- Email: [email protected]
Determination of submicrogram amounts of gallium by ion-exchanger fluorimetry Determination of gallium in natural waters
0039-9r40/90$3.00+ 0.00 Copyright 0 1990Pergamon Press plc
Tafanro, Vol. 37, No. 2, pp. 193-199, 1990 Printed in Great Britain. All rights reserved
DETERMINATION OF SUBMICROG~M AMOUNTS GALLIUM BY ION-EXCHANGER FLUORI~ETR~ DETERMINATION
OF GALLIUM
IN NATURAL
OF
WATERS
F. CAPITAN, A. NAVALON, J. L. VILCHEZand L. F. CAP~AN-VALERY* Department of Analytical Chemistry, University of Granada, 18071-Granada, Spain (Received
16 A4ay 1988. Revised 7 August 1989. Accepted
16 Augusf 1989)
Summa~--A method for microdetermination of gallium at ng/ml level has been developed, based on ion-exchanger fluorimetry. The gallium reacts with sahcylidene-o-aminophenol to give a highly fluorescent complex, which is fixed on a dextran-type cationic resin. The fluorescence of the resin, packed in a l-mm silica celi, is measured dire&y with a solid-surface attachment. The range of ~n~ntratio~ of the method is 2.0-10.0 ng/ml, the RSD 1.3% and the detection limit 0.3 ng/ml. The method has been applied to the determination of gallium in natural waters. The gallium content found in tap water was higher than that in raw water. This is related to the use of commercial aluminium salts in the water-treatment
Scanning of fluorescence emitted from solid surfaces has been widely used in the direct assay of organic substances held on such surfaces, such as paper chromatograms, thin-layer chromatoplates, KBr disks, electrophoresis strips, and silicone rubber pads.‘-3 The solid-surface fluorescence technique has been used as a detector system in planar chromatography or electrophoresis, or as a simple reagentless system for analysis. It has been applied in environmental research, especially on air pollution,4*s forensic science, e.g., in questioned-document work,6 pesticide analysis by thin-layer chromatography,‘*7 with previous derivatization in some cases,* and biochemistry and clinical chemistry for determination of enzymes, substrates, activators and inhibitors.3 Our aim is to combine the measurement of solid-surface fluorescence with use of a solid, e.g., an ion-exchanger resin, to preconc&rtrate the analyte, made fluorescent by use of an appropriate reagent. To the best of our knowledge this technique has so far been used only once, for preconcentration of aluminium from dilute solution by use of an azo dye immobilized on silica.9 This approach, named by us ion-exchanger fluorimetry (IEF), can be potentially useful for the analysis of very dilute solutions, as in water analysis. Furthermore the fluorescence measured is the diffuse transmitted fluorescence, *Author for correspondence.
which increases the signal intensity, compared to that in the diffuse reflected fluorescence methods. This is due to the fact that at the 1 mm thickness used, the dextran-type resin employed is more translucent than the silica gel or alumina more generally employed. A related technique is the solid phase spectrophotometry devised by Yoshimura,” and called ion-exchanger photometry, that measures directly the absorbance of a resin that contains the analyte fixed as a coloured chromogenic species. One advantage of IEF methods is their simplicity, since it is not necessary to immobilize a reagent on an insoluble substrate. The reagents are added to the sample and the fluorescent species are collected on the resin, which is then transferred to a l-mm silica cell and measured without previous drying. Other advantages are decreased interferences and increased sensitivity (because this is a function of the sample volume). Hence, we can design methods tailored to the problem in hand. Here we propose a new method for determination of gallium by the IEF method. Reports on the antitumour activity of gallium,” its use as a tumour-scanning agent,12 as well as other studies on the pharmacokinetics and toxicity of gallium”*‘* suggest there is a need for sensitive and reliable methods for its dete~ination, Salicylidene-~-aminophenol (SOAPh), a known Schiff’s base Auorimetric reagent for gallium,‘5@ can be employed for IEF determination of gallium in natural waters. 193
F. CAPITAN et al.
194
Surprisingly, the gallium content found in tap water by us is ten times higher than that in raw water. This can be related to the use of commercial aluminium salts as flocculants in the water treatment.
EXPERIMENTAL
Reagents Zon-exchanger. To avoid contamination, the Sephadex SP C-25 cation-exchange resin in the sodium form is used without previous treatment. Salicylidene-o-aminophenol droxyaniline-N-salicylidene).
(SOAPh,
2-hy-
Synthesized by condensation of salicylaldehyde with o-aminophenol by heating at 100” for 30 min,” purified by repeated recrystallization from ethanol and dried at room temperature over silica gel (yield 77%; m.p. 184.5-185”). Its purity was checked by TLC. The reagent was characterized by its infrared and ‘H-NMR spectra and elemental analysis (required for (&H,,O,N: C, 73.24%; H, 5.16%; N, 6.57%; found: C, 73.1%; H, 5.01% ; N, 6.4%). Used as a 0.1% solution in absolute ethanol; the solution is stable for at least 1 week. Standard gallium (ZZZ) stock solution, 1.0 mg/ml. Prepared from Ga(NO,), . 8H20 in 0.1 M nitric acid and standardized titrimetrically with EDTA (Xylenol Orange as indicator); diluted further with doubly-distilled water as required. Bu$%r solutions. Prepared from monochloroacetic acid, acetic acid and ammonia.” Unless otherwise stated, the reagents were of analytical grade. Apparatus
All fluorimetric measurements were performed with an LS 5 Perkin-Elmer spectrofluorimeter provided with a Quantic Rhodamine 101 counter to correct the excitation spectra, a Hamamatsu R928 photomultiplier, a Houston Omnigraphic X-Y recorder, a variable-angle solid-surface accessory, designed and constructed by us, and a Braun Melsungen Thermomix 1441 thermostat. A set of fluorescent polymer samples was used daily to adjust the spectrofluorimeter and compensate for changes in source intensity. A Crison 501 digital pH-meter with calomel and glass electrodes, and an Agitaser 2000 rotating agitator were also used.
Fluorescence measurements
The mixture of Sephadex, sample solution and reagent was shaken mechanically in a polyethylene bottle. The resin beads were filtered off under suction and packed together with a small volume of solution into a l-mm silica cell by means of a pipette. The relative fluorescence intensity (RFI) measured was the diffuse fluorescence transmitted through the resin at the unirradiated face of the cell. The optima1 angle between the plane of the cell and the excitation beam was 45” in all cases. Procedures Basic procedure. A 500-ml water sample containing l&5.0 pg of Ga(II1) was transferred to a l-litre polyethylene bottle, and 2 ml of pH 4.70 buffer so1ution,‘5 5 ml of 0.1% SOAPh solution and 100 mg of Sephadex SP C-25 resin were added. The mixture was shaken mechanically for 20 min. Afterwards, the resin beads were collected by filtration under suction and with the aid of a pipette were packed into a l-mm cell together with a small volume of the filtrate. A blank solution containing all the reagents but no gallium was prepared and treated in the same way as the sample. The intensity of fluorescence (at 20.0 + 0.5’) for the sample and blank was measured at A,,,,= 515 nm (with 1, = 405 nm) 20 min after loading of the samples. A calibration graph was constructed in the same way, with Ga(II1) solutions of known concentration. Procedure for tap water. To a 500-ml water sample in a I-litre polyethylene bottle, 2 ml of pH 4.70 buffer so1ution,‘5 7 ml of 0.1 g/l. sodium tetrafluoroborate solution, 5 ml of 0.1% SOAPh solution and 100 mg of Sephadex SP C-25 resin were added. The mixture was shaken mechanically for 20 min, and the subsequent steps were as in the basic procedure. The standard-addition method was used for calibration. Procedure for raw water. To a 6000-ml water sample in a 15-litre polyethylene container, 24 ml of pH 4.70 buffer solution,” 24 ml of 0.1 g/l. sodium tetrafluoroborate solution, 60 ml of 0.1% SOAPh solution and 100 mg of Sephadex SP C-25 resin were added. The mixture was shaken mechanically for 60 min, and the analysis continued as for tap water. The standardaddition method was used for calibration. Treatment of sample. Natural waters (preserved by addition of 0.25 ml of concentrated nitric acid per litre) were filtered through a 0.45~,um membrane filter (Millipore) into a polyethylene container carefully cleaned with
Determination
60
of gallium
195
-
-7
-60
1 -
60
-
40
-
20
/ 0
I
I
I
I
I
I
350
400
450
500
550
600
0
WaveLength (nm)
Fig. 1. Fluorescence spectra of SOAPh-Ga(III) complex.(-) resin phase: [SOAPh] 9.4 x IO-‘M; [Ga(III)] 2.9 x IO-‘M; pH = 4.00; 100 mg of Sephadex SP C-25; 500-ml sample; stirring time 15 min; solution: [SOAPh] &,, 405 nm; I, 515 nm; f, = 0.4, slit, = slit, = 2.5 nm; T 20.0 kO.5”. (-.-) 1.6 x 10-4M; [Ga(III)] 2.9 x 10A5M; pH 4.00; 1, 410 nm; &,,, 515 nm;f, = 0.7; slit,, = slif = 2.5 nm; T 20.0 f 0.5”; ethanol 20%, (f, = sensitivity factor).
nitric acid. The pH of an aliquot of the water was adjusted to 4-5 with sodium hydroxide solution immediately before analysis.
RESULTS AND DISCUSSION
Excitation and emission spectra of the resin and solution with reacts Salicylidene-o -aminophenol Ga(II1) to give a fluorescent 1:l chelate at pH 4. I6 In the presence of Sephadex cationexchanger, the complex is sorbed on the resin, so is probably cationic since it is not fixed by an anionic resin. An SP C-25 dextran-type resin is recommended, as it gives less background fluorescence. The peak wavelengths (515 nm) in the emission spectra of the SOAPh-Ga(III) system are identical for the immobilized and solvated systems (Fig. 1). The similarities in the emission spectra of the immobilized and solvated systems suggest that the complex is relatively insensitive to its environment. The maxima of the excitation spectra of the two systems differ. The maximum is at 410 nm for the solution and 405 nm for the resin phase. On the other hand, the most noticeable differences between the fluorescence spectra are the better resolution of the resin phase spectra, and the smaller peak
width of the spectra of the solution system. The opposite is observed in the fixation of aluminium chelates on silica gel.9 From the half-life of the excited state of the complex in the solid phase at different temperatures, we infer that the luminescence process is fluorescence (r < 5 psec). Optimization of variables Dependence on PH. First, we chose a buffer solution from those proposed in the literature on this system in solution. The ammonium acetate/hydrochloric acid bufferi cannot be used, because it alters the resin. The monochloroacetic acid, acetic acid and ammonia mixtureI was found to give the best results. The optimum pH value, adjusted with this buffer, for the formation and fixation of the species falls in the narrow range 4.60-4.80 (Fig. 2). At pH values below 3.5 or above 5.5 the complex is not fixed on the resin. We should point out that the optimum pH for use of the resin phase is higher than that for use of the solution system (4.0). The fluorescence is independent of ionic strength (,u), adjusted with the buffer solution, up to 7.5 x 10m3. Above this value there is a decrease in the fluorescence according to the equation RF1 = 0.56~~‘~(r = 0.992). This effect could be due to increased competition of
F. CAPITANet al.
196 70r
100 -
60
-
60 -
40 -
20 -
k1
I
I
I
I
I
0
5
4
Fig. 2. Influence of pH on RFI. [SOAPh] 9.4 x 10m5M; [Ga(III)] 2.9 x 10-‘&f; 100 mg of Sephadex SP C-25; 500-ml sample; stirring time 15 min; LcX,,405 nm; & 515 nm; f, = 0.4; slit,, = slit, = 2.5 nm; T = 20.05 k 0.5”.
hydrogen and ammonium ions with the complex for ionic sites on the resin, or to deactivation of excited SOAPh in the resin. SOAPh concentration. The dependence of the RF1 on SOAPh concentration (Fig. 3) can be described by the equation RF1 = 177.7 + 11.5 ln[SOAPh] (r = 0.996). As optimum SOAPh concentration we chose 4.7 x 10e5M. For maximum fluorescence intensity the [SOAPh]/[Ga(III)] ratio must be at least 160. Influence of temperature. The ion-exchange process is independent of temperature in the range 10-60”, the RF1 being measured at 20” (Fig. 4). On the other hand the RF1 decreases when the temperature of measurement increases, but this effect is completely reversible. The decreases in RF1 are 2% at 25”, 9% at 40” and 25% at 60”. This effect can be explained as
t
0;
9
(SOAPhI,
lo+
I
I
I
I
I
20
30
40
50
60
T (W
PH
10
I 10
10
M
Fig. 3. Effect of SOAPh concentration on RFI. [Ga(III)] 2.9 x lo-‘M; [SOAPh] from 4.7 x 10e6 to 9.4 x 10e5M; pH 4.70; 100 mg of Sephadex SP C-25; 500-ml sample; stirring time 15 min; I, 405 nm; J., 515 nm; f, 0.4; slit,, = slit, = 2.5 nm; T = 20.0 & 0.5”.
Fig. 4. Influence of temperature. (a): On RF1 (ionexchange at 20”). (A): On ion-exchange process (RF1 measured at 20”). [SOAPh] 9.4 x 10-sM; [Ga(III)] 5.8 x IO-‘M; pH 4.70; 100 mg of Sephadex SP C-25; 500-ml sample; stirring time 15 mitt; & = 405 nm; 1, = 515 nm; /; = 0.3; slit,, = slit, = 2.5 nm.
due to internal conversion processes as the temperature increases, facilitating non-radiative deactivation of the excited singlet state, because deactivation is the intersystem crossing nearly independent of temperature.‘* All other measurements reported here were made at 20.0 & 0.5”. Other experimental conditions. The stirring time necessary for maximum RF1 development is 20 min for a 500-ml sample and 60 min for a 6000-ml sample. As the use of a large amount of the resin lowers the RFI, for all measurements we used the optimum amount needed to fill the cell and make handling convenient, viz. 100 mg. The RF1 of the immobilized complex decreased slightly in the first 20 min, and then was constant for at least 3 hr. Volume e&t on the sensitivity. In ionexchanger fluorimetry the sensitivity should be increased if a larger amount of sample solution is taken for equilibrium with the solid. The increase in sensitivity can be estimated by measuring the RF1 of Sephadex equilibrated with different volumes of solutions containing the same concentration of Ga(II1) and proportional amounts of the other reagents. Figure 5 shows the increase of fluorescence signal with sample volume. The shape of the graph suggests a Langmuir-type isotherm, as is observed in some ion-exchanger photometry studies.” Calibration
and precision
The calibration graphs for standards treated by the procedure described above were linear over the concentration range 2.0-10.0 pg/l.
Determination
of gallium
197
Sensitivity and detection limit
60
20 t
I
0
500
I 1000
I 1500
V (ml) Fig. 5. Influence of the sample volume on RFI. [SOAPh] 4.7 x 10e5M; [Ga(III)] 2.9 x IO-‘M; pH 4.70; 100 mg of Sephadex SP C-25; sample volume from 100 to 1500 ml; stirring time 20 min; 1, 405 nm; A,, 515 nm; f. 0.4; slit, = slit, = 2.5 nm; T = 20.0 + 0.5”.
gallium for 500-ml samples. The equation RF1 = 2.68C (r = 0.999) was found for this range of concentrations, where C is the concentration of gallium (,ug/l.) in the sample solution. The reproducibility of the proposed method and of the packing of the resin into the l-mm cell was found by use of 500-ml sample solutions with a gallium concentration of 8 pg/l. For 10 determinations the relative standard deviation (RSD) was 1.3%; 10 measurements for replicate packing of the resin gave an RSD of 1.5%. It seems that the packing of the resin makes a major contribution to the overall error.
The method reported here is substantially more sensitive than the solution methods using SOAPh as reagent. A calibration graph for the determination of Ga(II1) with SOAPh in solution, by the method of Dagnall et al.” but under our experimental conditions, gave the equation RF1 = 0.1 + 0.04C (r = 0.999), so the new method is 67 times more sensitive. The standard deviation of the background fluorescence measured for 10 determinations of the blank was 0.3 RF1 units for 500-ml samples. The IUPAC detection limit is therefore 0.3 pg/l. Ga(II1) and the determination limit 1.1 /lg/l. The method is compared in Table 1 with chelate formation methods described in the literature for fluorimetric determination of Ga(II1). For comparison purposes we have chosen those methods which in our opinion could be considered as among the most sensitive devised so far. Effect of foreign ions A systematic study was made of the effect of foreign ions on the determination of 8 pg/l. Ga(II1). A loo-fold w/w ratio of potentially interfering ion to gallium was first tested and, if interference occurred, the ratio was reduced progressively until interference ceased. Higher ratios were not tested. The tolerance level is defined as the amount of foreign ion that produces an error equal to &5% in determination
Table 1. Methods for the fluorimetric determination of gallium(II1) Reagent 2-Hydroxy-5-methylbenzaldehyde-4-aminoantipyrine 1,5-Bis(salicylidene) thiocarbohydrazonet Salicylidene-o-aminophenol-IEF 4-(5-Chloro-2-hydroxyphenylazo)resorcinolt 4-Amino-2,4-dihydroxybenzil-2,3-dimethyl-l-phenyl-5-pyrazolone Dodecyl-lumogallion Hexyl-lumogalliont I-Hydroxyquinaldinet Lumogalliont Pyridine-2-aldehyde 2-furoylhydrazone Resorcylidene-4-aminoantipyrine 4-Salicylideneaminoantipyrine Benzyl 2-pyridyl ketone 2-pyridylhydrazone Rhodamine B + Cl-g Rhodamine 6G + Cl-8 o-(Salicylideneamino)-2-hydroxybenzene sulphonic acid N-Salicylidene-2-hydroxy-5-sulphoaniline 3,5,7,4’-Tetrahydroxyflavone 1,5_Bis(salicylidene) thiocarbohydrazone *Or minimum concentration used for calibration. TExtraction procedure. $Temary complex and extraction procedure.
Detection limit* NIll. 0.1 0.1 0.3 0.5 1 1 1 1
1 1 1 1 1.4 20 3 2 2 2 2
Reference 21 22 This paper 23 24 25 26 27 28 29 30 31 32 33 33 34 35 36 37
198
F. CAPITAN et al. Table 2. Effect of foreign ions on the determination of 8 pg/l. gallium Tolerance, @Il.
Foreign ion or species Ag, IL Na, n(I), Ba, Be, Ca, Cd, Cu(II), HI(H), Mg, Mn, Ni, Co, Pd, Sr, Zn, As(III), B, Bi, Eu(III), In, La, Sb(III), Y, Ce(IV), Ge, Se(IV), Th, MO, W, NO;, NO;, SO:-, Cl-, PO;BF; Cr(II1) Fe(III) FCr(VI) AlfIII~
of the analyte. The results are summarized in Table 2. To apply this system to determination of Ga(II1) in natural waters, we studied the interference of the ions commonly found in water. The relative errors found in the determination of 8 ,ug/l. gallium were 8% for SG.- (20 mg/l.), 2% for NO; (2 mg/l.), 11% for Cl- (10 mg/l.), 26% for Ca*+ (36 mg/l.), 20% for Mg2+ (12 mg/l.). In all cases the interference was manifested as a reduction in calibration slope, necessitating determination by the standardaddition method. The most serious interference is that of A13+, however. Better tolerances for it can be accomplished by two methods. The first uses the different effect of temperature on the two systems, e.g., the RF1 of the aluminium complex is decreased (by 78% at 50”) more than that of the gallium complex (16% at 500), and moreover the change in RF1 is fully reversible for the gallium complex on cooling, but not for the aluminium complex.‘* In practice we take advantage of both features, packing the analyte-loaded resin in the l-mm cell and heating it for 15 min in a water-bath at 70”, then cooling it to 20” for measurement of the RFI. By this method, a fivefold increase in tolerance to Al(II1) is obtained (20 pg/l. instead of 4 pg/l.). The second system for eliminating the interference of A13+ uses sodium tetrafluoborate as masking agent. 2oThe optimum amount needed is a function of the aluminium content, because the masking agent itself interferes (see Table 2). For an aluminium level of 200 pg/l. a 7-fold weight ratio is needed and for 20 pg/l. a 20-fold ratio. Determination waters
of gallium in tap and natural
The method was applied to the determination of gallium in water samples by the standard-
>800 710 670 240 140 90 A
addition method. As representative samples we selected tap water and raw water from Granada town supplies (Spain). The sample volume was a function of the gallium content: 500 ml for the tap water samples and 6000 ml for the raw water. The loss of sensitivity by the matrix effect can be evaluated from the ratio of the slopes of the standard calibration graph and the standard-additions calibration graph (9 in this case). The use of a high volume of sample increases the sensitivity. The experimental sensitivity ratio, calculated from the ratio of the slopes of the standard-additions calibration graphs for the two volumes used, is 6.7. Before the gallium determination we established the aluminium content in the waters by AAS, in order to select the amount of BF; necessary. The average (+ standard deviation, five determinations) gallium found was 63.0 f 0.9 pg/l. for the tap water and 6.5 f 0.2 pg/l. for the raw water from a nearly natural supply (Quentar dam). To check the accuracy of the proposed method, we used the water samples for a recovery study. The recoveries of 20.0, 30.0 and 40.0 pg/l. gallium added to 500 ml portions of the tap water were 96.5, 102.3 and 101.0% respectively, and for 4.0,6.0 and 8.0 pg/l. added to 6000 ml portions of raw water, 95.5, 98.2 and 101.8% respectively. The gallium content of the Granada town tap water is surprisingly high compared with the raw water supplied to the town. To establish the source of the gallium we have analysed all the chemicals used in the Granada water treatment plant. Only the commercial aluminium salts used as flocculants contained gallium. To analyse these salts we first established the aluminium content by AAS and then determined the gallium. The results obtained were: aluminium sulphate 4.34% Al and 0.048% Ga; aluminium chloride, 9.53% Al and 0.047% Ga; aluminium chlorosulphate, 5.13% Al and
Determination
0.035% Ga. This implies that the flocculation with commercial aluminium salts increases the gallium content in tap water. It is interesting that the Al/Ga ratio (w/w) in the tap water (2.86) is practically the same as the ratio (2.92) in the raw water. In view of the AI/Ga ratio in the flo~ulating agents, this means that in the water treatment relatively more gallium than aluminium is extracted from the chemicals used. The high content of gallium found in the tap water would justify the gallium concentration in human urine in Tokyo found by Nakamura et ~1.~’ (48 pg/l.), since gallium is readily excreted from the human body.14 Acknowledgement-This research was supported by the Comision Asesora de Inv~tigaci~n Cientifica y T&x&a (CAICYT) of Spain (Project No. 3082/83).
REFERENCES
1. R. J. Hurtubise, Solid Surface Luminescence Analysis: Theory, Instrumentation, Applications, Dekker, New York, 1981. 2. Idem, Trace Analysis by Luminescence Spectroscopy, in Trace Analysis: Spectroscopic Methods for Molecules, G. D. Christian and J. B. Callis (eds.), Wiley, New York, 1986. 3. G. G. Guilbault, Photochem. Photobioi., 1977, 25,403. 4. E. Sawicki, Talanta, 1969, 16, 1231. 5. C. R. Sawicki and E. Sawicki, in Progress in Thin Layer Chromatography and Related Methods, Vol. III, Chapter 6, A. Niederwieser and G. Pataki teds.), Ann Arbor Science, Ann Arbor, 1972. 6. E. P. Gibson, J. Forensic Sci., 1977, 22, 680. 7. J. F. Lawrence and R. W. Frei, J. Chromatog., 1974,98, 253. 8. J. G. Zakrevsky and V. N. Mallet, ibid., 1977, 132, 315. 9. M. A. Ditzler, G. Doherty, S. Sieber and R. Allston, Anal. Chim. Acta, 1982, 142, 305. 10. K. Yoshimura, H. Waki and S. Ohashi, Taianta, 1976, 23,449. 11. M. M. Hart and R. H. Adamson, Proc. Natl. Acad. Sci. USA, 1980, 68, 1623. 12. P. B. Hoffer, C. Bekerman and R. E. Henkin (eds.), Galhum- Imaging, Wiley, New York, 1978. 13. R. A. Newman, A. R. Brody and I. H. Krakoff, Cancer, 1979, 44, 1728.
of gallium
199
14. D. P. Kelsen, N. Alcock, S. Yeh, J. Brown and C. Young, ibid., 1980, 46, 2009. 15. R. M. Dagnall, R. Smith and T. S. West, Chem. fnd. London, 1965, 34, 1499. 16. K. Morishige, Anal. Chim. Acta, 1980, 121, 301. 17. D. C. Freeman and C. E. White, J. Am. Chem. See., 1956, 7It, 2678. 18. D. M. Hercules, Fhtoreseence and Phosphorescence Analysis: Principles and Application, Interscience, New York, 1966. 19. K. Yoshimura, M. Ishii and T. Tarutani, Anal. Chem., 1986, Ss, 591. 20. V. Patrovsky, Chem. Listy, 1954, 4% 537. 21. A. T. Pilipenko and A. I. Volkova, Ukr. Khim. Zh., 1977, 43, 536. 22. N. Scott, D. E. Carter and Q. Fernando, Anal. Chem., 1987, 59, 888. 23. L. Ngog Thu, R. M. Dranitskaya and V. A. Nazarenko, Ukr. Khim. Zh., 1968, 34, 186. 24. A. T. Tashkhodzhaev, L. E. Zel’tser, Sh. T. Talipov and Kh. Khikmatov, Zh. Vses. klhim. Obshch. L). I. ~en~le~a, 1976, 21, 114. 25. K. Kenyu, H. Houichi and I. Nob~iko, Runseki Kagaky 1977, 26, 246. 26. K. Kina, K. Shiraishi and N. Ishibashi, ibid., 1976, U, 501. 27. T. Shigematsu, ibid., 1958, 7, 787. 28. T. Imasaka, T. Harada and N. Ishibashi, Anal. Chim. Acta, 1981, 129, 195. 29. E. Requena, J. J. Lasema, A. Navas and F. GarciaSanchez, Analyst, 1983, 103, 933. 30. A. T. Tashkodzhaev, L. E. Zel’tser, Sh. T. Talipov and Kh. Khikmatov, Zavodsk. Lab., 1975, 41, 281. 31. Sh. T. Talipov, A. T. Tashkodzhaev, L. E. Zel’tser and Kh. Khikmatov, Otkr. Izobret, Prom, Obraztsy, Tovarnye Znaki, 1972, 15, 1109. 32. J. J. Lasema, A. Navas and F. Garcia-Sanchez, Anal. Chim. Acta, 1980, 121, 295. 33. A. P. Golovina, S. M. Sapezhinskaya and V. K. Runov, Tezisy tret ‘ei vsesoyuznoi konferentsii po analiticheskoi khimii, Minsk, 1979, Vol. 1, p. 262, quoted in Zavodrk. Lab., 1981, 47, No. 1, 17. 34. K. Morishige, J. Inorg. Nucl. Chem., 1978, 40, 843. 35. Idem, Anal. Chim. Acta, 1974, 72, 295. 36. Z. T. Maksimycheva, Sh. T. Talipov, Z. P. Pakudina and S. Sadykov, Izv. Vyssh. Ueheb. Zaved. Khim. Khim. Tekhnol., 1974, 17, 348. 37. E. U&a, A. Garcia de Torres, J. M. Cano-Pavon and J. L. Gomez-Ariza, Anal. Chem., 1985, 57, 2309. 38. M. J. Blanch, Tesis Licenciatura, Pub. Univ. Granada, 1986. 39. K. Nakamura, M. Fujimori, H. Tsuchiya and H. Orii, Anal. Chim. Acta, 1982, 138, 129.
Tafanro, Vol. 37, No. 2, pp. 193-199, 1990 Printed in Great Britain. All rights reserved
DETERMINATION OF SUBMICROG~M AMOUNTS GALLIUM BY ION-EXCHANGER FLUORI~ETR~ DETERMINATION
OF GALLIUM
IN NATURAL
OF
WATERS
F. CAPITAN, A. NAVALON, J. L. VILCHEZand L. F. CAP~AN-VALERY* Department of Analytical Chemistry, University of Granada, 18071-Granada, Spain (Received
16 A4ay 1988. Revised 7 August 1989. Accepted
16 Augusf 1989)
Summa~--A method for microdetermination of gallium at ng/ml level has been developed, based on ion-exchanger fluorimetry. The gallium reacts with sahcylidene-o-aminophenol to give a highly fluorescent complex, which is fixed on a dextran-type cationic resin. The fluorescence of the resin, packed in a l-mm silica celi, is measured dire&y with a solid-surface attachment. The range of ~n~ntratio~ of the method is 2.0-10.0 ng/ml, the RSD 1.3% and the detection limit 0.3 ng/ml. The method has been applied to the determination of gallium in natural waters. The gallium content found in tap water was higher than that in raw water. This is related to the use of commercial aluminium salts in the water-treatment
Scanning of fluorescence emitted from solid surfaces has been widely used in the direct assay of organic substances held on such surfaces, such as paper chromatograms, thin-layer chromatoplates, KBr disks, electrophoresis strips, and silicone rubber pads.‘-3 The solid-surface fluorescence technique has been used as a detector system in planar chromatography or electrophoresis, or as a simple reagentless system for analysis. It has been applied in environmental research, especially on air pollution,4*s forensic science, e.g., in questioned-document work,6 pesticide analysis by thin-layer chromatography,‘*7 with previous derivatization in some cases,* and biochemistry and clinical chemistry for determination of enzymes, substrates, activators and inhibitors.3 Our aim is to combine the measurement of solid-surface fluorescence with use of a solid, e.g., an ion-exchanger resin, to preconc&rtrate the analyte, made fluorescent by use of an appropriate reagent. To the best of our knowledge this technique has so far been used only once, for preconcentration of aluminium from dilute solution by use of an azo dye immobilized on silica.9 This approach, named by us ion-exchanger fluorimetry (IEF), can be potentially useful for the analysis of very dilute solutions, as in water analysis. Furthermore the fluorescence measured is the diffuse transmitted fluorescence, *Author for correspondence.
which increases the signal intensity, compared to that in the diffuse reflected fluorescence methods. This is due to the fact that at the 1 mm thickness used, the dextran-type resin employed is more translucent than the silica gel or alumina more generally employed. A related technique is the solid phase spectrophotometry devised by Yoshimura,” and called ion-exchanger photometry, that measures directly the absorbance of a resin that contains the analyte fixed as a coloured chromogenic species. One advantage of IEF methods is their simplicity, since it is not necessary to immobilize a reagent on an insoluble substrate. The reagents are added to the sample and the fluorescent species are collected on the resin, which is then transferred to a l-mm silica cell and measured without previous drying. Other advantages are decreased interferences and increased sensitivity (because this is a function of the sample volume). Hence, we can design methods tailored to the problem in hand. Here we propose a new method for determination of gallium by the IEF method. Reports on the antitumour activity of gallium,” its use as a tumour-scanning agent,12 as well as other studies on the pharmacokinetics and toxicity of gallium”*‘* suggest there is a need for sensitive and reliable methods for its dete~ination, Salicylidene-~-aminophenol (SOAPh), a known Schiff’s base Auorimetric reagent for gallium,‘5@ can be employed for IEF determination of gallium in natural waters. 193
F. CAPITAN et al.
194
Surprisingly, the gallium content found in tap water by us is ten times higher than that in raw water. This can be related to the use of commercial aluminium salts as flocculants in the water treatment.
EXPERIMENTAL
Reagents Zon-exchanger. To avoid contamination, the Sephadex SP C-25 cation-exchange resin in the sodium form is used without previous treatment. Salicylidene-o-aminophenol droxyaniline-N-salicylidene).
(SOAPh,
2-hy-
Synthesized by condensation of salicylaldehyde with o-aminophenol by heating at 100” for 30 min,” purified by repeated recrystallization from ethanol and dried at room temperature over silica gel (yield 77%; m.p. 184.5-185”). Its purity was checked by TLC. The reagent was characterized by its infrared and ‘H-NMR spectra and elemental analysis (required for (&H,,O,N: C, 73.24%; H, 5.16%; N, 6.57%; found: C, 73.1%; H, 5.01% ; N, 6.4%). Used as a 0.1% solution in absolute ethanol; the solution is stable for at least 1 week. Standard gallium (ZZZ) stock solution, 1.0 mg/ml. Prepared from Ga(NO,), . 8H20 in 0.1 M nitric acid and standardized titrimetrically with EDTA (Xylenol Orange as indicator); diluted further with doubly-distilled water as required. Bu$%r solutions. Prepared from monochloroacetic acid, acetic acid and ammonia.” Unless otherwise stated, the reagents were of analytical grade. Apparatus
All fluorimetric measurements were performed with an LS 5 Perkin-Elmer spectrofluorimeter provided with a Quantic Rhodamine 101 counter to correct the excitation spectra, a Hamamatsu R928 photomultiplier, a Houston Omnigraphic X-Y recorder, a variable-angle solid-surface accessory, designed and constructed by us, and a Braun Melsungen Thermomix 1441 thermostat. A set of fluorescent polymer samples was used daily to adjust the spectrofluorimeter and compensate for changes in source intensity. A Crison 501 digital pH-meter with calomel and glass electrodes, and an Agitaser 2000 rotating agitator were also used.
Fluorescence measurements
The mixture of Sephadex, sample solution and reagent was shaken mechanically in a polyethylene bottle. The resin beads were filtered off under suction and packed together with a small volume of solution into a l-mm silica cell by means of a pipette. The relative fluorescence intensity (RFI) measured was the diffuse fluorescence transmitted through the resin at the unirradiated face of the cell. The optima1 angle between the plane of the cell and the excitation beam was 45” in all cases. Procedures Basic procedure. A 500-ml water sample containing l&5.0 pg of Ga(II1) was transferred to a l-litre polyethylene bottle, and 2 ml of pH 4.70 buffer so1ution,‘5 5 ml of 0.1% SOAPh solution and 100 mg of Sephadex SP C-25 resin were added. The mixture was shaken mechanically for 20 min. Afterwards, the resin beads were collected by filtration under suction and with the aid of a pipette were packed into a l-mm cell together with a small volume of the filtrate. A blank solution containing all the reagents but no gallium was prepared and treated in the same way as the sample. The intensity of fluorescence (at 20.0 + 0.5’) for the sample and blank was measured at A,,,,= 515 nm (with 1, = 405 nm) 20 min after loading of the samples. A calibration graph was constructed in the same way, with Ga(II1) solutions of known concentration. Procedure for tap water. To a 500-ml water sample in a I-litre polyethylene bottle, 2 ml of pH 4.70 buffer so1ution,‘5 7 ml of 0.1 g/l. sodium tetrafluoroborate solution, 5 ml of 0.1% SOAPh solution and 100 mg of Sephadex SP C-25 resin were added. The mixture was shaken mechanically for 20 min, and the subsequent steps were as in the basic procedure. The standard-addition method was used for calibration. Procedure for raw water. To a 6000-ml water sample in a 15-litre polyethylene container, 24 ml of pH 4.70 buffer solution,” 24 ml of 0.1 g/l. sodium tetrafluoroborate solution, 60 ml of 0.1% SOAPh solution and 100 mg of Sephadex SP C-25 resin were added. The mixture was shaken mechanically for 60 min, and the analysis continued as for tap water. The standardaddition method was used for calibration. Treatment of sample. Natural waters (preserved by addition of 0.25 ml of concentrated nitric acid per litre) were filtered through a 0.45~,um membrane filter (Millipore) into a polyethylene container carefully cleaned with
Determination
60
of gallium
195
-
-7
-60
1 -
60
-
40
-
20
/ 0
I
I
I
I
I
I
350
400
450
500
550
600
0
WaveLength (nm)
Fig. 1. Fluorescence spectra of SOAPh-Ga(III) complex.(-) resin phase: [SOAPh] 9.4 x IO-‘M; [Ga(III)] 2.9 x IO-‘M; pH = 4.00; 100 mg of Sephadex SP C-25; 500-ml sample; stirring time 15 min; solution: [SOAPh] &,, 405 nm; I, 515 nm; f, = 0.4, slit, = slit, = 2.5 nm; T 20.0 kO.5”. (-.-) 1.6 x 10-4M; [Ga(III)] 2.9 x 10A5M; pH 4.00; 1, 410 nm; &,,, 515 nm;f, = 0.7; slit,, = slif = 2.5 nm; T 20.0 f 0.5”; ethanol 20%, (f, = sensitivity factor).
nitric acid. The pH of an aliquot of the water was adjusted to 4-5 with sodium hydroxide solution immediately before analysis.
RESULTS AND DISCUSSION
Excitation and emission spectra of the resin and solution with reacts Salicylidene-o -aminophenol Ga(II1) to give a fluorescent 1:l chelate at pH 4. I6 In the presence of Sephadex cationexchanger, the complex is sorbed on the resin, so is probably cationic since it is not fixed by an anionic resin. An SP C-25 dextran-type resin is recommended, as it gives less background fluorescence. The peak wavelengths (515 nm) in the emission spectra of the SOAPh-Ga(III) system are identical for the immobilized and solvated systems (Fig. 1). The similarities in the emission spectra of the immobilized and solvated systems suggest that the complex is relatively insensitive to its environment. The maxima of the excitation spectra of the two systems differ. The maximum is at 410 nm for the solution and 405 nm for the resin phase. On the other hand, the most noticeable differences between the fluorescence spectra are the better resolution of the resin phase spectra, and the smaller peak
width of the spectra of the solution system. The opposite is observed in the fixation of aluminium chelates on silica gel.9 From the half-life of the excited state of the complex in the solid phase at different temperatures, we infer that the luminescence process is fluorescence (r < 5 psec). Optimization of variables Dependence on PH. First, we chose a buffer solution from those proposed in the literature on this system in solution. The ammonium acetate/hydrochloric acid bufferi cannot be used, because it alters the resin. The monochloroacetic acid, acetic acid and ammonia mixtureI was found to give the best results. The optimum pH value, adjusted with this buffer, for the formation and fixation of the species falls in the narrow range 4.60-4.80 (Fig. 2). At pH values below 3.5 or above 5.5 the complex is not fixed on the resin. We should point out that the optimum pH for use of the resin phase is higher than that for use of the solution system (4.0). The fluorescence is independent of ionic strength (,u), adjusted with the buffer solution, up to 7.5 x 10m3. Above this value there is a decrease in the fluorescence according to the equation RF1 = 0.56~~‘~(r = 0.992). This effect could be due to increased competition of
F. CAPITANet al.
196 70r
100 -
60
-
60 -
40 -
20 -
k1
I
I
I
I
I
0
5
4
Fig. 2. Influence of pH on RFI. [SOAPh] 9.4 x 10m5M; [Ga(III)] 2.9 x 10-‘&f; 100 mg of Sephadex SP C-25; 500-ml sample; stirring time 15 min; LcX,,405 nm; & 515 nm; f, = 0.4; slit,, = slit, = 2.5 nm; T = 20.05 k 0.5”.
hydrogen and ammonium ions with the complex for ionic sites on the resin, or to deactivation of excited SOAPh in the resin. SOAPh concentration. The dependence of the RF1 on SOAPh concentration (Fig. 3) can be described by the equation RF1 = 177.7 + 11.5 ln[SOAPh] (r = 0.996). As optimum SOAPh concentration we chose 4.7 x 10e5M. For maximum fluorescence intensity the [SOAPh]/[Ga(III)] ratio must be at least 160. Influence of temperature. The ion-exchange process is independent of temperature in the range 10-60”, the RF1 being measured at 20” (Fig. 4). On the other hand the RF1 decreases when the temperature of measurement increases, but this effect is completely reversible. The decreases in RF1 are 2% at 25”, 9% at 40” and 25% at 60”. This effect can be explained as
t
0;
9
(SOAPhI,
lo+
I
I
I
I
I
20
30
40
50
60
T (W
PH
10
I 10
10
M
Fig. 3. Effect of SOAPh concentration on RFI. [Ga(III)] 2.9 x lo-‘M; [SOAPh] from 4.7 x 10e6 to 9.4 x 10e5M; pH 4.70; 100 mg of Sephadex SP C-25; 500-ml sample; stirring time 15 min; I, 405 nm; J., 515 nm; f, 0.4; slit,, = slit, = 2.5 nm; T = 20.0 & 0.5”.
Fig. 4. Influence of temperature. (a): On RF1 (ionexchange at 20”). (A): On ion-exchange process (RF1 measured at 20”). [SOAPh] 9.4 x 10-sM; [Ga(III)] 5.8 x IO-‘M; pH 4.70; 100 mg of Sephadex SP C-25; 500-ml sample; stirring time 15 mitt; & = 405 nm; 1, = 515 nm; /; = 0.3; slit,, = slit, = 2.5 nm.
due to internal conversion processes as the temperature increases, facilitating non-radiative deactivation of the excited singlet state, because deactivation is the intersystem crossing nearly independent of temperature.‘* All other measurements reported here were made at 20.0 & 0.5”. Other experimental conditions. The stirring time necessary for maximum RF1 development is 20 min for a 500-ml sample and 60 min for a 6000-ml sample. As the use of a large amount of the resin lowers the RFI, for all measurements we used the optimum amount needed to fill the cell and make handling convenient, viz. 100 mg. The RF1 of the immobilized complex decreased slightly in the first 20 min, and then was constant for at least 3 hr. Volume e&t on the sensitivity. In ionexchanger fluorimetry the sensitivity should be increased if a larger amount of sample solution is taken for equilibrium with the solid. The increase in sensitivity can be estimated by measuring the RF1 of Sephadex equilibrated with different volumes of solutions containing the same concentration of Ga(II1) and proportional amounts of the other reagents. Figure 5 shows the increase of fluorescence signal with sample volume. The shape of the graph suggests a Langmuir-type isotherm, as is observed in some ion-exchanger photometry studies.” Calibration
and precision
The calibration graphs for standards treated by the procedure described above were linear over the concentration range 2.0-10.0 pg/l.
Determination
of gallium
197
Sensitivity and detection limit
60
20 t
I
0
500
I 1000
I 1500
V (ml) Fig. 5. Influence of the sample volume on RFI. [SOAPh] 4.7 x 10e5M; [Ga(III)] 2.9 x IO-‘M; pH 4.70; 100 mg of Sephadex SP C-25; sample volume from 100 to 1500 ml; stirring time 20 min; 1, 405 nm; A,, 515 nm; f. 0.4; slit, = slit, = 2.5 nm; T = 20.0 + 0.5”.
gallium for 500-ml samples. The equation RF1 = 2.68C (r = 0.999) was found for this range of concentrations, where C is the concentration of gallium (,ug/l.) in the sample solution. The reproducibility of the proposed method and of the packing of the resin into the l-mm cell was found by use of 500-ml sample solutions with a gallium concentration of 8 pg/l. For 10 determinations the relative standard deviation (RSD) was 1.3%; 10 measurements for replicate packing of the resin gave an RSD of 1.5%. It seems that the packing of the resin makes a major contribution to the overall error.
The method reported here is substantially more sensitive than the solution methods using SOAPh as reagent. A calibration graph for the determination of Ga(II1) with SOAPh in solution, by the method of Dagnall et al.” but under our experimental conditions, gave the equation RF1 = 0.1 + 0.04C (r = 0.999), so the new method is 67 times more sensitive. The standard deviation of the background fluorescence measured for 10 determinations of the blank was 0.3 RF1 units for 500-ml samples. The IUPAC detection limit is therefore 0.3 pg/l. Ga(II1) and the determination limit 1.1 /lg/l. The method is compared in Table 1 with chelate formation methods described in the literature for fluorimetric determination of Ga(II1). For comparison purposes we have chosen those methods which in our opinion could be considered as among the most sensitive devised so far. Effect of foreign ions A systematic study was made of the effect of foreign ions on the determination of 8 pg/l. Ga(II1). A loo-fold w/w ratio of potentially interfering ion to gallium was first tested and, if interference occurred, the ratio was reduced progressively until interference ceased. Higher ratios were not tested. The tolerance level is defined as the amount of foreign ion that produces an error equal to &5% in determination
Table 1. Methods for the fluorimetric determination of gallium(II1) Reagent 2-Hydroxy-5-methylbenzaldehyde-4-aminoantipyrine 1,5-Bis(salicylidene) thiocarbohydrazonet Salicylidene-o-aminophenol-IEF 4-(5-Chloro-2-hydroxyphenylazo)resorcinolt 4-Amino-2,4-dihydroxybenzil-2,3-dimethyl-l-phenyl-5-pyrazolone Dodecyl-lumogallion Hexyl-lumogalliont I-Hydroxyquinaldinet Lumogalliont Pyridine-2-aldehyde 2-furoylhydrazone Resorcylidene-4-aminoantipyrine 4-Salicylideneaminoantipyrine Benzyl 2-pyridyl ketone 2-pyridylhydrazone Rhodamine B + Cl-g Rhodamine 6G + Cl-8 o-(Salicylideneamino)-2-hydroxybenzene sulphonic acid N-Salicylidene-2-hydroxy-5-sulphoaniline 3,5,7,4’-Tetrahydroxyflavone 1,5_Bis(salicylidene) thiocarbohydrazone *Or minimum concentration used for calibration. TExtraction procedure. $Temary complex and extraction procedure.
Detection limit* NIll. 0.1 0.1 0.3 0.5 1 1 1 1
1 1 1 1 1.4 20 3 2 2 2 2
Reference 21 22 This paper 23 24 25 26 27 28 29 30 31 32 33 33 34 35 36 37
198
F. CAPITAN et al. Table 2. Effect of foreign ions on the determination of 8 pg/l. gallium Tolerance, @Il.
Foreign ion or species Ag, IL Na, n(I), Ba, Be, Ca, Cd, Cu(II), HI(H), Mg, Mn, Ni, Co, Pd, Sr, Zn, As(III), B, Bi, Eu(III), In, La, Sb(III), Y, Ce(IV), Ge, Se(IV), Th, MO, W, NO;, NO;, SO:-, Cl-, PO;BF; Cr(II1) Fe(III) FCr(VI) AlfIII~
of the analyte. The results are summarized in Table 2. To apply this system to determination of Ga(II1) in natural waters, we studied the interference of the ions commonly found in water. The relative errors found in the determination of 8 ,ug/l. gallium were 8% for SG.- (20 mg/l.), 2% for NO; (2 mg/l.), 11% for Cl- (10 mg/l.), 26% for Ca*+ (36 mg/l.), 20% for Mg2+ (12 mg/l.). In all cases the interference was manifested as a reduction in calibration slope, necessitating determination by the standardaddition method. The most serious interference is that of A13+, however. Better tolerances for it can be accomplished by two methods. The first uses the different effect of temperature on the two systems, e.g., the RF1 of the aluminium complex is decreased (by 78% at 50”) more than that of the gallium complex (16% at 500), and moreover the change in RF1 is fully reversible for the gallium complex on cooling, but not for the aluminium complex.‘* In practice we take advantage of both features, packing the analyte-loaded resin in the l-mm cell and heating it for 15 min in a water-bath at 70”, then cooling it to 20” for measurement of the RFI. By this method, a fivefold increase in tolerance to Al(II1) is obtained (20 pg/l. instead of 4 pg/l.). The second system for eliminating the interference of A13+ uses sodium tetrafluoborate as masking agent. 2oThe optimum amount needed is a function of the aluminium content, because the masking agent itself interferes (see Table 2). For an aluminium level of 200 pg/l. a 7-fold weight ratio is needed and for 20 pg/l. a 20-fold ratio. Determination waters
of gallium in tap and natural
The method was applied to the determination of gallium in water samples by the standard-
>800 710 670 240 140 90 A
addition method. As representative samples we selected tap water and raw water from Granada town supplies (Spain). The sample volume was a function of the gallium content: 500 ml for the tap water samples and 6000 ml for the raw water. The loss of sensitivity by the matrix effect can be evaluated from the ratio of the slopes of the standard calibration graph and the standard-additions calibration graph (9 in this case). The use of a high volume of sample increases the sensitivity. The experimental sensitivity ratio, calculated from the ratio of the slopes of the standard-additions calibration graphs for the two volumes used, is 6.7. Before the gallium determination we established the aluminium content in the waters by AAS, in order to select the amount of BF; necessary. The average (+ standard deviation, five determinations) gallium found was 63.0 f 0.9 pg/l. for the tap water and 6.5 f 0.2 pg/l. for the raw water from a nearly natural supply (Quentar dam). To check the accuracy of the proposed method, we used the water samples for a recovery study. The recoveries of 20.0, 30.0 and 40.0 pg/l. gallium added to 500 ml portions of the tap water were 96.5, 102.3 and 101.0% respectively, and for 4.0,6.0 and 8.0 pg/l. added to 6000 ml portions of raw water, 95.5, 98.2 and 101.8% respectively. The gallium content of the Granada town tap water is surprisingly high compared with the raw water supplied to the town. To establish the source of the gallium we have analysed all the chemicals used in the Granada water treatment plant. Only the commercial aluminium salts used as flocculants contained gallium. To analyse these salts we first established the aluminium content by AAS and then determined the gallium. The results obtained were: aluminium sulphate 4.34% Al and 0.048% Ga; aluminium chloride, 9.53% Al and 0.047% Ga; aluminium chlorosulphate, 5.13% Al and
Determination
0.035% Ga. This implies that the flocculation with commercial aluminium salts increases the gallium content in tap water. It is interesting that the Al/Ga ratio (w/w) in the tap water (2.86) is practically the same as the ratio (2.92) in the raw water. In view of the AI/Ga ratio in the flo~ulating agents, this means that in the water treatment relatively more gallium than aluminium is extracted from the chemicals used. The high content of gallium found in the tap water would justify the gallium concentration in human urine in Tokyo found by Nakamura et ~1.~’ (48 pg/l.), since gallium is readily excreted from the human body.14 Acknowledgement-This research was supported by the Comision Asesora de Inv~tigaci~n Cientifica y T&x&a (CAICYT) of Spain (Project No. 3082/83).
REFERENCES
1. R. J. Hurtubise, Solid Surface Luminescence Analysis: Theory, Instrumentation, Applications, Dekker, New York, 1981. 2. Idem, Trace Analysis by Luminescence Spectroscopy, in Trace Analysis: Spectroscopic Methods for Molecules, G. D. Christian and J. B. Callis (eds.), Wiley, New York, 1986. 3. G. G. Guilbault, Photochem. Photobioi., 1977, 25,403. 4. E. Sawicki, Talanta, 1969, 16, 1231. 5. C. R. Sawicki and E. Sawicki, in Progress in Thin Layer Chromatography and Related Methods, Vol. III, Chapter 6, A. Niederwieser and G. Pataki teds.), Ann Arbor Science, Ann Arbor, 1972. 6. E. P. Gibson, J. Forensic Sci., 1977, 22, 680. 7. J. F. Lawrence and R. W. Frei, J. Chromatog., 1974,98, 253. 8. J. G. Zakrevsky and V. N. Mallet, ibid., 1977, 132, 315. 9. M. A. Ditzler, G. Doherty, S. Sieber and R. Allston, Anal. Chim. Acta, 1982, 142, 305. 10. K. Yoshimura, H. Waki and S. Ohashi, Taianta, 1976, 23,449. 11. M. M. Hart and R. H. Adamson, Proc. Natl. Acad. Sci. USA, 1980, 68, 1623. 12. P. B. Hoffer, C. Bekerman and R. E. Henkin (eds.), Galhum- Imaging, Wiley, New York, 1978. 13. R. A. Newman, A. R. Brody and I. H. Krakoff, Cancer, 1979, 44, 1728.
of gallium
199
14. D. P. Kelsen, N. Alcock, S. Yeh, J. Brown and C. Young, ibid., 1980, 46, 2009. 15. R. M. Dagnall, R. Smith and T. S. West, Chem. fnd. London, 1965, 34, 1499. 16. K. Morishige, Anal. Chim. Acta, 1980, 121, 301. 17. D. C. Freeman and C. E. White, J. Am. Chem. See., 1956, 7It, 2678. 18. D. M. Hercules, Fhtoreseence and Phosphorescence Analysis: Principles and Application, Interscience, New York, 1966. 19. K. Yoshimura, M. Ishii and T. Tarutani, Anal. Chem., 1986, Ss, 591. 20. V. Patrovsky, Chem. Listy, 1954, 4% 537. 21. A. T. Pilipenko and A. I. Volkova, Ukr. Khim. Zh., 1977, 43, 536. 22. N. Scott, D. E. Carter and Q. Fernando, Anal. Chem., 1987, 59, 888. 23. L. Ngog Thu, R. M. Dranitskaya and V. A. Nazarenko, Ukr. Khim. Zh., 1968, 34, 186. 24. A. T. Tashkhodzhaev, L. E. Zel’tser, Sh. T. Talipov and Kh. Khikmatov, Zh. Vses. klhim. Obshch. L). I. ~en~le~a, 1976, 21, 114. 25. K. Kenyu, H. Houichi and I. Nob~iko, Runseki Kagaky 1977, 26, 246. 26. K. Kina, K. Shiraishi and N. Ishibashi, ibid., 1976, U, 501. 27. T. Shigematsu, ibid., 1958, 7, 787. 28. T. Imasaka, T. Harada and N. Ishibashi, Anal. Chim. Acta, 1981, 129, 195. 29. E. Requena, J. J. Lasema, A. Navas and F. GarciaSanchez, Analyst, 1983, 103, 933. 30. A. T. Tashkodzhaev, L. E. Zel’tser, Sh. T. Talipov and Kh. Khikmatov, Zavodsk. Lab., 1975, 41, 281. 31. Sh. T. Talipov, A. T. Tashkodzhaev, L. E. Zel’tser and Kh. Khikmatov, Otkr. Izobret, Prom, Obraztsy, Tovarnye Znaki, 1972, 15, 1109. 32. J. J. Lasema, A. Navas and F. Garcia-Sanchez, Anal. Chim. Acta, 1980, 121, 295. 33. A. P. Golovina, S. M. Sapezhinskaya and V. K. Runov, Tezisy tret ‘ei vsesoyuznoi konferentsii po analiticheskoi khimii, Minsk, 1979, Vol. 1, p. 262, quoted in Zavodrk. Lab., 1981, 47, No. 1, 17. 34. K. Morishige, J. Inorg. Nucl. Chem., 1978, 40, 843. 35. Idem, Anal. Chim. Acta, 1974, 72, 295. 36. Z. T. Maksimycheva, Sh. T. Talipov, Z. P. Pakudina and S. Sadykov, Izv. Vyssh. Ueheb. Zaved. Khim. Khim. Tekhnol., 1974, 17, 348. 37. E. U&a, A. Garcia de Torres, J. M. Cano-Pavon and J. L. Gomez-Ariza, Anal. Chem., 1985, 57, 2309. 38. M. J. Blanch, Tesis Licenciatura, Pub. Univ. Granada, 1986. 39. K. Nakamura, M. Fujimori, H. Tsuchiya and H. Orii, Anal. Chim. Acta, 1982, 138, 129.