摘要:

April, 1983 SHORT PAPERS Spectrophotometric Determination of Some 531 Lanthanides as Tetraethylenepentamineheptaacetic Acid Chelates M. Tarek M. Zaki, Adel F. Shoukry and Mohamed B. Hafez Chemistry Department, Faculty of Sciences, United Arab Emirates University, P.O. Box 15551, Al-Ain, Abu-Dhabi, United Arab Emirates Keywords : Lanthanide determination ; spectrophotometry ; tetraethylene- pentamineheptaacetic acid Methylthymol blue,l arsenazo I11 ,2 dicarboxyarsenazo I11 [3,3'-( 1,8-dihydroxy-3,6-disulpho- 2,7-naphthylenebisazo)bis-(4-arsonobenzoic acid)],3 pyrogallol red, diphenylg~anidine,~ chlorosulphophenol M [3-(5-chloro-2-hydroxy-3-sulphophenylazo)-6-(3-sulphophenylazo)- chromotropic acid] ,5 gallocyanine (CI Mordant Blue 10) methyl ester,6 orthanilic B [3- (phenylazo)-6-(2-sulphophenylazo)chromotropic acid] ,' hydroxynaphthol blue* and calmagiteg have been applied as reagents for lanthanide determination.The determination of lanthan- ides in water using 2-hydroxy-2-methylpropionic acid has also been reported.1° Although arsenazo 111, hydroxynaphthol blue and calmagite are useful for the determination of lanthan- ides, they are pH dependent and the accuracy as a function of time is unsatisfactory. Although polyaminopolycarboxylic acids are known to form strong chelates with lanthan- ides,ll few data have appeared on their analytical application. Mishchenko and Poluetkov12 studied the absorption spectra of lanthanide - EDTA chelates. The optimum conditions for the determination were 0.1 M EDTA at pH 8-9 and a lanthanide concentration of 10 mg ml-l (as oxide).Tetraethylenepentamineheptaacetic acid (TPHA), which contains twelve donor groups, showed higher sensitivity with cerium(II1) than EDTA, which contains six donor groups.13 A spectrophotometric investigation of the interaction between Pr(III), Nd(III), Sm(III), Eu(III), Gd(III), Dy(II1) and Er(II1) and TPHA has been carried out and a method is pro- posed for the determination of these lanthanides using TPHA. The method is about 10 times more sensitive than that using EDTA. The average recovery of the measured lanthanides is 99.9% with a standard deviation of 0.1%. Experimental Reagents and Apparatus All chemicals were of analytical-reagent grade and doubly distilled water was used. Tri- valent lanthanide ion solutions were prepared by dissolving an accurately weighed amount of lanthanide chloride in water to give stock solutions of concentration 1 x M, except for Gd(III), for which the concentration was 5 x 10-3 M.The stock solutions of lanthanides were standardised with EDTA.l* Working solutions were obtained by dilution. A 0.1 M stock solution of TPHA (tetrasodium salt) was prepared and standardised against copper sulphate solution using murexide as indicator.l5 The optical measurements were carried out with a Prolabo ultraviolet - visible spectro- photometer using 10-mm quartz cells. The pH of the solutions was adjusted using a Type E 520 pH meter (Metrohm, Herisau) with a Metrohm AG 9100 combined glass - calomel electrode. Procedure Add 3 ml of M TPHA solution to a suitable volume of sample containing 0.1-1.0 mg of lanthanide ion.Make up the volume to 10 ml with water after adjusting its pH to 9.0 using carbonate-free sodium hydroxide solution. Measure the absorbance at 458,580,409,394,273, 351 and 258 nm for Pr(III), Nd(III), Sm(III), Eu(III), Gd(III), Dy(II1) and Er(III), respect- ively. Calculate the lanthanide ion concentration from a calibration graph prepared in the same manner.532 SHORT PAPERS Analyst, 1701. 108 Effect of pH The characteristic absorption bands of the lanthanide ions studied are located in the wave- length range from 250 to 700 nm (Fig. 1 and Table I). The absorbance of TPHA is negligible above 260 nm. The absorption bands of 0.5 x M trivalent lanthanide ions [2.5 x 10-4 M for Gd(III)] with 5 x M TPHA were shifted to longer wavelengths and the absorbance values increased as the pH increased.Within certain pH ranges, the position of the absorption maximum remains constant (Table I). The optimum pH for the formation of the erbium chelate was in a more acidic region than that of the praseodymium chelate, apparently because erbium is more basic than praseodymium. At pH >10.5 the absorbance decreased owing to the formation of hydroxy complexes. The molar absorptivity of each chelate is given in Table I. However, pH 9 was most convenient as interferences from other ions was minimal and non-complexed metal ions precipitated. Composition and Stability of the Chelates and stability of the chelates. Results and Discussion The proposed method was applicable in the pH range 2.5-10.5 (Table I).The molar ratio and continuous variation methods were applied to determine the composition The two methods confirmed the existence of a 1 : 1 metal - 0.2 0.1 s e $ n 4 0.2 Q 0.1 0.2 0.1 0.2 1.2 0.9 0.6 0.3 460 470 480 490 500 520 540 560 580 360 380 400 390 394 398 272 276 280 324 328" 348 356 364 I -- r"\ v 255 260 " 370 378 "480 488 I' 510 520 Wavelengthtnm 1 .o 0.5 1 .o 0.5 a C e 2 n 1.04 0.5 Fig. 1. Absorption spectra of (1) some lanthanide ions (5 x M) in 0.1 M hydrochloric acid and (2) [Ln3+ + TPHA] chelates (5 x lov4 M) at pH 9.0. Half the concentrations were used with gadolinium and its chelate.April, 1983 SHORT PAPERS 533 ligand chelate in each instance. The stability constants of the 1 : 1 chelates increased from praseodymium to erbium (Table I).It is sug- gested that the bonding in the chelates is purely electrostatic and therefore the stability of these complexes would be determined by the radius of the cation present.16 Therefore, the stability constant would then be expected to vary linearly with z2/r or, more generally, with l / r (where x is the atomic number and r is the radius of the cation present). No evidence was found for other ratios. TABLE I CHARACTERISTICS OF LANTHANIDE - TPHA CHELATES Ion Complex Ion Pr(II1) . . .. Nd(II1) . . .. Sm(II1) . . .. Eu(II1) . . .. Gd(II1) . . .. Dy(II1) . . .. Er(II1) . . .. r E x 10-21’ X/nm 1 mol-l cm-l 445 0.94 470 0.56 477 0.36 518 0.22 571 0.43 373 0.15 404 0.46 392 0.32 272 0.96 275 0.56 324 0.32 349 0.42 363 0.32 256 0.38 f X/nm 458 474 480 524 580 377 409 394 273 276 327 352 365 258 x 10-31 1 mol-1 cm-l 2.80 1.70 1.40 1.30 2.30 1.30 2.70 1.84 5.00 3.00 1.56 1.80 1.50 2.80 PH range 3.7-1 0.5 3.5-1 0.5 3.2- 10.5 3.0-1 0.5 2.7-1 0.5 2.5-10.5 2.5-10.5 PK 17.3 17.6 18.2 18.6 18.6 18.8 19.2 Effect of Other Ions The effect of other ions was investigated at pH 9.0.The general procedure was followed, except that the solutiom of foreign ions were added before the ligand solution. The results show that the method is selective for the ions studied and Sr(II), Ca(II), Co(II), Ni(II), Cu(II), Zn(II), Co(III), Ce(III), U022+, citrate, tartrate, phosphate, sulphate and nitrate do not interfere. The results showed that lanthanide ions do not interfere with each other with up to a 4-fold excess of the foreign ions.However, with europium, samarium interferes if present at twice the europium concentration. The lanthanum ion and its TPHA chelate do not absorb in the wavelength range studied and consequently there is no need to remove it before the analysis. Accuracy of the Method Beer’s law was obeyed in the pH range 2.5-10 at concentrations from 10 to 100 pg ml-l. The proposed method was applied at pH 9.0 with an average recovery of 99.9%. The standard deviation (12 determinations for each lanthanide ion) did not exceed 0.1%. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. References Zielinski, S., and Lomozik, K., Chem. Anal., 1978, 23, 815. Surin, I. G., Spitsyn, P. K., and Barkovskii, V. F., Zh. Anal. Khim., 1979, 34, 1103. Guan, T. Q., and Luo, M. Z., Huaxue Tongbao, 1974, 5, 26; Anal.Abstr., 1975, 28, 5B167. Kirillov, A. I., Makarenko, 0. P., and Vlasov, N. A., Izv. Vyssh. Ucheb. Zaved. Khim. Khim. Tekhnol., Akhmedi, M. K., Granovskaya, P. B., and Neimatova, K. A., Azerb. Khim. Zh., 1973, No. 2, 103; Kotoucek, M., Lasovsky, J., and Kucerova, M., Collect. Czech. Chem. Commun., 1979, 44, 1559. Kirillov, A. I., Shaulina, L. P., and Vlasov, N. A., Zh. Anal. Khim., 1974, 29, 62. Brittain, H. G., Anal. Chim. Acta, 1978, 96, 165. Brittain, H. G., Anal. Chim. Acta, 1979, 106, 401. Nevoral, V., Collect. Czech. Chem. Commun., 1978, 43, 2274. Pribil, R., “Analytical Applications of EDTA and Related Compounds,” First Edition, Pergamon Mishchenko, N. T., and Poluetkov, N. S., Zh. Anal. Khim., 1962, 17, 825. 1973, 16, 1150; Ref. Zh., Khim., 1974, 19GD, 2G179. Ref. Zh., Khim., 1974, 19GD, 8G50. Press, Oxford, 1972.534 SHORT PAPERS Analyst, Vol. 108 13. 14. 16. 16. Hafez, M. B., and Hafez, N., Ann. Chim., 1972, 2, 153. West, T. S., “Complexometry with EDTA and Related Reagents,” Third Edition, Broglia Press, Schwarzenbach, G., “Complexometric Titrations,” Methuen, London, 1960. Thopson, L. C., and Loraas, J. A., Inorg. Chew., 1963, 2, 89. London, 1969. Received May 24th, 1982 Accepted October 7th, 1982

ISSN:0003-2654    

DOI:10.1039/AN9830800531

出版商:RSC    

年代:1983

数据来源: RSC

摘要:

534 SHORT PAPERS Analyst, Vol. 108 Spectrophotometric Determination of Cerium with Methylthymol Blue in the Presence of Oxalate and Cyanide as Masking Agents Amal ia Ca brera- Martin, Roberto Izquierdo- Hornillos, Alberto J. Quejido-Cabezas and Jose L. Peral-Fernandez Defiartment of Analytical Chemistry, Facultad de Ciencias Quimicas, Universidad Complutense de Madrid, Ciudad Universitaria, Madrid-3, Spain Keywords ; Cerium determination ; Methylthymol Blue ; spectroplzotonzetvy ; eflect of masking agents The spectrophotometric determination of cerium can be carried out by several methods, which involve either the formation of complexes of cerium(II1) and cerium(1V) or the oxidation of suitable reagents by cerium(1V) and further measuring the intensity of the colour of the oxidised matter.The latter methods show a lack of selectivity and low sensitivity owing to the nature of the redox reaction. The methods that involve the formation of complexes have also been shown to have low selectivity and sensitivity.1,2 However, the most useful methods are those based on the com- plexes of cerium(II1) with Xylenol Orange3 and Methylthymol Blue (MTB),4 but they are affected by many interferences. In this work the reaction of cerium(II1) with MTB in the presence of oxalate and cyanide ions was studied at pH 10.2, which improves the sensitivity and the selectivity of the deter- mination of cerium. Experimental Apparatus was used. S@ectro$hotometer. pH meter. A Unicam SP8-200 spectrophotometer equipped with 1-cm glass cells Crison, Model Digit 501, with an Ingold 10 402 3253 combined electrode.Reagents All chemicals used were of analytical-reagent grade. Methylthymol BZue Solution, 1 mM. distilled water and dilute to 1000 ml in a calibrated flask. standard zirconium( IV) solution. distilled water and dilute to 1000 ml in a calibrated flask. Dissolve 0.757 g of Methylthmol Blue (Fluka AG) in Standardise this solution against a Dissolve 0.4342 g of cerium(II1) nitrate hexahydrate in Cerium(III) stock solution, 1 mM. Disodium oxalate solution, 0.01 M. Potassium cyanide solution, 0.5 M. Glycine stock solution, 0.1 M. Dissolve 7.507 g of glycine in distilled water; add 5.84 g of Sodium hydroxide solution, 0.1 M. Free from carbonate ion; this solution is prepared daily. Glycine - sodium hydroxide bufer.A 58.0-ml aliquot of glycine stock solution is mixed with 42.0 ml of sodium hydroxide solution. sodium chloride and dilute to 1000 ml in a calibrated flask. The mixture is adjusted to pH 10.2.April, 1983 SHORT PAPERS 535 Procedure To a solution of cerium(II1) add 5.0 ml of the glycine - sodium hydroxide buffer solution of pH 10.2, 8.0 ml of 0.01 M sodium oxalate solution, 1.0 ml of 0.5 M potassium cyanide solution and 2.0 ml of 1 mM Methylthymol Blue solution, and then dilute with distilled water to 25.0 ml. Measure the absorbance of this solution at 625 nm in 1-cm glass cells verstxs a blank prepared under the same conditions. Results and Discussion Absorption Spectra The absorption spectra of the complex containing MTB and cerium(II1) at pH 10.2 and in the presence of oxalate and cyanide have been obtained.A sharp maximum at 600 nm in the spectrum of the complex, which is hyperchromic as regards the reagent, has been observed. As was expected, the presence of these masking ions does not influence the main reaction, be- cause of their poor complexing action with ~erium(III).5~~ However, the most sensitive analytical measurements of the absorbance are obtained at 625 nm, as is observed in the difference spectrum of the complex verstxs MTB. Effect of pH In Fig. 1 the effect of pH on the complex formation at 600 and 625 nm is shown. It can be observed that the complex is formed at pH values in the range 9.0-10.8, reaching the maximum colour development at a pH of around 10.2. A glycine - sodium hydroxide buffer at pH 10.2 has been chosen as the most adequate for this reaction because of its poor complexing action with the cerium ions.'** Glycine has shown no competitive action in the cerium(II1) - MTB complex formation nor in the ternary com- plex formation.A minimum concentration ratio of 15% of the buffer (volume of buffer to final dilution volume of the sample) has been shown to be the most adequate for the analytical study of the reaction. 0.9 t A 9 10 PH 11 Fig. 1. Influence of pH on absorbance of Ce(II1) - MTB complex, measured at (A) 625 nm and (B) 600 nm. Influence of the Concentration of Oxalate, Cyanide and Tartrate as Masking Agents The influence of oxalate on the reaction has been studied. No effect was observed when the molar concentration of this ion was 1 000-fold greater than the cerium(II1) concentration.Cyanide also shows no effect on the measurements of the absorbance of the cerium(II1) - MTB complex when this ion is present in a molar excess of 10000-fold. Tartrate shows no effect up to a molar excess of 100-fold. Higher concentrations of tartrate decrease the values of the absorbance of the cerium(II1) - MTB complex because of its com- petitive action on the cerium(II1) cation.9 Study of the Stability The influence of time on the cerium(II1) - MTB system, at room temperature and in the presence or absence of light, has been studied using the experimental conditions given in the536 SHORT PAPERS Autalyst, Vol. 108 procedure. After the preparation of the complex, the absorbance decreases for 5 min. The comdex then remains stable for 25 min ; after this the complex starts to be degraded.The stability is not affected by the order of addition ofthe reagents or by-an excess of MTB. Effect of the Concentration of Ethanol absorbance maximum of the complex and the reagents at 600 nm is shifted to 605 nm. hypochromic effect is also observed, being more intense in the complex spectrum. stant (&) of the medium is observed. favour the decomposition of the complex by precipitation. If ethanol is added to the solutions prepared according to the given procedure, then the A A linear relationship between the decrease in the absorbance value and the dielectric con- Ethanol contents of higher than 25% (0, about 66) The decrease in the absorbance value can be expressed by the following equations.Reagent, MTB (correlation coefficient = 0.991) : A605 = -0.146 + 0.0140c Complex, cerium(II1) - MTB (correlation coefficient = 0.997) : A605 = -0.227 + 0.0200, Composition and Formation Constant of the Complex ratio, straight line and continuous variation methods. metry of the cerium(II1) - MTB complex was found to be 1 : 2. The stoicheiometric composition of the complex has been studied by application of molar At 625 nm and pH 10.2, the stoicheio- Assuming that the formation of the complex can be expressed by Ce(OH), + 2H2MTB4- e [Ce(OH)(HMTB),I8- + 2H20 the value of the formation constant of the 1 : 2 complex has been calculated from the data of the continuous variations applying the method of Likussar and BoltzlO through the expression IOgK,,, = 0.352 2 - 210gK - IOgY,,,.- 310g(l - Ymax,) where the value of K is 0.04 mM and Y is the normalised absorbance.ll tion constant of 3.5 x 1012 mol 1-2 has been obtained. Calibration Graph Beer’s law is obeyed at pH 10.2 and at 625 nm for cerium(II1) concentrations in the range 0.09-3.4 p.p.m. [b (internal path length) = 1.0 cm, glass cells], the molar absorptivity being 2.7 x lo4 1 mol-l cm-l. According to Ringbom and Ayres12 method, the range of the mini- mum spectrophotometric error for 1% T (transmittance) is 1.2 to 3.4 p.p.m. of cerium(II1). This method is more useful than the one discussed in reference 4, because of the molar absorptivity value, E . However, comparing this method with that using Xylenol Orange3 ( E = 3.4 x lo4 1 mol-l cm-l), the proposed method shows a lower sensitivity, but is more selective.Other methods used to determine cerium(III), such as the one which uses formal- doximel or the one based on the use of quinolin-8-ol,2 show values for the molar absorptivity of about lo3 1 mol-l cm-l, and therefore have a lower sensitivity. Statistics of the Results The statistical parameters of the proposed method are as follows: arithmetic mean (ft), 2.23 p.p.m. of cerium(II1) ; standard deviation (s), 0.047 p.p.m. of cerium(II1) ; relative stand- ard deviation ( s ) , 2.1%; standard deviation of the mean valte (slp), 0.014 p.p.m. of cerium- (111) ; relative error of the mean value, 1.4% ; confidence limit ( X & s~ t ) , 2.23 & 0.03 p.p.m. of cerium(II1) ; degrees of freedom, 10, t = 2.23 (95% probability level).Interferences The interference of foreign substances was studied by mixing cerium(II1) solution, the foreign substances, sodium oxalate, potassium cyanide, glycine buffer and reagent, in that order, and then diluting according to the procedure. A value for the forma-A$ril, 1983 SHORT PAPERS 537 Anionic substances were added at 100- and 10-fold molar excesses over that of cerium(III), in order to investigate their possible masking effect on this cation. No interference is shown by CO,”, tartrate, NO3-, NO2-, SO,2-, S,032-, SCN-, F-, C1-, B r , I-, BO,, Cr042- and Clop, but POZ- interferes. The study of the interference due to cations was carried out at a 10-fold molar excess over that of cerium(II1). The following ions show no interference: Ag(I), Zn(II), Hg(II), Sn(II), Co(II), Cd(II), AI(III), Th(IV), Bi(III), U(VI), Ni(II), Na(I), K(1) and NH,+.Interference is shown by Cu(II), Mg(II), Ca(II), Sr(II), Ba(II), Cd(II), Hg(I), Al(III), lanthanoids, Zr(IV), Th(IV), Pb(II), Bi(III), Cr(III), U(VI), Mn(II), Fe(I1) and Fe(II1). No effect of the ionic strength is shown by sodium perchlorate, sodium nitrate and potassium nitrate at concentration levels: p smaller than or equal to 0.5 M for the perchlorate and less than 0.6 M for both of the nitrates, p being the ionic strength of the whole solution. In this study an interference has been considered as any substance that shows any variation in the measurement of the absorbance that corresponds to a concentration of cerium(II1) (expressed as parts per million) greater than 2.5-fold the value of the standard deviation.This standard deviation has been calculated by multiple analyses of cerium(II1) solutions with no interferences. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. References Marczenko, Z., Anal. Chim. Ada, 1964, 31, 224. Roberts, J. E., and Ryterband, M. J., Anal. Chem., 1965, 37, 1585. Serduyk, L. S., and Smirnova, V. S., Zh. Anal. Khim., 1964, 19, 451. Cabrera-Martin, A., Cartagena, M. C., and Peral-Fernandez, J. L., An. Quim., 1980, 76B, 294. Ringbom, A., “Complexation in Analytical Chemistry,” Interscience, New York, 1963, p. 324. Sillen, L. G., and Martell, A. E., “Stability Constants of Metal-Ion Complexes, Section TI, Organic Cefola, M., Tompa, A. S., Celiano, A. V., and Gentile, P. S., Inorg. Chem., 1962, 1, 290. Tamer, S. P., and Choppin, G. R., Inorg. Chem., 1968, 40, 575. Kragten, J., “Atlas of Metal-Ligand Equilibria in Aqueous Solution,” Ellis Homood, Chichester, Likussar, W., and Boltz, D. F., Anal. Chem., 1971, 43, 1265. Momoki, K., Sekino, J., Sato, H., and Namaguchi, N., Anal. Chem., 1969, 41, 1286. Aynes, G. H., “Quantitative Chemical Analysis Part IV,” Second Edition, Harper and Row, New Received July 23rd, 1982 Accepted October 29th, 1982 Ligands,” The Chemical Society, London, 1964, p. 361. 1978, p. 170. York, 1968, Chapter 31.

ISSN:0003-2654    

DOI:10.1039/AN9830800534

出版商:RSC    

年代:1983

数据来源: RSC

摘要:

A$ril, 1983 SHORT PAPERS 537 Potentiometric Titration Method for the Simultaneous Determination of Two Monofunctional Weak Acids of Similar Strength: Titration to the Initial Potential Hilda Szalai and Tamas L. Paal National Institute of Pharmacy, POB 450, Budapest 5, H- 1372 Hungary Keywords : A cid-mixture determination ; Potentiometric titration ; titration to initial potential It is well known that only the total acid concentration of mixtures of weak acids with similar protonation constants ( ApK <4) can be calculated from the inflection point of their potentio- metric titration curve. However, using a numerical or a more sophisticated graphical evaluation, the concentration of the individual acid components can also be determined. A number of more or less difficult curve evaluation procedures have been described.l-* This paper describes a simple method for the determination of acid components of mixtures of two monofunctional weak acids (5>pK>1) of similar strength.538 SHORT PAPERS Experimental All chemicals were of analytical-reagent grade.Analyst, VoJ. 108 Instrumentation A precision digital pH meter, Type OP-208 (Radelkis, Hungary), an OP 7183 glass electrode (Radelkis), a saturated calomel electrode with a Wilhelm bridge and a Type U-10 ultrathermostat (VEB Priifgerate Werk, GDR) were used. Procedure A 20.00-ml volume (V,) of a mixture of two monofunctional weak acids, HA and HB, of total acid concentration (c,) 5-50m~ and ionic strength 0.10, adjusted with sodium per- chlorate, was placed in a thermostated (25 & 1 “C) vessel.The apparent pH of the solution shown by the instrument was measured with a precision of 0.001 pH unit (calibration of the pH meter with buffers was not necessary). The solution was diluted to 30.00 ml (V,’) with 0.10 M sodium perchlorate solution (the apparent pH changed). The mixture was then titrated with a standard solution (of concentration cHA M and ionic strength 0.10) of the HA acid component up to the initial apparent pH (VHA ml represents the volume of the acid solution consumed). The whole experiment was repeated with another volume V o of the mixture to be analysed but it was titrated with a standard solution (cHB M, ionic strength 0.10) of the other acid component HB ( V H B ml was consumed). The total acid concentration, c,, was determined by means of a common potentiometric titration with standard sodium hydroxide solution; C, was calculated on the basis of the inflection of the titration curve.Calculation of Results the sample mixture: Let cA and cB denote the unknown concentrations of the individual acid components of .. * ‘ (1) VHA CEA [c, ( V O - V O ’ - VHB) + VTHB CHB] CA = VHB cHB (Vo’ - Vo + VHA) - V/HA cHA (Vo‘ - Vo + V n B ) Equations (1) and (2) can be derived in the following way. The hydrogen-ion concentra- tion of the undiluted mixture, corresponding to the measured electrode potential, is given by where KA, KB, [A] and [B] are the dissociation constants and the equilibrium anion con- centrations of the two acid components, respectively. After dilution and titration with standard HA solution (subscript 1) or with standard HB solution (subscript 2) to the initial apparent pH, corresponding to the same [H+j0 we have From the law of electroneutralityA@ril, 1983 SHORT PAPERS Combination of equations (3), (4) and (5) yields V O cA v o + vHA CHA - [H+]O + ([H'IO - [Ale) vo' + vHA CA - - vO' + lJHA - CAI 0 'VO 539 On the basis of equations (7) and (8), equation (1) can be derived.Concentration of Standard Acid Titrant Solution The acid concentrations cHA and cHB should be chosen so as to ensure the attainment of the initial potential with a reasonable consumption of the titrating solutions. Naturally, a more concentrated standard solution of the weaker acid component should be used. The actual values of cHA and cHB depend on the expected individual acid component concentra- tions in the analyte and on the approximate ratio of the two protonation constants.Using the acid concentrations presented in Table I for nine acid mixtures for three different ApK series the consumptions of the standard acid solutions are between 3.5 and 15.0 ml. TABLE I RESULTS FOR pK values: acetic acid, Concentration in sample/mM & Mixture* CA CB I .. . . 10 10 1 10 1 I1 . . .. 10 10 2 20 10 1 I11 . . .. 5 5 2 20 5 0.5 ACID MIXTURES OF DIFFERENT COMPOSITION 4.645; chloroacetic acid, 2.666; formic acid, 3.90'; and trichloroacetic acid, 1 .30.8 Concentration in measuring standard solu tion/mM & CHA CHB 20 200 4 40 20 200 20 2 000 4 200 20 2 000 10 2 000 4 2 000 10 2 000 Results as yo of amounts added (92 = 5) & 99.5 100.5 100.2 100.0 99.4 105.7 100.4 99.6 99.3 100.1 99.8 101.9 99.7 100.3 100.2 100.0 99.9 100.5 CA CB Relative standard deviation (SD), % (n = 5) - 0.60 0.60 0.87 0.08 0.24 2.40 0.90 0.90 1-00 0.10 0.50 4.70 0.60 0.60 0.90 0.11 0.38 3.67 SDA SDB * I = Formic acid (A) - acetic acid (B), ApK = 0.74; I1 = chloroacetic acid (A) - acetic acid (B), ApK = 1.98; and I11 = trichloroacetic acid (A) - acetic acid (B), ApK = 3.34.Results and Discussion The examples in Table I illustrate the performance of the suggested method. Relatively high bias and poor precision were obtained only with mixtures containing the weaker acid component in a small proportion (less than 10%). However, the stronger component could even be determined using these mixtures.With other component ratios the method proved to be very accurate and precise. I t should be noted that in order to achieve the precision illustrated in Table I the equi- librium electrode potential really should be measured, taking into account the response time of the measuring cell.540 SHORT PAPERS Analyst, Vol. 108 Conclusion The proposed method can be used when the analyte contains only two known acids (or bases) and the task of the analyst is to determine their exact concentrations. Standard solutions of these acids are also necessary. (This situation generally occurs in pharma- ceutical analysis, in process control, etc.) The procedure is time consuming, involving three titrations at a constant temperature. However, the proposed simple method has advantages that can be particularly useful in routine work in most analytical laboratories.Unlike other potentiometric methods suggested for the solution of similar problems, our method requires neither the measurement of exact pH values (and consequently calibration of the electrodes) nor a knowledge of the exact values of the protonation constants. On the basis of the principle of the suggested procedure, the determination of the com- ponents of bifunctional acid or base mixtures and experiments in non-aqueous media also seem to be possible. Further investigations are in progress. 1. 2. 3. 4. 5. 6. 7. 8. References Ivaska, A., Talanta, 1974. 21, 1167 and 1175; 1975, 22, 995. Seymour, M. D., Clayton, J. W., Jr., and Fernando, Q., Anal. Chem., 1977, 49, 1429. Arena, G., Rizarelli, E., Sammartano, S., and Rigano, C., Talanta, 1979, 26, 1. Ivaska, A,, and NagypAl, I., Magy. Ke'm. Foly., 1980, 86, 84. Feldman, I., and Koval, L., Inorg. Chem., 1962, 1, 293. Ahrland, S., Acta Chem. Scand., 1949, 3, 783. Martin, D. L., and Rosotti, R. J. C., Proc. Chem. Soc., 1959, 60. Day, R. A., Jr., and Stoughton, R. W., J. Am. Chem. Soc., 1950, 72, 5662. Received August 2nd, 1982 Accepted October 29th, 1982

ISSN:0003-2654    

DOI:10.1039/AN9830800537

出版商:RSC    

年代:1983

数据来源: RSC

摘要:

540 SHORT PAPERS Analyst, Vol. 108 Rapid Routine Procedure for the Determination of Anhydrous and Hydrated Tripolyphosphate by means of X-ray Powder Diffraction Alessa ndro Ma ngia Istituto di Chimica Generale ed Inorganica, Universitic di Parma, V i a M . D’Azeglio 85, 43100 Parma, Italy Keywords : Tripolyphosphates determination ; X-ray powder diflraction Sodium tripolyphosphate (Na,P,O,,,) , one of the most important components of detergents, can exist as one hexahydrate and two anhydrous forms, one of which is stable over 400 oC.1$2 The relative content of these forms in detergent formulations influences the physical properties of the commercial product. Because of this, it is necessary to have a reliable analytical pro- cedure to determine the anhydrous form in the presence of the hydrated one, which can be used as a means of production control. The methods used are : ion-exchange chromatography with spectrophotometric dete~tion,~-~ exclusion chromatography with electrochemical detection7,* and flow injection analysi~.~,*~ These methods achieve the separation of polyphosphates from ortho- and pyrophosphates, but naturally do not allow the distinction between anhydrous and hydrated forms.The procedure of determination for the relative amount of anhydrous polyphosphate with respect to the hydrated one, which is usually applied in the control of industrial processes, is based on the determination of the total and of the adsorbed water content. The water corresponding to the difference is assumed to be bonded to the polyphosphate. The percentage of the hydrated form, and consequently of the anhydrous one, is deduced from the amount of bonded water and from the amount of total tripolyphosphate. Several examples of the determination of tripolyphosphates have been reported.Afiril, 1983 SHORT PAPERS 541 In this paper, a direct determination of the hydrated and anhydrous forms is proposed using X-ray powder diffraction. The procedure can be applied to the routine determination of both forms or to determine Na5P3010 in the presence of an excess of hexahydrated sodium tripoly- phosphate (Na5P,Ol0.6H2O).Experimental A Philips PW 1050 vertical goniometer, with a device for sample rotation, and a PW 1010 generator, Cu Ka (nickel filtered) radiation, were used at 40 kV, 20 mA; divergence lo, anti- scatter 0.2 mm, receiving slits 1".A proportional counter and a pulse-height discriminator were used. Intensities were calculated by integrating graphically the peak areas or were measured at the 20 value of the maximum for variable time in such a way as to have a standard deviation of about 1%. Results and Discussion Anhydrous and hydrated polyphosphates have the following crystal data : Na5P3Ol02'11 monoclinic, space group C2/c (a = 16.00, b = 5.24, c = 11.25 A, /3 = 93.0", 2 = 4); and Na,P,0,0.6H,012 triclinic, space group Pi (a = 9.799, 13 = 10.30, c = 7.562 A, a = 94.45, 18 = 92.20, y = 90.99", 2 = 2). 2e,o 8.70 11.75 12.60 14.05 15.10 16.55 16.80 17.30 18.10 TABLE I X-RAY POWDER DIFFRACTION DATA OF NA,P,O,~.GH~O Cu Kor radiation, leaks having I / I , > 20 are reported. dlA 1/11 hkl 28," d l ~ I l I , 10.2 100 010 20.10 4.42 30 7.55 22 001 22.11 4.02 28 7.03 24 110 26.78 3.46 20 6.30 90 o i i 28.15 3.17 25 5.87 22 011,101 29.30 3.06 60 5.35 25 1 ii 30.52 2.93 80 5.28 30 iii 32.18 2.78 40 5.13 30 020,ii i 34.10 2.63 50 4.90 70 200 34.96 2.57 55 hkl 0 2 i 201,12i 02z 151 202,022 230,320 252,32i 040 012 In Tables I and I1 the X-ray powder diffraction data of the two compounds are reported.Indexing of the anhydrous form has been obtained starting from published crystal data2 and from observed 20 values. A least-squares procedure of refinement was used, following the method of Vogel and Kempter,13 with the extrapolating function of Nelson and Riley.14 From the diffraction data it can be deduced that, even if some peaks overlap, the two forms can be identified easily.20," 10.80 18.65 19.40 19.75 23.55 24.02 24.75 28.02 29.55 TABLE I1 X-RAY POWDER DIFFRACTION DATA OF NA5P3010 (FORM 11) Cu Kct, peaks having 1/11 > 10 are reported. dlA 1/11 hkl 20," d / A 1/11 8.2 10 200 30.08 2.97 15 4.75 90 202 33.22 2.69 100 4.57 90 111 34.02 2.63 60 4.49 15 202 34.50 2.60 65 3.77 12 112 35.80 2.51 12 3.70 15 112 36.50 2.46 18 3.59 20 5 1 37.02 2.43 30 3.18 12 402 44.80 2.02 20 3.02 40 i i 3 47.50 1.91 20 - - - hkl 113 204 611 313 - 512 i T 0 2 , i ~ 114,221 422 024,614542 SHORT PAPERS Analyst, Vol. 108 As the most relevant problem from the point of view of application is the determination of the relative amounts of Na,P,O,, and Na,P,O,.GH,O, the relationship between X-ray diffrac- tion intensities and percentage of each form in a binary mixture was deduced.The most suitable ones appear to be those of 20 = 18.10" for the hexahydrate and 20 = 19.40" for the anhydrous form, even if they are not the most intense. These peaks, having near 20 values, allow the running of the spectrum in a narrow range of 20. Their intensities are affected in the same way by the instrumental conditions and they are free from interference by other usual com- ponents of commercial detergents such as sodium sulphate. The graphs of peak area (mm2) veysus percentage (m/m) for each form, expressed as P205 in a binary mixture, are linear for both substances. For Na,P,O,, (20 = 19.40') the data fit the equationy = 9 . 4 9 ~ + 35.69, with r2 = 0.99832 and standard deviation of the slope s = 0.19 for n = 7.= 0.99695 and the standard deviation of the slope, s = 0.18 for n = 7. Fig. 1 shows the diffraction spectrum in the range 20 = 15-22" of a mixture containing approximately 10% of Na,P,O,,. The relative amount of the anhydrous form in the total polyphosphate was determined in a powder obtained in the spray-drying process of the production of a commercial detergent. The determination was carried out on six samples, which had been dried in different conditions and this was repeated five times for each sample. The ratios of the integrated intensities of the peak at 18.10 and 19.40' were measured. These ratios were compared with those of a graph (Fig. 2) obtained by plotting the intensity ratios of the two peaks versus the percentage of the anhydrous form in a binary mixture.This graph, even if it does not add a particular meaning to the linear dependence on the percentage of each single compound, allows us to deduce directly the mass fraction of each form. It is to be noticed that the use of a relative measure of intensities makes the absorption effects negligible. The physical treatment to which the samples were subjected eliminates the effects of preferential orientation and good reproducibility was obtained ; the relative standard deviations of the results of the five independent determinations on each sample ranged between 6 and 3% at the minimum and maximum content of Na,P,O,,. Some peaks in the diffraction pattern of each compound were considered. For Na,P30,,.6H,0 (28 = 18.10"), y = 6.80~ + 1.13, This compound could be detected down to about 2%.B I I I I I I 21 20 19 18 17 16 28: Fig. 1. X-ray diffraction pattern of a spray-dried slurry containing 10% of anhydrous tripolyphos- phate: A, peaks of the anhydrous form; and B, peaks of the hexa- hydrate form. 3 0 2 2 -- 2 - - 1 0 20 40 60 80 100 [Na5P3Ol0.6H201, % m/m Fig. 2. Dependence of the ratio of the intensities of the peaks at 28 = 18.10" (Na,P,010.6H,0) and at 28 = 19.40" (Na,P3010) on the percentage (un/m) of Na,P3Ol0.6H,O expressed as PaOs. The results obtained with this procedure were checked by applying the standard additions method to the same samples. Three different amounts of anhydrous tripolyphosphate, corresponding to about 5, 10 and 20% of the total polyphosphate content, were added to different portions of each sample.The peak at 28 = 19.40" was used; results of the triplicate determinations were averaged and the comparison of the values obtained following the two procedures is shown in Table 111.A@&?, 1983 SHORT PAPERS TABLE I11 DETERMINATION OF ANHYDROUS TRIPOLYPHOSPHATE IN SPRAY-DRIED SLURRIES 543 Percentage of anhydrous form in total tripolyphosphate expressed as P,O, A I \ Intensity ratio Standard additions Sample procedure method 5 8 10 20 26 31 7 9 12 21 25 29 These results indicate that X-ray diffraction is a suitable method for the determination of anhydrous as well as hydrated polyphosphate; in particular the ratio of the intensities can be used as a very rapid procedure for the determination of anhydrous polyphosphate in routine production control.1. 2. 3. 4. 6. 6. 7. 8. 9. 10. 11. 12. 13. 14. References “Gmelins Handbuch der Anorganische Chemie,” 16, P, [c], Verlag Chemie, Weinheim, 1965, p. 239. Davies, D. R., and Corbridge, D. E. C., Acta Crystallogr., 1958, 11, 315. Pavanello, M., Bedendo, A., and Bertini, A., Riv. Ital. Sostanze Grasse, 1977, 54, 364. Gohla, W., Nielen, H. D., and Sorbe, G., GIT Fachz. Lab., 1979, 23, 89; Anal. Abstr., 1979, 37, 3C73. Hiray, Y., Yoza, N., and Ohashi, S., J . Liq. Chromatogr., 1977, 2, 677. Yoza, N., Ito, K., Hirai, Y., and Ohashi, S., J . Chromatogr., 1980, 196, 471. Tanaka, K., and Ishizuka, T., J . Chromabogr., 1980, 190, 77. Waki, H., and Tokunaga, Y., J . Chromatogr., 1980, 201, 259. Hirai, Y., Yoza, N., and Ohashi, S., Anal. Chim. Acta, 1980, 115, 269. Yoza, N., Kurokawa, Y., Hirai, Y., and Ohashi, S., Ana2. Chim. Acta, 1980, 121, 281. Powder Diffraction File, International Centre for Diffraction Data, JCPDS, Card 11-652. Powder Diffraction File, International Centre for Diffraction Data, JCPDS, Card 10-186. Vogel, R. E., and Kempter, C. E., Acta Crystallogr., 1961, 14, 1130. Nelson, J. B., and Riley, D. P., R o c . Phys. SOC., London, 1945, 57, 160. Received July 26th, 1982 Accepted September 20th, 1982

ISSN:0003-2654    

DOI:10.1039/AN9830800540

出版商:RSC    

年代:1983

数据来源: RSC

摘要:

A@&?, 1983 SHORT PAPERS 543 Catalimetric Determination of Iodine in Common Kambhampati Sriramam,* Brahmandam S. R. Sarma, Agnihotram R. K. Vara Prasad and Konidena Kalidas Department of Chemistry, Nagarjuna University, Nagarjunanagar 622 610 ( A .P.), India Keywords : Iodine determination ; common salt analysis ; catalysis ; hexa- chloroantimonate ( V ) hydrolysis For dietary and health reasons, it is important to have a sensitive method for the determina- tion of iodine in common salt. Spectrophotometric, activation and catalytic methods are well suited for the determination of iodine at the low concentration levels required (ca. 5 pg gl). We describe here a sensitive and specific catalytic method for the routine determination of iodine in common salt. The principle used is, however, different from the usual oxidation - reduction catalysis often employed in conjunction with indicator reactions such as that with arsenic(II1) - cerium(1V).It is based on the catalytic action of iodide on the hydrolysis of hexachloroantimonate(V) in dilute hydrochloric acid solutions. * To whom correspondence should be addressed.544 SHORT PAPERS Experimental Analyst, Vol. 108 Apparatus The apparatus used included a Beckman DU-2 spectrophotometer with matched silica cells of 1-cm path length. The spectrophotometer cell was thermostated by circulating water (30 & 0.2 "C) through the cell holder. Reagent solutions were kept in separate ground-glass stoppered vessels and placed in the thermostat for 30 min before they were mixed to start the reaction.The transfer of solution from the thermostated standard flask to the optical cell was done manually but as quickly as possible and no extra precautions were observed. The order of mixing the reagents does not affect the results of the kinetic experiments. Reagents distilled water prepared with an all-glass apparatus was used. antimonate [KSb(OH),] in 1 1 of concentrated hydrochloric acid. iodate solution according to Andrew's procedure, described by Voge1.l Analytical-reagent grade chemicals were used without further purification. Doubly HexachZoroantimonate( V ) solution, 0.005 M . Prepared by dissolving 1.3145 g of potassium Potassium iodide solution, 0.01 M. Standardised titrimetrically with standard potassium Sodium chloride solution, 2 M. Sample solution water and diluted to volume in a 100-ml calibrated flask.market were used without further treatment. An 11-12-g amount of common salt was accurately weighed and dissolved in doubly distilled Samples purchased from the local Results and Discussion Analysis of kinetic data revealed that the hydrolytic reaction of hexachloroantimonate(V) under the influence of iodide catalyst follows a different course from that of the uncatalysed reaction. The difference is reflected in the observed order of reaction with respect to SbCb-; its order was reported to be one in the uncatalysed reaction2 but it is zero in the iodine-catalysed reaction as the reaction rate is independent of [SbCl,-]. The order with respect to I- ion is one under our experimental conditions.Further, the pseudo-zero-order rate constant is inversely dependent on [H+] when [I-] is kept constant. The results of the kinetic study have been incorporated in Table I. The final reaction product is an equilibrium mixture of ions of the type Sb(OH),CI-,-,, the composition being a function of hydrochloric acid concentration. The observations are consistent with the following mechanism, which indicates that the reaction proceeds by both pH-dependent and -independent paths : SbC1,- + 1- __+ SbC1,I- + c1- .. .. ' (1) fast kl slow SbC1,I- + H,O SbC1,OH- + Hf + I- . . .. (2) k* slow SbC1,I- + OH- -+ SbC1,OH- + I- . . . . * * (3) SbClSOH- + (x-l)H,O - Sb(OH),CI-,-, + (x-~)H+ + (x-1) C1- . . fast (4) SbC1,I- is a precursor complex and the rate is determined by its hydrolytic fission.Step (1) is considered to be spontaneous and complete within the time of mixing so that at any time during the kinetic run [SbCl,I-] = [I-lO. Step (4) was considered to be fast in the mechanism of the uncatalysed reaction also. This leads to the following rate law: Rate = k,[SbCI,I-] + K,[SbCl,I-] [OH-]April, 1983 SHORT PAPERS 545 TABLE I RESULTS OF KINETIC STUDY (30 & 0.2 "c) Ion SbC1,- . . I-.. .. H+ .. Concentration Slope Concentration of other Type of of the range/ reactants/ Ionic Plot, graph Order of on mol 1-1 mol I-' strength y vs. n (linear) reaction y-axis 1.0 x 10-4 [I-] = 3.2 x 10-e 1.0 Absorbance -0,025 0 0.85 1.5 x 10-4 [H+] = 1.0 vs. 1.28 2.0 x 10-4 time (min) 1.70 Intercept . . 4 x 10-'-32 x 10-7 [SbCl -1 = 1 X lo-' 1.0 3 + logk 1 1 -0.06 [H+] "= 1.0 us. 7 + log[I-l, 0.3-1.0 [SbCI,-] = 1 x 1.0 k, x loa 0.64 -1 1.25 [I-] = 2.0 x 10-6 us.l/lH+I [SbCl,I-] [H+l = kl[SbC1,I-] + k2Kw = k, when [I-], and [H+] are constant. Treatment based on step (1) as an equilibrium reaction (with ke', the equilibrium constant at constant chloride concentration) would also lead to an identical result if ke' [SbC16-]>>1. In view of the linear dependence of the rate on [I-lO, as evident from the rate law expression, the reaction was tested to see if it can be used for the determination of iodide at low concentra- tions. When a graph of KO x lo2 vcysus [I-], x lo7 was used as the calibration graph, good results were obtained in the analysis of unknowns; the calibration graph can be constructed from the observed relationship k, = 8.25 (s.d.= 0.05) x 103[1-],, where k , = change of absorbance per minute, [I-], is in moles per litre and s.d. is standard deviation. The method is highly selective as most of the common ions do not interfere. The effect of foreign ions on the analytical result has been tested while following the kinetics at an iodide concentration of 10.27 pg per 50 ml. In particular it is free from interference from ions such as NO3-, SO,2-, PO,3-, Zn2+ and Pb2+ even when they are present in amounts as large as 1000 times the concentration of iodide being determined; bromide, however, interferes if it is present in more than a 100-fold excess and mercury(I1) even in trace amounts seriously inter- feres in the determination. The method has the advantage that the pseudo-zero-order reaction rate is insensitive to variation in the concentration of one of its reactants, viz., SbC16-.Interference from the simultaneous occurrence of background reaction is also virtu- ally negligible, as is evident from the observation that the line in the calibration graph passes through the origin. Iodine in the form of 103- does not catalyse the hydrolytic reaction, so it is necessary to reduce it to I- before it is introduced into the reaction mixture. The reduction is effected with zinc dust (150 mg) and 20 ml of 2 M hydrochloric acid in a stoppered iodine flask, the reaction being allowed to take place until the zinc has dissolved completely at room temperature (ca. 30 "C). The amounts of zinc and hydrochloric acid are taken into account in the calcula- tion of ionic strength and acid concentration for the kinetic run.If some I- is also present in the test solution, the final result includes this also, and the procedure works for total iodine concentrations of 2-20 pg per 50 ml. The prior reduction of iodate is also carried out, using sulphurous acid as reducing agent, and the procedure is similar to that described by Kolthoff . 3 9 4 The results are as accurate as those obtained with the zinc reduction method. The proposed method was found to be applicable to the determination of iodide in various samples of common salt, as shown in Table 11, and the procedure is as follows. Into each 50-ml calibrated flask place 6.3 ml of 6 M hydrochloric acid, 1 rnl of 5 x M hexachloroantimonate(V) and 20 ml of 2 M sodium chloride solution.To the above reaction mixture add an aliquot of iodide solution containing 2-20 pg and then dilute to volume with doubly distilled water. Follow the kinetics of the reaction at 30 & 0.2 "C by measuring the absorbance of the reaction solution at 270 nm against 1 M hydrochloric acid reagent blank at546 SHORT PAPERS TABLE I1 Analyst, Vol. 108 DETERMINATION OF IODIDE I N COMMON SALT Results are iodide found in micrograms per gram of sample and are the mean values of three determinations. Sample No. Fes+ - SCN- method5 Present method 1 2 3 3.02 3.56 3.78 3.06 3.58 3.80 regular time intervals. The graph of the pseudo-zero order rate constant, k, x lo2, against [I-], x 107 is used as the calibration graph and is constructed from the observed relationship k, = 8.15 (s.d.= 0.03) x 103[1-],, where k, = change of absorbance per minute and [I-], is in moles per litre. In the analysis of the sample by the above procedure, use 20 ml of test solution in place of standard iodide solution, but omitting the addition of sodium chloride reagent solution. Sea water is reported to contain both I- and 103-; however, the proportions vary according to location and depth.6 Because of the long and cumbersome process involved in making solar salt, the composition of the iodine forms in the final product is difficult to ascertain or to relate to that in the sea water from which it was made. However, our results indicate that in the batch of samples analysed, it is in the form of iodide as no prior reduction of iodate was effected in our method of determination. Further, the salt sample gave identical results for the content of iodine with and without the reduction step included in the procedure, again confirming our conclusion. The method is evidently applicable to the determination of iodine in iodised salt tablets recommended for medical use. Two of us (A.R.K.V.P. and K.K.) thank CSIR (India) for the award of research fellowships. References 1. 2. 3. 4. 6. 6. Vogel, A. I., “A Text Book of Quantitative Inorganic Analysis,” Longmans, London, 1975, p. 375. Neumann, H. M., and Ramette, R. W., J . Am. Chem. SOC., 1956, 78, 1848. Kolthoff, I. M., 2. Anal. Chem., 1924, 64, 260. Kolthoff, I. M., and Belcher, R., “Volumetric Analysis,” Volume 111, John Wiley, New York, 1957, Jwasaki, J., Usumi, S., and Ozawa, T., Bull. Chem. SOC. Jpn., 1953, 26, 108; Chem. Abstr., 1953, 47, Richard, A. B., and Thomas, G. T., Deefi-Sea Res., 1960, 7 , 24. p. 48. 12123. Received April 23rd, 1982 Accepted October 4th, 1982

ISSN:0003-2654    

DOI:10.1039/AN9830800543

出版商:RSC    

年代:1983

数据来源: RSC

摘要:

546 SHORT PAPERS Analyst, Vol. 108 Some Comments on Calibration Procedures Allan G. C. Morris ESAB Ltd., Beechings Way, Gillingham, Kent, ME8 6PU Keywords : Calibration firocedures ; X-ray fluorescence spectrometry There appears to be a misunderstanding among some authors of X-ray fluorescence (XRF) spectrometry publications1s2 as to which variable is to be taken as the independent variable for calibration purposes. The authors of both of the above-cited publications chose to take the relationship between concentration C of an element, which gives an XRF count intensity I , as C = mI + B, that is, they regressed C on to I and then used this equation for prediction purposes. Hence they chose, incorrectly, to take the concentration C as the dependent variable and the intensity I as the independent variable.The correct procedure, however, isA@&?, 1983 SHORT PAPERS 547 to do exactly the reverse, namely to choose the count intensity I as the dependent variable, the concentration C as the independent variable and to regress I on to C to form an equation I = mC + B and then, for predictive purposes, to transform the equation to C = ( I - B)/m. As this problem is of universal interest in the preparation of all types of calibration graphs, it seems appropriate, therefore, if the error in the above treatment is discussed in some detail and its mathematical fallacy pointed as it is, perhaps, unknown to those analysts without the necessary mathematical statistical training. His con- clusion was, for reasons not entered into here, that the fitting should be done in terms of the deviations that actually represent “error.” Thus, when the X values are selected in advance and kept to without error, then the errors are in the Y values and an equation of the form Y = a + bX should be fitted and the inverse of this relationship used to estimate X from a given Y .This method was therefore recommended in all subsequent standard text-books on regression analysis and, in fact, is still recommended today. In 1967, Krutchkoff entitled the method suggested originally1s2 the “inverse approach” and claimed that, based on the criterion of mean squared error, this approach was superior to that of Eisenhart, the standard or “classical” approach. However, for several reasons, both mathematical and practical and omitted here for brevity, this conclusion of Krutchkoff was refuted by various workers,7-1° who showed it to be incorrect and who concluded that the standard method was the correct one to use in calibration techniques.In the light of these facts, it is surprising that this alternative method propounded by Krutchkoff should now once more be advocated. Although, as the above ideas are far from new, it might be thought that they are common knowledge, in practice it appears not to be so and it is thought timely to reconsider the under- lying ideas behind the choice of the correct variable in calibration problems. In order for the mathematical theory of regression analysis to be valid, the dependent variable has to be subject to random error and not subject to selection, and a variable that is errorless, or that has been subject to selection, cannot be chosen as the dependent variable.The dependent variable is the one whose values are distributed at random, in the statistical sense, about the regression function, so that its expected value is some function of the observed values of the independent variable. The values of the independent variable need not be randomly distributed but may have been fixed or selected in any manner. It is not a random variable, in the statistical sense. It is also sufficient that the errors in X are small compared with the errors in Y . In other words, we assume that the random variation, if any, of the independent variable is so small compared with the range of that independent variable that we can effectively ignore the random variation.This assumption, rarely stated, is implicit in all least-squares work in which the independent variables are assumed “fixed.” The word “fixed” means “not random variables’’ in such a context ; it does not mean that the independent variables cannot take a variety of values or levels. For example, in the calibration of electrical moisture meters for the determination of moisture in wood, timber has to be conditioned to equilibrium moisture content under a number of different temperature and humidity conditions and its electrical resistance deter- mined. Then, as moisture content has been selected, even though subject to considerable variation at any one condition, it must be taken as the independent variable. The regression of resistance on moisture content will be determined, even though in using the equation for prediction purposes moisture content will be estimated from electrical resistance.When preparing a calibration graph we choose initially a series of fixed points, accurately known, in the range of elemental composition of interest and then determine how the X-ray counts vary at each of these fixed points. These counts include all the procedural errors, or influences, such as matrix effects, and also the random, statistical fluctuations of the X-ray photon counting process. In the light of all that has been said before, it is obvious that the counts must be chosen as the dependent variable, regressed on to the elemental composition values, and then the resulting equation transformed for prediction purposes. It is doubtful whether any X-ray spectroscopist would claim that the fixed, chosen, concentrations of the elements constituting the calibration samples depended on the X-ray counts obtained from those samples.A problem may appear to arise in the use of the standard method if a non-linear equation is the best fit to the calibration graph, as transformation of this equation for prediction purposes The problem itself is far from new and was first discussed by Eisenhart5 in 1939.548 SHORT PAPERS may involve more than one answer. However, as the original calibration graph will be avail- able for inspection, quick reference to it will enable the user to reject the unacceptable answers. It is well known that two regression lines coincide only when all the points fall exactly on a straight line, which is so when a linear functional relationship exists between two variables, both of which can be measured without error. Moreover, the greater the scatter of Y about the regressed line, the greater are the discrepancies between the two regressed lines of Y on X and X on Y .In the present instance it is assumed that a linear functional relationship exists and that the functional relationship, that is, the calibration graph, has been determined from a planned experiment , carried out under carefully controlled conditions. Under these circumstances the variation about the line is small and all three lines are virtually identical.11J2 This, however, does not detract from the fact that, strictly, from linear regression theory, the counts must be regressed on percentage composition when calibration is carried out.That this is not of mere academic interest is shown by dealing with the data taken from Table I in reference 2. If we proceed correctly and regress the intensity counts (Y) on %CaO ( X ) we find: Intensity (Y) = -3052.92029 + 283.226928 %CaO ( X ) which inverts to give %CaO ( X ) = 0.0035307377 x intensity ( Y ) + 10.77906085 For 11000 counts this gives %CaO = 49.617 compared with the 49.581 %CaO quoted. It is accepted that in this instance the difference is well within the precision of the analytical methods generally employed for calcium determinations at this level. Nevertheless, it shows the importance of using the correct regression model, as obviously, if both models were equiva- lent, then identical results should be obtained in each instance.It is possible, therefore, that many of the discrepancies quoted in the literature between values obtained by XRF spectroscopic methods and those obtained by “wet-chemical” methods are due to a large extent to the fact that the incorrect regression model has been used. Finally, the question may well be asked as to whether the computer calibration pro- grams of some commercial XRF spectroscopic systems, or indeed, other spectroscopic systems, are not based on an incorrect model. 1. 2. 3. 4. 5. 6 . 7 . 8. 9. 10. 1 1 . 12. References Jenkins, R., Gould, R. W., and Gedcke, D., “Quantitative X-ray Spectrometry,” Marcel Dekker, New Boinck, J., “Determination of a Linear Calibration Graph with the Aid of Regression Analysis,” Williams, E. J., “Regression Analysis,” John Wiley, New York, 1959. Draper, N., and Smith, H., “Applied Regression Analysis,’’ Second Edition, John Wiley, New York, Eisenhart, C . , Ann. Math. Stat., 1939, 10, 162. Krutchkoff, R. G., Technometrics, 1967, 9, 425. Williams, E. J., Technometrics, 1969, 11, 189. Berkson, J., Technometrics, 1969, 11, 649. Martinelle, S . , Technometrics, 1970, 12, 157. Halperin, M., Technometrics, 1970, 12, 727. Davies, 0. L., Editor, “Statistical Methods in Research and Production,” Third Edition, Oliver and Davies, 0. L., and Goldsmith, P., Editors, “Statistical Methods in Research and Production,” Fourth York and Basle, 1981, p. 440. Technical Bulletin, N. V. Philips Gloeilampenfabrieken, Eindhoven, 1968. 1981. Boyd for Imperial Chemical Industries, London, 1958, p. 153. Revised Edition, Longmans for Imperial Chemical Industries, London, 1977, p. 18 1 . Received September 13th, 1982 Accepted November 30th, 1982

ISSN:0003-2654    

DOI:10.1039/AN9830800546

出版商:RSC    

年代:1983

数据来源: RSC

摘要:

Analyst, A + d , 1983 Book Reviews 549 TOPICS IN ENZYME AND FERMENTATION BIOTECHNOLOGY. VOLUME 6. Edited by ALAN WISEMAN. Pp. 232. Ellis Horwood (distributed by John Wiley). 1982. Price L21.50. ISBN 0 85312 372 1 (Ellis Horwood) ; 0 470 27304 6 (Halsted Press) ; ISSN 0140 0835. The sixth in this series of volumes detailing recent research in biotechnology takes a similar inter- disciplinary approach to that of the earlier books. Several topics are singled out for a quite comprehensive discussion combined, where appropriate, with a review of the relevant literature. Topics covered range from antibiotics to brewing, taking in the use of yeast in the detection of mutagens and carcinogens and microbial oxygenases. Chapter 2 gets more solidly into the subject matter with a discussion of 4 hydroxycoumarin antibiotics.This includes the mode of action of these antibiotics and mechanisms of biosynthesis. In excess of 50 references are given. In Chapter 3, the ubiquitous group of compounds termed secondary metabolites are examined. The description “secondary metabolite’’ covers a wide range of materials, enzymes, pigments and toxins, but this work is concerned primarily with antibiotics. Screening procedures are described elegantly and in great detail and the nature of secondary metabolites is defined. Culture methods for selected organisms in a variety of media are described and some 86 references are presented. Chapter 4 is titled “New Approaches to Enzyme Stabilisation” and gives a brief description of possible stabilisation procedures.The author concludes that polysaccharide stabilisers have the most potential (37 references). Chapter 5 is concerned with developments in beer fermentation and is a fascinating account of current technology in the production of this sometime ambrosial but often insipid product. After a brief account of the position of beer in the market place, a description of the brewing process and yeast physiology is given. The factors affecting yeast growth (oxygen, nitrogen, wort concentra- tion, effect of lipid addition, carbon dioxide and ethanol concentration) are described in some detail. Methods of separating yeast cells from beer, by flocculation and yeast head formation are outlined. Types of fermentation vessel and control of flavour substances are discussed. In all, I found this chapter particularly informative and enjoyable if somewhat thirst making; 130 references are given.Chapter 6 is entitled the “Yeast Mitochondria1 System; A Test for Antimitochondrial Drugs and Mutagens.” I t does exactly as the title suggests, taking the reader through an introduction to the system, a description of genetic aspects of yeasts and a lengthy discussion of numerous antimito- chondrial agents. In Chapter 7, the author discusses a group of enzymes of considerable commercial importance and potential, microbial oxygenases. Both mono- and dioxygenases are discussed at some considerable length. The reaction mechanisms of several transformations catalysed by these enzymes are described. For dioxygenases, dihydroxylation of aromatic compounds and cleavage of aromatic rings are examined together with the action of sulphur dioxygenase and some intermolecular dioxygenases.A wide range of mono-oxygenase reactions are also covered and applications of oxygenase enzymes are noted; over 250 references are given. Chapter 1 is an introduction and survey of patent literature (46 references). The book also contains a cumulative index for the first five volumes of the series. In general, the book is well written and concise, while going into useful detail in some areas. Like the previous volumes in the series, it is a compilation of specialist biotechnological topics, which may prove useful both to biotechnologists and those involved in other fields. B. G. HENSHAW ADVANCES IN STEROID ANALYSIS. PROCEEDINGS OF THE SYMPOSIUM ON THE ANALYSIS OF STEROIDS, EGER, HUNGARY, MAY 20-22, 1981.Analytical Chemistry Symfiosia Series, Volume 10. Pp. xii + 551. Elsevier. 1982. Price $104.75; Dfl225. ISBN 0 444 99711 3 (Volume 10); 0 444 41786 9 (Series). Edited by S. Gorog. One of the many problems that organisers of scientific conferences or symposia have to face is the question of whether the proceedings should be published, and, if so, in what form, with or without financial support and by whom. Those interested in the conference subject would, no doubt, prefer published versions of the papers to appear in the journals they regularly read, and as soon550 BOOK REVIEWS Analyst, Vol. 108 after the event as possible. For a variety of reasons this rarely occurs. Indeed, there seems to be no completely satisfactory means whereby the scientific fruits of a conference may be disseminated quickly, widely and cheaply.Although hardly achieving these ends, publication in book form has the appeal of completeness, convenience and prestige, as well as providing a compact reference work and, perhaps, an elegant souvenir of the occasion. The book under review, Volume 10 in the series of conference proceedings published by Elsevier, covers the scientific programme of the Symposium on the Analysis of Steroids held at Eger, Hungary on May 20-22, 1981. The 69 papers, contributed by authors from 15 countries, cover the two main divisions of the subject, biological - clinical and pharmaceutical steroid analysis. The arrangement of the material in sections according to the analytical technique employed, emphasises the different objectives in the two fields.Although selectivity is important in both, the high, sometimes extreme, sensitivity vital in biomedical applications contrasts with the need for accuracy and precision in pharmaceutical analysis. Reviews or state-of-the-art surveys head some of the sections, but most of the papers are 4-8 page accounts of research work, new or in progress. After an introductory general section (6 papers), protein-binding methods ( 5 ) are grouped to- gether followed by immunological methods, a large section, sub-divided into reviews (2), radio- immunoassays (14) and enzymeimmunoassays (3). Two sections devoted to gas-chromatographic methods come next, one concerned with biological applications (6 papers), the other with the method in combination with mass spectrometry (4).Other chromatographic techniques are well represented, high-performance liquid chromatography in both biological (4 papers) and pharmaceutical (6) contexts and thin-layer chromatography by sub-sections on separation problems (6) and quantitative analysis (7). Short concluding sections cover spectroscopic methods (5) and polarography (2). Most steroids of biological importance are included. While hormones predominate, other groups of natural steroids, notably the D vitamins, bile-acids, cardiac glycosides and steroidal saponins are represented, together with a few biologically active synthetic steroids. The Editor of the volume has succeeded well in assembling the material on the various subjects and techniques, often of a complex nature, into an essentially readable form.The book is well bound and manifestly a prestige publication, though at a commercial price. Some may find the assortment of type faces and inconsistent reference systems a little unexpected in an otherwise well presented volume. Whether from a research or an industrial-development viewpoint, this publication is clearly of interest to all whose work concerns the determination of steroids in biochemical, clinical or pharma- ceutical contexts. Analysts in other fields may find the methodology rewarding, particularly that embraced by the larger sections dealing with immunological and chromatographic techniques. W. H. C. SHAW DETERGENT ANALYSIS. A HANDBOOK FOR COST-EFFECTIVE QUALITY CONTROL.By B. M. George Godwin (A Division of Longman MILWIDSKY and D. M. GABRIEL. Group Ltd.). 1982. Price &20. ISBN 0 7114 5735 2. Pp. xii + 291. Not everyone will agree with the concept of quality control as presented in this book, but there is a lot of merit in the approach adopted and in the cost-effective procedures described therein. The introduction rightly states that “routine analysis is as important as production” and makes it clear that the purpose of routine analysis is to provide quality assurance. However, quality itself is the responsibility of the production department working to an established standard or specifica- tion and it does seem contradictory to say that “unless quality is maintained, the best production techniques are valueless.” The book includes routine analysis schemes in detail and correctly emphasises the point that a minimum number of tests should be selected to ensure that the product meets the agreed specifica- tion, otherwise time and money may be wasted.Chapters concerned with apparatus, standard solutions and reagents, routine analysis, plant control, new and raw materials, environmental considerations and the analysis of unknowns are incorporated and it is particularly good to see the importance of correct sampling and storage stability stressed, including microbiological control. Also there is excellent advice in the introduc- tion to the chapter on the analysis of unknowns.A $ d , 1983 BOOK REVIEWS 551 When reviewing a book full of practical details such as this one, it is not difficult to find criticisms or to suggest personally favoured alternatives to some of the procedures put forward by the authors.For example, split-type volumetric flasks are recommended for the preparation of detergent solutions to overcome foam formation and no mention is made of the procedure recommended by IS0 for detergent powders, in which water is added from a volumetric flask to the sample contained in a large beaker. In the section describing Karl Fischer titrations, special apparatus is recommended but there is no mention of equipment using the coulometric principle. I t is claimed that infrared spectrometry has virtually no application in routine control work. This is surely not true when the analysis of raw materials is included. The chapter concerned with standard solutions and reagents does not mention the possibility of using standard solutions purchased ready for use.These are considered to be satisfactory in many control laboratories. When describing detergent formulations on page 87, optical brighteners and enzymes are included but analytical comment, qualitative and quantitative, is missing. Perfume is not mentioned at all. In the section on non-ionic detergents, the use of ion-exchange resins is not included. The titrimetric determination of chelating agents is not described. However, these may be personal preferences only and there is no doubt that this is a very useful book for the industrial laboratory. I t fills a gap in the existing literature and everyone concerned with quality control in the detergent industry should read it, particularly those operating with a limited capital expenditure budget.P. PLATT SEPARATION AND PRECONCENTRATION METHODS IN INORGANIC TRACE ANALYSIS. By J. MINCZEW- SKI, J. CHWASTOWSKA and R. DYBCZYI~SKI. Ellis Horwood Series in Analytical Chemistry. Pp. xii + 543. Ellis Horwood (distributed by John Wiley), Wydawnictwa Naukowo- Techniczne. 1982. Price J37.50. ISBN 0 85312 165 6 (Ellis Horwood); 0 470 27169 8 (Halsted Press). The vast expansion of trace analysis is familiar to us all. Most of the accolades have gone to the instrumental detection techniques, but there has been a gradual awakening to the importance of the other, less glamorous components of the analytical process, which include separation and pre- concentration, and to the value of developing complete analytical strategies.The appearance of this book at this time, therefore, is very opportune. It essentially provides comprehensive accounts of precipitation and coprecipitation, volatilisation (including ashing) , liquid - liquid extraction, ion- exchange chromatography and reversed-phase partition chromatography. In each instance, there is an extensive, but not overbearing, theoretical treatment, followed by much descriptive chemistry, comprehensive details of individual systems, an abundance of tabulated information and of refer- ences. For example, the 186 page extraction chapter has about 1350 references, whilst the ion- exchange section, even longer at 220 pages, has 1000 references. Either of these two chapters could stand as monographs in their own right.The main material of the book is prefaced by a useful description of sample handling and avoidance of contamination. The book is well written and superbly edited. I t is a mine of information, including a liberal sprinkling of references to Eastern Europe work. There is little to criticise, and such as it is, is mainly concerned with the introduction, where, for example, advocation of scales of working such as “submicro traces” goes against IUPAC recommendations and the dreaded milliard (log) makes an appearance, instead of units such as micrograms per kilogram being recommended. If all books were as thoroughly prepared as this one, the analytical chemist would be very well served. A. TOWNSHEND ION-SELECTIVE ELECTRODES, 3. THIRD SYMPOSIUM HELD AT MATRAFURED, HUNGARY, 13-15 OCTOBER, 1980.Edited by E. PUNGOR and I. BuzAs. Analytical Chemistry Sywzfiosia Series, Volume 8. Pp. xii + 428. Elsevier, 1981. Price $100; Dfl215. ISBN 0 444 99714 8 (Volume 8) ; 0 444 41786 9 (Series). The second MBtrafured Symposium was held three years earlier than this one and led to a book of The third Symposium appears to have been less well attended too, but with the Is interest in ion-selective over 600 pages. same percentage (20%) of the participants being from the West.552 BOOK REVIEWS Analyst, Vol. 108 electrodes declining after 20 years? I think not, although it is true there is less novelty in this volume. In the first category, Bates (Florida) discusses thermodynamic measurements, Buck (North Carolina) imped- ance measurements, Koryta (Prague) electrode kinetics a t aqueous - organic interfaces and Simon (Zurich) transport of ions by neutral carriers. From the other presentations, those of Vlasov from Moscow (ISFETs), Fjeldly from Trondheim (incorporation of an integrated circuit on top of the ion-selective electrode), Ebel from Vienna [XPS studies of Cu (11) ion-selective electrodes] may be singled out for mention.Clinical medicine applications were described by Simon (valinomycin in silicone rubber for K+ in urine), Pungor (glucose in blood and urine) and Umezara from Japan (immuno-micro-electrode) . This compares with five such presentations at the previous symposium but in the meantime, of course, ion-selective electrode based instrumentation has become an accepted feature of clinical laboratories. Each presentation is followed here by a discussion section ; in most instances, rather short and presumably edited, because a t the meeting the preface tells us that equivalent times were spent on presentation and discussion. There are many problems in the understanding of ion-selective electrodes and their use, which remain to be solved. It is interesting that the problem of the mechanism of neutral carrier electrodes in poly(viny1 chloride), thought to have been solved in 1977, still remains open and Simon’s ETH group were busy in 1980 purifying poly(viny1 chloride). Legibility remains good throughout in spite of variations in format and size. The price will deter most private purchasers but most libraries ought to add it to their analytical sections. Again presentations were divided into Plenary, Keynote and Discussion lectures. The format is again camera-ready typescript with figures at the end. A. K. COVINGTON

ISSN:0003-2654    

DOI:10.1039/AN9830800549

出版商:RSC    

年代:1983

数据来源: RSC

摘要:

552 BOOK REVIEWS Analyst, Vol. 108 Errata NOVEMBER (1981) ISSUE. BDH Chemicals, lot 239563.” Only glass beads should be used in order to avoid contamination from the anti-bumping granules which contain aluminium oxide.” Page 1176, Table 111: the values for iron and copper in the tomato leaves standard and pine needles standard should be reversed and the certified iron value for tomato leaves changed to 690; the correct version is shown below. Page 1174 under Reagents: for “Anti-bumping grandes. read “Anti-bumping granules. Concentration/pg g-l I - Sample Fe cu Zn A1 Cr As SRM 1571 orchard leaves . . .. Found 297 12 24 - 2.7 10.3 Certified 300 12 25 - 2.6 10 SRM 1573 tomato leaves . . .. Found 595 11 62.5 - 4.3 0.26 Certified 690 11 62 - 4.5 0.27 SRM 1575 pine needles . . . . Found 220 3 65 585 - 0.23 Certified 200 3 - 545 - 0.21 FEBRUARY (1982) ISSUE. -OCONH,; and the correct structure for Cephalonium is as shown below. Page 214, Table I : under R’ for Cefuroxime for -OOCNH, read

ISSN:0003-2654    

DOI:10.1039/AN9830800552

出版商:RSC    

年代:1983

数据来源: RSC