ANALYST OCTOBER 1986 VOL. 11 I 1159 A/-Oxalylamine(salicylaldehyde hydrazone) as an Analytical Fluorimetric Reagent for the Determination of Nanogram Amounts of Aluminium Fernando de Pablos Jose Luis Gomez Ariza and Francisco Pino Department of Analytical Chemistry Faculty of Chemistry University of Seville Seville Spain The synthesis characteristics and an a lytica I app I icat ions of N-oxa I yla m i ne( sa I icy1 a I de hyde hyd razon e) OSH, are described. The AI(III) - OSH system was studied and a spectrofluorimetric method is proposed for the determination of Al in a medium of water - dimethylformamide (1 + 1) at pH 3.7. Under these conditions the complex has excitation and emission maxima at 387 and 474 nm respectively. The detection limit is 5 p.p.b. and Al can be determined up to 160 p.p.b.Interferences were evaluated; Ga(lll) In(lll) Sb(lll) and Zr(lV) gave the greatest perturbations. The method has been applied to the determination of aluminium in mineral waters. Keywords N-Oxalylamine(salicylaldehyde hydrazone) reagent; aluminium determination; fluorimetry Derivatives of oxalylmonohydrazide and oxalyldihydrazide are structurally related to cuprizone but differ from it in the reactivity of the metal because the ring at the end of the cuprizone molecule is not aromatic. We began to carry out analytical investigations of these compounds in our Depart-mentl as previous work2-6 had shown the affinity of these reagents to metal ions of pseudo-noble gas electronic con-figurations such as Al(III) Ga(II1) and In(II1). Aluminium is an essential element with a low toxicity,7 but aluminium(II1) sulphate which is often added to drinking water in order to clarify it has proved to be highly toxic to patients undergoing dialysis treatment.8 For this reason drinking water standards have been fixed at very low levels; the aluminium concentra-tion in tap or mineral water should be about 50 p.p.b.In this paper the application of N-oxalylamine(salicyla1-dehyde hydrazone) OSH to the determination of aluminium in drinking waters has been tested. \ 0 0 OSH OH OSH exhibits a close structural relationship to N,N’-oxalylbis(salicyla1dehyde hydrazone) (OSBH) previously investigated by US,^ which may be considered to be a molecular duplicate of OSH. Both compounds are used as reagents for the determination of trace amounts of aluminium, gallium and indium as is N N’-oxalylbis(resorcyla1dehyde hydrazone) (OBRH).9 These reagents are therefore of interest owing to the toxicity of the elements concerned.Experimental Apparatus A Perkin-Elmer 554 spectrophotometer equipped with 1 .O-cm glass or quartz cells and a Crison-501 digital pH meter with a combined glass - Ag/AgCl electrode were used. The fluor-escence measurements were made with a Perkin-Elmer LS-5 spectrofluorimeter equipped with a xenon lamp source a Colora KS ultrathermostat and 1 .O-cm quartz cells. Solutions All chemicals were of at least analytical-reagent grade. A standard solution of aluminium(II1) was prepared from aluminium nitrate [Al(N03)3.9H*O] and was standardised gravimetrically with 8-hydroxyquinoline.Working solutions were prepared by appropriate dilution with distilled de-ionised water. Buffer solutions (pH 3.7) of potassium formate - formic acid (0.1 M) were used. Reagents Preparation of OSH Semioxamazide (1 .O g) was dissolved in 450 ml of ethanol and the solution was boiled under reflux. A 1.0-ml volume of salicylaldehyde was added and the mixture was refluxed for 1 h and then allowed to cool to room temperature. The yellowish white powder obtained was recrystallised from ethanol. The melting-point of the reagent was 277-280 “C. Elemental analysis gave a composition of C 51.70 H 3.94 and N 20.49% ; C9H903N3 requires C 52.17 H 4.34 and N 20.28%. The yield of the synthesis was 50%. Solutions of the reagent were prepared weekly in dimethylformamide (DMF).Properties of the reagent OSH is slightly soluble in water ethanol and chloroform and moderately soluble in DMF at room temperature. Infrared spectra (KBr discs) show the characteristic stretching vibra-tion bands of the structure assigned to OSH; N-H (3400 cm-I) 0-H (3250 cm-l) the amide I band and C=N (1665 cm-1) and C=C of the aromatic ring (1610-1475 cm-1). The ultraviolet spectrum of an aqueous 4% V/VDMF solution of the reagent in a neutral medium shows absorbance maxima at 330 and 280 nm; these bands undergo a hypo-chromic effect in acidic media and a bathochromic shift in alkaline media to 374 and 292 nm respectively. The ionisation constants were determined by classical spectro-photometric methods.lG-12 Only an acidic pK value at 2.3 could be accurately determined because the reagent under-goes moderate hydrolysis in alkaline media.The reagent exhibits fluorescence properties in ethanol solution (hex = 385 nm he = 470 nm) and also in aqueous 4% V/V DMF solutions at pH 5.0 (hex = 370 nm he = 470 nm) the wavelength of fluorescence being affected by the pH of the medium (hex = 385 nm he = 480 nm at pH 9.0). Fluorimetric determination of aluminium Transfer 1.0 ml of 0.2% m/V OSH solution in dimethylform-amide into a 25-ml calibrated flask add 11.5 ml of dimethyl 1160 ANALYST OCTOBER 1986 VOL. 111 formamide and adjust the pH of the solution with 1.0 ml of buffer solution (to give an apparent sample pH of 5.1). Add not more than 10 ml of a test solution containing 0.30-1.98 pg of aluminium and dilute to volume with distilled water.Measure the fluorescence at 474 nm with excitation at 387 nm. Determine the amount of aluminium in the sample from a calibration graph prepared under identical conditions. Alternatively a standard additions procedure based on the above can be applied when interfering ions are present and for samples with a high salt content. Results and Discussion Reactions with Metal Ions The chromogenic properties of OSH on reaction with metal ions were tested in both acetic acid - acetate and ammonia -ammonium buffered media; the best results were obtained in acetic acid - acetate media and these are summarised in Table 1. The reagent is most sensitive to AI(III) Ga(II1) and In(III), which exhibit a noble gas electronic configuration.This is in agreement with the presence of oxygen atoms in the structure of the compound. Reactions of OSH with Fe(III) Ni(I1) and Zr(1V) interfere with the determination of Group IIIA elements. The chelates of AI(III) Ga(II1) and In(II1) with OSH exhibit fluorescence properties at pH 4.5. The highest relative fluorescence intensity of the three is that of the aluminium chelate (44); the relative fluorescence intensities of Ga(II1) and In(II1) are 22 and 3.5 respectively. The relative fluor-escence intensity of 0.1 p.p.m. of quinine sulphate is 12.3. The wavelengths of maximum excitation and emission of Al(II1) and In(II1) chelates are at 387 and 474 nm respectively and those of Ga(II1) are at 395 and 480 nm respectively.Study of the Aluminium - OSH System Aluminium solutions rapidly form a yellow chelate in an excess of OSH; the majority of this chelate is formed instantaneously and the chelation reaction is completed within 20 min. The absorbance value only increases by about 7.6% in this time and then remains stable for at least 7 h. The fluorescence intensity of the complex rapidly reaches a high value increasing by about 2% in 20 min and then remains constant for at least 7 h. Influence of pH The effect of pH on the absorbance and fluorescence of the chelate determines the experimental conditions for the determination of aluminium as both the metal and the reagent are affected by the acidity of the medium. Fig. l(a) shows a Table 1. Photometric characteristics of the complexes (acetate-buffered medium) Cation A,,,,/nm E,, /1 mol-l cm-l AI(II1) .. Ga(IT1) . . In(II1) . Ti(1V) . . Zr(IV) . . Fe(I1) . . V(V) . . U(V1) . . Co(I1) . . Cu(I1) . . Ni(I1) . . Pd(I1) . . 373 382 380 362 380 392 365 375 380 382 380 392 1.2 x 104 1.7 x 104 1.6 x 104 3.7 x 103 1.2 x 104 8.7 x 103 7.3 x 103 5.8 x 103 3.2 x 103 1.0 x 104 9.7 x 103 4.8 x 103 narrow pH interval (between 7.0 and 8.0) on the absorbance versus pH graph in which the absorbance is independent of the pH; for this reason a succinic acid - succinate buffer solution (pH 6.3) was selected for the preparation of samples with a final pH of about 7.3. As can be seen by the fluorescence intensity versus pH graph the fluorescence is also independent of pH between pH 3.5 and 4.5 [Fig.1(6)] in a 10% DMF - water medium hence the pH is fixed in the optimum interval with a buffer solution (formic acid - formate) of pH 3.7. In addition the influence of pH on the fluorescence intensity in a 50% DMF - water medium has been tested; the results are analogous to those obtained previously although the optimum fluorescence intensity zone is slightly shifted to pH 5.1-5.7. The same buffer solutions are used to fix the pH in this interval. Stoicheiometry of the complex The molar ratio of OSH to aluminium was determined by the Asmusl3 and the modified Holme - Langmyhr14 methods, which are suitable for the dissociated chelate that OSH forms with the metal ion; other classical methods such as those of Yoe and Jones15 and Job16 do not yield reliable results.The results obtained which indicated a 1:l molar ratio are shown in Fig. 2. The Asmus and modified Holme - Langmyhr methods were also applied to the chelate in the excited state, measuring the fluorescence intensity at 474 nm with excita-tion at 387 nm. The results were in agreement with those of the chelate in the ground state. Fluorimetric Determination of Aluminium with OSH Aluminium can be determined fluorimetrically by following the recommended procedure described under Experimental. The most suitable concentration of the reagent was found to be 3.86 x 10-4 M (for 198 p.p.b. of aluminium) and the I 1 4 6 a 10 0.1 PH 70 .s 60 2 50 > C .-$ 40 $ 30 C 0 5 20 -10 2 3 4 5 6 7 8 9 PH Fig.1. Influence of pH on the aluminium - OSH system. (a) Graph of absorbance versus pH at 380 nm. CAI = 1 p.p.m.; CR = 1 X M in dimethylformamide (10%). ( b ) Grauh of fluorescence versus pH. C, = 1 p.p.m.; CR = 1 x 10-3 M in dimethylformamide (10%). he = 387 nm A, = 474 n ANALYST OCTOBER 1986 VOL. 111 1161 1 2 3 4 5 6 7 8 1 / A -4.6 - ! / 4 2 4.4 ' 3 - I 4.3 L- A 0.2 Log ( x - 1) Fig. 2. Determination of the composition of the aluminium - OSH chelate. (a) Asmus method. A n =2; B n = 1; C n = Yz. ( b ) Modified Holme - Langmyhr method. Conditions pH = 6.3 h = 380 nm. A = absorbance; L = free ligand concentration; V = volume of reagent stock solution taken; x = AmaX./A; n = ratio of ligand to metal ion in the complex Table 2.Effect of foreign ions on the fluorimetric determination of aluminium (39.7 p.p.b.) Foreign ion or species NO,- I- SCN- S2032- B4072-, Br0,- S042- SO3*- 104 ~ 10,-, acetate V(V) C103- Br- C032-, C104- Fe(CN),,- NH,(I) Mn(I1) . . . Hg(II) Bi(III) Ca(II) Ba(II) U(VI), Pb(II) Sr(II) TI(I) La(III) Ti(IV), Zn(II) Th(IV) Ni(I1) . . . . . . . Mo(VI) citrate Ag(I) Pd(II), C2042- Cu(II) Cr(III) Co(II), Mg(II) Au(III) W(VI) S2- Fe(III), Cd(II) As02- Fe(CN)64- . . . . . Sn(II) NO2- tartrate . . . . . . . , Ga(III) In(III) Sb(III) Zr(IV), . Be@) Ce(IV) F- P2072- As043-Tolerance, p.p.b. 1000 500 100 40 <40 Table 3. Determination of aluminium in mineral waters A1 found p.p.b. Sample OSH* AAS Bezoya . . . .. . 20.9 k 2.6 20 Bezoya (tetrabrik) . . 12.8 2 2.3 12 Lanjaron . . . . . . 14.1 k 1.3 16 Veri . . . . . . . . 10.9+ 2.6 10 Fonter . . . . . . 54.3 k 2.9 56 Lanjaron (carbonic) . . 10.4 2 3.1 10 * Average of three determinations. samples were prepared in DMF - water (1 + 1) media. There was a linear relationship between the aluminium concentra-tion and the fluorescence intensity over the range 0-160 p.p.b. The application of the recommended procedure to a series of 11 samples with an A1 content of 39.6 p.p.b. gave a relative error (P = 0.05) of 20.86%. Effect of foreign ions The cations or species that could affect the fluorescence emission of the A1 - OSH chelate were carefully examined. Table 2 summarises the results obtained. A remarkable tolerance of the method towards numerous anions and 12-fold amounts of Cu(II) Cr(III) Co(II) Hg(II) Zn(II) Ni(II), Pb(II) Ca(1I) and Sr(I1) was observed.However attempts to increase the selectivity of the procedure have been restricted to the determination of aluminium in water especially in table waters with some mineral ingredient to increase their digestive and diuretic action. This sample selection is based on the high sensitivity of the method which is in accordance with the low level of aluminium fixed in drinking water standards. In addition the alternative standard additions method proposed under Procedures increases the tolerance of the method for cations that are present in a large excess over aluminium in these types of waters e.g. Mg(I1) can be tolerated up to 10 p.p.m.against 19.4 p.p.b. of aluminium present (ratio 500 1) Ca(I1) up to 25 p.p.m. (ratio 1200 1) and Fe(II1) up to 1 p.p.m. (ratio 50 1). Determination of aluminium in mineral waters The concentration of aluminium in potable waters especially table water is often important in assessing both their quality and the performance of treatment plants. The proposed method was applied to six commercial bottled waters. In these the following ions are generally present; alkali metal ions [Na(I) 10 p.p.m.; K(I) 2 p.p.m.1; alkaline earth metal ions [Mg(II) 10 p.p.m.; Ca(II) 30 p.p.m.1; SO4*- 20 p.p.m.; C1- 5 p.p.m.; and HC03- 150 p.p.m. Veri water contained 0.09 p.p.m. of Fe(II1). The HC03- was removed by an acid and boiling treatment, and then an aliquot of 5.0 ml was used for the aluminium determination.The standard additions method17 was used in all instances. The results obtained are shown in Table 3. There is good accordance between the results obtained by this method and those obtained by atomic absorption spec-trometry with a graphite furnace. Conclusions Many organic reagents have been proposed for the fluori-metric determination of aluminium and Table 4 summarises some of the more important. With OSH it is possible to determine a very low concentration of aluminium and solvent extraction is not necessary. Most of the methods need a heating step to develop the fluorophore or alternatively a standing period; OSH does not require this time-consuming pre-treatment because the measurements can be carried out immediately.OSH is sensitive although it is surpassed by reagents such as 2-hydroxynaphthoic acid and lumogallion; however the OSH method is preferred as it is more rapid and involves fewer manipulations. OSH and OSBH exhibit similar sensitivities but the former is again preferred because in the OSBH method it is necessary to heat the reaction mixture at 60°C for 5 min and to cool it before measurement; a further advantage of OSH is its greater solubility than OSBH which makes the use of an ethanol - water medium possible although a DMF - water medium is recommended as the sensitivity o 1162 ANALYST OCTOBER 1986 VOL. 111 Table 4. Characteristics of reagents for the fluorimetric determination of aluminium Reagent 2-Hydroxy- 1 -naphthaldehyde benzoh ydrazone .. . . . . . . 2-Hydroxy-3-naphthoicacid . . . . . . 8-Hydrox yquinoline . . . . . . . . Lumogallion . . . . . . . . . . Lumogallion . . . . . . . . . . Morin . . . . . . . . . . . . . . Pontachrome BBR . . . . . . . . 2-Quinalizarinsulphonicacid . . . . . . OSBH . . . . . . . . . . . . OSH . . . . . . . . . . . . . . LJ nm 395 370 360 485 365 440 535 500 390 387 A e J nm 475 460 415 576 548 525 600 558 375 474 Experimental conditions pH 4.6; incubate at 40 "C pM5.8; activate after 1 h for 30 min pH 5.0; heat at 80 "C for 20 min With addition of Antarox CO 890 pH 3.0; activate after 20 min pH5.0; heat at 100 "C for 10 min pH 4.8; activate after 1 h pH 5.0; heat at 60 "C for 5 min pH 5.0 Solvent Ethanol - water Water Chloroform Water Water Ethanol - water Ethanol - water Water DMF - water DMF - water Sensitivity, p.p. b. References 100 18 2 19 20 20 4 21 0.5 22 23 50 24 20 25 11 26 5 1 5 This work the method is higher. Therefore the proposed method is fast, reliable and versatile and confirms the suitability of oxalyl-hydrazide derivatives as fluorimetric reagents for aluminium. The authors thank Dr. Domingo Martinez Ruiz and Dr. Miguel Lopez Artigues of the National Institute of Toxicology of Seville for the analysis by atomic absorption spectrometry. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. References G6mez Ariza J. L. Marques M. L. and Montaiia M. T., Analyst 1984 109 885.Narang K. K. and Yadav U. S. Indian J . Chem. 1980 19, 697. Narang K. K. and Bindal A . J . Sci. Res. Banaras Hindu Univ. 1978 28 1. Narang K. K. and Yadav U . S. Curr. Sci. 1980 49 852. Narang K. K. and Lal R. A. Curr. Sci. 1977 46 401. Narang K. K. and Dubey R. M. J. Sci. Res. Banaras Hindu Univ. 1979 30 173. Browning E. "Toxicity of Industrial Metals," Second Edition, Butterworths London 1969. Elliot H. L. Dryburgh F. Fell G. S. and Macdongall A. J . , Br. Med. J . 1978 1 1101. Pastor E. Pablos Pons F. and Gomez Ariza J. L. paper presented at the SAC 8613rd BNASS Conference Bristol UK, July 1986. Maroni P. and Calmon J. P. Bull. SOC. Chim. Fr. 1964,519. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. Agren A. Acta Chem. Scand. 1955 9 49. Sommer L. Folia Fac. Sci. Natur. Univ. Purkyn. Brno 1964, 5 1. Asmus E. Fresenius 2. Anal. Chem. 1960 178 104. Jimenez Sanchez J. C. Muiioz Leyva J. A. and Roman Ceba M. Anal. Chim. Acta 1977 90 223. Yoe J. H . and Jones A. L. Znd. Eng. Chem. 1944 16 111. Job P. Ann. Chim. 1928 10 9. Bader M. J. Chem. Educ. 1980 57 703. Uno T. and Taniguchi H. Bunseki Kagaku 1971 20 113. Kirkbright G. F. West T. S. and Woodward C. Anal. Chem. 1965,37 137. Rees W. T. Analyst 1962 87 202. Nishikawa T. Hiraki K. and Nagano N. Bunseki Kagaku, 1970 19 551. Ishibashi N. and Kina K. Anal. Lett. 1972 5 637. Nishikawa Y . Hiraki K. Morishige K. and Shigematsu T., Bunseki Kagaku 1967 16 692. Will F. 111 Anal. Chem. 1961 33 1960. Donaldson D. E . U.S. Geol. Surv. Pro5 Pap. No. 550-d, 1966 p. 258; cf. Possidoni J. F. An. Assoc. Quim. Argent., 1963 51 96. Capitan F. Roman Ceba M. and Guiraum A. An. Quim., 1974 70 508. Paper A51257 Received July 15th 1985 Accepted April 2nd I98