ISSN:0003-2654    

DOI:10.1039/AN98005FX009

出版商:RSC    

年代:1980

数据来源: RSC

ISSN:0003-2654    

DOI:10.1039/AN98005BX011

出版商:RSC    

年代:1980

数据来源: RSC

ISSN:0003-2654    

DOI:10.1039/AN98005FP025

出版商:RSC    

年代:1980

数据来源: RSC

ISSN:0003-2654    

DOI:10.1039/AN98005BP033

出版商:RSC    

年代:1980

数据来源: RSC

摘要:

March 1980 The Analyst Vol. 105 No. 1248 Spectrophotometric Determination of Microgram Amounts of Calcium in Waters and Foods Using Diphenylglyoxal Bis( 2-hydroxybenzoyl Hydrazone) M. Silva and M. Valc%rcel Department of Analytical Chemistry, Faculty of Sciences, University of Cdrdoba, Cdrdoba, Spain The synthesis, characteristics and analytical applications of diphenylglyoxal bis(2-hydroxybenzoyl hydrazone) are described. This compound reacts with calcium(I1) a t pH 12 to produce a yellow complex (Amax. = 432 nm, E = 1.76 x 104 1 mol-l cm-l); another complex (1 : 1) can be detected a t pH 7.80. Dipyridylglyoxal bis(2-hydroxybenzoyl hydrazone) also reacts with calcium and both reagents are compared. A sensitive and selective spectrophoto- metric method is proposed for the determination of calcium using diphenyl- glyoxal bis (2-hydroxybenzoyl hydrazone). Interferences have been investi- gated and when masking agents are added common cations do not interfere.The yellow calcium(I1) complex has been used for the determination of calcium in waters and foods. The results are compared with those obtained using glyoxal bis(2-hydroxyanil) . Keywords : Calcium determination ; spectrophotometry ; diphenylglyoxal bis- (2-hydroxybenzoyl hydrazone) ; waters ; foods This work forms part of a systematic investigation of the use of a-bisaroyl hydrazones as analytical reagents. Only the bis(4-hydroxybenzoyl hydrazones) of glyoxal, methylglyoxal and dimethylglyoxal have been described as colorimetric and fluorimetric reagents for the determination of calcium(II), cadmium(II), lanthanum(II1) and bismuth(III),l but the monoaroyl hydrazones have been more widely used in chemical analyses.2 Pyridine-2- carbaldehyde 2-hydroxybenzoyl hydrazone has been investigated as a spectrophotometric reagent for the determination of nickel and zinc,3 ~anadium(V)~ and i r ~ n ( I I ) .~ The preparation and properties of the bis(2-hydroxybenzoyl hydrazones) of diphenyl- glyoxal (BSHB) and dipyridylglyoxal (BSHP) are described in this paper. R Compound BSHB The calcium(I1) - BSHB complex is of great interest and several spectrophotometric methods for the determination of trace amounts of calcium in waters and foods have been reported. The over-all results are discussed and compared with those obtained using the classical spectrophotometric reagent for calcium, glyoxal bis(2-hydroxyanil) (GBHA) [di(2- h ydrox yphen ylimino) et hane] .6-10 A search of the Literature revealed few colorimetric methods for the determination of cal- cium in comparison with other metal ions. Marc- zenko12 reported only two main methods, using GBHA and murexide.l3-l5 Other methods for the spectrophotometric determination of calcium include the use of azo reagents : Chloro- 193 Sandell quoted five methods in detail.ll194 SILVA AND VALCARCEL : SPECTROPHOTOMETRIC DETERMINATION Analyst, Vol. 105 phosphonazo 11116,17 ; Calcichrome II118-20 ; Arsenazo 121,22; Arsenazo 11123~24; Eriochrome Black T25--27; calmagiteB; Acid Chrome Blue KZg; c a l ~ o n ~ ~ ; and Azo-azoxy BN.31 Some complexone derivatives have also been employed : ~ a l c e i n ~ ~ ; metal~hthalein~~ ; thymol- phthalex~ne~~ ; b r o m o - o ~ i n e ~ ~ , ~ ~ ; and TTA [4,4,4-trifluoro-1-(2-thienyl)butane-l,3-dione] .37 Experimental Apparatus The pH measurements were carried out with a Philips PW 9408 pH meter, equipped with a combined glass - calomel electrode.The absorption spectra were recorded on a Unicam SP 8000 spectrophotometer and measurements at fixed wavelengths were recorded on a Beckman DU spectrophotometer in the ultraviolet region and on a Coleman 55 (digital) spectrophotometer in the visible region. Silica or glass (according to the wavelength re- quired) 1-cm path length cells were used. The infrared spectra were recorded on a Perkin- Elmer, Model 577, spectrophotometer. Reagents All reagents were of analytical-reagent grade and all solutions were prepared using dis- tilled water.Diphenyl- and dipyridylglyoxal bis(2-hydroxybenzoyl hydrazone) reagent solutions. Prepare 0.1% m/V solutions in dimethylformamide - ethanol ( 1 + 1 ) . These solutions are stable for at least 1 week. Glyoxal bis(2-hydroxyanil) (Merck) stock solution. Prepare a o.05y0 m/V solution in meth- anol. Standard calcium solution, 1 mg ml-l. Dissolve 2.498 g of calcium carbonate (Carlo Erba), which has been dried at 110 O C , in 40 ml of 2 M hydrochloric acid. Remove any carbon dioxide by boiling, and dilute the solution to 1 1 with distilled water in a calibrated flask. Procedure Synthesis of bisaroyl hydrazones Equimolecular amounts of 2-hydroxybenzoylhydrazide (Ega Chemie) , diphenylglyoxal (Merck) or dipyridylglyoxal (Ega Chemie) were mixed in ethanol - water (1 + 1) ; several drops of concentrated hydrochloric acid were added and the mixture was refluxed for 30 min.The white BSHB and yellow BSHP compounds were separated by filtration. The products obtained were washed with hot ethanol - water ( 1 + 1). The elemental analysis results were as follows. Calculated for BSHB (C2,H2,N,0,) , C 70.29y0, H 4.60% and N 11.71%; found, C 70.0y0, H 4.6% and N 11.570 (yield 53% and melting-point above 300 "C). Calculated for BSHP (C2,HmN604), C 65.00~0, H 4.16% and N 17.50% ; found, C 64.9y0, H 4.2% and N 17.3% (yield 47% and melting-point 281-282 "C). Determination of calcium with BSHB Add 10 ml of the BSHB solution to the sample solution in a 25-ml calibrated flask and then 3 ml of 0.1 M sodium hydroxide solution and dilute with distilled water to the mark. Measure the absorbance of the solution at 432 nm against a reagent blank.If the analytical group 1-111 metals (hydrogen sulphide scheme) are present, add approxi- mately 10 mg of potassium cyanide or 0.2 ml of mercaptoacetic acid, in order to mask small amounts of these metals. Other experimental firocedures Calcium has been determined by the BSHB method in waters and foods. The pre-treat- ments required by the various samples are described below. Waters. Filter sample waters before the determination in order to remove any suspended matter. Egg. Add a sample of the egg yolk or white and 20 ml of 10% m/V sodium carbonate solution to a porcelain capsule and evaporate to dryness at 100-105 "C. Transfer the capsule, while hot, into a furnace at 500 "C for 1 h.Cool, add a few drops of distilled water, break The sample solution must be neutral (pH 6-8) and contain 12.5-50 pg of calcium.March, 1980 OF MICROGRAM AMOUNTS OF CALCIUM IN WATERS AND FOODS 195 up the residue with the flat end of a glass rod, and cover the capsule with a watch-glass; slowly add 10 ml of nitric acid (1 + 3), while stirring, filter, wash the charred material thoroughly with distilled water and collect the filtrate in a 100-ml calibrated flask. To 10ml of milk, in a 100-ml Erlenmeyer flask, add slowly and with constant stirring 10 ml of 20% m/V trichloroacetic acid solution. Place the flask in a boiling water- bath for 30min. Cool, filter, wash with distilled water and collect the filtrate in a 100-ml calibrated flask.To a 1.5-3.5-g sample of grated cheese in a 100-ml Erlenmeyer flask, add 10 ml of 40% m/V urea solution and 5 ml of 4 N hydrochloric acid and warm gently. Transfer the mixture into a porcelain capsule and evaporate to dryness on a sand-bath at 100-105 "C and proceed as for egg. Evaporate 50 ml of juice, in a porcelain capsule, to dryness on a sand-bath at 100-105 "C. Cool, add 10 ml of distilled water and 5 ml of concentrated hydrochloric acid and. evaporate to dryness. Extract the residue with 25 ml of hydrochloric acid (1 + 9), heat for 15 min and filter. Beer. Add 50 ml of beer and 4 g of sodium carbonate to a porcelain capsule and evaporate to dryness, then proceed as for orange juice.Milk. Cheese. Orange juice. Place the capsule in a furnace a t 500 "C for 1 h. Results and Discussion Analytical Properties of the Reagents The solubilities of the BSHB and BSHP reagents were measured in various solvents, and were found to be very low in organic solvents, except for dimethylformamide. In dimethyl- formamide the solubilities of BSHB and BSHP are 32.65 and 14.95 g l-l, respectively. In, for example, chloroform, ethanol, benzene and nitrobenzene the solubilities are less than The infrared spectra of the BSHB and BSHP were obtained using potassium bromide discs. Both spectra are analogous and selected infrared absorption bands were assigned to stretching vibrations of the NH bond (3200 cm-l), the C=N bond (1 660 cm-l), the OH bond (1160 cm-l) and the C=O bond (1 680 cm-l) for both bis(2-hydroxybenzoylhydrazones). When measuring the ultraviolet spectra a water - dimethylformamide (3 + 2) medium was used in order to prevent the precipitation of the reagents and their complexes.In this medium a BSHB solution of 1.4 x M shows absorption maxima at 305 and 325 nm. The ultraviolet spectrum of the BSHP for a 2 x M solution shows two analogous absorp- tion maxima at 305 and 320 nm. In both instances the ultraviolet spectra of these reagents show bathochromic shifts in an alkaline medium (Amax. = 370 nm). The Phillips and Merritt method38 was used for the determination of the ionisation con- stants ; the average pK value found for the diphenylglyoxal bis(2-hydroxybenzoylhydrazone) was 8.20 & 0.05. This behaviour may be caused by de-protonation of the hydroxyl groups.1.0 g 1-1. TABLE I CHARACTERISTICS OF BSHB COMPLEXES IN SOLUTION Buffer solution, pH 4.7 A I \ Amax./ Molar absorptivity/ Colour of Metal ion nm lo3 1 mol-l cm-l complex Pb(I1) . . 435 13.4 Yellow Cu(I1) . . 455-480 4.8 Brown Ni(I1) . . 425 3.1 Yellow Co(I1) . . 410 5.1 Orange Bi(II1) . . 470 17.7 Orange Sn(I1) .. 450 22.5 Orange - yellow Fe(I1) . . 425 3.5 Brown - red 19.6 Yellow ;!$!)I) :: 2; 15.3 Orange In(II1) . . 440 46.0 Yellow Ca(I1) . . Ba(I1) . . Eg;) : : Buffer solution, pH 9.8 Amax./ Molar absorptivity/ Colour of nm lo8 1 mol-1 cm-l complex 450 29.0 Yellow 430 4.2 Yellow 445 5.7 Yellow 465 1.6 Orange 460 20.5 Orange 450 10.5 Orange - yellow 450 2.3 Brown - red 445 19.5 Yellow 460 5.0 Orange 440 45.8 Yellow 432 17.5 Yellow 430 1.0 Yellow 430 5.5 Yellow 430 1.7 Yellow I h \196 SILVA AND VALCARCEL : SPECTROPHOTOMETRIC DETERMINATION Analyst, Vd.105 BSHP gave two dissociation constants in the medium used, with average values of 3.27 5 0.05 and 7.85 & 0.05. The first pK value may be that of the protonated pyridine nitrogen atom and the second that of the hydroxyl groups. Solutions (0.1% m/V) of the reagents were stable for at least 1 week. The C=N groups of both reagents are resistant to hydrolysis at any pH. The reactions of the reagents with 30cations at various pH values were investigated in a medium containing 60% V/V of dimethylformamide and 40% V/V of water. The charac- teristics of the most important BSHB and BSHP complexes in solution are shown in Tables I and 11.TABLE I1 CHARACTERISTICS OF BSHP COMPLEXES IN SOLUTION Buffer solution, pH 4.7 Buffer solution, pH 9.8 L A I \ I > Amex./ Molar absorptivity/ Colour of Amax./ Molar absorptivity/ Colour of Metal ion nm los 1 mol-l cm-1 complex nm l o 3 1 mol-l cm-1 complex Pb(I1) . . 385 14.9 Yellow 450 20.5 Yellow Cu(I1) . . 380 47.0 Yellow 445 51.5 Yellow Ni(I1) . . 380 46.4 Yellow 425 10.0 Yellow Co(I1) . . 375 96.6 Yellow 460 15.3 Orange Bi(II1) . . 390-470 12.2 Orange 465 20.6 Orange Sn(I1) . . 450 2.3 Orange - yellow 450 1.1 Yellow Fe(I1) . . 630 5.7 Green 600 1.7 Brown UO,(II) . . 450 14.3 Yellow 450 15.2 Yellow Ti(1V) . . 460 17.5 Orange 460 15.3 Orange - yellow In(II1) . . 440 45.5 Yellow 440 33.0 Yellow ca(I1) . . 435 10.0 Yellow 435 0.6 Yellow 435 3.0 Yellow 435 1.0 Yellow Ba(I1) .. z$;) : : Study of the Calcium(I1) - BSHB Complex in Solution When dilute solutions of calcium(I1) and BSHB were mixed, a soluble yellow complex was obtained. The absorption spectrum of the calcium(I1) - BSHB system (Fig. 1) shows that the wavelength of maximum absorbance is 432 nm. 0.6 W m 0.5 0.4 2 0.3 2 0.2 0.1 400 450 500 Wavelength/ n m Fig. 1. Absorption spectra of the calcium complexes. Concentration of calcium(II), 1.5 p.p.m. ; and concentration of reagents, 6.25 x lo-' M. A, Calcium(I1) - BSHB complex in dimethylformamide - ethanol (1 + 1) aqueous medium at pH 12; B, calcium(I1) - BSHP complex in dimethyl- formamide - ethanol (1 + 1) aqueous medium at pH 9.8; and C, reagent blank at the same concentration.March, 1980 OF MICROGRAM AMOUNTS OF CALCIUM I N WATERS AND FOODS Stability of the complex At alkaline pH in aqueous dimethylformamide - ethanol solution the complex forms im- mediately and the absorbance remains stable for at least 4 h.Glyoxal bis(2-hydroxyanil), one of the best spectrophotometric reagents for determining calcium, has been employed in this work as a standard reagent for the measurement of cal- cium in several samples; this compound forms a soluble red complex with calcium in alkaline media.6 In order to compare both complexes, the stability of the calcium(I1) - GBHA system was studied. The absorbance at 455 nm of a solution of the red complex decreases continually with time. The absorption spectrum of GBHA shows a bathochromic shift after 1-3 h.Using X-ray diffraction and infrared spectroscopy, Lindstrom and MilliganS9 have shown that instability of the calcium(I1) - GBHA complex is due to the following: the hydrolysis of GBHA, which is present in sufficient excess, the alkaline medium catalysing the reaction to yield glyoxal and o-aminophenol; the low stability constant of the calcium(I1) - GBHA complex; the formation of the anion of glycollic acid, OH-CH,-COO-, from glyoxal, which is catalysed by the OH- anion; and the formation of the calcium(I1) - glycolate complex. The greater stability of the calcium(I1) - BSHB system is due to the stability of the chelate, and to the resistance of the C=N bonds to hydrolysis. 197 Fig. 2 shows this behaviour for both reagents and their complexes with calcium.1 2 3 Time/h Fig. 2. Stability of diphenylglyoxal bis(2-hydroxybenzoyl- hydrazone), glyoxal bis (2-hydroxyanil) and their calcium complexes : a, BSHB solution, concentration 1.4 x M (A = 370 nm); 0, calcium(I1) - BSHB complex, calcium concentration 3.12 x 1 0 - 5 ~ (A = 432 nm); A, GBHA solution, concentration 1 0 - 3 ~ ( A = 455 nm) ; and A, calcium(I1) - GBHA complex, calcium concentration 2.5 x M (A = 516 nm). Infiuence of $H The absorbance was maximum and constant over the alkaline region and showed two possible optimum pH ranges of 7.4-8.2 and 9.0-12.5 (Fig. 3). The latter range was chosen because of the greater sensitivity it offered. The influence of pH was studied using a series of solutions in the pH range 7-13. Efect of reagent concentration, ionic strength and proportions of the dimethylformamide - ethanol mixture The absorbance of the complex was studied as a function of the molar ratio of BSHB to calcium(I1).A 15-fold molar excess of the reagent over calcium was necessary in order to obtain the maximum absorbance value.198 SILVA AND VALCARCEL : SPECTROPHOTOMETRIC DETERMINATION Analyst, ‘Vd. 105 The ionic strength of the solution does not affect the absorbance of the calcium(I1) - BSHB system. The same constant absorbance measurements were obtained when 5 ml of 0.5 M potassium nitrate solution or 5 ml of 0.5 M potassium chloride solution were added. At alkaline pH the solubility of BSHB is greater, and a medium containing 40% V/V of dimethylfonnamide - ethanol (1 + 1) and 60% V/V of water was chosen for further experi- mental work as the yellow complex does not precipitate in this medium.The results obtained are in total agreement if dimethylformamide is used alone in place of the dimethyformamide - ethanol (1 + 1) mixture. Ethanol cannot be used alone because the complex and the excess of reagent precipitates. Hence, it is proposed to use the dimethyl- formamide - ethanol (1 + 1) solvent mixture, which will also reduce the cost owing to the use of the cheaper ethanol as part of the mixture in place of the more expensive dimethyl- formamide alone. 5 6 7 8 9 10 11 12 13 PH Fig. 3. Effect of pH on the formation of calcium complexes. Concentration of calcium(II), 3.12 x M: 0, with BSHB at 432 nm; and e, with BSHP at 435 nm. Nature of the complex The stoicheiometry of the chelates in solution has been determined by Job’s (Fig.4), Yoe JonesJs and slope-ratio methods at both pH ranges in which the absorbance remains constant (see Fig. 3). At pH 7.80 (triethanolamine - hydrochloric acid buffer) the formation of two complexes with stoicheiometric ratios of metal to ligand of 2: 3 and 1: 1 has been observed. However, at pH 12 only one complex species, with a metal to ligand ratio of 2 : 3, has been found, which is probably favoured by an excess of reagent and by the high pH. A study of the retention of these chelates on an anion-exchange resin showed that, under these experimental conditions, the complexes were anionic. In both stoicheiometric complexes the reagent acts as a quadrivalent ligand. The presence of the double aroylhydrazone contiguous chain was essential for the formation of the cal- cium(I1) chelates with BSHB, as the 2-hydroxybenzoyl hydrazone of benzaldehyde does not give rise a complex with calcium(I1).Study of the Calcium(I1) - BSHP Complex in Solution The absorption spectrum of the calcium(I1) - BSHP complex is illustrated in Fig. 1. The absorption maximum of the complex occurs at 435nm. The yellow complex is formed immediately at alkaline pH and the absorbance at 435 nm is stable for at least 4 h. Fig. 3 shows the influence of pH on the calcium(I1) - BSHP system. The optimum pH range is 8.5-10.5. A volume of not less than 5 ml of 0.1% m/V reagent solution is required in order to obtain the maximum development of the yellow complex. The ionic strength does not affect the absorbance and the same proportions of dimethyl- formamide and ethanol were used as for the calcium(I1) - BSHB complex.In order to determine the composition of the yellow complex, continuous-variation and molar-ratio methods were carried out. The results indicated a calcium to BSHP ratio of 2:3- (Fig. 4). The complex was retained by anion-exchange resin, and therefore this system is anionic in alkaline solution.March, 1980 OF MICROGRAM AMOUNTS OF CALCIUM I N WATERS AND FOODS 199 0.7 0.6 0.5 0 C 5 0.4 2 a 0.3 a 0.2 0.1 I I t l l , , I , 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 [Ca2+] [Ca2+] + [Reagent] Fig. 4. Stoicheiometry of the com- plexes. Calcium(I1) - BSHB complex ( A = 432 nm) : concentration of calcium(II), 6 x 1 0 - 4 ~ ; 0, pH 12; a, pH 7.80.A, Calcium(I1) - BSHP complex ( A = 435nm) a t pH 9.8: concentration of calcium(II), 5 x 1 0 - 4 ~ . Spectrophotometric Determination of Calcium with BSHP In an alkaline medium, the calcium(I1) - BSHP system obeys Beer’s law from 0.75 to 3.00 p.p.m. of calcium. The molar absorptivity is 1.03 & 0.05 x lo4 1 mol-l cm-l. A Ringbom plot shows that the range for minimum error in the determination is 1.25-2.50 p.p.m. of calcium. The mean values of results from 11 samples each containing 2.0 p.p.m. of calcium gave the relative error (P = 0.05) of the method as *o.32y0. The effect of diverse ions on the determination of 2.0 p.p.m. of calcium using BSHP was studied under the experimental conditions used. Strontium, barium, magnesium and most of the analytical group 1-111 metals (hydrogen sulphide scheme) interfere down to 2 p.p.m.The anions EDTA and oxalate also give rise to serious interferences. The use of masking agents in the BSHP method to eliminate interferences has not been introduced because the interferences of other species were much more significant when masking agents were used than when they were not and because the BSHB method has a greater sensitivity and is more selective. Spectrophotometric Determination of Calcium with BSHB Characteristics of the method Under the optimum conditions used for the formation of the calcium complex, Beer’s law is obeyed between 0.5 and 2.0 pg ml-l of calcium and the molar absorptivity at 432 nm is 1.76 & 0.14 x lo4 1 mol-l cm-l. The sensitivity of the method, according to Sandell, is 0.002 3 pg cm-2.The optimum concentration range, evaluated by Ringbom’s method, is 0.5-1.5 p.p.m. of calcium. The mean values of results from 11 samples each containing 0.5 p.p.m. of calcium gave the relative error (P = 0.05) of the method as & 0.34%. Table I11 shows a comparative study of the calcium(I1) - BSHB and calcium(I1) - BSHP chelates in solution. The properties are analogous, although the sensitivity of the reaction with BSHB is better. The interferences are not shown in this table, but the calcium(I1) - BSHB system is subject to less interferences than the calcium - BSHP system.200 SILVA AND VALCARCEL : SPECTROPHOTOMETRIC DETERMINATION Analyst, VoZ. 105 TABLE I11 COMPARISON OF CALCIUM COMPLEXES WITH BOTH BISHYDROXYBENZOYL HYDRAZONES Stoicheiometry, Molar absorptivity/ Compound Amax,/nm Optimum pH metal t o ligand ratio lo3 1 mol-1 cm-1 BSHB .... 432 9.0-1 2.5 1: 1, 2: 3 17.6 BSHP .. . . 435 8.5-10.5 2: 3 10.4 Efect of diverse iorts The effect of various amounts of 50 diverse ions on 12.5 pg of calcium per 25 ml of solution was investigated. The tolerance limits showed that calcium can be determined in the pre- sence of a large number of ions (Table IV). TABLE IV TOLERANCE LIMITS I N THE DETERMINATION OF 12.5 pg PER 25 ml OF CALCIUM(II) WITH BSHB Ions tolerated K(I), Na(I), NH,(I), Mg(II), Sr(II),* Sb(III),t Al(III),t W(VI),t Ni(II), I-, SCN-, B,0,2-, CN-, SO,2-, NO,-, Br-, tartrate, ClO,-, Br0,-, Sa-, ASOa3-, COSa- . . .. . . UO,(II) PO,3-, citrate, F-, SO,%-, Mo(V1) .. . . .. ..Li(I), Rb(I), S,0a8- . . .. Pb(II), Cd(II), La(II1) . . .. .. . . .. . . .. . . .. . . .. . . .. . . .. . . .. Ba(II), Mn(IIi,* V(Vi . . Hg(II), Sn(II), Bi(III), Be(11). Ag(I), Tl(i), C,O>- . . .. * Centrifuge the sample prior t o measuring the absorbance. t 0.25% m/V BSHB solution. Tolerance limit/ pg per 25 ml 2 500 1250 625 250 125 50 25 Ascorbic acid, mercaptoacetic acid (MAA) and triethanolamine (TEA) could be used as Ascorbic acid (100 g 1-l), mercapto- masking agents, so their tolerance limits were studied. acetic acid (10 g 1-l) and triethanolamine (20 g 1-l) did not interfere. TABLE V ELIMINATION OF INTERFERENCES OF FOREIGN IONS I N THE DETERMINATION OF CALCIUM WITH BSHB BY ADDITION O F MASKING AGENTS Foreign ion Tl(1) . . Bi (111) Pd(I1) Cd(I1) Sn(I1) Pt(1V) Mo(V1) V(V) .. WIV) La(II1) Fe(I1) Be(I1) In(II1) Zn(I1) Li(1) . . ::g/ &Y& RW) . . .. .. . . . . . . .. .. .. .. .. .. .. .. . . .. .. . . . . .. .. .. Without masking agent .. 50 .. 25 .. 50 .. 50 .. 50 .. 25 .. 50 .. 250 .. 625 .. 25 .. 50 .. .. .. .. .. .. .. .. .. .. 125 .. 125 With masking agent 2 500 2 500 1250 2 500 2 500 2 500 625 2 500 2 500 625 1250 2 500 2 500 1250 625 1250 2 500 2 500 250 2 500 2 500 2 500 Tolerance limitlpg per 25 ml Masking agent CN-, 100 pg ml-1 TEA, 20 g 1-l CN-, 100 p g ml-1 MAA, 10 g 1-1 MAA, 10 g 1-l CN-, 100 p g ml-1 MAA, 10 g 1-1 MAA, 10 g 1-1 CN-, 100 p g ml-l Ascorbic acid, 100 g 1-1 CN-, 100 p g ml-l MAA, 10 g 1-1 MAA, 10 g 1-1 CN-, 100 pg ml-l F-, 10 pg ml-1 MAA, 10 g 1-1 MAA, 10 g 1-1 TEA, 20 g 1-1 MAA, 10 g 1-1 MAA, 10 g 1-1 MAA, 10 g 1-1 CN-, 100 p g ml-1March, 1980 mercaptoacetic acid (Table V).interfered at the 1.0 p.p.m. level. OF MICROGRAM AMOUNTS OF CALCIUM IN WATERS AND FOODS 201 Good results were obtained by masking the interfering ions with potassium cyanide and The most serious interference was from EDTA, which Determination of calcium in waters and foods The proposed method was satisfactorily applied to the determination of calcium in waters and foods. Aliquots of the prepared solutions, different in each sample according to the concentration of calcium, were analysed as described. The results were compared (Table VI) with those obtained by the glyoxal bis(2-hydroxyanil) spectrophotometric method. TABLE VI RESULTS OF THE DETERMINATION OF CALCIUM IN WATERS AND FOODS BY THE USE OF BSHB AND GLYOXAL BIS(2-HYDROXYANIL) (GBHA) METHODS Sample Mineral water .. . . Beer . . . . .. .. Swamp water . . . . Orange juice .. . . Egg white . . .. . . Egg yolk . . .. . . Milk . . .. . . . . Cheese .. .. .. Calcium found* BSHB method GBHA method A r 5 20.8 f 0.1 p.p.m. 19.9 f 0.1 p.p.m. 59.0 f 0.1 p.p.m. 33.0 f 0.1 p.p.m. 115.0 f 0.2 p.p.m. 0.152 f 0.002% 0.701 -j= 0.002% 0:109 f 0.001% 20.2 f 0.1 p.p.m. 19.6 f 0.1 p.p.m. 58.0 f 0.1 p.p.m. 32.0 f 0.1 p.p.m. 118.0 f. 0.2 p.p.m. 0.150 f 0.001% 0.108 0.001% 0.698 f 0.002% * Average of 5 separate determinations. Conclusions It has been shown that the sensitivity of both spectrophotometric techniques is similar ( E = 1.8 x lO41mol-1cm-1). The interferences of the BSHB method are similar and in many instances much more advantageous than those from the GBHA method, which requires extraction with Azo-azoxy BN.Therefore, the BSHB procedure does not require a pre- liminary extraction step. The calcium(I1) - BSHB system offers a greater stability than that of the calcium(I1) - GBHA complex. The analytical data provided by our method show this to be the case. The stability of the BSHB solutions permits the same reagent blank solution to be used for at least 1 d. With the GBHA method a blank solution has to be prepared at the same time each sample is made. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. References Lever, E., Anal. Chim. A d a , 1973, 65, 311. Katyal, M., and Dutt, Y., Talanta, 1975, 22, 151. Gallego, M., Garcia-Vargas, M., Valcarcel, M., and Pino, F., Microchem. J., 1978, 23, 353.Gallego, M., Garcia-Vargas, M., Valchrcel, M., and Pino, F., Microchem. J., 1979, 24, 143. Gallego, M., Garcia-Vargas, M., and Valchrcel, M., Analyst, 1979, 104, 613. Umland, F., and Meckenstock, K. U., 2. Anal. Chem., 1960, 176, 96. Ken, J. R., Analyst, 1960, 85, 867. Williams, K. T., and Wilson, J. R., Anal. Chem., 1961, 33, 244. Kuczerpa, A. V., Anal. Chem., 1968, 40, 581. Hunter, G., Analyst, 1972, 97, 233. Sandell, E. B. , “Colorimetric Determination of Traces of Metals,” Third Edition, Interscience, New Marczenko, Z., “Spectrophotometric Determination of Elements,” First Edition, John Wiley, New Tammelin, L. E., and Mogensen, S., Acta Chem. Scand., 1952, 6, 988. Williams, M.B., and Moser, J. H., Anal. Chem., 1953, 25, 1414. Pollard, F. H., and Martin, J. V., Analyst, 1956, 81, 348. Ferguson, J . W., Richard, J. J., O’Laughlin, J. W., and Banks, C . V., A.naZ. Chem.. 1964, 36, 796. Lukin, A. M.. Smirnova, K. A., and Vysokova, N. N., Zavod. Lab., 1968, 34, 1436. West, T. S., Analyst, 1962, 87, 630. Close, R. A,, and West, T. S . , Talanta, 1960, 5, 221. York, 1959, p. 366. York, 1976, p. 182.202 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. SILVA AND VALCARCEL Herrero-Lancina, M., and West, T. S., Anal. Chem., 1963, 35, 1231. Polyak, L. Ya., Zavod. Lab., 1961, 27, 803. Vlasov, N. A., and Morgen, E. A., Zh. Prikl. Khim., 1965, 38, 998. Zhukova, M. P., Matrosova, T. V., and Yakovlev, P. Ya., Zh. Anal. Khim., 1971, 26, 2231. Howell, D. S., Pita, J. C., and Marquez, J. F., Anal. Chem., 1966, 38, 434. Michaylova, V., and Kouleva, N., Talanta, 1974, 21, 523. Menon, V. P., and Das, M. S., Analyst, 1958, 83, 434. Impedovo, S., Traini, A., and Papoff, P., Talanta, 1971, 18, 97. Ingman, F., and Ringbom, A., Microchem. J., 1966, 10, 545. Goryushina, V. G., and Archakova, T. A., Zavod. Lab., 1962, 28, 796. Reilley, C. N., and Hildebrand, G. P., Anal. Chem., 1959, 31, 1763. Dziomko, V. M., and Dunaevskaya, K. A., Zh. Anal. Khim., 1960, 15, 661. Robinson, C., and Weatherell, J. A., Analyst, 1968, 93, 722. Bosholm, J., Anal. Chim. Acta, 1966, 34, 71. BezdekovA, A., and Budesinsky, B., Collect. Czech. Chem. Commun., 1966, 30, 818. Umland, F., and Meckenstock, K. U., 2. Anal. Chem., 1959, 165, 161. Luke, C. L., Anal. Chim. Acta, 1965, 32, 221. Poluetkov, N. S., and Bel’tyukova, S. V., 2%. Anal. Khim., 1970, 25, 2106. Phillips, J. P., and Merritt, L. L., J. Am. Chem. Soc., 1948, 70, 410. Lindstrom, F., and Milligan, C. W., Anal. Chem., 1967, 39, 132. Received July 20th, 1979 Accepted October loth, 1979

ISSN:0003-2654    

DOI:10.1039/AN9800500193

出版商:RSC    

年代:1980

数据来源: RSC

摘要:

Analyst, March, 1980, Vo2. 105, pp. 203-208 Spectrophotometric Determination of Cobalt with 203 Biacetyl Mono( 2-pyridy1)hydrazone A. G. Asuero and M. M. Rodriguez Department of Applied Chemical Analysis, Faculty of Pharmacy, University of Seville, Seville-4, Spain The absorbance versus pH graphs and stoicheiometry of the complexes formed by biacetyl mono(2-pyridy1)hydrazone (BPH) with cobalt(II), nickel(II), copper(I1) and iron(I1) have been studied. A method for the spectrophoto- metric determination of cobalt has been devised. The orange - red chelate is formed at acidities between 2 and M (pH S), in aqueous ethanolic solution, and has A,,,. a t 505 nm with a molar absorptivity of 2.3 x 1 0 4 1 mol-1 cm-1. The method has been applied to the determination of cobalt in nitrates and has been compared with other pyridylhydrazone procedures for the spectro- photometric determination of cobalt.Keywords : Biacetyl mono (2-pyridy1)hydrazone reagent ; cobalt determination: spectrophotometry The analytical use of pyridylhydrazone compounds in colour-forming reactions with metal ions is well kn~wn,l-~ and these compounds are very useful as spectrophotometric and spectrofluorimetric reagents and as acid - base indicators. After a preliminary study of pyridylhydrazones derived from a-diketones as analytical reagent^,^-^ we selected biacetyl mono(2-pyridy1)hydrazone (BPH) for the spectrophotometric determination of cobalt for several reasons: (i) the high solubility of BPH and its complexes in aqueous media; (ii) the fact that the absorption of BPH at the Amax.of the cobalt complex is negligible even for a great excess of reagent; (iii) the reactions of BPH with cobalt ion are superior in sensitivity to those of biacetyl bis(2-pyridy1)hydrazone (BBPH) and biacetyl monoxime (2-pyridy1)- hydrazone (BMPH); (iv) BPH is the most selective of the reagents tested in colour re- actions; and (v) the BPH method is the most economical. Therefore, in this paper, BPH is introduced as a valuable reagent for the spectrophotometric determination of cobalt. Experimental Solutions Biacetyl mono(2-pyridy1)hydrazone was used as a 0.25% m/V solution in ethanol. A standard solution of cobalt(I1) (4.9283 mg ml-l) was prepared from cobalt nitrate hexahydrate and was standardised by EDTA titration.6 Working solutions were prepared by appropriate dilution.The other standard metal-ion solutions were available from previous investigations in this laboratory . The solutions of metals for interference tests were prepared by weighing the required amounts of the nitrates and adding water and the appropriate amount of nitric acid until complete dissolution was obtained. Apparatus The apparatus used was identical with that described previ~usly.~ Procedures Absorbance versus PH graphs Absorbance versus pH graphs for the complexes were obtained with the reagent in excess. The solutions were always prepared with the same order of operations: the metal, addition of reagent, adjustment of ionic strength (2.5 ml of 1 M potassium chloride solution per 25 ml) and adjustment of acidity. No buffer was used.Stoicheiometry of the complexes were used to try to ascertain the nature of the complexes formed a t pH 7 and 10. The molar-ratio' and slope-ratio* methods and the method of continuous variation9J0 The204 ASUERO AND RODRIGUEZ : SPECTROPHOTOMETRIC DETERMINATION Analyst, Vd. 105 medium was 20% VlV ethanol; the concentration of the stock solutions of metal ions and BPH was 1-2 x ~ O - * M , various volumes being taken, and the pH was controlled by the addition of 5 ml of 1 M ammonium acetate solution or 5 ml of pH 10 borate buffer solution. The order of addition of reagents was metal ion, reagent, buffer and diluent. As a buffer was used, the ionic strength was not adjusted. Photometric determination of trace amounts of cobalt with BPH Transfer up to 20 ml of the sample solution, previously adjusted to pH 4.5 (not more than 65pg of cobalt), by pipette into a 25-ml calibrated flask.Add 1-5ml of the 0.25% m/V solution of BPH in ethanol. Mix the solutions thoroughly and allow the mixture to stand for 10 min (unless otherwise stated). Dilute the mixture to the mark with appropriate amounts of ethanol, perchloric acid and water to give final concentrations within the ranges 0.1-1 M in acid and 440% V/V in ethanol. Measure the absorbance at 505 nm in a glass cell of 10-mm path length against distilled water. 3 5 7 9 11 13 2 4 6 8 10 12 A B C D Fig. 1. Absorbance vevsus pH graphs for the metal ion - BPH complexes. (a) Ni(II), 2 p.p.m. : absorbance at (A) 425 nm and (B) 445 nm. C, D and E: reagent a t 430,440 and 470 nm, respectively.CR = 5.65 x M in 4% aqueous ethanol. (b) Cu(II), 3 p.p.m.: absorbance a t (A) 430 nm, (B) 450 nm, (C) 470 nm, (D) 490 nm, (E) 510 nm and (F) 530 nm. CB = 5.65 x (c) Fe(II), 10 p.p.m., plus 0.1 g of ascorbic acid: absorbance a t (A) 600nm, (B) 615nm, (C) 625nm, (D) 650nm, (E) 675nm and (F) 685nm. CB = 8.47 x ( d ) Co(II), 2 p.p.m.: absorbance at (A) 440 nm, (B) 465 nm, (C) 505 nm and (D) 535 nm. M in 4% aqueous ethanol. M in 6% aqueous ethanol. CR = 5.65 x 1O-O M in 4% aqueous ethanol.Mnvch, 1980 OF COBALT WITH BIACETYL MONO(^-PYRIDYL)HYDRAZONE 205 A calibration graph prepared using the same procedure (1 ml of the 0.25% mlV BPH solution, 20% V/V ethanol, 1 M perc.hloric acid) was rectilinear, passing through the origin, in the range 5-65 pg of cobalt, with a slope of 0.008 absorbance unit per microgram of cobalt.The effect of various ions was studied. The general procedure was followed except that the solutions of foreign ions were added before the reagent solution. When a large amount of interfering cation was used, the samples were first prepared as nitrates in a 50-ml beaker because of precipitation. After adding acid the precipitate dissolved and the samples were then transferred into a 25-ml calibrated flask and diluted to the mark with water, followed by spectrophotometric measurement. 0.8 Results and Discussion (a 1 r Absorbance versus pH Graphs From the shapes of absorbance versus pH graphs (Fig. 1) for the metal ion - BPH complexes, it seems that two different complexes are formed with each of nickel, copper and cobalt.The apparent pH values for the development of maximum absorbance, determined by spectrophotometric tit ration^,^ are approximately 4.8, 5.6, 7.1 and 7.6 for the cobalt (11), copper(II), nickel(I1) and iron( 11) complexes, respectively. The apparent pH for the formation of the cobalt (11) benzil mono(2-pyridy1)hydrazone complexll is approximately 7.5. However, on adding acid only the cobalt complex remained unchanged. Exploratory experiments indicated that the reagent forms precipitates with copper(1) and palladium(I1). Solvent-extraction Tests Solvent-extraction tests indicated that the complexes formed in basic medium are neutral. Q) 0.6 3 0.4 0.2 -t3 a 8 0.5 0.4 2 C n a 0.3 0.2 0.1 0 2 4 6 8 10 0 0.25 0.50 0.75 1 0 0.25 0.50 0.75 1 Ratio [BPHI /[CO! Ratio [Col /([Col + [BPHI I Ratio [Col/([Col + [BPHI 1 Fig.2. Composition of cobalt - BPH complex a t pH 7 (ammonium acetate). (a) Composition deter- mined by the molar-ratio method; cobalt(I1) concentration 3.4 x M. Absorbance a t (A) 465 nm and (B) 505 nm. (b) Composition determined by Job's method; concentration of cobalt(I1) plus ligand 7.63 x 10-5 M . (c) Composition determined by the slope-ratio method: A, concentration of cobalt(I1) variable and ligand 1.27 x M and B, con- centration of ligand variable and cobalt 1.27 x Absorbance at (A) 440 nm, (B) 465 nm and (C) 505 nm. M. Absorbance a t (0) 505 nm and (.) 535 nm. Stoicheiometry of the Complexes The method of continuous variation failed with the complexes of nickel(II), copper(I1) and iron(I1) at pH 7 and also failed for iron and zinc at pH 10.The slope-ratio method also failed with iron(I1) and zinc(I1) because the complexes were not formed when the metal ion was in excess. The data obtained for cobalt(I1) (Fig. 2) indicated a ratio of metal to ligand of 1:2, the same as that found by Pflaum and Stuckerll for the cobalt(I1) b e n d mono(2-p>Tridyl) hydrazone complex. The formation constant of the cobalt complex was not calculated by the Harvey and Manning methods because the complex is very strong. If EDTA was added after cobalt complexation no interference was observed. At higher pH,206 ASUERO AND RODRIGUEZ : SPECTROPHOTOMETRIC DETERMINATION Analyst, Vol. 105 a new cobalt complex appeared in the solution with maximum absorption at 450 nm and a very broad shoulder between 460 and 530 nm.The molar-ratio method indicated a 1 : 4 Ni - BPH complex at pH 7, whereas the slope-ratio method was not conclusive. Both the molar-ratio and continuous-variation methods indicated a 1 : 2 Ni - BPH complex at pH 10. The molar-ratio method for iron(I1) was not conclusive at pH 10. Spectrophotometric Determination of Cobalt with BPH M (pH 6) (in 20% V/V ethanol) indicated the formation of a single complex having maxima at 440 and 505 nm. The complex was stable for at least 24 h in 1 M perchloric acid (in 20% ethanol), for a ratio of cobalt to BPH of 1 : 35. The spectrum of the cobalt - BPH complex was recorded in various solvents miscible with water (Fig. 3). In a basic solvent such as dimethylfonnamide, the complex formed at high pH in aqueous medium was present.Spectra measured at acidities between 2 and 0.8 0.6 W C m 0.4 0, a a 0.2 nn (a 1 L / 3 V." 350 400 450 500 550 600 350 400 450 500 550 600 Wavelength/nm Fig. 3. Spectrum of the cobalt - BPH complex in various solvents miscible with water. (a) Solvent: 1, dimethylformamide; 2, formic acid; 3, acetic acid; 4, tetrahydrofuran. ( b ) Solvent: 1, acetone; 2, ethanol; 3, methanol; 4, dioxan. Cobalt, 2 p.p.m.; CR = 5.65 x M . The optimum range for accurate determination, as evaluated from a Ringbom plot, is 0.3-1.2 p.p.m. of cobalt. The sensitivity of the colour reaction is 0.0025 pg cm-2 (at 505 nm) and the molar absorptivity is 2.35 x lo4 1 mol-l cm-l. The precision of the method was checked by measuring the absorbance of ten samples, each series containing a final concentra- tion of 0.3, 0.8 and 1.2 p.p.m.of cobalt. The method gave relative errors of 0.54, 0.48 and 0.39%, respectively. TABLE I EFFECT OF VARIOUS IONS ON DETERMINATION OF COBALT BY THE RECOMMENDED PROCEDURE Concentration of cobalt: 0.3 p.p.m. No interference up to Interfering ions and approximate tolerance levels, p.p.m. 1.2 x 104p.p.m. - r -I Ag4: (2) Br-, I-, C1-, s2°32- (5) Ga3+ (30) &+ (( 100) * Tartrate, SO,2-, C2042- (10) Sb3+ (6) Po43-, h o d 3 - , SCN- (10) A13+ (12) Cr3+ (240) Ni2+ (loo)* V(V) (100) Fez+ (loo)* Mn2+ (1 200) E$I)(:kO) alkali metals, alkaline earth metals P20,4- (60) F- (5000) * 5 mlO.25% BPH.March, 1980 OF COBALT WITH BIACETYL MONO(2-PYR1DYL)HYDRAZONE 207 Results of the interference studies are given in Table I.The tolerance towards a foreign ion was taken as the largest amount that caused an error of not more than 5% in the absorbance of cobalt alone (0.3 p.p.m.). However, greater amounts of these metal ions can be tolerated if a larger amount of reagent is employed and/or the solutions are allowed to stand at room temperature long enough for the displacement reaction to occur. As expected, negative errors in the determination of cobalt occurred in the presence of larger amounts of nickel, copper and iron than indicated in Table I. As no standard samples were available, a series of recovery experiments were carried out by adding standard cobalt solution to samples of various nitrates and carrying out the analysis as described under Procedures, with the exception that the samples were allowed to stand for 2.5 h before perchloric acid was added.Small amounts of cobalt (7.5-60 pg) were correctly determinecbwhen added to nitrates (Table 11). Cation* Zinc Cadmium . Mercury Lead Amount added, p.p.m. 6 000 6 000 12 000 12000 12000 12000 6 000 12 000 12 000 TABLE I1 ANALYSIS OF COBALT IN NITRATES Co added, p.p.m. 0.32 1.19 0.32 1.19 0.32 1.19 0.32 1.19 0.32 Approximate metal to Co ratio 1 : 20 000 1 : 5000 1 : 40 000 1 : 10000 1 : 40000 1 : 10000 1 : 20 000 1: 10000 1 : 40 000 Absorbance values (samples) 0.124, 0.127, 0.117 0.461, 0.476, 0.478 0.145, 0.127, 0.127 0.489, 0.490, 0.488 0.125. 0.123, 0.127 0.484, 0.456, 0.478 0.127, 0.125, 0.125 0.470, 0.471, 0.477 0.124 Co found, p.p.m.Comments 0.31 1.17 0.33 1.22 0.31 1.17 0.35 1.17 0.34 2 ml of 0.25% BPH Uranyl . . 1200 0.27 1 : 4000 0.107 0.26 5 ml of 0.25% BPH 6 000 0.27 1 : 20000 0.103, 0.100, 0.097 0.25 6 000 1.08 1:6000 0.383, 0.396, 0.402 0.97 Iron(II1) . . 900 0.27 1:3000 0.099, 0.108, 0.112 0.26 900 1.08 1 : 800 0.338, 0.340, 0.352 0.85 * With the exception of mercury all of the precipitates initially formed were dissolved on adding perchloric acid. Conclusion Many organic reagents have been proposed for the absorptiometric determination of cobalt. Recent papers have recommended the use of several pyridylhydrazones (Table 111) (which have been claimed to offer advantages over nitroso reagents) and some azo compounds containing halogen-substituted pyridine.I6 The reagents containing sulphur as a ligand atom, most of which are less sensitive than the above, include quinoxaline-2,3-dithiol.17 With such reagents it could be said that the spectrophotometric determination of cobalt in the presence of large amounts of nickel and other transition metal ions is not an especial TABLE I11 CHARACTERISTICS OF COBALT - PYRIDYLHYDRAZONE COMPLEXES Reagent Optimum pH Pyridine-2-aldehyde 2-Benzoylpyridine Benzil mono(2-pyridy1)hydrazone .. Pyridine-2-aldehyde 2-pyridyl- Biacetyl .sono(2-pyridyl)- 2-quinolylhydrazone* . . . . 6.5-1 1.5 2-pyridylhydrazone . . . . 3.85-11.6 2-pyridylhydrazone * . . { 20% HClO, hydrazoneandeosin . . . . 5.4-5.8 hydrazone . . .. .. 2 MHCI-~H 6 Included for the sake of comparison. t Chloroform - acetone. 2 M HCl-pH 8 2,2'-Dipyridyl 3-11 510 478 535 480 500 547 505 Concentration in sample, % Molar absorptivity/ of ethanol I mol-l cm-l x lo-' 3.00 20 2.93 50 2.74 60 3.20 4.20 } 7.80 7+3t 2.35 4-25 Relative reagent cost Reference 5 12 3.5 13 0.6 11 19 14 1.5 15 1 Thiswork208 ASUERO AND RODRIGUEZ problem nowadays, which makes the justification for another method for cobalt difficult.Results of spectrophotometric measurements of the cobalt chelates show that the reactions with higher relative molecular mass hydrazones are more sensitive than those with BPH, as they possess a more extended n-system. The values of A,,,. for the orange - red chelates, however, do not differ significantly. In spite of its greater molar absorptivity, benzil mono- (2-pyridy1)hydrazone is not superior to BPH as a reagent for cobalt, and there are three good reasons for preferring the latter.Firstly, the cobalt - BPH complex has a low apparent pH, which leads to greater selectivity, secondly, a lower excess of reagent is required for complex formation, and thirdly, BPH and its metal complexes have high solubilities in aqueous ethanolic media. Further advantages of BPH include its solubilities in water (0.01%) and ethanol (greater than a%, at room temperature), its relatively low cost and its high selectivity and sensitivity. One of us (M.M.R.) expresses his gratitude to Professor Dr. F. Pino for giving him the opportunity to carry out these investigations. 1. 2. 3. 4. 5 . 6. 7 . 8. 9 . 10. 11. 12. 13. 14. 15. 16. 17. References Katyal, M., and Dutt, Y., Talanta, 1975, 22, 151. Asuero, A. G., Microchem. J . , 1978, 23, 390. Asuero, A. G., and Rodriguez, M. M., Mzcrochem. J., in the press. Rodriguez, M. M., Maqueda, M. J., and Asuero, A. G., Ajinidad, in the press. Asuero, A. G., Afinidad, in the press. Bermejo, F., and Prieto, A., “Aplicaciones Analiticas del AEDT y Analogos,” Departamento de Yoe, J . H., and Jones, A. L., Ind. Eng. Chem., Anal. Ed., 1944, 16, 11. Harvey, A. E., and Manning, D. L., J . Am. Chem. Soc., 1950, 72, 4488; 1952, 74, 4744. Job, P., Ann. Chim. (Paris), 1928, 9, 113. Vosburg, W. C . , and Cooper, R. G., J . Am. Chem. SOC., 1941, 63, 437. Pflaum, R. T., and Stucker, E. S., 111, Anal. Chem., 1971, 43, 458. Singhal, S. P., and Ryan, D. E., Anal. Chim. Acta, 1967, 37, 91. Going, J. E., and Pflaum, T., Anal. Chem., 1970, 42, 1098. Vasilikiotis, G. S., Kouimtzis, Th., Apostoloupoulou, C., and Voulgaropoulus, A., Anal. Chim. Haddad, P. R., Alexander, P. W., and Smythe, L. E., Talahta, 1976, 23, 275. Shibata, S., Furukama, M., Tshiguro, Y., and Sasaki, S., Anal. Chim. Acta, 1971, 55, 231. Burke, R. W., and Beardorff, F. R., Talanta, 1970, 17, 255. Quimica Analitica, Santiago de Compostela, 1960. Acta, 1974, 70, 319. Received July 23rd, 1979 Accepted September 5th, 1979

ISSN:0003-2654    

DOI:10.1039/AN9800500203

出版商:RSC    

年代:1980

数据来源: RSC

摘要:

Analyst, March, 1980, Vol. 10.5, $9. 209-216 209 Precision Spectrophotometric Determination of Chromium in Chromite Ores and Ferro-chrome D. Thorburn Burns and M. E. M. Abdel Aziz Department of Analytical Chemistry, The Queen's University of Belfast, Belfast, BT9 5AG, Northern Ireland The chemical and manipulative variables have been examined in detail in order to develop a precise and direct spectrophotometric determination of chromium in chromite ores and in ferro-chrome. Samples are sintered with sodium peroxide in zirconium crucibles for 3 h at 510 & 10 "C, then leached, filtered, acidified with perchloric acid and reduced with hydrazine hydrate. Chromium is determined from the absorbance of the chromium(II1) aquo ion at 410 and 578nm. The procedure has been evaluated using standard chrome ores, ferro-chrome alloys and a series of Sudanese and industrial ferro- chrome samples.The results are in good agreement with, and of comparable precision to, published data and to those obtained by titrimetric assay. Keywords : Precision spectrophotometry ; chromium determination ; chromiie ores ; ferro-chrome alloys Accurate assays of chromite ores are necessary because discrepancies of the order of 1% on a single 25-tonne truck load have significant financial implications for both the buyer and seller. At present there is pressure to change from the accepted classical gravimetric and titrimetric assay procedures for major constituents of ores owing to skilled manpower shortages and to rising costs, even in developing countries. Earlier work on the precision spectrophotometric determination of major metallic components in oxides and oxide mixtures1 indicated the potential of the technique, but only a few applications have been r e p ~ r t e d .~ , ~ Direct spectrophotometric methods of analysis using conventional equipment are rapid and simple but give precisions below those of classical procedures, Differential spectro- photometric methods can, under favourable circumstances, give results similar to those attainable by classical methods but they are susceptible to stray light problems and require considerably more time and greater operator skills4 Dinnin5 failed in an attempt to apply differential spectrometry to chromite analyses and expressed doubts if such a method could be made to work.This paper describes the direct precision spectrophotometric determination of chromium in chrome ores and ferro-chrome and the precautions necessary at each stage of the analysis in order to achieve the required over-all precision and accuracy. The main problems were encountered in the decomposition and filtration stages and in optimising the solution con- ditions prior to spectrophotometric measurement. To allow the accurate and precise assay of ores the component of interest must be dissolved completely and any matrix effects overcome. A survey of procedures for the decomposition of chromite ores showed that many were unnecessarily complicated and time consuming6+ and others were inefficient .9-14 Sodium peroxide is an excellent general-purpose oxidising alkaline flux, its main disadvantage being its corrosive properties.P ~ r c e l a i n , ~ ~ , ~ ~ iron,13,15J7-20 silver,21 n i ~ k e 1 , ~ ~ , ~ ~ s ~ ~ ~ i r c o n i u r n ~ , ~ ~ and vitreous graphite25 crucibles have been used for the sodium peroxide fusion of chromite ores. Studies of their relative resistance to corrosion has shown that zirconium is the best material for peroxide fusions at temperatures above 550 OC.2637 The decomposition of minerals can also be effected by sintering with a reagent at lower temperatures, which reduces attack on crucible^.^^,^^ Differential thermal analytical studies have shown that most minerals decompose with sodium peroxide between 250 and 300 "C. For chromite and other reasonably resistant materials the size of the endothermic peak indicates their resistivity.The efficiency of sintering is controlled by the relative amount of reagent to sample, the temperature and the time of sintering. RafterB suggested that 1.2-3.0 g of sodium peroxide was suitable for the decomposition of 0.2-1.0 g of sample, with heating a t 480 -+ 10 "C for 7 min in platinum crucibles; these conditions have not been found satisfactory for the decomposition of chromite ores.210 BURNS AND MI2 PRECISION SPECTROPHOTOMETRIC Analyst, 'Vd. 105 Problems were encountered at the filtration stage. Apart from excessive filtration times (under gravity, 10 h), small amounts of solid were present in the filtrate, which caused light scatter at short wavelengths, even when using anti-creep filter-~aper.~O In common with earlier workers, difficulties were encountered when using chromium (VI) solutions acidified with sulphuric a~id31,~~ or perchloric a ~ i d ~ ~ s ~ ~ owing to complex solution equilibria. More recent work,35 however, has indicated isosbestic points at 320 and 345 nm so that the quantitative spectrophotometric use of acidic solutions of chromium(V1) may still be possible.The final procedure is based on the absorption of chromium(II1) following reduction of chromium(V1) with hydrazine hydrate, which was used to avoid the adventitious addition of complexing anions. Experimental Reagents and Apparatus Except when unavailable, all reagents were of analytical-reagent grade. Sodium peroxide. Sulphuric acid, 97.5-100yo mlm. Perchloric acid, 72% mlm. Hydrazine hydrate, NHPH2.H2O, 99-100yo.Laboratory grade. Potassium dichromate. Ammonium dichromate. Volumetric glassware. This was of grade A specification and calibration. Absorbance Measurements Spectra were recorded in the ultraviolet and visible ranges using a Unicam SP 8000 spectro- photometer. The quantitative measurements were made using a Unicam SP 3000 spectro- photometer and 4-cm silica cells. The instrumental absorbance discrimination is 0.001 up to 1.2 absorbance unit and 0.01 absorbance unit thereafter. The linearity of response was checked by using calibrated neutral density filters (C. Davis Keller Ltd.) (in 0.1 absorbance unit steps from 0.20 up to 0.80 absorbance unit). Solutions were kept at 25 & 0.1 "C until required for measurement. TABLE I SUMMARY OF SAMPLE DECOMPOSITION EXPERIMENTS Amount of Crucible Na,O,/g Ternperaturel'C Zirconium .. 5 600-700 Platinum or zirconium . . 3 Platinum or zirconium . . 3 Platinum or zirconium . . 3 Platinum or zirconium . . 5 Zirconium . . 5 480 f 10 (sintering) 480 f 10 (sintering) 480 f 10 (sintering) 370 f 10 (sintering) 510 f 10 (sintering) Time 5 min 7 min 3 x 7min 3 h 6 h 3 h Recovery, yo Comments due to attack of zirconium crucible. (2) Violent reactions with ferro-chrome samples, particularly those with high carbon content. 100 (1) Light scatter problems 75 95 98 100 (1) Efficient decomposition but very slow. platinum crucibles even when left overnight. 100 Platinum crucibles (2) Negligible attack on severely attacked.March, 1980 DETERMINATION OF CHROMIUM IN CHROMITE ORES AND FERRO-CHROME 211 Preliminary Studies It was initially intended to determine both chromium and iron in a single solution obtained by acid dissolution of the sintered samples.The absorbance due to iron(I1) species depended critically on the solution conditions, including the excess of reducing agent, and it was therefore necessary to separate iron and chromium prior to their determination. \ r' Fig. 1. Filtration adaptor. 4' Filtration Following sintering, it is necessary to leach the samples and separate the insoluble hydrous iron oxide from the chromium(V1) solution. This precipitate is difficult and time consuming to filter and wash by gravity (10 h), despite the use of anti-creep (edge-silanised) filter-papers, and light scatter remained a problem.Glass-fibre papers (GF/C) and a special Buchner adapter (Fig. 1) allowed rapid filtration and washing directly into the volumetric flasks used for reduction, etc. TABLE I1 EFFECT OF SCATTERED LIGHT Decomposition process : fusion at 600-700 "C. Filter-paper : Whatman 542. Sample BCS 366 BCS 20414 Reg 1330 Reg 1830 Reg 1841 Reg 3283 Lot 167 G 129 Chromium content, Experimental result % a t 578 nm, yo .. 74.6* 75.1 .. 71.9* 72.5 .. 58.2t 58.7 .. 57.7t 58.6 .. 68.8t 70.1 . . 68.8t 69.1 .. 59.9t 60.4 .. 67.0t 68.6 Calculated result a t 410 nm, % 75.3 72.7 58.9 59.0 70.7 69.2 60.7 69.1 Experimental result at 410 nm, % 75.2, 72.7 58.9 58.7 70.4 69.20 60.6 69.0 * Chromium content : certificate value. t Chromium content: supplied value.212 BURNS AND AZIZ : PRECISION SPECTROPHOTOMETRIC Analyst, Vol.105 The higher apparent chromium contents as calculated from absorbance measurements a t 410 nm compared to those from measurements at 578 nm can be attributed to a wavelength- dependent absorbance from light scattering. The variation of the contribution by scatter to the absorbance of a solution with wavelength is complex.37 For small particles of diameter (d) less than X/3, the scattered intensity is proportional to l / X 4 (Rayleigh scattering) and for dmX the scattered component is proportional to 1/X2 (Clausius scattering). For very large particles the absorbance due to the particles is independent of A. The present results showed a l/h dependence, indicating d>h, which is consistent with the fact that the GF/C glass-fibre filters can pass particles up to 1.2 pm, i.e., 2-3X.1.1 1.0 7 - - 0 10 20 30 40 5 0 x 2.1298~ Concentration of chromium( I I I) ions/g ml-' Fig. 2. Beer's law graph for chromium(II1): A, 410 nm; and B, 578 nm. As absorbances are additive the apparent chromium content a t 410~nm (R410) can be calculated from the certificate value for a sample (v) and the assay result at 578 nm (R578) as follows: The data in Table 11, obtained using the finalised solution conditions, indicate that the apparent high results when filtration is not efficient are consistent with large particle scatter. Determination at two wavelengths serves as a check on filtration efficiency; the results should agree within two standard deviations on a single set. TABLE I11 VARIATION OF SPECIFIC ABSORPTIVITY OF CHROMIUM(III) IN [Cr(H,0),I3+ WITH DIFFERENT CONCENTRATIONS OF K2Cr,07 (7.2% m/m HClO, AND 0.15% V/V N2H4.H20) Specific absorptivity/ ml g-1 cm-1 1 K,Cr,O,/g ml-l 410 nm 578 n m 1.6040 x 301.3 259.0 8.5191 x lo-* 299.8 256.9 6.3880 x 299.9 256.4 4.2502 x 298.2 255.7 2.1215 x 10-4 298.6 257.9 Solution Conditions During initial experiments it was noted that chloride ion, added with the reducing agent, hydroxylammonium chloride, caused anomalous results that were additional to the effect of scattered light noted when filtration was inefficient. This problem was circumvented by using hydrazine hydrate.kf&?'Ch, 1980 DETERMINATION OF CHROMIUM I N CHROMITE ORES AND FERRO-CHROME 213 Beer's law was obeyed up to an absorbance value of 1.2 (Fig.2 and Table 111). The specific absorptivity of the chromium(II1) aquo ion was found to be independent of perchloric acid concentration above 14.4% m/m (Table IV).The amount of hydrazine hydrate is not critical provided that it is greater than 0.075 Y/V (Table V). TABLE IV WITH DIFFERENT CONCENTRATIONS OF HClO, (0.15% V/V N2H,.H,0 AND 1.5050 x lo-, g ml-l OF K,Cr,O,) VARIATION O F SPECIFIC ABSORPTIVITIES OF CHROMIUM(II1) I N [Cr(H,0)6]3+ Specific absorptivity/ 1 HClO,, yo mlm 410 nm 578 nm ml g-l cm-1 1.8 308.1 262.0 3.6 302.6 260.1 7.2 296.9 256.2 14.4 298.9 256.4 18.0 298.9 256.4 Specific Absorptivity of Chromium(II1) in 14.4% m/m Perchloric Acid at 25 OC It was observed that the specific absorptivities determined by reduction of potassium dichromate were slightly higher and considerably more precise (six results) than those deter- mined starting with Specpure chromium metal (Table VI).The errors were considered to be due to the loss of chromium as chromyl chloride during the dissolution of the metal in hot concentrated perchloric acid. TABLE V VARIATION OF SPECIFIC ABSORPTIVITY OF CHROMIUM( 111) I N [Cr(H2O)6I3+ WITH DIFFERENT CONCENTRATIONS OF N,H,.H,O [1.505 8 x g ml-l OF K,Cr,O, AND 7.2 m/m HClO, (72%)] Specific absorptivity/ ml g-l cm-1 r - 1 N,H,.H,O (5%). % V/V 410 nm 578 nm 0.050 297.1 258.3 0.075 298.9 258.3 0.100 289.0 258.3 0.200 299.0 258.4 0.250 298.9 258.3 As potassium perchlorate is not very soluble, it may precipitate on long standing. This problem can be overcome by the use of ammonium dichromate, which is available in AnalaR grade. The standard solutions and solutions obtained from chromite ores were stable and mean results (six values) identical to within &O.lyo of chromium were attained for solutions stored at ambient temperature for 1 month (Tables VII and VIII).TABLE VI SPECIFIC ABSORPTIVITIES OF CHROMIUM( 111) AQUO ION IN 14.4% m/m HC10, AT 25 "C Specific absorptivity/ml g-1 cm-1 r A b Standard reagent 410 nm 578 nrn Chromium beads . . . . . . 295.2 f 1.6 253.3 f 1.2 Potassium dichromate . . . . 299.0 f 0.1 257.0 f 0.1214 BURNS AND AZIZ : PRECISION SPECTROPHOTOMETRIC TABLE VII STABILITY OF AMMONIUM DICHROMATE STANDARD SOLUTIONS Analyst, Vol. 105 Specific absorptivity of chromium(III)/ml g-1 cm-l f L 7 Absorbance measurements 410 nm 578 nm After 1 h .. . . . . 298.1 f 0.4 256.1 f 0.3 After 1 month . . . . 298.2 f 0.4 266.3 f 0.3 Recommended Method A . Chromium ores Finely grind the sample to pass a test sieve of nominal aperture size 100 pm and dry for 1 h at 110 "C. Transfer 0.6-0.8 g of the sample, weighed to the nearest 0.01 mg, into a 25-ml zir- conium crucible. Mix thoroughly with 5 g of dry sodium peroxide, cover the crucible and sinter the contents at 510 & 10 "C for 3 h in an electric muffle furnace. Allow the crucible to cool, place it in a 250-ml tall-form Pyrex beaker, leach the sintered mass with 50ml of water, cover the beaker immediately with a watch-glass, and leave for about 1 h for the hydrogen peroxide to be decomposed. Filter the solution, under a differential pressure of approximately 6 cmHg, through a 5.5 cm diameter glass-microfibre GF/C filter placed in a Buchner funnel connected with an adaptor to a calibrated 250-ml volumetric flask in which 50 ml of perchloric acid (72% m/m) had previously been placed.Rinse the crucible, beaker and watch-glass and scrub them with a rubber policeman, then filter and wash the total residue on the filter thoroughly with hot distilled water. Reduce the filtrate with 4ml of hydrazine hydrate (5% V / V ) , swirl the volumetric flask until the vigorous evolution of nitrogen bubbles has ceased, thermostat at 25 & 0.1 "C and dilute with distilled water to the mark. Measure the absorbance immediately after stabilisation at 25 & 0.1 "C at 410 and 578 nm against distilled water using a 4-cm cell. Carry out fusion and reagent blank measurements.Calculate the chromium(II1) oxide content of the ore using a pre-determined specific absorptivity (ml g-l cm-l) of chromium(II1) at 410 and 578 nm and express the result as chromium(II1) oxide. 1. 2. 3. 4. 5. 6. 7. Mix thoroughly. 8. 9. 10. TABLE VIII STABILITY OF CHROMITE SAMPLE SOLUTIONS : DETERMINATION OF CHROMIUM(III) OXIDE CONTENTS Sample Chromium(II1) oxide content, yo 410 nm 578 nm f L 7 A f > I A \ After 2 h After 1 month After 2 h After 1 month Ingessana Hills (111) . . . . 55.5 f 0.3 55.7 f 0.3 55.6 f 0.2 55.4 f 0.2 Ingessana Hills (D) . . . . 57.9 f 0.2 58.0 f 0.3 57.8 f 0.1 57.8 f 0.1 Chromite sample (11) . . . . 49.1 & 0.1 49.3 f 0.1 49.2 f 0.1 49.2 f 0.2 B. Ferro-chrome alloys 0.01 mg, into a zirconium crucible.(ml g-l cm-l) of chromium(II1) at 410 and 578 nm and express the result as chromium. 1. 2. 3. Transfer 0.3-0.4 g of ferro-chrome flakes or, preferably, powder, weighed to the nearest Continue steps 3 to 9 as described in A. Calculate the chromium content of the alloy using a pre-determined specific absorptivityMarch, 1980 DETERMINATION OF CHROMIUM IN CHROMITE ORES AND FERRO-CHROME 216 Results and Discussion The assay results are shown in Tables IX and X. In each instance the standard deviation is based on six results. TABLE IX RESULTS OF ANALYSIS OF ALLOYS AND ORES Sample type Sample Ferro-chrome alloys . . BCS 366 Chrome ore . . . . BCS 308 BCS 20414 Student sample 49g G Student sample 49f Ferro-chrome . . . . Reg 1330 Reg 1830 Reg 1841 Reg 3283 Lot 167 G129 * Certificate value.t Supplied value. Stated chromium , content, yo 74.6 f 0.1* 71.9 f 0.04* 41.5 f 0.2* 53.4* 43.1* 58.2t 57.7t 68.8t 68.8t 59.9t 67.0t ExperimentaI result, yo 3 410 nm 578 nm 74.6 f 0.2 71.5 f 0.1 41.6 f 0.1 54.5 f 0.2 43.4 f 0.1 59.0 f 0.2 57.9 f 0.3 69.1 f 0.3 68.8 f 0.3 60.4 f 0.2 67.1 f 0.1 74.8 f 0.3 71.4 f 0.1 41.9 f 0.2 54.8 f 0.2 43.5 f 0.1 59.5 f 0.4 57.8 f 0.2 69.1 f 0.4 68.9 f 0.4 60.6 f 0.2 67.4 f 0.1 The precisions of the results are in all instances comparable to those attained by careful titrimetric procedures for similar analyses, where a coefficient of variation of 0.32% was found for a single analyst and 0.61% for between-laboratory results.% The accuracy of the present results is in accord with certificate values except for one “Student sample,” which is not a reference material. Results are determined at two wavelengths as a check on filtration efficiency; the results on chrome ores should agree to within &O.lyo of chromium. TABLE X RESULTS OF ANALYSIS OF SUDANESE CHROMITE SAMPLES Chromium content, yo I Sample Ingessana Hills (11) .. .. Ingessana Hills (D) . . .. Chromite sample (111) . . .. Gabaneit Mine NT1 . . .. Gabaneit Centre NT5 . . .. Gabaneit South NT14 . . .. Jebel Jam Mine JC40 . . . . Bayomi Lease BMC2 . . .. Jebel El Tawila T .. .. 410 nm 49.1 f 0.1 57.9 f 0.2 55.5 f 0.3 56.7 0.2 56.3 f 0.2 57.3 f 0.3 52.9 f 0.2 54.6 f 0.3 45.5 f 0.3 578 nm 49.2 f 0.1 57.8 f 0.1 55.6 f 0.2 56.5 f 0.1 56.2 f 0.1 57.2 f 0.3 52.7 f 0.2 54.5 f 0.2 45.6 f 0.2 The solution conditions adopted avoid the problems of sulphate-bridged polymeric chromium species39 and spectrochemical effects of anions coordinated to monomeric chromium(III).40 Hydrazine hydrate has the further advantage that neither it nor its decomposition products absorb in the visible or ultraviolet regions.It is possible to determine iron after dissolution of the hydrous oxide in concentrated hydrochloric acid.41 We thank the Industrial Research and Consultancy Institute (Sudan) for financial support for M. E. M. Abdel Aziz, the British Steel Corporation for the gift of analysed samples of ferro-chrome and Dr. I. M. Babiker (Geological Department, Sudan) for providing samples of Sudanese chromite ores.216 BURNS AND A212 References Johns, P., and Price, W. J., Analyst, 1970, 95.138. Johns, P., Spectrovision, 1971, (26), 8. Docherty, A. C., Farrow, S. G., and Skinner, J. M., Analyst, 1972, 97, 36. Svehla, G., Talanta, 1966, 13, 641. Dinnin, J . I., U.S. Geol. Surv. Bull., 1959, No. 1084(B), 31. Ito, J., Bull. Chem. Soc. Jpn., 1962, 35, 225. Smith, G. F., McVickers, L. D., and Sullivan, V. R., J . Soc. Chem. Ind., London, 1935, 54, 369. Cresser, M. S., and Hargitt, R., Anal. Chim. Acta, 1976, 82, 203. Beyermann, K., 2. Anal. Chem., 1962. 190, 4. Brunck, O., and Holtje, R., Angew. Chew., 1932, 45, 331. Sarudi, I., 2. Anal. Chem., 1958, 163, 34. Furness, W., Analyst, 1950, 75, 2. Theobald, L. S., Analyst, 1942, 67, 287. Harpham, E. W., Metallurgia, 1955, 52, 93. Majdel, I., Arh. Hem. Farm., 1930, 4, 8 ; Chem. Abstr., 1930, 24, 2397. Franke, A., and Dworzak, R., 2.Angew. Chem., 1926, 39, 642. de Sousa, A., Chemist-Analyst, 1961, 50, 9. Riddell, W. C., and Kittredge, E., Min. Sci. Press. 1918, 117, 558. Cunningham, T. R., and McNeill, T. R., Ind. Eng. Chem., Anal. Ed., 1929, 1, 70. Moir, J., J. S . Afr. Assoc. Anal. Chem., 1919, 2, 916; Chem. Abstr., 1919, 13, 2648. Dittler, E., Tschermaks Mineral. Petrogr. Mitt., 1929. 40, 189. Balyuk, S. T., Taukchi, V. N., and Demina, 2. I., Ogneupory, 1969, 34, 59. Berthet, Chim. Ind., 1928, Special No. 133; Chem. Abstr., 1928, 22, 4407. Bunch, K., N.S.W. Dep. Mines, Chem. Lab. Rep., 1962, 1963, 1964, No. 10, 29; Chem. Abstr., 1967, Jecko, G., and Royer, R., Chim. Anal., 1970, 52, 1109. Blake, H. E., and Holbrook, W. F., Chemist-Analyst, 1957, 46, 42. Belcher, C. B., Talanta, 1963, 10, 75. Rafter, T. A., Analyst, 1950, 75, 485. Seelye, F. T., and Rafter, T. A., Nature (London), 1950, 165, 317. “Whatman Laboratory Filter Papers,” Appendix Filtration Section, Technical Bulletin F2, H. Reeve Angel and Co. Ltd., London, p. 48. Kortum, G., 2. Phys. Chem., 1936, B33, 243. Johnson, E. A., Photoelectr. Spectrom. Group, 1967, (17), 505. Tong, J. Y., and King, E. L., J . Am. Chem. Soc., 1953, 75. 6180. Davies, W. G., and Prue, J. E., Trans. Faraday SOC., 1955, 51, 1045. Burke, R. W., and Mavrodineanu, R., J . Res. Nut. Bur. Stand., Sect. A , 1976, 80, 631. Yoshimori, T., Talanta. 1975, 22, 827. Partington, J . R., “An Advanced Treatise on Physical Chemistry,” Longmans, London, 1953, Hartford, W. H., Anal. Chem., 1953, 25, 290. Kuntzel, A., Erdmann, H., and Spahrkas, H., Leder, 1952, 3, 73. Harvey, K. B., and Porter, G. B., “An Introduction to Physical Inorganic Chemistry,” Addison- Abdel Aziz, M. E. M., and Thorburn Burns, D., Proc. Anal. Div. Chem. Soc., 1979, 16, 25. 66, 25829~. Volume 4, Section 10B, $ 44, 45, 46 and references cited therein. Wesley, Reading, Mass., 1963. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. Received July 27th. 1979 Accepted September 25th, 1979

ISSN:0003-2654    

DOI:10.1039/AN9800500209

出版商:RSC    

年代:1980

数据来源: RSC

摘要:

Analyst, March, 1980, Vol. 105, PP. 217-221 217 Selective Colorimetric Detection of Carboxylic Acids H. S. Rathore, M. N. Akhtar and S. K. Sharma Chemistry Section, Zakir Husain College of Engineering and Technology, A ligarh Muslim University, A ligarh-202001, India A selective colour reaction for the detection of carboxylic acids has been developed. Acetic anhydride in the presence of ammonium carbonate, sodium carbonate and the sodium salts of acetic, formic, oxalic and pyruvic acids is used as the reagent. Amongst 39 acids, citric, isocitric, a-keto- glutaric and oxaloacetic acids can be detected selectively. Keywords Carboxylic acid detection ; colorimetry ; acetic anhydride reagent A solution of citric acid in acetic anhydride has been used for the selective detection of tertiary amines.1 Feig12 remarked that the chemistry of the colour test seems to be complicated and has not been elucidated.The solution of citric acid in acetic anhydride also gives colours with salts of tertiary amines, quaternary ammonium salts, salts of potassium, rubidium, caesium, strontium and barium, and salts of alkali and alkaline earth metals with organic acids. However, the colour reaction between citric acid - acetic anhydride and salts of alkali metals has not previously been studied for the detection or determination of carboxylic acids. An attempt has been made to use this reaction for the selective detection of certain carboxylic acids and a-ketoglutaric acid in particular. Experimental and Results Material s An aluminium heating block maintained at 165 & 2 "C was employed.Aqueous or ethanolic solutions of the test substances (1%) were prepared. When it was not possible to prepare a 1% solution, a saturated solution was used. Dowex 1-X8 anion-exchange resin (20-50 US mesh) was converted into the carbonate, formate, acetate and oxalate forms by treatment with a 1 M aqueous solution of sodium carbonate, formic acid, acetic acid or oxalic acid, respectively. The resins obtained were washed with distilled water and dried at room temperature (25 "C). All chemicals and reagents were of analytical-reagent grade. General Procedure Two or three drops of the solution of the test substance (1 mg) were evaporated to dryness in a micro test-tube by cautious direct heating in order to eliminate all moisture, cooled to room temperature and then two or three drops of acetic anhydride were added. The colour that developed was noted, then the contents of the tube were heated at 165 2 "C in the heating block for 5 min and the colour that subsequently developed was also noted.The results obtained are given in Table I. Two or three drops of the solution of the test substance and an equal volume of a 1% aqueous solution of ammonium carbonate, sodium carbonate, sodium pyruvate, sodium formate, sodium acetate or sodium oxalate were evaporated to dryness in a micro test-tube and the above procedure was followed. The results obtained are given in Table I. Limits of detection are given in Table 11. The above procedure was then carried out but in place of the aqueous solutions of salts, 8-10 anion-exchange resin beads in the carbonate, formate, acetate or oxalate form were used.The limits of detection are given in Table 11. Detection and Semi-quantitative Determination of Carboxylic Acids in the Presence of Foreign Substances As a result of this study, it was concluded that sodium acetate - acetic anhydride was the best reagent for the detection of carboxylic acids. It was used for the detection of carboxylic acids, using a-ketoglutaric acid as a representative compound in the presence of large amounts of foreign substances.21 8 RATHORE et al. : SELECTIVE COLORIMETRIC TABLE I Analyst, Vol. 105 DETECTION OF ACIDS USING DIFFERENT REAGENTS AT 165 & 2 "c Colours: NC = no colour; V = very; L = light; 0 = orange: R = red; Y = yellow; G = green; Acid Acetic .. Ascorbic . . Benzoic . . Cinnamic . . Formic . . Gallic . . Glyoxalic . . Lactic . . Nicotinic . . m-Nitrobenzoic p-Nitrobenzoic Pyruvic . . Salicylic . . Adipic . . Fumaric .. Glutaric . . ct- Ke toglu t aric Maleic . . Malic .. Malonic . . Oxalic . . Oxaloacetic Phthalic . . Succinic . . Tartaric . . Citric . . Isocitric . . Alanine . . Argenine.HC1 Aspartic . . I-Cystine . . Cysteine. HC1 Glutamic . . Glycine . . I-Lysine . . Barbituric Boric . . Sulphamic uric .. .. .. .. .. . . .. .. .. .. . . .. .. .. .. .. . . .. .. .. .. . . . . . . .. .. .. . . . . . . .. .. .. . . . . .. .. .. . . . . . . .. .. .. .. . . . . .. .. .. .. .. .. .. . . .. .. .. .. . . . . .. .. . . .. .. .. . . . . .. .. .. . . . . . . . . .. .. .. Ac20 NC LY NC NC NC NC NC NC NC NC NC NC NC NC NC NC LG NC NC LO NC LBr NC NC LY LR LR NC LY NC NC LY NC LY NC VLY NC LR NC I .B = bl<e; Br = brow;. Ac2O .+ Ac2O + ammonium sodium carbonate carbonate NC 0 NC NC NC NC LO NC NC NC NC NC LY NC NC NC VLY - LY* VLY NC 0 NC Br NC NC NC R R NC Y NC NC NC NC Y NC VLY NC NC NC NC R NC VLY NC 0 VLY NC NC NC NC NC NC NC LY NC RG NC VLY VLY NC Br NC NC LY R R NC NC NC NC NC NC Y NC VLY NC NC NC Ac,? + sodium p yruvate LY R 0 LY Y LY VLY LO LY LY VLY NC LY 0 R LY GB R LY Y VLY Br LY Y 0 R R NC 0 LR 0 LY NC LY LY R NC LY - o* VLY Ac29 + sodium formate NC R NC NC NC NC NC NC NC NC NC NC NC NC NC NC RG LO VLY LY NC Br Y Y 0 R R NC 0 LR 0 LY NC LY LY LY - Y* NC 0 VLY Ac20 + sodium acetate NC R NC NC NC NC NC NC NC NC NC NC NC NC NC NC GB NC NC LY NC Br NC NC Y R R NC NC NC NC NC NC NC NC NC NC NC NC Ac29 + sodium oxalate NC 0 NC NC NC NC NC NC NC NC NC NC VLY NC NC NC GR NC NC 0 NC Br NC NC Y R R NC NC NC NC NC NC NC NC NC NC NC NC * The former refers to the colour at room temperature (25 "C) and the latter at 165 & 2 "C.Detection of a-ketoglutaric acid A known volume of the aqueous test solution (250 pg), two or three drops of sodium acetate solution (2 mg) and a known volume of a solution of the interfering substances (as listed in Table 111) were evaporated to dryness in a micro test-tube in order to eliminate all moisture, then cooled to room temperature. Two or three drops of acetic anhydride were added and the solution was heated at 165 2 "C. The results are given in Table 111. Semi-quantitative determination of a-ketoglutaric acid Known volumes of the solution of the test substance and 8-10 resin beads in the acetate form were placed in different test-tubes and the above procedure was followed.The results obtained are given in Table IV. The colours of solutions of unknown concentration were developed in the same way and then the nature and intensity of the colours were compared with those obtained from the solutions of known concentration.March, 1980 DETECTION OF CARBOXYLIC ACIDS TABLE I1 LIMITS OF DETECTION OF SOME ACIDS (IN MICROGRAMS) USING DIFFERENT REAGENTS AT 165 & 2 "C 219 The colour developed was the same when either the salt solutions or the ion-exchange resin beads were used. Ac,O + sodium Ac,O + sodium Ac,O + sodium Ac,O + sodium carbonate formate acetate oxalate 5 z z r - G z - - m Acid phase phase phase phase phase phase phase phase Ascorbic . .. . 50 40 50 50 100 100 100 100 Barbituric . . . . 250 250 100 100 100 100 100 100 10 5 20 50 60 60 20 50 50 50 50 20 20 50 50 50 Citric . . .. . . 20 Isocitric .. . . 100 100 a-Ketoglutaric . . . . 50 40 20 Oxaloacetic . . . . 80 100 70 70 100 100 100 100 Tartaric . . . . 500 500 100 100 150 150 1 50 150 2o 20 Behaviour of Other Compounds Various organic compounds (1-2 mg) were tested by the recommended procedure and were found not to interfere. The following compounds were tested : amines, aniline, diethanol- amine, dimethylamine, diphenylamine and trimethylamine ; amides, acetamide, benzamide, and salicylamide ; alcohols, butan-1-01, ethanol, methanol and propan-1-01; aldehydes, benzaldehyde, o-nitrobenzaldehyde, salicylaldehyde and vanillin ; carbohydrates, glucose, lactose and sucrose ; esters, ethyl formate ; ethers, diethyl ether; heterocyclic bases, indole, nicotine and pyridine ; hydrocarbons and their derivatives, benzene, carbon tetrachloride, chlorobenzene, nitrobenzene, light oil and o-toluidine ; ketones, acetone and acetophenone ; Phenols and their derivatives, catechol, hydroquinone, quinolin-8-01, phenol and resorcinol.TABLE I11 DETECTION OF a-KETOGLUTARIC ACID (250 /..Lg) IN THE PRESENCE OF INTERFERING SUBSTANCES Amount Substance added1 pg Bovine serum albumin (BSA) . . . . 50 150 500 Alanine . . .. . . .. .. 50 150 1000 Aspartic acid . . . . . . . . 50 150 1000 Glutamic acid .. .. .. . . 50 150 1000 Oxaloacetic acid . . . . . . . . 50 150 1000 Pyruvic acid . . . . . . . . 50 150 1000 * Abbreviations as in Table I. Colour* BrG BIG BrG YG JIG YG RG RG RG G G BrG BrG BrG BrG BrG BIG BrG The following compounds (1-2 mg) were found to interfere; the colours developed are given in parentheses : cinnamaldehyde (light yellow), P-nitrophenol (light yellow), pyrrole (red), thiourea (yellow) and urea (yellow).220 RATHORE et al. : SELECTIVE COLORIMETRIC Analyst, Vol. 105 Discussion From Table I it can be seen that acetic anhydride alone gives colours with various acids, but the sensitivity of these colour reactions is low and they cannot be used for analytical purposes. Acetic anhydride in the presence of ammonium carbonate gives colours with 12 of the acids.Acetic anhydride in the presence of sodium carbonate gives colours with 14 of the acids, with higher intensity than those with ammonium carbonate. Acetic anhydride in the presence of sodium pyruvate gives colour with all of the acids except pyruvic, glutamic and boric acids and it therefore cannot be used for the selective detection or determination of the acids. Acetic anhydride in the presence of sodium formate gives colours with 20 of the acids, in the presence of sodium acetate with 7 of the acids and in the presence of sodium oxalate with 8 of the acids. Hence acetic anhydride in the presence of sodium acetate or sodium oxalate is more selec- tive than in the presence of sodium formate. Table I1 shows the following or'der of sensitivity of the colour reactions of acetic anhydride in the presence of the different sodium salts: sodium formate > sodium carbonate > sodium acetate > sodium oxalate.Thus, acetic anhydride in the presence of sodium formate can be used for sensitive detection and in the presence of sodium acetate or oxalate it can be used for selective detection. For example, ketodicarboxylic acids (a-ketoglutaric and oxaloacetic) can be distinguished from keto- monocarboxylic acids (glyoxalic and pyruvic) . Perhaps the ketomonocarboxylic acids do not react under the conditions used. Similarly, hydroxytricarboxylic acids (citric and iso- citric) acids can be distinguished from hydroxydicarboxylic acids (malic and tartaric). Acetic anhydride in the presence of sodium acetate gives a greenish blue colour with a-keto- glutaric acid, which is different to those of the remaining 38 compounds (Table I).The limit of detection of a-ketoglutaric acid is 20 pg (Table 11). At low concentrations (less than 100 pg) it gives a brownish blue colour instead of greenish blue, and then cannot be distin- guished from oxaloacetic acid by this procedure. However, aspartic, glutamic, glutaric and pyruvic acids, alanine and BSA give no colour and do not interfere in the procedure (Table Table IV shows that anion-exchange resin beads in the acetate form in the presence of acetic anhydride can be used for the semi-quantitative determination of a-ketoglutaric acid. It is of interest that pyruvic acid does not interfere with this determination of a-keto- glutarate, nor does oxaloacetate when the keto acid concentration is not extremely low.These observations, together with the finding that substantial amounts of BSA also do not interfere, suggests the possible use of this reaction in the assay of transaminases such as glutarate oxaloacetate transaminase and glutarate pyruvate transaminase. 111). TABLE IV SEMI-QUANTITATIVE DETERMINATION OF a-KETOGLUTARIC ACID BY THE USE OF ANION-EXCHANGE RESIN BEADS IN THE ACETATE FORM AND ACETIC ANHYDRIDE AT 165 & 2 "C Amount of a-ketoglutaric acid added/ pg Colour of resin beads* 50 LBr 150 Br 250 BrG 350 BrG 500 G * Abbreviations as in Table I. The results show that acetic anhydride alone reacts with acids very slowly, even at high temperature (165 "C), whereas in the presence of salts it reacts very rapidly at 165 "C.Higher acids, such as ketodicarboxylic and hydroxytricarboxylic acids, give dark colours. It is known3 that the higher acid anhydrides can be prepared by heating a mixture of the acid and acetic anhydride. These observations suggest that in the colour reaction higher acids react with acetic anhydride to give a higher acid anhydride. The anhydride so formedMarch, 1980 DETECTION OF CARBOXYLIC ACIDS 22 1 reacts with acetic anhydride to give a coloured product and salts act only as catalysts. The following reaction scheme may be proposed for the acetic anhydride - cc-ketoglutaric acid reaction in the presence of a salt: (CH-&0)20+ HOOCCH2CH2COCOOH ___+ (OCCH2CH2COCO)20 + 2CH3COOH I 0 0 The large conjugated system of structure I1 seems to be responsible for the colour. We thank Professor M. Qureshi, Head, Chemistry Section, for providing research facilities M. N. A. and Dr. Saleem Uddin, Reader, Biochemistry Department, for helpful suggestions. is grateful to U.G.C. (India) for providing financial assistance. References 1. 2. 3. Ohkuma, S., J . Pharm. SOC. Jpn., 1955, 75, 1124. Feigl, F., “Spot Tests in Organic Analysis,” Seventh Edition, Elsevier, Amsterdam, 1966, p. 252. Finar, I. L., “Organic Chemistry,” Volume I, Sixth Edition, Longmans, London, 1973, p. 260. Received, August 29th, 1979 Accepted October l s t , 1979

ISSN:0003-2654    

DOI:10.1039/AN9800500217

出版商:RSC    

年代:1980

数据来源: RSC

摘要:

222 Analyst, March, 1980, Vol. 105, f$. 222-226 Use of an Orthogonal Function in Differential Spectrophotometry M. Abdel-Hady Elsayed," Yousry M. Elsayed and Hassan Abdine Department of Pharmaceutical Analytical Chemistry, Faculty of Pharmacy, University of Alexandria, Alexandria, Egypt An orthogonal function method has been applied to the differential spectro- photometric determination of acetazolamide and hydrochlorothiazide in tablets. The error resulting from setting the wavelength scale during the application of the orthogonal function to conventional spectrophotometry is thereby minimised. The results of assay using the proposed method [&) method] are more precise than those obtained with the p , method. Keywords : Pharmaceutical analysis ; differential ultraviolet spectrophotometry ; orthogonal function Differential spectrophotometric methodP3 were developed in order to improve the precision of the results of determinations of the main components of given samples.This improve- ment can be achieved through the appropriate expansion of the absorbance scale by using a solution of suitable concentration in the reference cell. In this instance the relative absor- bance value (A,), assuming the validity of Beer's law, is directly proportional to the con- centration : .. * . (1) A , = a (C, - C,) . . * . . . C, and C, are the concentrations of the analysed solution and the reference solution, res- pectively, and a, a constant, is the specific absorbance of the blank. The orthogonal function method4 has been used for the correction of the interfering sub- stances in the determination of the main components of pharmaceutical preparation^.^ In the application of the orthogonal function method to spectrophotometric analysis, the absor- bance, A , is replaced by the coefficient of orthogonal function, pi.Hence, subject only to the usual Beer - Lambert law limitations, a given coefficient is exactly proportional to con- centration : .. .. * ' (2) aj = ajc .. .. .. In order to extract the coefficient of a given polynomial from an absorption curve, it is necessary to obtain absorbances at a number of equally spaced wavelengths. Therefore, to extract the coefficient of the quadratic polynomial p 2 , we need, say, six absorbance mea- surements at six equally spaced wavelengths. By plotting the quadratic coefficient 9, at different intervals veysus A, (the mean of the set.of wavelengths) a convoluted curve is ob- tained.6 The optimum wavelength range is selected in order to maximise the coefficient p,.Also, Am is better sited on a broad peak or minimum than on a narrow peak. The sensitivity of the coefficient to over-all shifts in the spectrophotometer wavelength calibration is thereby minimised. On application of the orthogonal function to differential spectrophotometry, A , and a are replaced with and aj, respectively. Thus, where Pj(,) is the coefficient obtained from the convoluted differential curve. constitutes the basis of this paper. This theory * Present address : Department of Pharmacy, University of Nigeria, Nsukka, Nigeria.ELSAYED, ELSAYED AND ABDINE 223 Experimental Reagents Cairo, Egypt) in ethanol.(Elnile Company for Pharmaceutical and Chemical Industries) in ethanol. (50 mg of hydrochlorothiazide and 300 mg of potassium chloride) (Elnile). Acetazolamide standard solution. A 1 mg ml-l solution of acetazolamide (Cid Company, Hydrochlorothiaxide standard solution. A 1 mg ml-l solution of hydrochlorothiazide Tablets. Cidamox (250 mg of acetazolamide per tablet) (Cid Company) and Hydrazide K Instrument A Prolabo photoelectric spectrophotometer with l-cm silica cells was used. Determination of Optimum Concentration of Reference Solution 100 ml of acetazolamide or hydrochlorothiazide. each compound, 100 ml for the differential absorbance measurements. Prepare a set of solutions in 0.1 N sulphuric acid containing 0.1, 0.3, 0.5 and 1.9 mg per Carry out five sets of measurements for Select reference solutions of concentrations 0.7, 0.9, 1.1, 1.3 and 1.5 mg per Preparation of Standard Calibration Graphs Prepare a set of solutions in 0.1 N sulphuric acid containing 0.3, 0.5, 0.7, .. ., 1.8 mg per 100 ml of acetazolamide or hydrochlorothiazide. Measure the absorbance ( A value) of these solutions at equally spaced wavelengths ( p , method), using 0.1 N sulphuric acid as the blank. Measure the absorbance (A,. value) at Amax. ( A , method) or at equally spaced wavelengths [$,(,.) method], using a solution containing 1.3 mg per 100 ml (for acetazolamide) or 0.9 mg per 100 ml (for hydrochlorothiazide) as the blank. Assay for Tablets extract the active ingredient with ethanol (4 x 25 ml).Powder and mix 20 tablets. Weigh accurately an amount equivalent to one tablet. Next Dilute the ethanolic solution with 210 250 290 330 W avelen gthh m Positive scale: A curve (solid line); and A, curve (broken line). Negative scale: p , curve (solid line) ; and pzcr) curve (broken line) derived therefrom. Sample, 2.3 mg-% of acetazolamide in 0.1 N sulphuric acid. Fig. 1. - 20 210 250 290 330 Wavelength/nm Fig. 2. Positive scale: A curve (solid line); and A, curve (broken line). Negative scale: p , curve (solid line); and p,(,) curve (broken line) derived therefrom. Sample, 2 mg-yo of hydrochlorothiazide in 0.1 N sul- phuric acid.224 ELSAYED et al. : USE OF AN ORTHOGONAL Analyst, Vol. 105 0.1 N sulphuric acid as necessary in order to carry out the spectrophotometric measurements.Measure the A value and the A , value as described above for the different methods. Results and Discussion Differential Spectrophotometric Method (A, Method) In the differential spectrophotometric determination the A , values of acetazolamide and hydrochlorothiazide solutions in 0.1 N sulphuric acid were measured at the corresponding Amax. values, as shown in Figs. 1 and 2 (acetazolamide a t 264nm, hydrochlorothiazide at 270 nm). In each of the five sets of experiments the relative absorbance values (Ar), when plotted against concentration, yielded linear graphs. By use of the method of least squares,' the equations of the best lines fitting the experimental points were determined, together with the standard deviation of the slopes, Sb.' The best results when employing differential spectro- photometry can be obtained by using the reference solution for a set of experiments that gives the lowest s b value.2 Therefore, the concentrations of the reference solutions used for the differential measurement of acetazolamide and hydrochlorothiazide were 1.3 and 0.9 mg per 100 ml, respectively (Tables I and 11).To check for the validity of the equations (with lowest S , value, Tables I and 11) the laboratory-made tablets of acetazolamide and commercial tablets for both compounds were analysed using the A , method. The results are given in Table 111. TABLE I DETERMINATION OF OPTIMUM CONCENTRATION OF REFERENCE SOLUTION FOR ACETAZOLAMIDE Concentration in reference solution/ Linear calibration graph* 1 0.7 -0.3215 + 0.4486C 5.59 2 0.9 -0.4433 + 0.4750C 3.31 3 1.1 -0.4977 + 0.45506 3.98 4 1.3 -0.5824 + 0.45206 2.50 5 1.5 -0.6766 + 0.451 6C 3.24 * In this investigation, the reproducibility of any calibration graph was checked by taking three different amounts, from each of which three different concentrations were prepared.Experiment No. mg per 100 ml f A , = u + bC Sb x 10s Orthogonal Function Method (pz Method) For the application of an orthogonal function to the spectrophotometric analysis of acet- azolamide and hydrochlorothiazide the quadratic polynomial p , was chosen (Figs. 1 and 2). The absorbance values for different concentrations in 0.1 N sulphuric acid were measured over the wavelength range 256-276 nm at 4-nm intervals (Am, 266 nm) for acetazolamide, and over the wavelength range 268-278nm at 2-nm intervals (Arn, 273nm) for hydro- TABLE I1 DETERMINATION OF OPTIMUM CONCENTRATION OF REFERENCE SOLUTION FOR HYDROCHLOROTHIAZIDE Concentration in reference solution/ Linear calibration graph* Experiment No.mg per 100 ml f A , = a + bC s b x 10' 1 0.7 -0.4336 + 0.6277C 3.37 2 0.9 -0.5454 + 0.6284C 2.63 3 1.1 -0.6623 + 0.6180C 5.28 4 1.3 -0.7948 + 0.6175C 2.91 5 1.5 -0.9357 + 0.6170C 3.54 * See footnote to Table I.March, 1980 FUNCTION IN DIFFERENTIAL SPECTROPHOTOMETRY 225 chlorothiazide. The wavelength range mentioned above was chosen as the analytical set of wavelengths because the corresponding 9, value is a maximum. The quadratic coefficient, p,, was calculated from $2 == [(+5) A , + (-1) A2 + (-4) A3 + (-4) A , + (-1) A , + (+5)A,]/84 in which A,, A,, A,, .. . are the absorbance values for the compound at the corresponding wavelengths mentioned above, the numbers in parentheses are obtained from the work of Fisher and Yatesa and the divisor is the normalising factor. Within the concentration range 0.3-1.8 mg per 100 ml for acetazolamide and hydrochloro- thiazide, p , versus C shows a linear relationship. The corresponding calibration graphs can be described by the following equations : p , x lo3 = +0.08357 - 9.1587C for acetazolamide p2 x lo3 = -0.0479 - 9.4800C for hydrochlorothiazide The validity of the above equations was tested by analysing laboratory-made tablets of acetazolamide and commercial tablets of both compounds (Table 111). TABLE I11 ASSAY RESULTS FOR ACETAZOLAMIDE AND HYDROCHLOROTHIAZIDE TABLETS Mean* recovery? f standard deviation, % Tablet A, method Ace t azolamide (laboratory prepared) .. 101.57 It 1.39 Cidamox . . . . . . 100.74 =k 1.21 (4.23)§ (2.29) (7.52) Hydrazide K . . . . 96.38 & 1.SO ps method 100.78 2.34 (0.94) 99.69 3 2.78 (0.32) 96.99 f 3.21 (2.65) 99.92 + 0.81 (0.28) 100.67 f 1.03 (1.84) 95.79 3 0.95 (12.53) 1 A method 101.68 -f 0.90 (5.28) 102.10 1.27 (4.68) 98.68 =k 0.90 (3.03) F-value$ 8.35 7.28 11.41 * Mean of 14 determinations (A, method) and 8 determinations (other methods) t Recovery for commercial tablets is expressed as a percentage of the label claim. $ The calculated F-value for which the theoretical value at the 5% level equals 3.79; the F-value is determined to compare the p , and 3 The figures in parentheses are the calculated t values, for which the theoretical value ( a = 0.05) for seven degrees of freedom = 2.365 p2(1) methods.and for thirteen degrees of freedom = 2.160. Application of Orthogonal Functions to Differential Spectrophotometry : the Proposed Method Method] The error involved in the p , method, as indicated by the relatively high standard deviation, can be attributed to the wavelength-setting errors as absorbance measurements are usually made on the slopes of the absorption curve^.^ Moreover, for a given error in setting the wavelength scale, the resultant absorbance error increases with the slope of the absorption curve at the wavelength of measurement. Therefore, any procedure that diminishes the slope will also reduce such error.The coefficient 92(r) was calculated by measuring the absorbance values of a sample solution at a set of wavelengths equally spaced against a reference solution of optimum concentration. When the concentration of reference solution becomes equal to that of sample solution negli- gible slope is encountered on the spectrum and this is termed the balance point.1° In fact, the slope of the spectrum decreases with increasing concentration of reference solution until the balance point is reached. In the course of determining the orthogonal function coefficient ( p , method) most of the absorbance measurements are made on the slope of the absorption curves, so that a major proportion of the variance of a coefficient must arise from wavelength- setting error.An optimum set of experimental conditions must be worked out to find the optimum concentration of the reference solutions. With the optimum concentration of the reference solutions known (Tables I and 11), the orthogonal function can be applied to differen- tial spectrophotometry. The optimum experimental conditions for the application of the orthogonal function method to the differential spectrophotometric determination of acet- azolamide and hydrochlorothiazide are presented in Table IV. In order to measure the accuracy of each method [A,, p , and$,,,, methods] the true amount of the drug should be known. This amount can be found by the analysis of tablets prepared226 ELSAYED, ELSAYED AND ABDINE TABLE IV OPTIMUM EXPERIMENTAL CONDITIONS FOR THE APPLICATION OF AN ORTHOGONAL FUNCTION TO THE DIFFERENTIAL SPECTROPHOTOMETRY [pj(,,)] METHOD Experimental parameter Acetazolamide H ydrochlorothiazide Reference solution concentration/ mg per 100 ml .. . . .. 1.3 0.9 Wavelength interval/nm .. 4 2 Wavelength rangelnm . . .. 256-276 268-278 Median wavelength/nm . . .. 266 273 Concentration rangelmg per Regression equation . . . . p2(7) x lo3 = 11.4898-8.9204C x lo3 = 7.6634-8.5884C 100ml . . . . . . . . 1.5-2.7 1.1-2.3 in the laboratory by adding a specified amount of tablet base (composed of commercial lactose 60 g, calcium carbonate 30 g, starch 4 g, gelatin 3 g and talc 3 g ) to a known amount of the drug. The laboratory-prepared tablets were also assayed by the A method (traditional spectrophotometric method).Commercial tablets (Cidamox and Hydrazide K) were analy- sed after powdering and mixing 20 tablets of only one batch by each of the four methods (Table 111). The accuracy of each method is determined by calculation of the t value. For laboratory-prepared tablets and Cidamox, the calculated value of t in the 9, and p2(,) methods does not exceed the theoretical value at the 95% confidence level, whereas in the A , and A methods the calculated value is higher than the theoretical value. While the p , and P2(,) methods give assay results not significantly different from the true or label value, indicating their high accuracy, the A , and A methods give assay results significantly different from the true value. Therefore, the irrelevant absorbance is corrected mathematically by the use of either the p , or All of the methods for Hydrazide K give results significantly different from the true value, which suggests that the tablet does not really contain 100% of the label claim. With the P2(,) method a low standard deviation is obtained. The F-test shows that the precision of the +,(,,) method is better than that of the 9, method. method. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. References Svehla, G., Talanta, 1966, 13, 641. Pall, A., Svehla, G., and Erdey, L., Talanta, 1964, 11, 1383. Abdine, H., Elsayed, M. A., and Jbrahim, S. A., J . Pharm. Pharmacol., 1974, 26, 382. Glenn, A. L., J . Pharm. Pharmacol., 1963, 15, Suppl., 123T. Abdine, H., Wahbi, A. M., and Korany, M. A., J . Pharm. Pharmacol., 1971, 23, 444. Blackman, R. B., ‘and Tukey, J. W., “The Measurement of Power Spectra,” Dover Publications, Bauer, E. L., “Statistical Manual for Chemists,” Academic Press, London, 1971, pp. 61 and 109. Fisher, R. A:, and Yates, F., “Statistical Tables for Biological, Agricultural and Medical Research,” Ismail, M., and Glenn, A. L., Pharm. J., 1964, 16, Suppl., 150T. Agwu, I. U., and Glenn, A. L., J . Pharm. Pharmacol., 1967, 19, Suppl., 76s. New York. 1958, p. 72. Fourth Edition, Oliver and Boyd, Edinburgh, 1953, p. 80. Received May 22nd, 1979 Accepted September 26th. 1979

ISSN:0003-2654    

DOI:10.1039/AN9800500222

出版商:RSC    

年代:1980

数据来源: RSC

摘要:

Analyst March 1980 Vo,?. 105 $9 227-233 227 Tungsten Filament Vaporiser and Oxyhydrogen Flame for Optical-emission Spectrometry R. L. Warren Institute of Nuclear Medicine Middlesex Hospital Medical School Mortimer Street London W1 N 8A A The vaporisation of a sample from a tungsten filament into a hydrogen atmosphere with subsequent combustion in oxygen provides a novel source for optical-emission spectrometry. The burner and ancillary equipment are described. Some of the factors affecting sensitivity particularly in relation to flame background are discussed. Results are presented for calcium, chromium copper iron lead lithium magnesium manganese nickel silver, strontium and thallium. The precision is about 5% for a 3-111 sample of total element content between 10 pg (for lithium) and about 7 ng (for lead).Keywords Flame-emission spectrometry ; tungsten filament vaporiser ; micro-samples ; oxyhydrogen flame In the determination of sodium and potassium a t the picogram level it was found that the introduction of such a small sample into a physically small flame provided a greater proportion of emission of sample radiation over the emission of the flame itself than was the case when the flame was larger. The problem of the introduction of the sample into a small flame was solved by volatilising the sample from an electrically heated filament into the air supply of a pre-mixed a i r -hydrogen flame. The brief emission produced by the sample in the flame was integrated by means of a monochromator - photomultiplier combination monitoring the appropriate spectrum line in the flame the period of integration being controlled by a timer circuit.This method and its application to the analysis of cochlear endolymph was described by Bosher and Warren.l Subsequent similar applications of this technique have been reported by Haljamoe and Wood2 and Grime and Vickers.3 Certain advantages of the technique suggested a possible wider application even where sample size was not a restriction. Linearity can be assessed using only one concentration of standard solution by the application of multiples of the measured aliquot to the filament. Similarly sample matrix effects can be investigated by comparison of the level of signal obtained when sample and standard are simultaneously vaporised from the filament as opposed to the sum of the signals that they produce separately.The introduction of the sample vapour into the flame does not cause any significant alteration in the combustion processes for example by dilution and consequent cooling as occurs during the aspiration of aqueous aerosols. Advantage can also be taken of the fact that in some instances the temperature at which different elements are evaporated from the filament is sufficiently marked that by appropriate adjustment of the filament current a thermal fractionation can be obtained. Initial drying of the sample and subsequent ashing if required can be effected by similar control and following vaporisation of the element any sample residue can be removed by a final higher temperature heating of the filament.The initial limitation on the application of the original method to other elements was imposed by the use of a platinum filament and an air - hydrogen flame; the melting-point of the former would not allow vaporisation of such elements as calcium or magnesium nor would the temperature of the latter be suitable for their excitation. Experimental A new burner - filament combination has been developed together with a four-channel filament power supply and timer circuit to control the duration of heating in each channel and the period of integration of the sample emission. The system uses a tungsten filament operating in an atmosphere of pure hydrogen. The stream of hydrogen passing over the filament is ignited by a pilot oxyhydrogen flame as i 228 WARREN TUNGSTEN FILAMENT VAPORISER AND Analyst Vol.105 emerges from the burner. The pilot hydrogen supply is then cut off and the flame supported by the oxygen supply that had previously supplied the pilot flame. Access to the filament for application of the sample is provided by a hinged door fitted with a transparent silica window which permits observation of the behaviour of the sample on the filament. The door is sealed with a rubber O-ring and secured with a screw latch. Before opening the door the hydrogen supply to the filament is cut off then nitrogen is substituted and allowed to flush out the filament compartment. Fig. 1 shows a photograph of the burner with the door open and a cross-sectional diagram is shown in Fig. 2. The filament F is formed from a 65 mm length of 0.04 x 0.7 mm tungsten ribbon bent in the form shown.A 12-mm length a t each end of the ribbon is bent back on itself to ensure that the maximum voltage drop occurs across the centre V portion. These double thickness ends of the filament are clamped between split 8 mm diameter brass terminals (E) set 20 mm apart which pass through gas-tight PTFE insulated bushes in the back plate of the filament compartment. The terminals are connected to the power supply by cable rated a t 20 A a.c. current. In a pure hydrogen atmosphere the filament operates a t white heat with a current of about 16 A. Filament life has not been assessed; it is in excess of 1000 heating cycles and its destruction has so far been caused only by inadvertent overheating in an oxidising atmosphere or by mechanical mishandling.The burner body is screwed into the top of the filament compartment and is constructed so that the main stream of hydrogen emerges at the centre of three pilot flame jets (B) a t the top of the burner. These jets are angled towards the centre line and are formed from 10 mm long 0.6 mm internal diameter stainless-steel tubes pressed into the body of the burner. Oxygen and by-pass hydrogen are supplied through the concentric tube mixer as is shown in Fig. 2. A water-cooled jacket surrounds the upper 27 mm of the burner body. A 160 x 33 mm diameter transparent silica chimney is placed 25 mm above the top of the burner. This chimney has a steadying influence on the flame as well as providing a safe-guard against inadvertently reaching across the flame which is not readily seen being small and relatively non-luminous.By lowering the chimney and providing an adequate pure air supply the flame can be enclosed and isolated from the laboratory atmosphere as in the original design. Gas flow-rates are monitored and adjusted by means of flow meters fitted with needle valves. The consumption of gas depending on the particular element being deter-mined is oxygen 0.6-0.8 pilot hydrogen 0.15-0.20 main hydrogen 0.35-0.45 and nitrogen 0.8 1 min-1. The burner is positioned about 60 mm before the entrance slit of the monochromator and a t the centre of curvature of a 56 mm focal length concave mirror placed on the opposite side of the burner to the entrance slit. The aperture of this mirror and the position of the flame then ensure that the monochromator aperture is adequately filled.The monochro-mator was adapted from a design by Perry4 and is of the Littrow type using a 60" quartz prism with a refracting face 17.8 cm long by 11.4 cm high. The relevance of the use of a monochromator of such aperture is discussed in a later section. The photomultiplier is an EMI Type 6256 SA the output of which is connected to the input of a Model 6lOC Keithley electrometer which in the coulombmeter mode is used to integrate the flame emission. A Honeywell Electronik 194 pen recorder is driven by the output of the electrometer in order to provide a permanent record of the integrated signals. As a coulombmeter the electrometer functions as a current integrator or charge measuring device capable of integrating current of pulse width down to about 100 ns.The duration of the input to the electrometer during integration (channel 3) is controlled by the timer circuit shown in Fig. 3. It also controls the duration of heating of the filament in the other channels. The required current in each is provided by four separate variable transformers, TI which regulate the input to a single low voltage high current transformer T, the secondary winding of which is connected to the filament. The duration of each stage of heating is determined by four separate variable resistance - capacitor combinations (R and C,) in the timer circuit. For the purpose of determining the required level and duration of passage of filament current in each channel appropriate switching is incorporated to enable the power supplies and timer circuits to be manually adjusted separately independently of each other.Similarly only one channel is shown in the circuit diagram. This has only proved desirable for very low level sodium determinations. For clarity this switching is omitted from Fig. 3 March 1980 OXYHYDROGEN FLAME FOR OPTICAL-EMISSION SPECTROMETRY 229 Following the application of the sample to the filament the door of the compartment is closed and the nitrogen to hydrogen change-over effected; after re-ignition of the main flame the pilot hydrogen supply is turned off. This switching on and off of the pilot hydrogen and the change-over and reversal of the nitrogen and main hydrogen supplies are effected by use of electromagnetic valves that are switched in sequence by microswitches actuated by motor driven cams.The same motor also drives an additional set of cams and microswitches, which in turn energise the coils of four separate five-pole relays. Each relay selects the variable transformer (contact A) timer resistance and capacity (contacts B and C) and energises a single-pole relay RL (contact D) to trigger the timer. The timer then energises a three-pole relay RL, which stops the cam drive motor for the duration of the timing interval connects the mains supply (contact E) to the selected variable transformer and energises a single-pole change-over relay RL, via contact F to switch the photomultiplier anode from earth to the input of the electrometer. B i 1 cm I--+ D Fig. 2.Schematic cross-section of vaDoriser and burner " A = cooling water B = pilot flame jets C = oxygen D = hydrogen E = filament terminals F = filament G = hydrogen or nitrogen. The electrometer input El is normally earthed through the contacts of a single-pole relay, RL, but just prior to vaporisation of the sample (channel 3) this relay is de-energised by a cam-opening microswitch MS so that the flame emission is integrated when the sample is vaporised. At the expiration of the timing interval the filament current is switched off and the photomultiplier anode re-earthed by de-energising relay 4. The integrated charge stored in the input capacitor of the electrometer corresponds to the total photomultiplier current for the period defined by the setting of the timer and consists of the dark current and the current due to flame background plus the sample emission.As the recorder response lags behind that of the electrometer a sufficient time interval, defined by the dwell angle of the lobes of the cam controlling relay 5 allows the recorder pen to reach the correct indication of the magnitude of the charge after which the electro-meter is discharged by re-earthing the input and the pen returns t o zero. Twin lobes o 230 WARREN TUNGSTEN FILAMENT VAPORISER AND AnaZyst VoZ. 105 this cam and the cam actuating the microswitch of channel 3 repeat this channel and thus provide a recording of the flame background plus dark current. Following channel 4 and cleaning of the filament the gas solenoid valves are operated in order to re-light the pilot flame and change over the filament hydrogen supply to nitrogen.To limit the recorder chart drive to the appropriate part of the cycle chart drive switches are incorporated in the five-pole relays and together with over-ride switches these allow the chart drive to be on or off in any channel. + Fig. 3. Timer,* control circuit and filament supply A B C D = five-pole relay contacts; E F = three-pole relay contacts; RL = three-pole relay; RL and RL = one-pole relay; RL = one-pole relay change-over; R and C,* = timer resistance and capacitor; ZN 1034E = I.C. timer*; T = variable transformer 0-250 V 0.7 A; T = 240V/6V 100VA transformer; MS = micro-switch; E = electrometer input; E = electro-meter earth; PM = photomultiplier anode. * R.S. Components Ltd.Data sheet R/2466 for component types values and timing data. The sample volume as applied to the filament is of the order of 0 . 3 4 ~ 1 and provided this is identical for the standard and the sample it is not necessary to know the actual volume. The sample is pipetted with a small Pasteur pipette drawn from approximately 2 mm 0.d. silica tubing. For volumes below 1 pl it is drawn out to about 0.25 mm i.d. with the tip drawn to a diameter of 0.1 mm. Depending on the volume required a mark is made with waterporoof ink 10-20 mm from the tip. For volumes over 1 pl the pipette is drawn out so as to form a small bulb of such a volume as to contain the major portion of the sample, as in a conventional bulb pipette and with a mark on the constricted portion of the neck above the bulb.Either type of pipette is attached to a small rubber teat and is filled to the mark mainly by capillary effect pressure on the teat being used to stop the meniscus at the required level and later to expel the contents gently so that the liquid hangs as a single drop near the tip. The use of a magnifying glass is recommended when pipetting. Being made of a silica these pipettes can be thoroughly cleaned in a Bunsen burner flame after rinsing with distilled water. In the It can then be transferred to the filament. Drying of the sample requires a current of about 4.0 A to pass for about 45 s March 1980 OXYHYDROGEN FLAME FOR OPTICAL-EhlISSION SPECTROMETRY 23 1 presence of protein this drying time may need to be extended. It is important that visible bubbling or boiling of the solution on the filament should be avoided otherwise loss may occur.When required the removal of sodium and potassium is achieved at a dull red heat, i.e. at about 750 "C. This or a slightly lower temperature is also suitable for ashing of protein and depending on the nature of the sample the duration of treatment at this temperature is usually about 25 s with a filament current of about 8 A. The evaporation of elements such as calcium and magnesium requires a current of about 12 A but this need only be applied for about 0.3 s. The optimum evaporating current varies for different elements and lies between 9.0 and 12 A for the elements listed in Table I below. The final clean-up of the filament requires the maximum power available i .e . about 16A, which is applied for approximately 4 s. The reproducibility of the timer - integrator part of the circuit was measured by means of a constant current source being substituted for the photomultiplier and integrated for a period of 0.35s. The peak to peak difference for 24 separate integration3 was less than 0.3% of the recorder full-scale deflection when this deflection was equal to 3 x The photometric precision was assessed from a series of integrations of the 404.64nm line of a low pressure mercury lamp a t a band width of 0.1 nm. The photomultiplier dynode voltage was 1200 and the integration period and electrometer sensitivity were as above. The level of illumination was set to produce a photomultiplier anode current of about 8 x 10-8A; the dark current was 3 x 10-lOA.These conditions correspond closely to those employed in measuring the elements listed in Table I. The mean and relative standard deviation of the integrated charge were 2.24 x 10-8 C & 1.6% (n = 11). This degree of variation is principally a result of photomultiplier noise which can be reduced by increasing the light level and reducing either the dynode voltage or the input sensitivity of the electro-meter. For example with the band width increased to 1 nm and the dynode voltage reduced to 900 the relative standard deviation is reduced to 0.6%. C. TABLE I BACKGROUND EQUIVALENT CONCENTRATION CBeq AND DETECTION LIMITS cL Element Calcium Chromium Copper Iron . . Lead . . Lithium Magnesium Manganese Nickel Silver Strontium Thallium .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . Analytical linelnm 422.67 425.43 327.39 371.99 405.78 670.78 285.21 403.08 341.48 338.29 460.73 377.57 (a) (b) (4 Band width/ (-Ap-, nm p.p.m. Sr % C B ~ ~ p.p.m. CL p.p.m. 0.11 0.04 5.4 0.005 3 0.000 8 0.11 0.20 4.0 0.065 0.008 5 0.04 0.30 4.4 0.035 0.009 0.08 1.00 4.4 0.100 0.020 0.10 2.50 3.5 0.600 0.040 0.40 0.004 5.8 0.000 5 0.0002 0.03 0.20 3.6 0.025 0.006 0.10 0.05 5.6 0.008 5 0.001 4 0.06 0.50 3.0 0.075 0.010 0.06 0.10 5.5 0.008 0.003 0.15 0.05 6.0 0.008 0.001 0.08 0.20 4.8 0.030 0.004 (a) Sr calculated from 11 replicate determinations of each element. (b) ground. (c) of the background noise.Concentration of element used to determine CBeq and calculate CL. Relative standard deviation, CBeq = concentration of each element required to produce a signal level equal to the flame back-CL = concentration of element required to give a signal equal to three times the standard deviation Results and Discussion The instrument has been applied to the measurement of calcium in adrenal cells by Mackie et aZ.,5 and to the analysis of cochlear endolymph for calcium and magnesium by Bosher and Warren.6 The elements so far investigated in addition to sodium and potassium are shown in Table I. Fig. 4 is reproduced from recordings made during the compilation of the data recorded in Table I and illustrates the deflections obtained from eleven vaporisations of 3 p1 of a 0.05 p.p.m.manganese solution. Adjacent to each peak is the corresponding backgroun 232 WARREN TUNGSTEN FILAMENT VAPORISER AND Analyst Vol. 105 deflection. The mean and standard deviation for the line and the background were 36.00 & 2.00 and 6.2 0.34 respectively the units being divisions of the chart where 100 is equal to full-scale deflection. To the right are the deflections obtained from 1 2 3 and 4 drops of the same manganese solution applied sequentially to the filament dried and vaporised. For these last recordings the input sensitivity for full scale was reduced from 3 x 10-8 to 10-7c. n, 1 Fig. 4. Reproducibility and linearity for a 0.05 p.p.m. manganese solution (see text). Wavelength = 403.08 nm. If the band widths given in Table I are increased a reduction in sensitivity occurs.With strontium if the entrance and exit slits are increased from 0.020 mm to 0.20 mm the flame background signal increases by a factor of 100 but that from the line increases only about 16 times and the background equivalent concentration is increased from 0.008 to 0.045 p.p.m. When the element concentration is higher the need to differentiate between the spectrum line and the flame background is reduced the band width transmitted by the monochromator can then be increased and the precision improved by a reduction in the dynode voltage of the photomultiplier as is shown above. By using a band width of 1 nm a dynode voltage of 800 and an input sensitivity of the electrometer of 3 x lo-* C a manganese solution of 2 p.p.m.was analysed 11 times (22 evaporations) which gave a mean and standard deviation of 2.019 & 0.030 p.p.m. A t this level the flame background following evaporation was not discernible nor could it be detected adjacent to the spectral line during evaporation. If the flame background does increase during evaporation of the sample it must be measured and the level subtracted from the sample signal. This measurement is made adjacent to the spectral line and requires the wavelength setting of the monochromator to be altered accordingly. The present monochromator can be set with the necessary precision to ensure that the wavelength reproducibility is within the required limits i.e. at 404.35 nm to within 0.004 nm. When the concentration of the element is such that the flame background is not discernible, it is advantageous to increase the band width.This may not be practicable if it allows molecular bands or lines of other elements to overlap the analytical line as was described by Warren7 for the hydroxyl band overlapping magnesium and copper lines. The analysis of magnesium at 285.21 nm at a concentration below 1 p.p.m. and with the band width greater than about 0.06 nm is complicated by the overlapping of the 285.27 nm hydroxyl band and as the concentration increases the sensitivity is proportionally reduced as self-absorption increases. With copper the more intense persistent line at 324.75 nm is superimposed on the profile of a hydroxyl band and is therefore incapable of separation by increased instrument resolution whereas the 327.40 nm line is usable down to at least 0.1 p.p.m.provided that the band width does not exceed about 0.03 nm. It has been shown that when the element concentration is increased the band width can also be increased allowing the dynode voltage or the input sensitivity of the electrometer t March 1980 OXYHYDROGEN FLAME FOR OPTICAL-EMISSION SPECTROMETRY 233 be reduced. Alternatively a smaller monochromator could be employed. By use of a Hilger Uvispec monochromator with a 55 x 86 mm 30" glass prism and adapted to use a 13-stage photomultiplier a 0.1 p.p.m. calcium solution was measured with a background equivalent concentration and a limit of detection of 0.028 and 0.004 p.p.m. respectively. This monochromator equipped with a quartz prism is less suitable for the analysis of magnesium and copper for the reasons already given.The potential of electrothermal evaporation in flame emission spectrometry has been described and demonstrated. Normally one sample per minute can be measured as no cool-down period is required between samples. The burner is silent in operation and gas consumption is very low compared with other flame methods. The application of this method is not confined to the twelve elements listed here and probably extends to a further ten. The most notable feature of the method is the very small sample size the repeated measurement of which does not appear to introduce an unacceptable error as indicated by the relative standard deviation for the 2 p.p.m. manganese solution (1.5%) as opposed to the photometric precision (0.6%).The absolute amount of element applied to the filament and vaporised is between 12 pg (for lithium) and 7 ng (for lead). As with other emission methods it is also potentially capable of simultaneous multi-element analysis. Additionally the prior removal and analysis of the alkali metals by thermal fractionation is possible where a sufficiently marked difference in evaporating temperature exists. This can be useful in reducing the flame background by eliminating the continuous emission of these elements. The technique has so far been applied in the analysis of calcium and magnesium. The original construction of the monochromator was financed by a grant from the Paul Construction of the timer and filament supply and Instrument Fund of the Royal Society. development of the burner was financed by the Smith Kline and French Foundation. References 1. 2. 3. 4. 5. 6. 7. Bosher S. K. and Warren R. L. Proc. H . SOC. London Ser. B 1968 171 227 Haljamoe H. and Wood D. C . Anal. Bioclzem. 1971 42 155. Grime J. K. and Vickers T. J. Anal. Chem. 1975 47 432. Perry J. W. Trans. Opt. SOC. 1931 33 159. Mackie C. Warren R. L. and Simpson E. R. J . Endocrinol. 1978 77 119. Bosher S. K. and Warren R. L. Nature (London) 1978 273 377. Warren R. L. Analyst 1965 90 549. Received August 23rd 1979 Accepted September 17th 197

ISSN:0003-2654    

DOI:10.1039/AN9800500227

出版商:RSC    

年代:1980

数据来源: RSC