ANALYST, JANUARY 1991. VOL. 116 27 Determination of Titanium(1v) in River Water by Ion-pair Reversed-phase High-performance Liquid Chromatography With 4,4' -Dia n ti pyry I met ha ne Nobuo Uehara, Kazuhiro Morimoto and Yoshio Shijo Department of Applied Chemistry, Faculty of Engineering, Utsunomiya University, Ishii-cho, Utsunomiya 32 I , Japan An ion-pair reversed-phase hig h-performance liquid chromatographic method for the selective determina- tion of Tiiv with 4,4'-diantipyrylmethane (DAM) is described. The TiIV-DAM complex was separated on an ODS column using acetonitrile-water (30 + 70) containing 1 x 10-4 mol dm-3 DAM, 0.01 mol dm-3 ammonium iodide and 0.02 mol dm-3 chloroacetate (pH 2.25). The detection limit for TiIV with the proposed method is 1.8 pg I-'. Titanium(iv) in river water can be determined without the interference of foreign ions after pr e-co ncen t rat ion.Keywords : Titanium( I v) ; 4,4 '-dian tip yrylm ethane; ion -pair re versed-p hase high -perf0 rmance liquid chroma t- ography; river water; pre-concentration 4,4'-Diantipyrylmethane (DAM) is the most popular reagent for the spectrophotometric determination of Ti" because of its selectivity.1J The reagent has been applied to the deter- mination of Ti1" in cements,3 ferro-niobium,4 silicate rocks5 and plant materials.6 However, it is difficult to use DAM for the determination of low levels of Ti" because the molar absorptivity of the Ti"-DAM complex is not very large. Recently, more sensitive spectrophotometric reagents for Ti'" have been reported. Inoue et al.7 described an extraction spectrophotometric method for Ti1' with N-p-octyloxy- benzoyl-N-phenylhydroxylamine and phenylfluorone. Marini et al. 8 used 2-(5-chioro-2-pyridylazo)-5-dimethylaminophenol and hydrogen peroxide for the determination of Ti". Gregorowicz el aZ.9 determined Ti'" in steel with Eriochrome Azurol G. However, these reagents are not as selective as DAM. Most of the work combining high-performance liquid chromatography (HPLC) with spectrophotometric detection has been aimed at the development of sensitive and selective analytical methods for determining metal ions. However, there are relatively few reports on the application of HPLC to Ti". Main and Fritz15 determined Ti'" by HPLC with bis(quaternary ammonium hydrazones) of 2,6-diacetylpyri- dine. The detection limit of Ti'" was 8 x 10-8 mol dm-3 when 2,6-diacetylpyridinebis(N-methylenepyridinohydrazone) was used as the chelating agent.However, the molar absorptivities of these Ti"-bishydrazone complexes are smaller than that of the Ti'"-DAM complex. Therefore, DAM appears to be more sensitive than these bishydrazones for the determination of Ti'" by HPLC. This paper describes the use of DAM as a pre-column chelating agent for the sensitive determination of Ti" by ion-pair reversed-phase HPLC. The proposed method was applied to the determination of Ti'" in river and rain water after pre-concentration using a simple and rapid evaporation of the sample solution. Experimental Apparatus The liquid chromatographic system consisted of a Nihon Seimitsu Kagaku (Tokyo, Japan) NSP-800-3U pump, a Japan Spectroscopic (Tokyo, Japan) hexane dumper, a Japan Spectroscopic 870-UV spectrophotometric detector, a Rheodyne 7125 loop injector with a 20 p1 sample loop and a Shimadzu (Kyoto, Japan) U-125 MU recorder. A Yamamura Kagaku (Tokyo, Japan) YMC R-ODs-5 column (250 X 4.6 mm i.d.) was used for all experiments.The simple laboratory- built evaporator shown in Fig. 1 was used for pre-concentra- tion; it was of the same type as that described in a previous paper. 16 Reagents Distilled, de-ionized water was purified further by means of a Millipore Milli-Q system. Analytical-reagent grade aceto- nitrile and methanol were filtered through a Millipore filter (0.45 pm) after distillation. The DAM was obtained from Dojindo (Kumamoto, Japan).The TitV standard solution (1000 mg 1 - 1 ) for atomic absorption spectrometry was obtained from Wako Pure Chemicals (Osaka, Japan). All other chemicals were of guaranteed-reagent grade. Evaporation Glass cylinder U Cooling t t /water Fig. 1 Schematic diagram of the laboratory-built evaporator28 ANALYST, JANUARY 1991, VOL. 116 Eluent and Chromatographic Conditions The eluent used was acetonitrile-water (30 + 70) containing 1 x 10-4 rnol dm-3 DAM, 0.01 mol dm-3 ammonium iodide and 0.02 mol dm-3 chloroacetate buffer (pH 2.25). The pH of the eluent was adjusted before addition of acetonitrile. The flow-rate of the eluent was 1.0 ml min-1. The eluate was monitored at 390 nm. Procedure Ten millilitres of 60% nitric acid were added to 1.0 1 of sampled river water and the solution was filtered through a Millipore filter (0.45 pm).To 10 ml of river water sample, 1.0 ml of 60% nitric acid and 0.1 ml of 3 mol dm-3 sulphuric acid were added. The solution was evaporated to dryness with the laboratory-built evaporator. The residue was dissolved in 0.4 ml of 2% DAM solution (0.5 mol dm-3 HC1) by shaking for 2 min and the solution was allowed to stand for 20 min. To a 0.3 ml aliquot of the solution, 0.1 ml of 4 mol dm-3 sodium chloroacetate solution and 0.1 ml of methanol were added. An aliquot of the solution (20 pl) was injected on to the HPLC column. Results and Discussion Derivatization Studies The TP-DAM complex was formed within 15 min in an acidic medium of 0.5 rnol dm-3 HCI. However, direct injection of 0.5 mol dm-3 HC1 solution into the chromatograph is inadequate.Therefore, the pH of the solution was adjusted to 2.25, which is a suitable pH for the ODS column. The solution must be injected into the chromatograph immediately after adjustment of the pH because the complex decomposes gradually at pH 3 9c L.LJ. HPLC Studies Fig. 2 shows a typical chromatogram of the TP-DAM complex. All the other metal ions studied, viz., A P , Crvl, CuII, Fell1, InlI1, NP, Mn", Vv and Z P , gave no peaks on the chromatogram under the conditions used. Iodide ion, which has been used for the ion-pair extraction of TitV with DAM was examined as a counter anion for HPLC of the TP-DAM complex, as the complex has a positive charge.2 Fig. 3 shows the effect of ammonium iodide concentration on the mass distribution ratio of the Ti'"-DAM complex.As the ammonium iodide concentration increased the mass distribution ratio of the complex also increased, which indicated that ammonium iodide acted as an ion-pair t v) 0 n F E 8 a L al I I I 1 I L 0 2 4 6 8 10 Time/m i n Fig. 2 Chromatogram of the TP-DAM complex. Column, YMC R-ODS-5 (250 x 4.6 mm i.d.); eluent, acetonitrile-water (30 + 70) containing 1 x mol dm-3 DAM, 0.01 mol dm-3 ammonium iodide and 0.02 mol dm-3 chloroacetate (pH 2.25). Ti'", 0.1 mg 1-1; flow-rate, 1.0 ml min-1; detection wavelength, 390 nm; detector sensitivity, 0.01 a.u.f.s.; and injection volume, 20 p1 reagent. The peak of the complex was well resolved from the solvent front above an ammonium iodide concentration of 1 X 10-2 mol dm-3.Acetonitrile was a good organic modifier for the retention of the TP-DAM complex. The mass distribution ratio of the complex decreased as the content of acetonitrile in the mobile phase increased, as shown in Fig. 4. An acetonitrile content of 30% v/v was selected in order to obtain a suitable chromato- gram for the detection of TiIV. The retention behaviour of the TP-DAM complex was investigated for various pH values of the eluent. As the pH of the eluent increased, the mass distribution ratio of the complex also increased and the peak broadened, as shown in Fig. 5 Sach behaviour appears to be due to hydrolysis of the complex and interaction between the residual silanol groups of the ODS column and the complex. The DAM concentration of the eluent also affected the retention of the TP-DAM complex.The peak height of the complex increased and the mass distribution ratio decreased with an increase in the DAM concentration. The addition of DAM to the eluent prevented the decomposition of the complex and allowed its rapid elution. 8 t 6 QE 4 2 0 -3 -2 -1 Log (cNH,,/mol dm-3) Fig. 3 distribution ratio. For chromatographic conditions see Fig. 2 Effect of concentration of ammonium iodide on the mass 4 I\ d 2 0 ' I 20 30 40 Acetonitrile content (% v/v) Fig. 4 For chromatographic conditions see Fig. 2 Effect of acetonitrile content on the mass distribution ratio. 2 & 1 1 lo0g .- a r Y I : 2.0 3.0 4.0 PH Fig. 5 peak height. For chromatographic conditions see Fig. 2 Effect of pH of the eluent on the mass distribution ratio andANALYST, JANUARY 1991, VOL.116 29 Table 1 Effect of foreign ions on the determination of TitV. Concentration of Ti1V added, 0.1 mg 1-l Concentration added/ Error* Foreign ion mgl-l W) 100 100 100 100 50 50 20 20 20 20 20 20 20 20 20 20 0.5 +1.3 +2.5 +2.2 -3.1 +2.9 -0.9 +1.3 +0.2 +0.9 +2.9 -3.3 -3.6 -0.9 -3.5 -1.2 +1.8 -4.9 * A plus sign indicates a positive error, a minus sign indicates a t Ascorbic acid was added as a masking agent. negative error. Calibration Graph and Detection Limit The calibration graph of peak height versus TiIV concentration was a straight line in the concentration range 0.01-0.3 mg 1-1 TiLV with 20 p1 injections. The detection limit for TiIV was 1.8 pg 1-1 at a signal to noise ratio of 3. The reproducibility of the method is 3.9%, expressed as the relative standard deviation for ten replicate analyses of solutions containing 0.1 mg 1 - 1 TiIV .Interferences The effect of foreign ions on the determination of TiIV with HPLC was studied. Table 1 shows the results. None of the metal ions tested interfered with the determination of TitV at concentrations commonly found in river water.17 Only Bi"' interfered seriously with the determination of TiIV. However, the interference from Bi"' would not be a problem in practice because the concentration of Bi"' in river water is very low. The presence of a large amount of FelI1 gave a broad peak which interfered with the determination of T P . However, the Fellr peak disappeared completely on adding ascorbic acid to the sample solution, because Fe" does not react with DAM.Applications The simple evaporation system used in a previous paper16 was employed for pre-concentration because the HPLC method Table 2 Results for the determination of TiIV in river and rain water Concentration of Ti1"*/ I - % - ' Sample Watarase river 1.85?0.10(n=4) Kinu river 0.76 k 0.09 ( n = 3) Rain water? 2.66 k 0.27 ( n = 3) * Mean k standard deviation. t Sampled at Utsunomiya University. was not sufficiently sensitive to determine Ti'" in river water directly. The 15-fold enrichment achieved by pre-concentra- tion enabled TitV to be determined in river water by HPLC. The water sampled at the Watarase and Kinu rivers (both in Tochigi, Japan) and also rain water were analysed by the standard additions method. To 10 ml of the water sample, 0-80 pl of 1 mg 1-1 TiIV standard solution were added.The results are summarized in Table 2. The determination of TiIV with the proposed method is sensitive and free from interfer- ences. The technique can be used for the determination of pg 1-l levels of TiIV in the presence of large amounts of foreign ions after pre-concentration. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 References Cheng, K. L., Ueno, K., and Imamura, T., Handbook of Organic Analytical Reagents, CRC Press, Boca Raton, FL, 1982, p. 23. Ishii, H . , Bunseki Kagaku, 1972, 21, 665. Ishii, H., Bunseki Kagaku, 1967, 16, 110. Kajiyama, R., and Yamaguchi. K., Bunseki Kagaku, 1967, 16, 908. Chung, C., Anal. Chirn. Acta, 1983, 154, 259. Tusl, J . , Chem. Listy, 1988, 82, 1303. Inoue, S., Takahashi, T., Hoshi, S . , and Matsubara, M., Bunseki Kagaku, 1988, 37, 316. Marini, H. J., Anton, R. I . , and Olsina, A., Bull. Chem. SOC. Jpn., 1987, 60, 2635. Gregorowicz, Z., Cebura, J., Gorka, P . , and Kowalski, S., Chem. Anal. (Warsaw), 1987, 32, 505. Nickless, G., J. Chrornatogr., 1985, 313, 129. O'Laughlin, J. W., J. Liq. Chrornatogr., 1984, 7, 137. Willeford, B. R., and Veening, H., J. Chromatogr., 1982, 251, 61. Suzuki, N., and Saitoh, K., Kagaku no Ryoiki Zokan, 1983, 138, 127. Yotsuyanagi, T., and Hoshino, H . , Bunseki, 1983, 556. Main, M. V., and Fritz, J. S., Anal. Chern., 1989, 61, 1272. Uehara, N . , Morimoto, K., Shimizu, T., and Shijo, Y., Chern. Lett., 1989, 411. Ogura, N., and Fukushima, K., Bunseki, 1990, 181. Paper OlO33.536 Received July 24th, 1990 Accepted September loth, 1990