PVC Membrane and Coated Graphite ... - ACS Publications

Anal. Chem. 2003, 75, 5680-5686

PVC Membrane and Coated Graphite Potentiometric Sensors Based on Et4todit for Selective Determination of Samarium(III) Mojtaba Shamsipur,*,† Morteza Hosseini,‡ Kamal Alizadeh,‡ Zahra Talebpour,‡ Mir Fazlollah Mousavi,‡ Mohammad Reza Ganjali,§ Massimiliano Arca,| and Vito Lippolis|

Departments of Chemistry, Razi University, Kermanshah, Iran, Tarbiat Modarres University, Tehran, Iran, Tehran University, Tehran, Iran, and Universita` degli Studi di Cagliari, Cagliari, Italy

Solution studies on the binding properties of 4,5,6,7tetrathiocino[1,2-b:3,4-b′]diimidazolyl-1,3,8,10-tetraethyl2,9-dithione (Et4todit) toward a number of cationic species including some lanthanide ions revealed the occurrence of a selective 1:1 complexation of the ligand with Sm3+ ion. Consequently, Et4todit was used as a suitable neutral ionophore for the preparation of novel polymeric membrane (PME) and coated graphite (CGE) Sm3+selective electrodes. The electrodes exhibit a Nernstian behavior for Sm3+ ions over wide concentration ranges (1.0 × 10-5-1.0 × 10-1 M for PME and 1.0 × 10-71.0 × 10-1 M for CGE) and very low limits of detection (8.0 × 10-6 M for PME and 1.6 × 10-8 M for CGE). The proposed potentiometric sensors manifest advantages of relatively fast response, and, most importantly, good selectivities relative to wide variety of other cations, including other lanthanide ions. The selectivity behavior of the proposed Sm3+-selective electrodes revealed a great improvement compared to the best previously reported electrode for samarium(III) ion. The potentiometric responses of the electrodes are independent of the pH of the test solution in the pH range 4.0-6.5. The electrodes were successfully applied to the recovery of Sm3+ ion from tap water samples and also, as an indicator electrode, in potentiometric titration of samarium(III) ions. Despite the exciting advantages of potentiometric methods based on ion-selective electrodes (ISEs), including simple design and operation, wide linear dynamic range, relatively fast response, reasonable selectivity, and low cost,1-3 there have been very limited studies on the ISEs for lanthanide ions.4-10 Thus, the * Corresponding author. E-mail: [email protected]. † Razi University. ‡ Tarbiat Modarres University. § Tehran University. | Universita` degli Studi di Cagliari. (1) Umerzawa, Y., Ed. CRC Handbook of Ion-Selective Electrodes: Selectivity Coefficients; CRC Press: Boca Raton, FL, 1990. (2) Bakker, E.; Bu ¨ hlmann, Pretsch, E. Chem. Rev. 1997, 97, 3083. (3) Bu ¨ hlmann, P.; Pretsch, E.; Bakker, E. Chem. Rev. 1998, 98, 1593. (4) Harrell, J. B.; Jones, A. D.; Choppin, G. R. Anal. Chem. 1969, 41, 1459. (5) Takasaka, Y.; Suzuki, Y. Bull. Chem. Soc. Jpn.1979, 52, 3455. (6) Zhang, Y.; Wu, J.; Wang, E. Electroanalysis 1993, 5, 868. (7) Chowdhury, D. A.; Ogata, T.; Kamata, S. Anal. Chem. 1996, 68, 366. (8) Shamsipur, M.; Yousefi, M.; Ganjali, M. R. Anal. Chem. 2000, 72, 2391.

5680 Analytical Chemistry, Vol. 75, No. 21, November 1, 2003

development of new ion-selective electrodes for lanthanide ions is still a challenging task. It is well known that the sulfur-containing ligands can selectively coordinate different transition and heavy metal ions.11-13 In recent years, we have successfully used some thia-substituted crown ethers as neutral ionophores in construction of ion-selective membrane electrodes for Hg2+,14 Ag+,15 Cd2+,16 and Cu2+ ions.17 In 1998, Masuda et al. reported the successful extraction and separation of landanide ions with a synergistic extraction system consisting of lauric acid and a thiacrown ether (1,4,10,13-tetrathia7,16-diazacyclooctadecane).18 Recently, we have also used 1,3,5trithiane (as ionophore) in conjunction with oleic acid (as additive) for the preparation of a highly selective PVC membrane sensor for Ce3+ and La3+ ions.8,10 In continuation of our work on the use of S-containing neutral ligands in PVC membrane electrode studies, in this paper we employed 4,5,6,7-tetrathiocino[1,2-b:3,4-b′]diimidazolyl-1,3,8,10-tetraethyl-2,9-dithione (Et4todit) as a very suitable neutral ionophore for the preparation of polymeric membrane (PME) and coated graphite electrodes (CGE) for selective and sensitive determination of Sm3+ ion. Et4todit belongs to an interesting class of polyfunctional ligands containing two equivalent thiocarbonyl groups located in positions far apart in the molecule, so that they may act as bridging ligands in the preparation of polynuclear metal complexes.19-23 It has been clearly shown that the resulting complexes of Et4todit with Cu(II)19 and Cu(I)20 are of 1:1 stoichi(9) Amarchand, S.; Menon, S. K.; Agrawal, Y. K. Electroanalysis 2000, 12, 522. (10) Shamsipur, M.; Yousefi, M.; Hosseini, M.; Ganjali, M. R. Anal. Chem. 2002, 74, 5538. (11) Cooper, S. R. Acc. Chem. Res. 1988, 21, 141. (12) Blake, A. J.; Schroder, M. Adv. Inorg. Chem. 1990, 35, 1. (13) Housecroft, C. E. Coord. Chem. Rev. 1992, 92, 141. (14) Fakhari, A. R.; Ganjali, M. R.; Shamsipur, M. Anal. Chem. 1997, 69, 3693. (15) Mashhadizadeh, M. H.; Shamsipur, M. Anal. Chim. Acta 1999, 381, 111. (16) Shamsipur, M.; Mashhadizadeh, M. H. Talanta 2001, 53, 1065. (17) Shamsipur, M.; Javanbakht, M.; Mousavi, M. F.; Ganjali, M. R.; Lippolis, V.; Garau, A.; Tei, L. Talanta 2001, 55, 1047. (18) Masuda, Y.; Zhang, Y.; Yan, C.; Li, B. Talanta 1998, 46, 203. (19) Bigoli, F.; Pellinghelli, M. A.; Deplano, P.; Trogu, E. F.; Sabatini, A.; Vacca, A. Inorg. Chim. Acta 1991, 180, 201. (20) Bigoli, F.; Pellinghelli, M. A.; Deplano, P.; Trogu, E. F.. Inorg. Chim. Acta 1991, 182, 33. (21) Bigoli, F.; Deplano, P.; Devillanova, F. A.; Lippolis, V.; Lukes, P. J.; Mercuri, M. L.; Pellinghelli, M. A.; Trogu, E. F. J. Chem. Soc., Chem. Commun. 1995, 371. (22) Bogoli, F.; Deplano, P.; Mercuri, M. L.; Pellinghelli, M. A.; Sabatini, A.; Trogu, E. F.; Vacca, A. Can. J. Chem. 1995, 73, 380. 10.1021/ac0205659 CCC: $25.00

© 2003 American Chemical Society Published on Web 10/03/2003

ometry with a crystalline structure packing of [Cu(Et4todit)Cl2]n‚ nTHF, in which the coordination around the metal involves two Cl- ions and two S-thioamide atoms of two different ligand molecules related by a screw axis.

EXPERIMENTAL SECTION Reagents. Reagent grade dibutyl phthalate (DBP), benzyl acetate (BA), 2-nitrophenyl octyl ether (NPOE), oleic acid (OA), sodium tetraphenylborate (NaTPB), tetrahydrofuran (THF), and high relative molecular weight PVC were purchased from Merck and used as received. Reagent grade nitrate salts of all cations used (all from Merck) were of the highest purity available and used without any further purification except for vacuum-drying. Reagent grade acetonitrile (Merck) was purified and dried as described elsewhere.24 All other reagents needed were purchased from Merck and used as received. Synthesis of Et4todit.23 A mixture of N,N′-diethylimidazolidine-2-thione-4,5-dione (1.19 g, 6.4 × 10-3 mol) and Lawesson’s reagent (2.60 g, 6.4 × 10-3 mol) in toluene (100 mL) was refluxed for 3 h. The solvent was removed under reduced pressure and the crude product crystallized from dichloromethane (36% yield). Mp: 238 °C. Elemental analysis, Found (calcd for C14H20N4S6): C, 38.57 (38.51); H, 4.83 (4.62); N, 12.82 (12.83); S, 43.50 (44.05). IR bands (KBr pellets, cm-1): 2945 m, 1623 w, 1430 msh, 1399 vs, 1367 ssh, 1328 m, 1259 s, 1145 m, 1092 m, 1047 w, 987 w, 955 w. Preparation of [Sm.Et4todit][NO3]3. A mixture of Et4todit (43.6 mg, 0.1 mmol dissolved in 2 mL of CHCl3) and Sm(NO3)3 (33.6 mg, 0.1 mmol dissolved in 2 mL of acetonitrile) were stirred together for 2 h, after adding 6 mL of methanol. The complex [Sm.Et4todit][NO3]3 was obtained as pale yellow crystals by slow evaporation of the solvent (70% yield). Mp: >400 °C. Elemental analysis, Found (calcd for SmC14H20O9N7S6): Sm, 18.48 (19.43); C, 21.22 (21.76); H, 2.51 (2.59); O, 18.42 (18.65); N, 12.44 (12.69); S, 24.50 (24.87). IR bands (KBr pellets, cm-1): 2945 m, 1618 mbr, 1432 msh, 1397 vs, 1372 vs, 1328 ssh, 1260 s, 1145 m, 1092 m, 1046 w, 979 w, 953 w. Preparation of Electrodes. Membrane solutions were prepared by thoroughly dissolving 3 mg of ionophore Et4todit, 30 mg of powdered PVC, 57 mg of plasticizer BA, and 10 mg of additive OA in 3 mL of freshly distilled THF. The resulting clear mixture was evaporated slowly at ambient temperature until an oily concentrated mixture was obtained. A Pyrex tube (5-mm i.d.) was dipped into the mixture for ∼10 s so that a nontransparent membrane of ∼0.3-mm thickness was formed. The tube was then pulled out from the mixture and kept at room temperature for ∼1 h. The tube was then filled with an internal solution (1.0 × 10-3 M Sm(NO3)3). The electrode was finally conditioned for 24 (23) Aragoni, M. C.; Arca, M.; Demartin, F.; Devillanova, F. A.; Garau, A.; Isaia, F.; Lelj, F.; Lippolis, V.; Verani, G. J. Am. Chem. Soc. 1999, 121, 7098. (24) Greenberg, M. S.; Popov, A. I. Spectrochim. Acta, Part A 1975, 31, 697.

h by soaking in a 1.0 × 10-3 M samarium nitrate solution. A silver/ silver chloride electrode was used as the internal reference electrode. To prepare the coated graphite electrodes, spectroscopic grade graphite rods 10 mm long and 3 mm in diameter were used. A shielded copper wire was glued to one end of the graphite rod, and the electrode was sealed into the end of a PVC tube of about the same diameter with epoxy resin. The working surface of the electrode was polished with fine alumina slurries on a polishing cloth, sonicated in distilled water, and dried in air. The polished graphite electrode was dipped into the membrane solution mentioned above, and the solvent was evaporated. A membrane was formed on the graphite surface, and the electrode was allowed to stabilize overnight. The electrode was finally conditioned soaking in a 1.0 × 10-3 M solution Sm(NO3)3 for 48 h. Emf Measurements. The emf measurements with the polymeric membrane electrodes and the coated graphite electrodes were carried out with the following cell assemblies:

Ag-AgCl || KCl (3 M) | internal solution 1.0 × 10-3 M Sm(NO3)3 | PVC membrane| sample solution || Hg-Hg2Cl2, KCl (satd) (PME) graphite surface | PVC membrane | sample solution || Hg-Hg2Cl2, KCl (satd) (CGE)

The emf observations were made relative to a double-junction saturated calomel electrode (SCE, Philips) with the chamber filled with an ammonium nitrate solution. A double-junction silver/silver chloride electrode (Metrohm) containing a 3 M solution of KCl was used as the internal reference electrode. Activity coefficients were calculated according to the Debye-Huckel procedure, using following equation:25

log γ ) -0.511z2[µ1/2/(1 + 1.5µ1/2 ) - 0.2 µ]

(1)

where µ is the ionic strength and z the valency. Procedures. Conductivity measurements were carried out with a Metrohm 660 conductivity meter. A dip-type conductivity cell, made of platinum black with a cell constant of 0.8310 cm-1, was thermostated at the 25.00 ( 0.05 °C using a Phywe immersion thermostat. In a typical run, 15 mL of a metal nitrate solution in acetonitrile (8.0 × 10-5 M) was placed in a water-jacketed cell equipped with a magnetic stirrer and connected to the thermostat circulating water at the desired temperature. Then, a known amount of the macrocyclic solution was added in a stepwise manner using a calibrated micropipet. The conductance of the solution was measured after each addition. Addition of the ligand was continued until the desired ligand-to-cation mole ratio was achieved. The 1:1 binding of the metal ions (Mn+) with Et4todit (L) can be expressed by the following equilibrium: Kf

Mn+ + L y\z MLn+

(2)

(25) Kamata, S.; Bhale, A.; Fukunaga, Y.; Nurata, A. Anal. Chem. 1988, 60, 2464.

Analytical Chemistry, Vol. 75, No. 21, November 1, 2003

5681

The corresponding equilibrium constant, Kf, is given by

Kf ) ([MLn+]/[Mn+][L]) (γ(Mn+)/γ(Mn+)γ(L))

(3)

where [MLn+], [Mn+], [L], and γ represent the equilibrium molar concentration of complex, free cation, and free ligand and the activity coefficients of the species indicated, respectively. Under the dilute conditions used, the activity coefficient of uncharghed ligand, γ(L), can be reasonably assumed as unity.26-29 The use of the Debye-Huckel limiting law30 leads to the conclusion that γ(Mn+) ∼ γ(MLn+), so the activity coefficients in eq 3 cancel out. Thus, the complex formation constant in terms of molar conductance, Λ, can be expressed as31,32

Kf ) [MLn+]/[Mn+][L] ) (ΛM - Λobs)/(L)

(4)

where

[L] ) CL - (CM(ΛM - Λobs)/(ΛM - ΛML))

(5)

Here, ΛM is the molar conductance of the metal ion Mn+ before addition of ligand, ΛML the molar conductance of the complexed metal ion, Λobs the molar conductance of the solution during titration, CL the analytical concentration of the ligand added, and CM the analytical concentration of the metal nitrate. The complex formation constant, Kf, and the molar conductance of the complex, ΛML, were obtained by computer fitting of eqs 4 and 5 to the molar conductance-mole ratio data using a nonlinear least-squares program KINFIT.33 The stoichiometry of the Sm3+-Etbtodit complex was investigated by the absorbance measurements at three wavelengths of the spectra of solutions in which varying concentrations of the metal ion (1.0 × 10-2 M) were added to a fixed concentration of the ligand (5.0 × 10-5 M) until a desired metal-to-ligand mole ratio was achieved. Attainment of equilibrium was checked by the observation of no further change in the spectra after several hours. The influence of the pH of a test solution on the potential response of the PME was studied as follows. A 25.0-mL portion of a 1.0 × 10-4 M solution of samarium nitrate was taken, and its pH was adjusted by dropwise addition of a 0.1 M solution of either HCl or NaOH, and the emf of the electrode was measured at each pH value, in a pH range of 2.5-10.0. The influence of the concentration of internal solution of the PME was studied as follows. Four similar PMEs were prepared under optimal membrane composition, and each electrode was filled with an internal solution of varying samarium nitrate concentrations of 1.0 × 10-1, 1.0 × 10-2, 1.0 × 10-3, and 1.0 × 10-4 M. The electrodes were then conditioned for 24 h by soaking in a 1.0 × 10-3 M samarium nitrate solution. Finally, the emf (26) Hasani, M.; Shamsipur, M. J. Inclusion Phenom. 1993, 16, 123. (27) Hasani, M.; Shamsipur, M. J. Solution Chem. 1994, 23, 721. (28) Tawarah, K. M.; Mizyed, S. A. J. Solution Chem. 1989, 18, 387. (29) Smetana, A. J.; Popov, A. I. J. Chem. Thermodyn. 1979, 11, 1145. (30) Debye, P.; Huckel, H. Phys. Z. 1928, 24, 305. (31) Takeda, Y. Bull. Chem. Soc. Jpn. 1983, 56, 3600. (32) Zollinger, D. P., Bulten, E.; Christenhuse, A.; Bos, M.; van der Linden, W. E. Anal. Chim Acta 1987, 198, 207. (33) Nicely, V. A.; Dye, J. L. J. Chem. Educ. 1971, 48, 443.

5682

Analytical Chemistry, Vol. 75, No. 21, November 1, 2003

versus pSm3+ plot for each electrode was constructed in a pSm3+ range of 2-5. To determine the selectivity coefficients of different interfering ions for the PME and CGE, specified amounts of samarium ion (A) (in the range of 1.0 × 10-2-1.0 × 10-5 M for PME and 1.0 × 10-3-1.0 × 10-7 M for CGE) were added to a reference solution (1.0 × 10-6 M for PME and 1.0 × 10-8 M for CGE) and the potential was measured. In a separate experiment, the interfering ions (B) (in the range of 1.0 × 10-2-1.0 × 10-3 M for PME and 1.0 × 10-2-1.0 × 10-4 M for CGE) were successively added to an identical reference solution until the measured potential matched that obtained before by adding the samarium ions. The MPM matched potential method selectivity coefficient, KA,B , is then given by the resulting primary ion to interfering ion activity MPM (concentration) ratio, KA,B ) aA/aB. RESULTS AND DISCUSSION Preliminary Studies. A solid-state 1:1 complex [Sm.Et4todit][NO3]3 was prepared and characterized by elemental analysis, as described in the Experimental Section. The molecular structures of the uncomplexed Et4todit and its complex with the lanthanide ion were built with the Hyperchem program. The structure of free ligand was optimized using the 6-31G* basis set at the restricted Hartree-Fock (RHF) level of theory. The optimized structure of the ligand was then used to find out the initial structure of its lanthanide complex. Finally, the structure of the resulting 1:1 complex was optimized using Lanl2mb basis set at the RHF level of theory. No molecular symmetry constraint was applied; rather, full optimization of all band lengths, bond angles, and torsion angles was carried out using the Gaussian 98 program. The optimized structures are shown in Chart 1. As is obvious from Chart 1, in the resulting 1:1 complex, the two CdS groups and two adjacent nitrogen atoms of the two thioimidazole rings of the ligand are nicely involved in bond formation with the central lanthanide ion, while the four sulfur atoms of the upper cycle of the ligand will remain unattached. It is noteworthy that such an involvement of the two S atoms of the thionic group of the ligand molecule in bond formation with the central metal ion has already been observed in the crystalline structures of 1:1 complexes of Cu(I) and Cu(II) with Et4todit.19,20 In preliminary experiments, the complexation of Et4todit with La3+, Cd3+, Ce3+, and Sm3+ ions and, for comparison purposes, with Ca2+, Cu2+, and Cd2+ was studied conductometrically in acetonitrile solution at 25.00 ( 0.04 °C,26-28 in order to obtain a clue about the stoichiometry and stability of the resulting complexes. The resulting molar conductance, Λ, versus [Et4todit]/ [Mn+] mole ratio plots are shown in Figure 1. As seen, in all cases, addition of the ligand to the cation solution caused a continuous increase in the molar conductance, indicating the higher mobility of the complexed cations compared to the solvated ones. The increased molar conductance of the metal nitrates in acetonitrile solution upon addition of the ligand can also be related to some extent to the dissociation of some ion-paired species usually present in acetonitrile, as a solvent of intermediate dielectric constant and solvating ability, as a result of the metal ion complexation with Et4todit.34 The conditional complex formation constant of the resulting 1:1 complexes were evaluated by (34) Amini, M. K.; Shamsipur, M. Inorg. Chim. Acta 1991, 183, 65.

Chart 1. Optimized Structures of Free (A) and Complexed (B) Et4todit

Figure 1. Molar conductance vs [Et4todit]/[Mn+] for different cations in acetonitrile solution at 25 °C. Inset shows the computer fit of the molar conductance-mole ratio data for the resulting samarium complex. Table 1. Formation Constants of 1:1 Complexes of Et4todit with Different Cations in Acetonitrile Solution at 25 °C

computer fitting of the conductance-mole ratio data to appropriate equations, as given in Experimental Section, and the results are summarized in Table 1. A sample computer fit of the molar conductance-mole ratio data for the resulting samarium complex is shown in the inset of Figure 1. It is interesting to note that the formation of Et4todit complexes of 1:1 stoichiometry with copper ions in crystalline state have also been reported in the literature.19,20 The formation of a 1:1 Sm3+-Et4todit complex in acetonitrile solution was further supported by the resulting absorbance-mole ratio plots obtained at three different wavelengths of the ligand spectrum, during its titration with Sm3+ ions (Figure 2). It should be noted that the UV spectrum of Et4todit in acetonitrile possesses two absorbance maximums at 267.0 and 330.3 nm, the molar absorptivity of the first absorption being some 4 times greater than the second one. The absorbance bands are most probably due to the two substituted thioimidazole presented in the ligand’s structure. To obtain further information about the stoichiometry of the Sm3+-Et4todit complex in acetonitrile solution, the circular dichroism (CD) spectra of the ligand (with five asymmetric centers in its molecular structure) in the absence and presence of Sm3+ having varying [Sm3+]/[Et4todit] were obtained using a Jasco J-715 spectropolarimeter, and the results are shown in Figure 3. As is obvious from Figure 3, the well resolved CD spectra of the ligand

cation

log Kf

cation

log Kf

Sm3+

4.60 ( 0.02 3.63 ( 0.03 3.59 ( 0.03 3.51 ( 0.4

Cu2+

3.18 ( 0.05 3.02 ( 0.03 2.80 ( 0.05 2.78 ( 0.03

Gd3+ La3+ Ce3+

Cd2+ Ca2 Tl+

are significantly changed upon addition of the metal ion at metalto-ligand mole ratios of >0 until a molar ratio of 1 is reached. Further addition of Sm3+ resulted in no significant change in the CD spectra of the system, clearly indicating the formation of a 1:1 complex in solution. From the data given in Table 1, it is immediately obvious that Et4todit may act as a selective ionophore for Sm3+ ion in a PVC membrane electrode. Thus, in preliminary experiments it was found that, while the use of an ionophore-free PVC membrane resulted in no measurable response with respect to Sm(III), the addition of Et4todit shows a near-Nernstian response for the cation in the range of 1.0 × 10-1-1.0 × 10-5 M. Membrane Composition. The performance characteristics reported for a given ionophore-incorporated PVC membrane may vary depending on electrode composition and the nature of solution (e.g., ionic strength, pH) to which the electrodes are exposed.18,14-17,35 Thus, several membranes of varying nature and ratios of plasticizer/PVC/ionophore/additive were prepared for the systematic investigation of the membrane composition, and the results are summarized in Table 2. As expected, the amount of ionophore Et4todit was also found to affect the PVC membrane Analytical Chemistry, Vol. 75, No. 21, November 1, 2003

5683

Table 2. Optimization of Membrane Ingredients

Figure 2. Absorbance vs [Sm3+]/[Et4todit] in acetonitrile solution at 25 °C: (A) 267.0, (B) 268.5, and (C) 264.5 nm. The dash lines represent the extrapolation of linear segments of the mole ratio plots.

no.

PVC

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24

30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30

a

composition (%) plasticizer additive DBP, 70 DBP, 69 DBP, 68 DBP, 67 DBP, 66 DBP, 64 DBP, 62 DBP, 57 DBP, 52 DBP, 66 DBP, 65 DBP, 64 BA, 62 BA, 57 BA, 52 BA, 54 BA, 66 BA, 65 BA, 64 NPOE, 62 NPOE, 57 NPOE, 52 NPOE, 65 NPOE, 64

OA, 5 OA, 10 OA, 15 TPB, 1 TPB, 2 TPB, 3 OA, 5 OA, 10 OA, 15 OA, 10 TPB, 1 TPB, 2 TPB, 3 OA, 5 OA, 10 TPB, 1 TPB, 2 TPB, 3

ionophorea

slope (mV decade-1)

1 2 3 4 6 3 (0.39) 3 (0.19) 3 (0.13) 3 (2.4) 3 (1.2) 3 (0.80) 3 (0.39) 3 (0.19) 3 (0.13) 6 (0.39) 3 (2.4) 3 (1.2) 3 (0.8) 3 (0.39) 3 (0.19) 3 (0.13) 3 (1.2) 3 (0.8)

6.2 ( 0.3 12.4 ( 0.2 14.4 ( 0.1 13.8 ( 0.2 13.7 ( 0.3 15.1 ( 0.3 17.2 ( 0.3 17.2 ( 0.3 15.9 ( 0.2 16.8 ( 0.2 16.8 ( 0.3 18.1 ( 0.2 19.6 ( 0.3 19.6 ( 0.3 16.4 ( 0.2 16.1 ( 0.3 16.2 ( 0.3 16.2 ( 0.3 17.8 ( 0.2 19.5 ( 0.2 17.3 ( 0.2 19.6 ( 0.3 18.7 ( 0.3

Values in parentheses are the ionophore-to-additive molar ratios.

sensitivity (nos. 2-6). The calibration slope increased with increasing ionophore content until a value of 3% was reached. However, further addition of Et4todit resulted in a diminished response slope of the electrode, most probably due to some inhomogenities and possible saturation of the membrane.36 As seen in Table 2, among the three different plasticizers used, BA and NPOE were found to be the most effective solvent mediators in preparing the Sm3+ ion-selective electrodes (i.e., nos.

14 and 21, respectively). It is noteworthy that the nature of the plasticizer influences both the dielectric constant of the membrane and the mobility of ionophore and its complex.2,35,37 We have recently shown that a change in the nature of the plasticizer will even result in altering the selectivity of the same membrane system toward Ce3+ and La3+ ions.8,10 In this work, we also found that, despite the proper slope and linear range of the PVC membrane containing NPOE for samarium ion, the use of this plasticizer resulted in increased interfering effects of some other cations of higher charge density such as Cu2+ and Gd3+. However, the use of 57% BA in conjunction with 10% OA resulted in improved response characteristics and selectivity of the resulting samarium ion-selective electrodes. It should be noted that, in both membranes 14 and 21, 10% OA as a polar additive is present which, in turn, can also modify the polarity of the PVC membranes prepared, so that both membranes response well to the concentration of Sm3+ in solution. The data given in Table 2 clearly demonstrate the important role of lipophilic additives in the improvement of the response behavior of the proposed Sm3+-selective membrane sensor. In this table, the carrier-to-additive molar ratio in each membrane is also given in parentheses. As is obvious, the presence of 10% OA or 2% NaTPB as suitable lipophilic additives will improve the sensitivity (and the selectivity) of the resulting Sm3+ sensors. It is well known that the presence of lipophilic anionic sites improves the potentiometric behavior of certain cation-selective electrodes by reducing the ohmic resistance and improving the response behavior and selectivity38-40 and, in some cases, by catalyzing the exchange kinetics at the sample-membrane interface.41 As can

(35) Yang, X.; Kumar, N.; Chi, H.; Hibbert, D. D.; Alexander, P. N. W. Electroanalysis 1997, 9, 549. (36) Ammann, D.; Pretsch, E.; Simon, W.; Lindner, E.; Bezegh, A.; Pungor, E. Anal. Chim. Acta 1985, 171, 119.

(37) Norov, S. K.; Gulamova, M. T.; Zhukov, A. F.; Borova, G. M.; Mamediva, Y. G. Zh. Anal. Khim. 1988, 43, 777. (38) Eugster, R.; Gehrig, T. M.; Morf, W. E.; Spichiger, U. E.; Simon, W. Anal. Chem. 1991, 63, 2285.

Figure 3. CD patterns of Et4todit in acetonitrile solution in the presence of varying amounts of Sm3+. [Sm3+]/[Et4todit]: (A) 0.00, (B) 0.27, (C) 0.70, (D) 1.00, (E) 1.21, and (F) 1.70.

5684 Analytical Chemistry, Vol. 75, No. 21, November 1, 2003

Table 3. Response Characteristics of the Sm3+ Ion-Selective Electrodes slope (mV electrode decade-1) PME CGE

Figure 4. Calibration graphs for the PME and CGE at pH 5.0. The PVC membrane composition used is that given for membrane 14, Table 2.

be seen in Table 2, membrane 14 with a PVC/BA/Et4todit/OA percent ratio of 30:57:3:10 resulted in Nernstian behavior of the membrane electrode over a very wide concentration range. Effect of Internal Solution. In accord with generally adopted ISE response formalism,2 the internal solution may affect the ISE response when the membrane internal diffusion potential is appreciable. Thus, the influence of the concentration of the internal solution on the potential response of the polymeric membrane electrode was also studied. The Sm(NO3)3 concentration was changed from 1.0 × 10-4 to 1.0 × 10-1 M, and the emf-pSm3+ plots were obtained. It was found that the concentration of the internal solution does not affect the potential response of the electrode, except for an expected change in the intercept of the resulting plots. A 1.0 × 10-3 M concentration of the reference solution is quite appropriate for a smooth Nernstian function of the polymeric membrane system. It should be noted that a change in concentration of the internal solution is expected to change the potential difference on the internal solution/membrane interface and, consequently, the overall emf of the cell assembly. Response Characteristics of PME and CGE. The critical response characteristics of the proposed PME and CGE were assessed according to IUPAC recommendations.42 The emf response of the polymeric membrane and coated graphite electrodes (Figure 4) indicates their Nernstian behavior over a wide concentration range. The slopes and linear ranges of the resulting emf-pSm3+ graphs are given in Table 3. The limits of detection, defined as the concentration of samarium ion obtained when the linear regions of the calibration graphs are extrapolated to the baseline potentials, are also included in Table 3. The improved performance characteristics of the CGE over those of the PME presumably originate from coated graphite (39) Rosatzin, T.; Bakker, E.; Suzuki, K.; Simon, W. Anal. Chim. Acta 1993, 280, 197. (40) Schaller, U.; Bakker, E.; Spichiger, U. E.; Pretsch, E. Anal. Chem. 1994, 66, 391. (41) Gehrig, P. M.; Morf, W. E.; Welti, M.; Prersch, E.; Simon, W. Helv. Chim. Acta 1990, 73, 203. (42) IUPAC Analytical Chemistry Division. Commission on Analytical Nomenclature. Recommendations for Namenclature of Ion-Selective Electrodes. Pure Appl. Chem. 1976, 48, 127.

linear range (M)

limit of detection (M)

response time (s)

19.6 ( 0.5 1.0 × 10-5-1.0 × 10-1 8.0 × 10-6 19.3 ( 0.3 1.0 × 10-7-1.0 × 10-1 1.6 × 10-8

20 18

technology, where an internal 1.0 × 10-3 M Sm(NO3)3 solution, in the case of PME, has been replaced by a copper wire, in the case of CGE. It is well known that the higher limit of detection of PME compared to that of the CGE is mainly due to some leakage of internal solution into the test solution via the polymeric membrane.2 Meanwhile, the much higher electrical conductivity of copper wire (in CGE) than that of the internal solution (in PME) is expected to result in lower response time of the CGE in comparison with the PME. It is noteworthy that, in the case of CGE, the potential is not expected to be stabilized at the graphite/ membrane interface if only the exchange of Sm3+ is considered. However, samarium is known to exist in two valences Sm3+/Sm2+ that may form a suitable determining couple and stabilize the potential at this interface. This was verified by obtaining a quasireversible cyclic voltammogram for a 1.0 × 10-3 M solution of Sm(NO3)3 at the surface of a bare graphite electrode. The average time required for the electrodes to reach a potential response within (1 mV of the final equilibrium value after successive immersion in a series of Sm3+ solutions, each having a 10-fold difference in concentration, was investigated. The resulting response times of the electrodes for concentrations of e1.0 × 10-3 M are also included in Table 3. A comparison of the data given in Table 3 revealed that, while both electrodes show a Nernstian behavior with relatively fast response, the linear range and limit of detection of the coated graphite electrode are greatly improved relative to those of the polymeric membrane electrode. Furthermore, the membrane electrodes prepared could be used for at least 2 months without any measurable divergence, the lifetime of the CGE being longer than that of the PME. After this period of time, the calibration slope of the proposed electrodes was found to diminish by at most 5%, while the linear range remained almost unchanged. The time of contact and concentration of equilibrating solution were optimized so that the sensors generated stable and reproducible potentials at relatively short response times. The optimum equilibration time in a 1.0 × 10-3 M Sm(NO3)3 for the PME and CGE was 24 and 48 h, respectively. Effect of pH of the Test Solution. The influence of pH of the test solution on the potential response of the membrane sensor was tested in the pH range 3-10, and the results are shown in Figure 5. As seen, the potential remained constant from pH 4.0 to 6.5, beyond which the potential changed considerably. The observed large decrease in potential at higher pH values could be due to the formation of some hydroxyl complexes of Sm3+ in solution. At low pH, the potentials increased, indicating that the membrane sensor responds to hydrogen ions. Selectivity Coefficients. Potentiometric selectivity coefficients, describing the preference of the ion-selective electrodes for an interfering ion, B, relative to Sm3+ ion, A, were determined by the matched potential method, which is recommended43 to Analytical Chemistry, Vol. 75, No. 21, November 1, 2003

5685

Figure 5. Effect of pH of test solution on the potential response of the PME at a Sm3+ concentration of 1.0 × 10-4 M. The PVC membrane composition used is that given for membrane 14, Table 2.

Figure 6. Potentiometric titration curve of 15.0 mL of 1.0 × 10-4 M solution of Sm3+ with 1.0 × 10-2 M EDTA at pH 6.5. Table 5. Recovery of Sm3+ Ions from Tap Water Samples by the Proposed PME and CGE

Table 4. Selectivity Coefficients of Various Interfering Ionsa KMPM A,B

KMPM A,B

Mn+

PME

CGE

Mn+

PME

CGE

Ce3+ La3+ Gd3+ Cu2+ Pb2+ Zn2+ Cd2+ Ni2+ Co2+

8.1 × 10-2 5.4 × 10-2 8.6 × 10-2 3.1 × 10-3 9.2 × 10-2 7.3 × 10-3 1.9 × 10-3 4.2 × 10-3 5.2 × 10-3

7.6 × 10-4 6.2 × 10-4 7.2 × 10-4 4.2 × 10-4 6.1 × 10-3 1.2 × 10-4 2.1 × 10-4 4.0 × 10-4 5.0 × 10-4

Hg2+ Ba2+ Ca2+ Mg2+ Ag+ Tl+ K+ Na+ Li+

8.9 × 10-2 2.2 × 10-3 1.2 × 10-3 1.9 × 10-3 3.9 × 10-3 6.1 × 10-4 4.1 × 10-4 3.2 × 10-4 4.1 × 10-4

6.2 × 10-3 3.4 × 10-4 3.8 × 10-4 2.6 × 10-4 4.2 × 10-4 5.2 × 10-5 2.7 × 10-5 8.2 × 10-5 1.5 × 10-5

a Conditions: reference solution, 1.0 × 10-6 M for PME and 1.0 × 10-8 M for CGE; Sm3+, 1.0 × 10-2-1.0 × 10-5 M for PME and 1.0 × 10-3-1.0 × 10-7 M for CGE; Mn+, 1.0 × 10-2-1.0 × 10-3 M for PME and 1.0 × 10-2-1.0 × 10-4 M for CGE.

overcome the difficulties associated with the methods based on the Nicolsky-Eisenman equation.43,44 According to this method, the specified activity (concentration) of the primary ion (A) is added to a reference solution and the potential is measured. In a separate experiment, the interfering ions (B) are successively added to an identical reference solution until the measured potential matched that obtained before by adding the primary ions. MPM The matched potential method selectivity coefficient, KA,B , is then given by the resulting primary ion to interfering ion activity MPM (concentration) ratio, KA,B ) aA/aB. The resulting values for both samarium(III) ion-selective electrodes are listed in Table 4. From Table 4, it is immediately obvious that the proposed samarium(III) ion-selective electrodes are highly selective with respect to other common cations, including members of the lanthanide family other than Sm3+. The selectivity coefficients are in the order of 10-2 (for PME) and 10-3 (for CGE) or lower, which (43) Umezawa, Y.; Umezawa, K.; Sato, H. Pure Appl. Chem. 1995, 67, 507. (44) Bakker, E. Electroanalysis 1997, 9, 7.

5686 Analytical Chemistry, Vol. 75, No. 21, November 1, 2003

Sm3+ (M) electrode

added

found

% recovery

PME

5.0 × 10-3 6.6 × 10-4 5.0 × 10-3 6.6 × 10-5

(5.1 ( 0.2) ×10-3 (6.6 ( 0.2) × 10-4 (4.9 ( 0.2) × 10-3 (6.7 ( 0.3) × 10-5

102.0 100.0 98.0 101.5

CGE

seem to indicate that these cations have negligible impact on the functionality of the Sm3+ sensors. Meanwhile, the data given in Table 4 revealed that, in all cases, the selectivity coefficients obtained for the coated graphite electrode are lower than those for the polymeric membrane electrode, emphasizing the superiority of the CGE in this respect as well. It is interesting to note that a comparison of the selectivity coefficients obtained with the proposed PME and, especially, CGE with those reported before by Chowdhury et al.7 for their proposed coated graphite Sm3+ ionselective electrodes based on bis(thialkylxanthato) alkenes (which are actually the most selective samarium sensors available) clearly indicated a large enhancement in the selectivity behavior of the proposed electrodes for Sm3+ ion. Applications. The proposed membrane sensors were found to work well under laboratory conditions. The CGE was used as an indicator electrode in the successful titration of a Sm3+ ion solution (1.0 × 10-4 M) with EDTA (1.0 × 10-2 M) at pH 6.5. The resulting titration curve is shown in Figure 6, indicating that the amount of Sm3+ can be accurately determined with the electrode. Both the PME and CGE were also used for the recovery of Sm3+ ion from tap water samples, and the results are summarized in Table 5. As is obvious, in the case of both electrodes, quantitative recovery of Sm3+ ions from tap water is achievable. Received for review September 11, 2002. Accepted August 14, 2003. AC0205659