Thiazole-Containing Benzo-Crown Ethers: A New Class of Ammonium

1−3 have the same thiazole-containing dibenzocrown ether (TDB18C6) frame except that 2 ..... Ghauri, M. S.; Thomas, J. D. R. Analyst 1994, 119, 2323...
0 downloads 0 Views 151KB Size
Anal. Chem. 2000, 72, 4683-4688

Thiazole-Containing Benzo-Crown Ethers: A New Class of Ammonium-Selective Ionophores Hong-Seok Kim,*,† Hyun Joo Park,‡ Hyun Joon Oh,‡ Young Kook Koh,† Jun-Hyeak Choi,† Dong-Hoon Lee,† Geun Sig Cha,‡ and Hakhyun Nam*.‡

Chemical Sensor Research Group, Department of Chemistry, Kwangwoon University, Seoul 139-701, Korea, and Department of Industrial Chemistry, Kyungpook National University, Taegu 702-701, Korea

A new class of crown ethers containing heterocyclic units has been designed to develop synthetic NH4+-selective ionophores whose resulting potentiometric properties in solvent polymeric membranes are superior or at least comparable to those of nonactin. It was found that the derivatives of thiazole containing dibenzo-18-crown-6 (1; TDB18C6) incorporated in PVC-based membranes provide enhanced NH4+ selectivities over alkali metal cations, especially over Na+, compared to those of the nonactin-doped membranes: for example, the selectivity POT (J ) Li+, Na+, K+), for the 2-nicoefficients, log KNH + 4 ,J trophenyloctyl ether (NPOE) plasticized PVC membranes doped with the hexyl chain-substituted TDB18C6 (2) were -4.9, -3.9, and -1.3, while those for the same type of membranes with nonactin -4.4, -3.0, and -1.0, respectively. Unlike other synthetic NH4+-selective neutral carriers reported to date, TDB18C6-type compounds result in potentiometric performance highly comparable to that of the nonactin-based ones except their slightly higher detection limits (∼3 × 10-6 vs 7 × 10-7). The complex formation between TDB18C6 and NH4+ was identified from 1H NMR spectra. The 1:1 complex formation constant of TDB18C6 with NH4+ in solvent polymeric membranes estimated using the method suggested by Bakker et al. and the NH4+ selectivity coefficient over K+ POT (KNH + +) are ∼200 times smaller than that of nonactin 4 ,K (1.8 × 107 vs 3.6 × 109). NH4+-selective electrodes based on a solvent polymeric membrane doped with nonactin are routinely used for the direct determination of ammonia in most aqueous samples.1 They also find wide applications in biosensor construction for the substrates (e.g., urea, creatinine, L-phenylalanine, and adenosine) that produce NH4+ as a detectable species through the enzymecontaining layer overcoated on them.2 Despite their versatile analytical utility,3-8 however, the applicability of the nonactin-based †

Kyungpook National University. Kwangwoon University. (1) Standard Methods for the Examination of Water and Wastewater, 19th ed.; Eaton, A., Clesceri, L. S., Greenberg, A. E., Eds.; American Public Health Association, American Water Works Association, Water Environment Federation: Alexandra, VA, 1995: Part 4000. (2) Hall, E. A. H. Biosensors; Open University Press: Buckingham, U.K., 1990: Chapter 9. (3) Fraticelli, Y. M.; Meyerhoff, M. E. Anal. Chem. 1981, 53, 992-997. ‡

10.1021/ac000177b CCC: $19.00 Published on Web 08/23/2000

© 2000 American Chemical Society

polymer membranes have been severely limited if the sample contains high concentration of K+ or Na+. It is because the NH4+ selectivities of the nonactin-based polymer membranes over K+ POT -1) and over Na+ (KPOT -3 (KNH + + ∼ 1.0 × 10 NH4+,Na+ ∼ 1.0 × 10 ) 4 ,K + are not high enough to determine the NH4 concentrations in salty samples (e.g., physiological fluids and seawater).9-14 To overcome such limitations, there have been some attempts to develop synthetic neutral carriers that are highly selective for binding NH4+ over K+ and Na+. For example, the Suzuki group reported that the PVC membranes plasticized with glycol dibenzyl POT ethers exhibit enhanced NH4+ selectivity over K+ (KNH + + ) 4 ,K -2 2.0 × 10 ) compared to that of the nonactin-based membranes.15 However, other important potentiometric properties (e.g., response slope, detection limit, lifetime, and stability) of glycol dibenzyl ether-based membranes were not evaluated in their work. Moriuchi-Kawakami et al. synthesized a series of pyrazol-containing crown ethers and evaluated their potentiometric properties with the PVC-based membranes;16 it was reported that the membranes formulated with dibenzyl ether as a plasticizer and sodium tetrakis[3,5-bis(trifluoromethyl)phenyl]borate (NaTFPB) as an additional salt exhibit slightly improved NH4+ selectivity over POT -2), but not over Na+ (KPOT K+ (KNH + + ) 6.3 × 10 NH4+,Na+ ) 1.3 × 4 ,K 10-2). On the other hand, the same ionophore-doped membranes with 2-nitrophenyl octyl ether (NPOE) plasticizer responded more selectively to K+, indicating it is not genuinely selective toward NH4+ without the aid of solvent. Recently, Kim et al. introduced a rationally designed NH4+-selective receptor, 1,3,5-tris(3,5-dimethylpyrazol-1-ylmethyl)-2,4,6-triethylbenzene, in which the tripodal (4) Pranitis, D. M.; Meyerhoff, M. E. Anal. Chem. 1987, 59, 2345-2350. (5) Butt, S. B.; Camman, K. Anal. Lett. 1992, 25, 1597-1615. (6) Yasuda, K.; Miyagi, H.; Hamada, Y.; Takata, Y. Analyst 1984, 109, 61-64. (7) Liu, D.; Meyerhoff, M. E.; Goldberg, H. D.; Brown, R. B. Anal. Chim. Acta 1993, 274, 37-46. (8) Shin, J. H.; Yoon, S. Y.; Yoon, I. J.; Choi, S. H.; Lee, S. D.; Nam, H.; Cha, G. S. Sens. Actuators B 1998, 50, 19-26. (9) Scholer, R. P.; Simon, W. Chimia 1970, 24, 372-374. (10) Davies, O. G.; Moody, G. J.; Thomas, J. D. R. Analyst 1988, 113, 497-500. (11) Ghauri, M. S.; Thomas, J. D. R. Analyst 1994, 119, 2323-2326. (12) Ghauri, M. S.; Thomas, J. D. R. Anal. Proc. Incl. Anal. Commun. 1994, 31, 181-183. (13) Bratov, A.; Abramova, N.; Munoz, J.; Dominguez, C. J. Electrochem. Soc. 1997, 144, 617-622. (14) Buhlmann, P.; Pretsch, E.; Bakker, E. Chem. Rev. 1998, 98, 1593-1687. (15) Siswanta, D.; Hisamoto, H.; Tohma, H.; Yamamoto, N.; Suzuki, K. Chem. Lett. 1994, 945-948. (16) Moriuchi-Kawakami, T.; Nakazawa, S.; Ota, M.; Nishihira, M.; Hayashi, H.; Shibutani, Y.; Shono, T. Anal. Sci. 1998, 14, 1065-1068.

Analytical Chemistry, Vol. 72, No. 19, October 1, 2000 4683

pyrazol groups on benzene ring form an ideal binding site for NH4+ ion.17 The NH4+ selectivity of the PVC membrane based on POT -3) was ∼40 times higher this ionophore (KNH + + ) 2.5 × 10 4 ,K than that based on nonactin. However, its detection limit for NH4+ was over 10-4 M near the neutral pH region, which is 100 times higher than that of the nonactin-based membranes.18 In the past few years, we also have searched for synthetic ionophores whose NH4+ selectivity and resulting potentiometric performance in the solvent polymeric membranes is superior or at least comparable to that of nonactin. Reviewing the structures of nonactin and other NH4+-selective ionophores in the literature, we presumed that the macrocyclic polyethers having two conformational features may exhibit enhanced NH4+ selectivity: crown ether-type compounds with right cavity size for NH4+, and those containing symmetrically positioned heterocylic units in the structure. It was reasoned that NH4+ may be nested in or perched on the cavity of a cyclic polyether and bound tightly via balanced hydrogen bonding to the ligand donor atoms of the heterocylic units. Of the many available crown ether derivatives, dibenzo-18crown-6 (DB18C6) seemed to be a good frame molecule since the PVC-based membranes prepared with it provides high K+ and NH4+ selectivity, indicating that the compound has about the right cavity size for those ions.19 Hence, we designed two types of heterocyclic unit containing DB18C6-type compounds (1-4) as shown in Figure 1. In this contribution, we report the potentiometric evaluation for those ionophores incorporated in the PVCbased membranes. It is demonstrated here that the rationally designed heterocyclic unit containing DB18C6-type compounds indeed exhibit enhanced NH4+ selectivity with other potentiometric properties comparable to those of nonactin. EXPERIMENTAL SECTION Reagents. Poly(vinyl chloride) (PVC), potassium tetrakis(4chlorophenyl)borate (KTpClPB), NPOE, nonactin, and 9-(diethylamino)-5-octadecanoylimino-5H-benzo[R]phenoxazine (ETH 5294) were purchased from Fluka Chemie AG (Buch, Switzerland). Synthetic procedure and the method of purification for 1-5 shown in Figure 1 are provided separately in the Supporting Information. Reagents used for the synthesis of 1-5, i.e., catechol, 3,4dihydroxybenzaldehyde, 2,3-dihydroxynaphthalene, 1,3-bis(bromomethyl)benzene, and 2,6-pyridinedimethanol, were purchased from Aldrich Chemical Co. All other chemicals used were analytical reagent grade. Standard solutions and buffers were prepared with freshly deionized water (18 MΩ‚cm). Preparation of Electrodes and Their Potentiometric Evaluation. Nine different ion-selective membrane cocktails were prepared by dissolving the carriers 1-5 or nonactin, ETH 5294, or KTpClPB, together with PVC and plasticizer (1:2 by weight), to give a total cocktail mass of 200 mg, in 5 mL of THF. The specific compositions of membranes 1-9 (denoted as membrane (17) Chin, J.; Walsdorff, C.; Stranix, B.; Oh, J.; Chung, H. J.; Park, S.-M.; Kim, K. Angew. Chem., Int. Ed. 1999, 38, 2756-2758. (18) Recently, in the 84th Annual Korean Chemical Society Meeting (22 October, 1999) Kim, K.: et al. showed that the detection limit of the PVC-based membrane doped with 1,3,5-tris(3,5-dimethylpyrazol-1-ylmethyl)-2,4,6-triethylbenzene could be extended to ∼10-6 M, but only in a high-pH (>9.0) region. (19) Hong, U. S.; Kwon, H. K.; Cha, G. S.; Nam, H.; Chang, S. H.; Chung, K. B. J. Korean Chem. Soc. 1995, 39, 698-704.

4684 Analytical Chemistry, Vol. 72, No. 19, October 1, 2000

Figure 1. Structures of NH4+-selective neutral carriers examined in this work.

no. [ionophore/ETH 5294/KTpClPB] and the unit in parentheses is mmol/kg) given in the order of their respective number are as follows: membrane 1 [1 (22.8)/(0)/(2.3)]; membrane 2 [2 (19.1)/ (0.0)/(1.9)]; membrane 3 [3 (21.0)/(0.0)/(2.1)]; membrane 4 [4 (23.4)/(0.0)/(2.3)]; membrane 5 [5 (23.6)/(0.0)/(2.4)]; membrane 6 [nonactin (13.6)/(0.0)/(1.4); membrane 7 [(0.0)/(8.6)/(7.7)]; membrane 8 [1 (30.1)/(8.6)/(7.7)]; membrane 9 [nonactin (18.0)/ (8.6)/(7.7)]. The cocktail solutions were then poured into a glass ring (i.d. 22 mm) placed on a slide glass and dried over a day at room temperature. Small disks were punched from the cast films and mounted in Philips electrode bodies (IS-561; Glasblaserei Mo¨ller, Zu¨rich, Switzerland). For all electrodes, 0.1 M KCl was used as the internal reference electrolyte. All electrodes were presoaked in distilled water for 4 h before use. Potential differences between the ISEs and the Orion sleeve-type double-junction Ag/AgCl reference electrode (model 90-02) were measured using an IBM AT-type computer equipped with a homemade highimpedance input 16-channel analog-to-digital converter. The dynamic response curves were obtained at room temperature by adding standard solutions to 200 mL of magnetically stirred background electrolyte (0.05 M Tris-HCl, pH 7.2) every 100 s to vary the concentrations of each ionic species stepwise from 10-6 to 10-1 M, and the potentials were measured every second at room temperature. Selectivity coefficients were estimated according to the separate solution-matched potential method (IUPAC SSM II method) by comparing the activity of an interfering cation that induced the same potential change as that induced by NH4+ activity of 1.0 × 10-2 M.20 The detection limits of the electrodes were also determined following the IUPAC recommendation.20 To determine the complex formation constants of 1 and nonactin, the pH responses of the electrodes based on membrane

Figure 2. Dynamic potentiometric responses of the electrodes based on 1-6 and nonactin to NH4+ and alkali metal cations (background electrolyte: 0.05 M Tris-HCl, pH 7.2).

7-9 were measured by titrating the 0.1 M KCl and 0.1 M NaCl solutions, each buffered to pH 3.0 in 1.0 mM citric acid and 1.0 mM boric acid, with a concentrated LiOH solution, while simultaneously monitoring the sample pH with a pH glass electrode. The lifetime of the 1-3-based electrodes, which were stored in magnetically stirred distilled water, were measured by measuring their response to NH4+ for three months. Semiempirical calculations for the free and NH4+-complexed 1 were made using Gaussian 94 and Chem3D programs (CambridgeSoft, MA). RESULTS AND DISCUSSION Five compounds shown in Figure 1 were prepared to test our hypothesis that the heterocyclic unit containing DB18C6-type (20) IUPAC Recommendations for Nomenclature of Ion-Selective Electrodes. Pure Appl. Chem. 1994, 66, 2527-2536. (b) IUPAC Selectivity Coefficients for Ion-Selective Electtrodes: Recommended Methods for Reporting KPOT AB values. Pure. Appl. Chem. 1995, 67, 507-518.

crown ethers has an appropriate cavity size while providing enhanced hydrogen bonding for NH4+. 1-3 have the same thiazole-containing dibenzocrown ether (TDB18C6) frame except that 2 and 3 have hexyl chain and naphthalene to increase their lipophilicity. 4 was synthesized to contain pyridines to examine the effect of different heterocyclic units in the structure. 5 was prepared to examine the role of ligand-donating atoms in heterocyclic units. The NPOE-plasticized PVC-based ion-selective membranes were prepared with 1-5 and nonactin; their compositions are given in the Experimental Section. A small amount of lipophilic additive, 10 mol % KTpClPB with respect to the ionophore, was added to each membrane to catalyze the ion-exchange process at the sample/membrane interface. Figure 2 shows the dynamic response curves of the electrodes based on membranes 1-6 to NH4+, K+, Na+, and Li+ ions. It clearly demonstrates that the electrodes based on 1-3 provide enhanced NH4+ selectivity over K+ and Na+ compared to that of the nonactin-based electrode, Analytical Chemistry, Vol. 72, No. 19, October 1, 2000

4685

Table 1. Potentiometric Properties of the NH4+-Selective Membranes POT selectivity coefficients (log KNH + ) 4 ,J

membrane no.

Li+

Na+

K+

Rb+

Cs+

Mg2+

Ca2+

Sr2+

Ba2+

detectn limita

slopeb

1 2 3 4 5 6

-4.8 -4.9 -4.9 -2.6 -2.4 -4.4

-3.7 -3.9 -4.4 -0.4 -1.6 -3.0

-1.1 -1.3 -1.1 0.5 0.2 -1.0

0.7 0.6 0.6 0.5 1.1 -1.4

1.1 1.1 1.1 1.1 2.2 -2.4

-4.5 -4.6 -4.4 -2.5 -2.7 -4.2

-2.6 -2.7 -2.6 -2.7 -2.8 -2.5

-1.6 -1.7 -1.6 -1.6 -1.5 -1.6

-3.7 -3.8 -3.9 -2.0 -2.5 -4.5

-5.5 -5.6 -5.3 -3.8 -3.8 -6.1

59.4 57.7 58.2 41.4 42.8 59.3

a

Logarithmic scale. b Slopes from 10-5 to 10-1M (mV/decade).

Figure 3. Stereoview of the optimized structures for the free and NH4+-complexed 1 [bis(5,16-thiazolyl-1,2,10,11-benzene)coronand-18crown-6; TDB18C6]: (A) free, (B) 2:1 complex, and (C) 1:1 complex. Calculations were made with the semiempirical PM3 method.

supporting the validity of our presumption. While the enhancement in NH4+ selectivity of the 1-3-based electrodes over K+ is POT marginal compared to that of the nonactin-based (KNH + + ) 7.4 4 ,K -2 -2 -2 -1 × 10 , 5.0 × 10 , and 7.9 × 10 vs 1.0 × 10 , respectively), POT that over Na+ was substantially improved (KNH + + ) 2.2 × 4 ,Na -4 -4 -5 -3 10 , 1.3 × 10 , and 3.6 × 10 vs 1.1 × 10 , respectively). The enhanced NH4+ selectivity over Na+ would be a great advantage in the determination of NH4+ in physiological fluids, e.g., serum, whose Na+ concentrations are over 100 mM. On the other hand, the 1-3-based electrodes exhibited lower NH4+ selectivities over Rb+ and Cs+ than the nonactin-based electrode. However, those ions are not major interferences in the NH4+ determination. In 4686 Analytical Chemistry, Vol. 72, No. 19, October 1, 2000

one experiment, we examined the potentiometric performance of the 1-based membranes plasticized with dibenzyl ether to further enhance the NH4+ selectivity.16 However, the resulting membranes yield oily surfaces with no apparent improvement in NH4+ selectivity and other potentiometric properties. As shown in Figure 2, the electrodes based on 4 and 5 exhibited poor potentiometric responses to all cations examined with little improvement in its NH4+ selectivity. These results indicate that the pyridine units in the DB18C6-type structure hardly contribute to cation binding and the enhanced NH4+ selectivity of 1-3 is induced primarily by the thiazole units in them. As listed in Table 1, the NH4+ selectivities over other cations

Table 2. Experimental EMF Differences between NPOE/PVC Membranes Containing the Proton-Selective Carrier ETH 5294 and Anionic Site Additive KTpClPB and Ones with 1 and Nonactin, and Their Effective Complex Formation Constants ion M+ K+ Na+ NH4+ K+ Na+ NH4+

∆EM (mV) Nonactin (n ) 1) 388.7 322.3 1 (n ) 1) 264.2 160.2

logβM+L 8.56 7.44 9.56* 6.12 4.36 7.25*

* The 1:1 complex formation constants for NH4+ are estimated from POT the selectivity coefficients, KNH + +, assuming they are the ratio of 4 ,K experimental formation constants.

and the response slopes of the 1-3-based electrodes are also comparable to that of the nonactin-based one. To examine the structure of free 1, the base molecular frame of the TDB18C6-type compounds, and its NH4+ complex, we performed the semiempirical calculation with the PM3 method using Gaussian 94 interfaced to Chem3D programs. Figure 2 is the optimized structures of the free and 2:1 and 1:1 complexes: for the 2:1 complex, the distances between the four ether oxygens and the center of NH4+ [R(O‚‚‚NH4+)] are in the 2.63-2.80-Å range and the distances between the thiazole nitrogen and the nearest NH4+ hydrogens [R(N‚‚‚H4N+)] in the 3.43-3.63-Å range; and for the 1:1 complex, R(O‚‚‚NH4+) in the 1.77-1.78-Å range and R(N‚‚‚H4N+) in the 3.0-3.02-Å range. Although the PM3 method often results in too short hydrogen bond lengths (∼0.1 Å) and the wrong geometry for the weakly bound complexes, this result suggests that 1 forms stable 1:1 and 2:1 complexes with NH4+ through electrostatic interaction and hydrogen bonding. On the other hand, the same calculation fails to obtain a 2:1 complex with K+ or Na+; alkali metal cations are nested in the free TDB18C6 structure forming 1:1 complexes. The 1H NMR spectra of the free and NH4+-complexed 1 support our conclusion, exhibiting an interesting trend in their chemical shift as the molar ratios of the ligand to NH4+ (rNH4+/L) increase from 0 to 2. The thiazole proton peak at 7.38 ppm is gradually downfield shifted to 7.49, 7.53, and 7.57 ppm corresponding to the rNH4+/L values of 0.5, 1.0 and 2.0, respectively. This result indicates that the degree of interaction between the thiazole nitrogens and the NH4+ hydrogens increases with increasing NH4+ concentrations. The two pairs of closely spaced benzene protons, centered at 7.10 and 7.00 ppm, respectively, split into two separate pairs of multiplets centered at 6.78 and 7.02 ppm when rNH4+/L ) 0.5 and coalesced back at ∼6.95 ppm with increasing rNH4+/L values. This observation suggests that the chemical environments around the two benzene units, which are slightly different for the free ligand, become distinct when the rNH4+/L e 0.5 and approximately equivalent in the presence of an excess amount of NH4+. The structures of free and 2:1 and 1:1 complexes of 1 shown in Figure 3 agree well with this trend: the two benzene units, tilted outward forming a partial cone shape in the free 1, assume an approximately L-shaped conformation in the 2:1 complex and a nearly symmetric saddlelike conformation in the 1:1 complex.

Figure 4. 1H NMR (400 MHz, CD2Cl2/acetone, 1:1 v/v) spectrum of TDB18C6 (7.6 mM) (A) and spectral changes upon addition of 0.5 (B), 1.0 (C), and 2.0 equiv (D) of ammonium triflate (asterisk indicates solvent peak).

The methylene proton peaks show a large upfield shift (∆δ ) -0.5 ppm) when rNH4+/L ) 0.5, and shift back to downfield with increasing NH4+ concentration. The electrostatic interaction between the ether oxygens and NH4+ and the varying anisotropic effect due to large conformational changes around the benzene groups as depicted in Figure 3 may explain the observed trend. In summary, the H1 NMR results indicate that 1 forms a 2:1 complex when rNH4+/L < 0.5, but a 1:1 complex with an excess amount of NH4+ in a CD2Cl2/acetone solution. The cation complex formation constants of 1 in solvent polymeric membranes have been examined using the method proposed by Bakker and Pretsch:21 two different types of membranes, one containing a highly proton-selective ionophore (ETH 5294) with KTpClPB and another containing the same components with NH4+-selective neutral carriers (1 and nonactin), were prepared for this purpose (membranes 7-9). Varying the pH of the buffer solution (a mixture of 10-3 M citric acid and boric acid) (21) Bakker, E.; Pretsch, E. Anal. Chem. 1998, 70, 295-302.

Analytical Chemistry, Vol. 72, No. 19, October 1, 2000

4687

have not measured the complex formation constants for NH4+ as its concentration varies at high pH, evolving ammonia gas. Instead, we estimated the 1:1 complex formation constant for NH4+ using POT its selectivity coefficients, KNH + +, by assuming they are ap4 ,K proximately the ratio of the two formation constants, βNH4+,L and βK+,L, and both ions have similar partition constants between the sample and organic phase.22 The βK+,L for nonactin estimated in this manner is ∼3.6 × 109 and for 1 ∼1.8 × 107. The formation constant for the 2:1 complex has not been estimated, as we have no substantial evidence that 1 forms a 2:1 complex with alkali metal cations. The lifetime of a solvent polymeric membrane depends on the leaching rate of the ionophore incorporated in it into the sample phase. Dinten et al. have shown that the leaching rate constant is inversely proportional to the lipophilicity of an ionophore.23 It was observed that the response slopes (mV/decade) of the electrode based on 1 rapidly decreased from 59.4 to 43.0 mV in a month, indicating its lipophilicity is not sufficient. On the other hand, the electrode based on 2, which would exhibit higher lipophilicity than 1 due to the extra hexyl chain on the TDB18C6 molecular frame, provided potentiometric performance similar to that of the nonactin-based electrode for more than three months (57.0 mV/ decade on the average). The lifetime of the 3-based electrode was also similar to that of the 2-based one.

Figure 5. pH response functions of PVC/NPOE membranes (no. 7-9 in the Experimental Section) containing the H+ carrier ETH 5294 and KTpClPB (a) and, in addition, either TBD18C6 (b) or nonactin (c) as cation-selective carrier in 0.1 M NaCl (A) and 0.1 M KCl (B).

containing 0.1 M K+ or Na+, from 3 to 13, the potentiometric responses of the electrodes based on those two types of membranes have been recorded. It has been shown that the measured potential difference (∆EM) between the membranes with and without neutral carrier at high pH is related to the n:1 complex formation constant (βM+Ln+, where M+ denotes the cation and L the ligand) with the equation

∆EM )

RT ln((LT - RT )nβM+Ln) zF

(1)

where LT and RT represent the total concentration of ionophore and lipophilic additive doped in the membrane, respectively. The pH response functions for the nonactin- and 1-containing membranes in buffered KCl solution is shown in Figure 5, while the estimated complex formation constants are given in Table 2. We (22) Bakker, E.; Buhlmann, P.; Pretsch, E. Chem. Rev. 1997, 97, 3083-3132. (23) Dinten, O.; Spichiger, U. E.; Chaniotakis, N.; Gherig, P.; Rusterholtz, B.; Morf, W. E.; Simon, W. Anal. Chem. 1991, 63, 596-603.

4688

Analytical Chemistry, Vol. 72, No. 19, October 1, 2000

CONCLUSION In this report, we have shown that the TDB18C6-type compounds incorporated in PVC-based membranes provide enhanced NH4+ selectivities over other alkali metal cations compared to those of nonactin-containing membranes. The 1H NMR spectra of 1 suggest that TDB18C6 forms a 2:1 or 1:1 complex with NH4+ in the solution phase. Unlike other synthetic NH4+-selective neutral carriers reported to date, TDB18C6-type compounds result in potentiometric performance highly comparable to that of the nonactin-based ones except at their detection limits (∼3 × 10-6). Their NH4+ complex formation constants estimated in solvent polymeric membranes are ∼200 times smaller than that of nonactin. To further improve their NH4+ selectivity and binding strength, the same type of compounds with different functional groups are now being investigated. ACKNOWLEDGMENT This research was supported by the Korea Science and Engineering Foundation (KOSEF Project 96-0501-05-01-3) and in part by the Kwangwoon University Research Fund. H.-S.K. also gratefully acknowledges the Center for Biofunctional Molecules (1CB9902106) for the financial support. SUPPORTING INFORMATION AVAILABLE The detailed synthetic procedures and purification methods for s 1-5. This material is available free of charge via the Internet at http://pubs.acs.org. Received for review February 11, 2000. Accepted July 11, 2000. AC000177B