ANALYST, APRIL 1989, VOL. 113 435 Synthetic Inorganic Ion-exchange Materials Part XLIX.* Adsorption and Desorption Behaviour of Heavy Metal Ions on Hydrated Titanium Dioxide Mitsuo Abe, Peng Wang,t Ramesh Chitrakar and Masamichi Tsuji Department of Chemistry, Faculty of Science, Tokyo Institute of Technolog y, 2- 12- I , Ooka yama, Meguro-Ku, Tokyo 152, Japan The ion-exchange selectivity of a number of divalent metal ions was studied as a function of pH in nitrate and chloride media. The order of selectivity was Pb2+ > Hg2+ > Cd2+ > Mg2+ in nitrate solution and Pb2+ > Cd2+ > Ca2+ > Mg2+ > Hg2+ in chloride solution. A good linear relationship between the logarithm of the distribution coefficient of the heavy metal ions and their effective ionic radius was found.On the basis of the Kd values, the separation of a mixture of Cd2+, Hg2+ and Pb2+ and the group separation of Hg2+ and Pb2+ from several common metal ions were achieved on an ion-exchange column containing amorphous hydrated titanium dioxide. Keywords ; Hydrated titanium dioxide; ion exchange; chromatographic separation; transition metals; lead separation Many useful inorganic ion-exchange materials have been synthesised during the last two decades and many have found application in the fields of analytical chemistry, radio- chemistry, environmental chemistry and biochemistry. 1 Inor- ganic ion exchangers possessing high selectivities for certain ions or groups of ions*-' can be utilised for the chromato- graphic separation of many elements. Hydrated titanium &oxide (HTDO) has attracted considerable attention as an ion exchanger owing to its high sorption selectivity for certain metal ions.Many reports have described the basic characteris- tics and sorption properties of this material.&-11 Three types of HTDO are known: amorphous, anatase and rutile."Il Several new types of HTDO, e.g.. crystalline HTDO fibres"-13 and crystalline HTDO with a layered structure, have been prepared recently. 14-16 The applications of HTDO have been mainly concerned with the recovery of uranium from sea- water."-1') Little effort has been devoted to the chromato- graphic separation of heavy metal cations, except for the separation of 1 % ~ - YSr and 1 W s - '"Pr 20 and chromate and phosphate ions.21 More recently, separations of Ni2+ - Co2+ and Zn*+ - Cd'+ - Cu2+,11 Cu*+ from Mg2+ - Ca2+ - Sr2+, Cu2+ from Ni2f - Co*+, Mg*+ from Ba2+ - Zn2+ 22 and Cu2+ from Zn?+ - Mn2+ - Co2+ - Ni2+ 23 have been achieved with an HTDO column.Heavy metal ions such as Hg2+ and Pb*+ are hazardous and are frequently discharged in industrial wastewater; much attention has been directed towards studying their effect as environmental pollutants. Hence there is a need to develop a method to separate these biologically toxic elements from common metal ions also present in contaminated natural w a t e r s . This paper describes the adsorption and desorption behav- iour of a number of heavy metal ions on amorphous HTDO (Am-HTDO) at different pH values and in different media. Their selectivities and the applicability of some chromato- graphic separations of these heavy metal ions are also discussed.Part XLVIII of this series has been submitted for publication in i Present address: Harbin Institute of Technology, Department of Solveizt Extr. Ion Exch. Applied Chemistry, Harbin, China. Experiment a1 Reagents The reagents used were supplied by Wako Pure Chemical Industries (Japan). Titanium tetrachloride was of the highest available grade (>99% as metal); all other chemicals were of analytical-reagent grade. Standard solutions of the metal ions were prepared by dissolving the high-purity metals (>99.9%) in the minimum amount of nitric acid. A solution of Hg2+ was prepared by dissolving mercury(I1) nitrate in 5% (VIV) nitric acid in order to prevent hydrolysis of the Hg*+ ion.Preparation of the Am-HTDO Ion Exchanger This was prepared as described previously. I1.Z3 Water (150 ml) was added slowly to TiCI4 (SO ml) and a 2 . 8 ~ sodium hydroxide solution (800 ml) was added to the resulting solution. The precipitate formed was washed with de- mineralised water, left for 2 d and filtered under suction. This procedure was repeated until the pH of the filtrate reached a constant value (about pH 12). The precipitate was then air-dried at room temperature until it had been transformed into a semi-transparent glassy gel. This gel was immersed in dG-mineralised water to break it down into fine particles. The product was ground and sieved to 100-200 mesh size. In order to obtain the hydrogen ion form, the sample was conditioned as follows.About 10 g of the sample were placed in a column (10.0 X 1.0 cm i.d.) and 0.1 M hydrochloric acid was passed through the column until the amount of Na+ ion in the effluent was negligible ("a+] < 1 0 - j ~ ) . The sample was then washed with de-mineralised water until the pW of the supernatant solution was in the range 4-S. The resulting product was air-dried to a constant mass. Apparatus A Varian-Techtron 1100 atomic absorption spectrometer was used for the determination of the metal ion content in the aqueous phase. A Model HM-5B TOA pH meter was used for all pH measurements. X-ray powder diffraction patterns were recorded using a JEOL Model JDX-7E X-ray diffractometer with Ni-filtered Cu K a radiation. A Rigaku Denki Model 8001 Thermoflex was used for the thermal analysis studies at a heating rate of 10°C min-1 by employing a-A1203 as the reference material.Infrared spectra were measured by the436 ANALYST, APRIL 1989. VOL.. 113 KBr disc method with a JASCO Model DS-701G infrared spectrometer. Equilibrium Studies The equilibrations were carried out batchwise. A 0.10-g amount of Am-HTDO in the hydrogen ion form was treated with 10.0 ml of a 1.0 x 1 0 - 4 ~ solution of the metal ions at various pH values with a constant ionic strength of 0.1. The mixtures were shaken intermittently at 30 i 0.5"C. After equilibrium had been attained, the pH and metal ion content of the supernatant solutions were determined. The distribu- tion coefficient ( K d ) was calculated from the following equation: amount of metal ion in exchanger amount of metal ion in solution Kd = X volume of solution (ml) mass of exchanger (g) The concentrations of the metal ions in the exchanger were calculated from the differences between the initial and final concentrations in the aqueous phase.The separation factor, a(AIB), was determined from the equation where &A and Kc,B are the distribution coefficients of ions A and B, respectively. Chromatographic Separation Separations of mixtures of metal ions were carried out on a column ( 5 X 0.5 cm i.d.) of Am-HTDO in the hydrogen ion form. A mixed solution containing 1 pmol of each metal ion was loaded on the top of the column and then eluted with different eluents. The eluents were charged continuously with a high-pressure pump (Nihon Seiinitsu Kagaku, Model NP-DX-2).The effluents were collected by using a drop- counting type fraction collector (Ohtake Works, Model UM-160). The effluent fractions were analysed for their metal ion concentration and the recoveries of the metal ions were calculated from the difference between the initial and equilib- rium concentrations. Results and Discussion Characterisation of Am-HTDO Amorphous HTDO is a semi-transparent glassy gel. The samples were characterised both before and after conditioning with 0.1 M I1C1 by X-ray diffraction, infrared spectra and thermal analysis. The results showed good agreement with those reported previously.11 25 The composition of the gel in the hydrogen ion form can be written as TiO?. 1 .6H20. Rate of Metal Ion Sorption The rate of exchange of different metal ions was measured qualitatively in order to determine the equilibrium di\tribu- tion coefficients ( K d values).Fig. 1 shows the rate of sorption of the metal ions on Am-HTDO from their nitrate or chloride solutions. The ion-exchange reactions of these metal ions are relatively fast. An exchange equilibrium appears to be reached within 4 d; about 4-7 d are required to determine the equilibrium KCi valueu. The K,, values were therefore deter- mined by equilibrating the solid with the metal ion solution for 7 d. 100 ( a ) Final pH I 0 2 4 6 PH 3.20 K+ 5.10 -x Cd2+ Hg2+ 1.75 20 Fig. 2. Values of K,, for metal ions in nitrate media. Exchanger, 0.10 g; total volume, 10.0 ml; concentration of metal ion, 1 x 10 A M ; temperature. 30 IL 0.5 "C; and ionic strength.0.1 ki (NaNO, + FINO, or NaOH) 104 I 1 I 2 4 6 8 I I I 0 2 4 6 8 0 ' ' Timeid Fig. 1. Rate o f metal ion sorption. Exchanger, 0.25 g; total volume, 25.0 ml; concentration of metal ion, 1 x 10-4 M ; ionic strength, 0.1 M ; and temperature, 30 IL 0.5 "C. (a) Nitrate solution (NaNO? + HNO, or NaOH); and ( b ) chloride solution (NaCI + HCI or NaOH) I I I U" 0 2 4 6 8 PH Fig. 3. 0.1 M (NaC1 + HCI or NaOH). Other conditions as in Fig. 2 Values of K,, for metal ions in chloride media. Ionic strength,ANALYST. APRIL 1089, VOL. 114 437 Table 1. Values of the distribution coefficient (K,/ml 8-1) and separation factor (a) on Am-HTDO and Bio-Rad AG 50W-X8’7-3 Ions Ion exchanger Solution Am-HTDO . . . . 0.IMNaCl (pH 5 ) Am-HTDO . . . . 0.1 M N ~ N O ~ (PH 3.5) Bio-Rad AG50W-X8 .. . . O.2MHCI Bio-Rad AGSOW-X8 . . . . 0 . 2 ~ H N 0 3 Hg2- 33 2.42 Mg’+ 8 1.8 Hg’ 0.9 68.9 ME’+ 295 1.3 Mg’+ 80 Ca’ t 15 5.38 10.7 PW 1 62 1.35 Cd’+ 392 1.2 Ca7+ 430 4.65 Cd’+ 160 43.8 Cd2 + 84 6.3 Ca’ 480 2.3 Cdz+ 2000 >5 Hg?+ 7000 >1.5 Mg2- 530 1.5 Hg2 1 1090 1.3 PW+ > 104 Pb’ + > 103 Ca‘ + 790 Pb’ + 1420 103 r I 0 - E , Ic“ 102 I 0 1 2 3 4 PH Fig. 4. Conditions as in Fig. 2. 0, NO,-; A , CI-; and 0. Br- Values of Kd for Pb’+ on Am-HTDO in different media. Hg2+ - Mn2+ I I 0 0.5 1.0 1.5 EWA Fig. 5 . Relationshi between log Kd and effective ionic radius (EIR) of heavy metal ions ?correlation coefficient, 0.9964). Conditions as in Fig. 2 Distribution Coefficients (Kd Values) A graph of the logarithm of the Kd values of the metal ions against the final pH values (Figs.2 and 3) shows a linear relationship with a slope of about unity (1 J-1.3) for divalent ions and 0.33 for the K+ ion for all the systems studied at an initial concentration cf metal ion of 1 X 10-4 hi. The Kd value for the K+ ion was included as being representative of those alkali metal ions that are usually present in actual water samples. The slopes do not correspond to the valency of the ions exchanged. Similar results were obtained for the adsorp- tion properties of metal ions on amphoteric inorganic ion exchangers. 11 This might be due partly to the uptake of anions (e.g., Cl-) because of the amphoteric behaviour of Am- HTDO and partly because the amount of dissociated H+ responsible for ion exchange was not constant because of the weakly acidic nature of Am-HTD0.l6 For Am-HTDO, the following order of selectivity was observed: Pb*+ > Hg2+ >> Cd*+ > Ca*+ > Mg*+.Of these ions Pb2+ and Hg*+ exhibit very high Kd values in nitrate media (Fig. 2), whereas the selectivity order was Pb2+ >> Cd2+ > Ca2+ > Mg2+ > Hg2+ in chloride media (Fig. 3). In the latter instance a very low Kd value for Hg2+ was observed due to the formation of a chloride complex. The large difference in the values of Kd for Hg2+ in the two media should make it possible to separate Hg2+ from other matrix ions. Fig. 4 shows the Kc1 values for Pb*+ on Am-HTDO in different media. Lower Kd values were found in bromide solution, which may be due to the formation of a bromide complex. The Kd values and separation factors, a(A/B), are sum- marked in Table 1.the values obtained on Bio-Rad AG 50W-X827-29 being included for comparison. Selectivity for Heavy Metal Ions The selectivity order of Am-HTDO for heavy metal ions was found to be Pb*+ > Hg*+ > Cd*+ (Fig. 2). The log Kd values at pH 3.15 for the heavy metal ions correlated very closely to their effective ionic radius (EIR) (correlation coefficient, 0.9964) as shown in Fig. 5 . The EIR values used were those reported by Shannon.”) A similar selectivity order has been reported previously. 11 This selectivity order might be explained as follows. For the heavy metal ions, the observed order of selectivity was the same as that of their increasing ionic radii, i.e., their decreasing hydrated ionic radii.This suggests that the energy required for the dehydration of the metal ions so that they can occupy a site in the exchanger plays an important role in determining the selectivity series for the transition metal ions.22 On the other hand, according to the principle of hard and soft acids and bases (the HSAB principle), hard acids prefer to bind to hard bases and soft acids to soft bases.31 The hydrated titanium dioxide (HTDO) can act as a Lewis base and the heavy metal ions as Lewis acids. The softness of the cations increases as the ionic radius of the heavy metal ions increases. The interaction of transition metal ions with438 ANALYST, APRIL 1989, VOL. 114 E + 16 20 24 28 32 36 Fraction No. Fig. 6. Chromatographic separation of heavy metal ions.Exchanger, 1.25 g; column, 5 x 0.5 cm i.d.; loading, 1 pmol of each metal ion; and flow-rate, 0.15 ml min-1. Volume of each fraction, 7 ml r- 0 . 5 ~ HBr 16 20 24 28 32 36 * 0.002 M HNO, Fraction No. Fig. 7. common metal ions. Conditions as in Fig. 6 Chromatographic separation of Hg2+ and Pb2+ from some HTDO, acting as a soft base, could be expected to increase with an increase in EIK. Ion-exchange Chromatographic Separation It is evident from studies of the Kd values on Am-HTDO that some selective separations should be feasible for these heavy metal ions. Hence a number of separations were attempted on a column of this material by using various eluents. Separation of cadmium, mercury and lead On the basis of the distribution coefficients, the separation of a mixture of Cd*+, Hg2+ and Pb2+ was attempted; their elution curves are shown in Fig.6. A 0.01 M nitric acid medium was used as the eluent for Cd2+, 0.01 M hydrochloric acid for Hg’& and 0 . 5 ~ hydrobromic acid for Pb2+. Complete separation was achieved with recoveries of 100,98, 98.8% for Cd2+, Hg2+ and Pb2+, respectively. Separation of mercury and lead f r o m common metal ions This was performed with 0 . 0 0 2 ~ nitric acid for Na+, K+, Mg2+ and Ca2+, 0.01 M hydrochloric acid for Hg2+ and 0.5 M hydrobromic acid for Pb2+ (Fig. 7). Recovery of the common metal ions was quantitative. The yields of Hg2f and Pb2+ were quantitative (99’/0) with complete separation. The excellent separation achieved on Am-HTDO can be applied to the separation of highly toxic metal ions, e.g., Hg2+ and Pbzf, in drinking water. In conclusion, Am-HTDO is useful for the concentration and chromatographic separation of highly toxic heavy metal ions such as Hg2+ and Pb2+ from common metal ions and can be applied to the determination of very low concentrations of Hg2+ and Pb2+ in natural water.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. References Clearfield, A . , Editor, “Inorganic Ion Exchange Materials,” CRC Press, Boca Raton, FL, 1982. Vesely, V., and PakBrek, V., Talanra, 1972, 19, 219. Abe, M., and Ito, T., Bull. Chem. SOC. Jpn., 1967, 40, 1013. Abe, M., and Uno, K., Sep. Sci. Technol., 1979, 14, 355. Abe, M., and Ito, T., Kagaku Zasshi, 1967, 70, 440. Abe, M., and Ito, T., Nippon Kagaku Zasshi, 1965, 86, 814.Amphlett, C. B., McDonald, L. A . , and Redman, M., J . Inorg. Nucl. Chem., 1958, 6 , 236. Babyrenko, Yu. Ya., Domaltov, Yu. D., and Bragina, M. I . , Zh. Prikf. Khim. (Leningrad), 1970, 43, 1152. Heitner-Wirguin, C . , and Albu-Yaron, A., J . Inorg. Nucl. Chem., 1966, 28, 2379. Levi, H. W . , and Schiewer, E., Radiochim. Acta, 1966, 5 , 126. Abe, M., Tsuji, M., Qureshi, S. P., and Uchikoshi, H., Chromatographia, 1980, 13, 626. Fujiki, Y., Komatsu, Y . , Sasaki, T., and Ohta, N., Nippon Kagaku Kaishi, 1981, 1656. Sasaki, T., Kornatsu, Y., and Fujiki, Y., Chem. Lett., 1981, 957. Marchand, R . , Brohan, L., and Tournoux, M., Mater. Res. Bull., 1980, 15, 1129. Ohta, N., and Fujiki, Y., Yogyo Kyokai Shi, 1980, 88, 1. Izawa, H . , Kikkawa, S . , and Koizumi, M., J . Phys. Chem., 1982, 86, 5023. Ogata, N., Nippon Kaisui Gakkai-Shi, 1971, 24, 197. Keen, N. J., J. Br. Nucl. Energy SOC., 1968, 7, 178. Yamashita, H., Ozawa, Y., Nakajirna, F., and Murata, T.. Nippon Kagaku Kaishi, 1978, 1057. Lavrukhina, A. K., Malyshev, V. V., and Rodin, S . S., Zh. Vses. Khim. Ova., 1963, 8, 227. Kraus, K. A., Phillips, H. O., Carlson, T. A., and Johnson, J. S . , “Proceedings of the 2nd International Conference on the Peaceful Uses of Atomic Energy,” IAEA, Geneva, 1958, Volume 28, p. 3. Sasaki, T . , Komatsu. Y., and Fujiki, T . , Solvent Extr. Zon Exch., 1983, 1, 775. Sasaki, T . , Kornatsu, Y., and Fujiki, T., Sep. Sci. Technol.. 1983, 18, 49. Inoue, Y., and Tsuji, M., Bull. Chem. SOC. Jpn., 1976.49. 11 1. Tsuji, M., and Abe, M., J . Radioanal. Nucl. Chem., Articles, 1986, 102, 283. Inoue, Y., andTsuji, M., Bull. Chem. SOC. Jpn., 1978,51,479. Strelow, F. W. E., Rethemeyer, R., and Bothma, C. J. C., Anal. Chem., 1965, 37, 106. Strelow. F. W. E., Rethemeyer, R . , and Bothma, C. J. C., Anal. Chem., 1960, 32, 1185. Strelow, F. W. E . , Rethemeyer, R . , and Bothma, C. J . C . , Anal. Chem., 1971,43, 7. Shannon, R . D., Acta Crystallogr., Sect. A , 1976, 32, 751. Pearson, R. G., J. Chem. Educ., 1968, 45, 581 and 643. Paper 8104367A Received November 2nd, I988 Accepted December 14th, I988