00 20.00 15.00 7.50 10 10 100 - 25 10 - 10.00 5 75 100 - - - - 75.00 15.00 7.5 10 50 100 - - 50 - 25-00 20.00 7.50 10 150 100 2 75 40 25 30-00 10.00 5 100 100 - 7.5 2 3 10.00 5 100 100 - 7.5 2 3 10.00 5 100 100 - 12.5 40 40 15.00 10.00 5 25 100 - 25 5 5 25.00 10.00 10 200 100 - 125 50 10 25.00 10.00 10 100 100 - 150 50 10 10.00 I L 3 _ _ _ _ - - - - I___- - - _ _ _ _ - - __. _ _ _ _ _ _ - - - - - - - - - - _ _ _ _ - - - _ _ _ - - - - _ _ - - - - - _ _ _ _ _ - - - _ _ _ _ - - - _ _ _ - - - - Mag- nesium found, p.p.m.20.11 15-17 7.54 10-45 14-93 7.47 20.14 7.65 9.80 10.09 10-15 10.08 9.86 9.84 9.80 - - - I - - - acid Un- corrected error, % + 0.6 + 1.1 + 0.5 + 4.5 - 0.5 - 0.4 - - - -+- 0.7 + 2.0 - 2.0 + 0.9 + 1.5 -t 0.8 - 1.4 - - I - - 1.6 - 2.0 [Vol.82 Cor- rected* error, Yo - - 0.2 + 0.3 - 0.3 + 2-5 - 0.8 - 0.7 - 0.1 + 1.2 - 2.7 - 1.1 - 0.5 - - - - - 0.5 - 2.2 - 2.4 - 4.0 - - * Corrected for magnesium present in the added calcium. manganese content was comparatively high. No appreciable interference by manganese could be traced in the results of these tests. The interference tests would in that case have further reduced the magnesium content. The close agreement between the added and the found values for the magnesium could be explained only by the damping effect of the calcium on the interference by phosphate and sulphate. Comparative chemical and jiame-spectrophotometric analysis-B y way of comparison, gravimetric determinations of magnesium in birch leaves were carried out.About 5g of finely powdered leaves were weighed out for each test. The procedure used was that described on p. 246 up to the dissolution of the sample in 0.02 f ! hydrochloric acid. The magnesium was determined as the pyrophosphate by the method due principally to Kolthoff and Sande11.26 The precipitate was heated to constant weight in a porcelain filtering crucible at 1100" C. Aluminium, iron, manganese, phosphate, calcium and magnesium were then precipitated. Ammonium salts were removed before precipitation of the magnesium. Table XI1 gives the results of the comparative gravimetric and flame-spectrophotometric determinations of the magnesium in birch leaves. Manganese was present in considerable quantities.Removal of phosphate and sulphate ions had only a slight effect. The reason for this was perhaps the damping action of the calcium present on the interference. In the study of the precision a standard deviation of the single determination of & 1.7 per cent. was found for 7 series of determinations of about 9 p,p.m. of magnesium. Gravimetric determinations of magnesium are time-consuming. Examples of the pre- cision and accuracy are given by Hillebrand and L ~ n d e l l . ~ ~ $27 At smaller magnesium percentages the accuracy was poor. Comparison with the precision .fig.ures 0 f Beckman and others-The precision of the mag- nesium determination was studied, 10 readings being made symmetrically in relation to the standard solutions in 6 series of measurements for 10.00 p.p.m.A standard deviation of 0.9 per cent. was found for the mean of a series of 10 readings. The corresponding standard deviation for a solution containing 2-00 p.p.m. was 2 0.8 per cent. (10 readings in 5 series). For between 2 and 10 p.p.m. of magnesium a standard deviation of less than 1 per cent. was found. The background reading was equivalent to about 6 p.p.m. of magnesium. The detection limit, in the sense in which the term is used by Beckman, is then 0.06 p.p.m.April, 19571 MAGNESIUM I N PLANT MATERIAL 253 Beckman2s gives 0.2 p.p.m. at 2852 A in an oxygen - hydrogen flame. Hence the sensitivity of the method is improved, the signal-to-background ratio being 3 to 4 times greater with a corresponding reduction of the detection limit.TABLE XI1 ANALYSIS OF BIRCH LEAVES BY GRAVIMETRIC AND FLAME-PHOTOMETRIC METHODS Organic substance and silica were removed from all samples. The values given are on a dry-weight basis Mag- nesium found gravi- Sample metric number method, * Mean, 2071 t:!;i7 } 0.380, 0.377, 2072 0.370, 0.374 0.374 0.276, 0.273 by. Y O Yo 0-275 I 2075 Magnesium found by flame-photometric method A f before after removal removal of phos- of phos- phate and phate and Cal- Man- Phos- sulphate sulphate cium ganese phorus Sulphur ions, Mean, ions, Mean, found, found, found, found, Yo Yo Y O Y O % Yo Y O % 0*371J } 0.373 0.361t } 0.369 1.38 0.33 0.77 - 1.54 0.35 1.14 - 0.374 0.377 0.388, 0.368, 0-377 0.387, 0.385 0.381 0.267, 0.275 0.388, 0.374 0.262, 0-261 0.256, 0.266 0.258 0.264 * By the method of Kolthoff and Sande11.26 1-21 0-20 0-12 0.16 I- I- 1 1 In the determination of sodium and potassium Solomon and C a t ~ n ~ ~ obtained a precision that was slightly lower than Beckman’s optimal value. They used a Beckman mono- chromator and the flame attachment, a multiplier phototube, an Applied Physics Corporation model 30 vibrating-reed electrometer as an amplifier and a Brown recording potentiometer.SUMMARY The determination of magnesium at 2852 A, when the emission was derived from the neutral magnesium atom and when a highly reducing carburising flame was used, showed a considerably higher sensitivity, whereas the interference by aluminium, phosphate and sulphate ions was lower. In weakly and strongly reducing flames the calcium ions reduce the interference of the phosphate and sulphate ions.These ions may be removed by adsorption in an alumina column. For a concentration of 2 p.p.m. of magnesium, the interference by aluminium, phosphate and sulphate ions is less than for 10 p.p.m. of magnesium, with the same proportions of magnesium to each of the foreign ions. The study shows that the flame-photometric determination of magnesium in plant material may be performed with satisfactory precision and accuracy in the region of 2 to 10 p.p.m. I express my appreciation to Dr. C. 0. Tamm, for valuable discussions during the course of the work and in the preparation of this paper, to T. A. Bengtsson, Fil. lic., for helpful criticism, and to Dr. F. Nydahl, Associate Professor of Analytical Chemistry, Uppsala University, for valuable advice.REFERENCES The cathode follower was built by T. Hanaas. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. Brown, J . G., Lilleland, O., and Jackson, R. K., Proc. Amer. Soc. Hod. Sci., 1950, 56, 11. Cholak, J., and Hubbard, D. M., Ind. Eng. Chem., Anal. Ed., 1944, 16, 728. Fieldes, M., King, P. J. T., Richardson, J . P., and Swindale, L. D., Soil Sci., 1951, 72, 219. Kick, H., 2. PjlErnahr. Dung., 1954, 67, 53. Kuemmel, D. F., and Karl, H. L., Anal. Chem., 1954, 26, 386. LundegArdh, H., “Die quantitative SpectraJanalyse der Elemente,” Volume 11, Jena, 1934. Burstrom, H., Svensk Kem. Tidskr., 1935, 17, 12. Close, P., Smith, E. W. M., and Watson, M. T., jun., Anal. Chem., 1953, 25, 1022. Rich, C. I., Proc. Soil Sci. Amer., 1952, 16, 51. Vallee, B. L., Nature, 1954, 174, 1050.254 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. BRADFIELD : AN IMPROVED FORMALDOXIME METHOD FOR THE p o l . 82 Shoij, K., and Nakata, S., Proc. Arner. SOC. Hort. Sci., 1954, 64, 299. Stace, H. C. T., J . Coun. Sci. Ind. lies. Australia, 1948, 21, 305. “Instruction Manual 334-A Models DU and B Flame Spectrophotometers,” Beckman Instruments Spiegelberg, H., “Svets- och skarteknik rned acetylen-syrgas,” AGA, Stockholm-Lidingo, 1944, “Hzitte, Des Ingenieurs Taschenbuch I,” Edition 27, Akademischen Verein Hutte, Berlin, 1942, Beckman Instruments Inc., ofi. cit., p. 19. Smith, G. F., Anal. Chim. Acta, 1953, 8, 397. Milton, R. F., and Waters, W. A., “Methods of Quantitative Micro-Analysis,” Edward Arnold Stugart, R., Ind. Eng. Chem., Anal. Ed., 1931, 3, 390. Ehrlin-Tamm, G., Acta Chem. Scand., 1950, 4, 1317. Hutchinson, G. E., Soil Sci., 1945, 60, 29. Nydabl, F., Anal. Chem., 1954, 26, 580. Gjems, O., and Lydersen, D., 2. PjZErnGhr. Dilng., 1954, 64, 36. Feigl, F., “Qualitative Analyse rnit Hive von Tupfel-reaktionen,” Third Edition, Akademische Kolthoff, I. M., and Sandell, E. B., “Textbook of Quantitative Inorganic Analysis,” Third Edition, Hillebrand, W. F., and Lundell, G. E. F., “Applied Inorganic Analysis with Special Reference Beckman Instruments Inc., op. ~it., p. 13. Solomon, A. K., and Caton, D. C., Anal. Chem., 1955, 27, 1849. Inc., South Pasadena, California, 1954, p. 4. p. 45. p. 584. & Co., London, 1949, p. 102. Verlagsgesellschaft m.b.H., Leipzig, 1938, p. 334. The Macmillan Company Inc., New York, 1952. to Metals, Minerals and Rocks,” John Wiley & Sons Inc., New York, 1929, p. 879. -- J , op. cit., Second Edition, 1953, pp. 3 and 804. THE FOREST RESEARCH INSTITUTE OF SWEDEN STOCKHOLM 51, SWEDEN August 13th, 1956

ISSN:0003-2654    

DOI:10.1039/AN9578200241

出版商:RSC    

年代:1957

数据来源: RSC