Anal. Chem. 1998, 70, 4631-4635
Discrimination between Butylammonium Isomers by Calix[5]arene-Based ISEs Marco Giannetto,† Giovanni Mori,*,† Anna Notti,‡ Sebastiano Pappalardo,§ and Melchiorre F. Parisi‡
Dipartimento di Chimica Generale, Inorganica, Analitica e Chimica Fisica, Universita` di Parma, Viale delle Scienze, I-43100 Parma, Italy, Dipartimento di Chimica Organica e Biologica, Universita` di Messina, I-98166 Messina, Italy, and Dipartimento di Scienze Chimiche, Universita` di Catania, I-95125 Catania, Italy Penta-O-alkylated p-tert-butylcalix[5]arenes 1-5 (R ) benzyl, isohexyl, isopropoxyethyl, isopropoxycarbonylmethyl, and tert-butoxycarbonylmethyl, respectively) in a fixed C5v cone conformation have been studied as ionophores in liquid membrane ion-selective electrodes (ISEs) for n-butylammonium against the other branched butylammonium isomers, alkali metals, and ammonium ions, in terms of detection limits, sensitivity, and selectivity. The highest levels of potentiometric selectivity and detection limits up to 3 × 10-6 M are observed with ISEs based on ionophore 2, where selectivity follows the order n-BuNH3+ . i-BuNH3+ > s-BuNH3+ > t-BuNH3+. The lower potentiometric selectivity displayed by ISEs based on ionophores 3-5 is ascribed to their affinity for the Na+ ion of the lipophilic salt present in the membrane, as evidenced by appropriate 1H NMR competition experiments with Na+ and n-BuNH3+ ions. Further investigation on the selectivity mechanism of ionophore 2 by means of frequency response analysis shows that the interaction of the linear butylammonium ion with membranes containing 2 involves a lower resistance process than that occurring with the other branched isomers, thus suggesting the presence of a favorable kinetic-controlled mechanism.
formationally preorganized polyfunctional derivatives, whose convergent ligating groups can act cooperatively to selectively bind ions or neutral molecules.5 For instance, the attachment of ester, keto, and amide groups to the lower rim of calix[n]arenes has produced a series of lipophilic cation receptors, which display remarkable selectivities for alkali metal cations in complexation, extraction, and transport experiments.5 This has led to the development of calixarene-based ISEs for alkali metal cations.6-8 On the other hand, the architecture of calixarenes is such that they also possess hydrophobic cavities generated by the aromatic walls of phenol residues, which are potentially useful for the inclusion of alkylammonium ions.9 In principle, a variety of size and shape of their cavities is available, depending on the number of constituent phenol residues and on the conformation assumed by the calixarene. Whereas the calix[4]arene cavity in the cone conformation is too narrow to accommodate even small amine salts, that of calix[6]arene is large enough to include small to medium-sized ammonium ions. However, due to their larger dimensions and considerable flexibility, it is rather difficult to obtain permanent cavities from calix[6]arenes, unless multiple intrabridging or appropriate scaffolding elements are introduced at the lower or upper rim to suppress, or at least reduce, the interconversion processes among the various conformers.10
In recent years there has been considerable interest in the development of molecular sensors for the on-line detection of chemical and biochemical species.1 Liquid membrane ion-selective electrodes (ISEs) provide one of the most popular and versatile sensing devices, and interesting selectivity matching with target analytes can be achieved by choosing suitable host molecules as sensing elements.2 However, there are still limited examples of ISEs, mainly based on crown ethers3 or natural ionophores,4 which strongly respond to and discriminate between organic ammonium ions. Calixarenes present attractive possibilities in host-guest chemistry as molecular frameworks for the preparation of con-
(3) Maeda, T.; Ikeda, M.; Shibahara, M.; Haruta, T.; Satake, I. Bull. Chem. Soc. Jpn. 1981, 54, 94-98. Shirahama, K.; Kamaya, H.; Ueda, I. Anal. Lett. 1983, 16(B19), 1485-1494. Bochenska, M.; Biernat, J. F. Anal. Chim. Acta 1984, 162, 369-371. Moody, G. J.; Owusu, R. K.; Thomas, J. D. Analyst 1987, 112, 121-127. Hassan, S. S.; Elnemma, E. M. Anal. Chem. 1989, 61, 21892192. Thoma, A. P.; Viviani-Nauer, A.; Schellenberg, K. H.; Bedekovic, D.; Pretsch, E.; Prelog, V.; Simon, W. Helv. Chim. Acta 1979, 62, 2303-2316. Bussmann, W.; Lehn, J.-M.; Oesch, U.; Plummer, P.; Simon, W. Helv. Chim. Acta 1981, 64, 657-661. Yasaka, Y.; Yamamoto, T.; Kimura, K.; Shono, T. Chem. Lett. 1980, 769-772. Shinbo, T.; Yamaguchi, T.; Nishimura, K.; Kikkawa, M.; Sugiura, M. Anal. Chim. Acta 1987, 193, 367-371. (4) Maruyama K.; Sohmiya, H.; Tsukube, H. J. Chem. Soc., Chem. Commun. 1989, 864-865. Tsukube, H.; Sohmiya, H. Tetrahedron Lett. 1990, 31, 7027-7030. Tsukube, H.; Sohmiya, H. J. Org. Chem. 1991, 56, 875-878. (5) Gutsche, C. D. In Calixarenes; Stoddart, J. F., Ed.; Monographs in Supramolecular Chemistry 1; Royal Society of Chemistry: Cambridge, 1989. Calixarenes, a Versatile Class of Macrocyclic Compounds; Vicens, J., Bo ¨hmer, V., Eds.; Kluwer: Dordrecht, 1991. Bo ¨hmer, V. Angew. Chem., Int. Ed. Engl. 1995, 34, 713-745. (6) For a recent review on calixarene sensing agents, see: Diamond, D.; McKervey, M. A. Chem. Soc. Rev. 1996, 15-24. (7) Careri, M.; Casnati, A.; Guarinoni, A.; Mangia, A.; Mori, G.; Pochini, A.; Ungaro, R. Anal. Chem. 1993, 65, 3156-3160. (8) Bocchi, C.; Careri, M.; Casnati, A.; Mori, G. Anal. Chem. 1995, 67, 42344238. (9) Araki, K.; Shimizu, H.; Shinkai, S. Chem. Lett. 1993, 205-208.
* To whom correspondence should be addressed. Phone +39 521 905431. Fax: +39 521 905557. E-mail:
[email protected]. † Universita ` di Parma. ‡ Universita ` di Messina. § Universita ` di Catania. (1) Odashima, K.; Yagi, K.; Tohda, K.; Umezawa, Y. Anal. Chem. 1993, 65, 1074-1083. (2) Jaramillo, A.; Lopez, S.; Justice, J. B., Jr.; Salamone, J. D.; Neil, D. B. Anal. Chim. Acta 1983, 146, 149-159. 10.1021/ac9803840 CCC: $15.00 Published on Web 09/25/1998
© 1998 American Chemical Society
Analytical Chemistry, Vol. 70, No. 21, November 1, 1998 4631
with the composition of the solution under analysis, it was calculated using the Henderson equation:15
∑z u (a - a ′) RT ∑z i i
Ej )
Recently, we have shown11 that penta-O-alkylated p-tert-butylcalix[5]arenes, frozen in a C5v-symmetric cone conformation, are probably the most efficient and selective synthetic neutral molecular receptors for linear alkylammonium ions reported to date. These compounds present a cavity size of intermediate dimensions between those of calix[4]arenes and calix[6]arenes and are able to form inclusion complexes with linear alkylammonium ions with remarkable selectivity against branched isomers. We have now prepared a number of calix[5]arene-based ISEs, and in this paper we show one of these host compounds to be selective and potentiometrically responsive to the binding of linear alkylammonium guest ions, even in the presence of inorganic interfering ions. Selected receptors (1-5) and guest analytes (isomeric butylammonium ions, 6a-d) are shown in Figure 1. EXPERIMENTAL SECTION Calix[5]arenes 1,12 2-4,11 and 513 were prepared according to literature procedures. All chemicals were of analytical reagent grade. Dibutyl sebacate (DBS), high-molecular-weight poly(vinyl chloride) (PVC), and sodium tetraphenylborate (Na(TPB)) were purchased from Fluka-Selectophore (Milan, Italy). The composition of membranes employed in measurements was as follows: PVC (33% w/w), Na(TPB) (0.2% w/w) as lipophilic anion salt, calixarenes 1-5 (0.8% w/w) as ionophores, and DBS (66% w/w) as plasticizer. Liquid membrane electrodes 1-5, based on calix[5]arenes 1-5, were prepared according to a previously reported procedure.7,8 Electrodes were conditioned for 1 h in deionized water before each measurement. The solution employed in the salt bridge applied to Ag/AgCl reference electrode was 1 M CaCl2. All concentration values were converted into activities before calculations, using the Guggenheim-Davies14 relation, and each potentiometric response was corrected for the junction potentials Ej at the salt bridge-analyte solution interface. Since Ej varies (10) Ikeda, A.; Shinkai, S. Chem. Rev. 1997, 97, 1713-1734. Otsuka, H.; Suzuki, Y.; Ikeda, A.; Araki, K.; Shinkai, S. Tetrahedron 1998, 54, 423-446. (11) Arnaud-Neu, F.; Fuangswasdi, S.; Notti, A.; Pappalardo, S.; Parisi, M. F. Angew. Chem., Int. Ed. Engl. 1998, 37, 112-114. (12) Stewart, D. R.; Krawiec, M.; Kashyap, R. P.; Watson, W. H.; Gutsche, C. D. J. Am. Chem. Soc. 1995, 117, 586-601. (13) Barrett, G.; McKervey, M. A.; Malone, J. F.; Walker, A.; Arnaud-Neu, F.; Guerra, L.; Schwing-Weill, M.-J.; Gutsche, C. D.; Stewart, D. R. J. Chem. Soc., Perkin Trans. 2 1993, 1475-1479. (14) Stokes, R. H. J. Am. Chem. Soc. 1950, 72, 763-767.
4632 Analytical Chemistry, Vol. 70, No. 21, November 1, 1998
i
i
∑ i
Figure 1. Structures of calix[5]arene-based hosts 1-5 and isomeric butylammonium guests 6a-d.
i
zi2ui(ai - ai′)
F
ln
uiai′
2 i
i
∑
(1)
zi2uiai
i
where zi is the charge, ui is the absolute mobility, ai is the activity in the sample solution, and ai′ is the activity in the salt bridge solution for a given ion i. Selectivities and slopes were determined by the fixed interfering (0.01 M) method and calculated using a nonlinear regression on the Nikolsky-Eisenman equation. n-BuNH3+ was chosen as the primary ion, and selectivity constants toward i-BuNH3+, s-BuNH3+, t-BuNH3+, Na+, K+, and NH4+ were then measured. For the determination of selectivity constants among butylammonium isomers 6a-d, solutions of hydrochlorides of the respective amines were employed. All potentiometric calibrations were performed in the 10-6-10-1 M range at 25 °C, using an AMEL 338 potentiometer. Most calculations were performed with the Statgraphics (Manugistic Inc., Rockville, MD) statistical package. 1H NMR spectra were recorded at 300 MHz, at a room temperature of 22 ( 1 °C. Chemical shifts (δ) are given in ppm from (CH3)4Si, used as an internal standard. Complexation experiments were carried out by first adding sodium picrate and then n-BuNH3+ picrate (and vice versa) to hosts 1-5. NMR samples were prepared by mixing stock solutions of hosts (6.6 × 10-3 M in CDCl3) and guests (2.0 × 10-1 M in CD3OD) to a final CDCl3-CD3OD (9:1 v/v) solvent mixture. For a given set of measurements (see above), equimolar amounts of hosts and guests were used, and concentrations were kept in the 4.0 × 10-35.0 × 10-3 M range. Frequency response analysis (FRA) experiments were conducted on Pt “coated wire” 16 electrodes in the 10-1-104 Hz range. Measurements were performed at 25 °C at the open cell potential, using an AUTOLAB PGSTAT 20 frequency response analyzer, controlled by AUTOLAB FRA software (Utrecht, The Netherlands). RESULTS AND DISCUSSION Potentiometric Measurements. Figure 2 shows the potentiometric selectivity constants of electrodes 1-5 with the inorganic and organic ions under investigation, choosing n-BuNH3+ (6a) as the primary ion. These data clearly indicate that only electrode 2, based on ionophore 2, is analytically useful for the selective detection of 6a in the presence of the other branched isomers 6b-d. In the absence of any interfering ion, electrode 2 displays a detection limit of 3 × 10-6 M for 6a. The selectivities of ionophores 2-5 for linear vs branched butylammonium ions, obtained by UV techniques in a previous study,11 are reported in Table 1. The potentiometric (pKpot values in Table 2) and the UV selectivity data, evaluated as the difference between the log(Kf) for the primary and the interferring ions, display qualitatively similar trends, but the potentiometric values are all at least 1 order of magnitude lower than those obtained by (15) Bard, A. J.; Faulkner, L. R. Electrochemical Methods: Fundamentals and Applications, John Wiley & Sons: New York, 1980; Chapter 2. (16) James, H.; Carmack, G.; Freiser, H. Anal. Chem. 1972, 44, 856-857.
Figure 2. Potentiometric selectivity constants of electrodes 1-5 (primary ion n-BuNH3+, specified interfering ion). Table 1. UV n-BuNH3+/Specified Ion Selectivity Constants (As Calculated from Stability Constants, Kst) UV selectivity constants (∆log Kst) ionophore
i-BuNH3+
s-BuNH3+
t-BuNH3+
2 3 4 5
2.08 ( 0.02 1.42 ( 0.06 1.63 ( 0.02 2.38 ( 0.08
2.12 ( 0.02 1.44 ( 0.03 1.7 ( 0.1 2.7 ( 0.2
2.64 ( 0.03 1.76 ( 0.04 1.8 ( 0.2 2.98 ( 0.08
Table 2. Selectivity Parameters (pKpot sel ) and Slopes (mV/pn-BuNH3+) of Electrode 2 with and without Ionophore (in Parentheses) in the Membrane interfering ion
pKpot sel
slope (mV/dec)
t-BuNH3+
1.65 ( 0.05 (0.9 ( 0.2) 1.31 ( 0.03 (0.4 ( 0.1) 1.20 ( 0.05 (0.41 ( 0.07) 2.26 ( 0.04 (0.85 ( 0.04) 2.26 ( 0.05 (0.92 ( 0.04) 1.24 ( 0.04 (1.06 ( 0.04)
49 ( 1 (12 ( 1) 51.7 ( 0.9 (17 ( 1) 46 ( 1 (17 ( 1) 45 ( 1 (48.8 ( 0.7) 47 ( 1 (1.3 ( 0.9) 46 ( 1 (51 ( 1)
s-BuNH3+ i-BuNH3+ K+ Na+ NH4+
UV techniques. Even taking into account the possibility of ion pairs formation between the alkylammonium and tetraphenylborate ions within the membrane and differences of solvents and anions in solution (UV measurements were performed on picrates in CHCl3), surprisingly low potentiometric selectivity levels were observed for electrodes 3-5, which are based on ionophores 3-5. The latter are potentially ditopic receptors, since they possess not only a π-rich hydrophobic cavity at the upper rim, which is capable of molecular recognition via specific cation/π interactions, but also a hydrophilic cavity at the lower rim, defined by the five phenoxy groups and pendant ethereal (ionophore 3) or carbonyl ester functionalities (ionophores 4 and 5), which is well suited for the complete encapsulation of alkali metal ions. In fact, McKervey et al. have reported very high stability constants for the alkali metal complexes with 5 (log β values in the range 5.1-6.1 M-1).13 Furthermore, the crystal structure of the sodium complex with a
structurally related p-tert-butylcalix[5]arene penta-tert-butyl ketone has shown the cation to be deeply accommodated inside an eightcoordinate enclathration cage, composed of four carbonyl groups, three phenolic oxygens, and a fourth oxygen atom of a methanol solvent molecule.17 Therefore, we thought that the remarkable lowering of selectivity observed when using calix[5]arenes 3-5 as sensing molecules in ISEs might be due to competition from the sodium ions present in the lipophilic salt of the electrode membrane (Na(TPB)). This hypothesis has been fully verified by the results of appropriate NMR experiments (see below). As a result, our subsequent potentiometric studies were focused on electrode 2, based on pentaisohexyl derivative 2, which does not suffer from interference by the sodium ion. A problem, which should not be underestimated, in the study of the selectivity mechanisms of ionophores in liquid membrane ISEs, sometimes arises from an intrinsic affinity of the primary ion for the plasticizer solvent of the membrane. For instance, high selectivity values have been observed for acetylcholine, even in the absence of a sensing agent (resorcinol/acetaldehyde cyclic tetramer) in the membrane.18 With the electrodes under study here, it is important to take into account the lipophilicity of the ions investigated. Among butylammonium ions, n-BuNH3+ is the most lipophilic and could, therefore, preferentially interact with a lipophilic medium such as the electrode membrane. To rule out this possibility, we have double-checked the efficiency/selectivity of our liquid membrane electrodes in the absence of ionophore 2. The data reported in Table 2 show that, with all interfering ions investigated, very low selectivity and sensitivity values were observed on membranes without any ionophore. 1H NMR Experiments of n-BuNH + Inclusion in the 3 Presence of Interfering Ions. 1H NMR competition experiments of derivatives 1-5 with Na+ and n-BuNH3+ picrates have clearly shown that 1 and 2 form preferentially an inclusion complex with the alkylammonium ion, whereas 3-5 exhibit a marked affinity for the inorganic ion. A comparison of the 1H NMR spectra (not shown) of 3-5, obtained in CDCl3-CD3OD (9:1 v/v), prior to and after addition of an equimolar amount of sodium picrate, has revealed in all instances the formation (100%) of the corresponding Na+ complexes (i.e., downfield shift of the singlet of the aromatic protons and smaller separation of the AX components arising from the bridging CH2 groups).19 Subsequent addition of 1 equiv of n-butylammonium picrate to the solutions of these preformed complexes produced no further spectral changes, indicating that the Na+ ion binds to 3-5 more tightly than the n-BuNH3+ ion. Although the 1H NMR data have shown a very low affinity of pentaether 1 for the Na+ ion (slight broadening of the signals with no hint of spectral changes upon addition of 1 equiv of Na+), only moderate efficiency was observed for n-BuNH3+ (38% endocavity complexation), likely due to overcrowding of the rigid benzyl substituents at the lower rim. In contrast, when sodium (1 equiv) and n-butylammonium (1 equiv) picrates were added in sequence to ionophore 2, the Na+ ion was first complexed, (17) Bell, S. E. J.; Browne, J. K.; McKee, V.; McKervey, M. A.; Malone, J. F.; O’Leary, M.; Walker, A.; Arnaud-Neu, F.; Boulangeot, O.; Mauprivez, O.; Schwing-Weill, M.-J. J. Org. Chem. 1998, 63, 489-501. (18) Macca`, C., personal communication. (19) Arduini, A.; Pochini, A.; Reverberi, S.; Ungaro, R.; Andreetti, G. D.; Ugozzoli, F. Tetrahedron 1986, 42, 2089-2100.
Analytical Chemistry, Vol. 70, No. 21, November 1, 1998
4633
Figure 3. Displacement of Na+ cation (loosely bound at the lower rim of ionophore 2) by n-BuNH3+ via formation of a more stable endocavity complex 2‚n-BuNH3+.
Figure 5. FRA diagram of an interfering ion solution superimposed on the diagram relevant to interfering and primary ions at different concentrations.
Figure 4. 1H NMR spectra (CDCl3-CD3OD, 9:1 v/v) of ionophore 2 (trace a) and spectral changes after sequential addition of sodium picrate (1 equiv) (trace b) and n-BuNH3+ (1 equiv) (trace c).
presumably by the phenolic oxygens of the lower rim, and then fully displaced by the inclusion of the n-BuNH3+ ion into the preorganized π-rich calix[5]arene cavity. This process is depicted schematically in Figure 3. The doubling of the 1H NMR signals observed in trace b of Figure 4 indicates the formation of a loose (broad signals) 2⊂Na+ complex (∼25%), which is disrupted by the subsequent addition of 1 equiv of n-BuNH3+ ion in favor of the more stable 2⊂n-BuNH3+ complex (trace c). The formation of such a 1:1 inclusion complex is strongly supported by the appearance of the n-butylammonium resonances at very high field (δ values in the range from -0.29 to -1.90 ppm) and by the presence of a new singlet (δ 7.25 ppm) and an AX system (δ 4.45 and 3.50 ppm) for the aromatic and bridging methylene protons, respectively. The reverse experiment, in which 2 was first exposed to 1 equiv of n-BuNH3+, gave rise to a 1H NMR spectrum (not shown) similar to that reported in Figure 4 (trace c). The 2⊂n-BuNH3+ complex formed (95%) was unaffected by subsequent addition of Na+ ions (up to 2 equiv), confirming once again the low affinity of 2 for this inorganic ion. FRA Measurements. The relationship between UV and potentiometric selectivities was investigated by means of FRA experiments.20-23 For this purpose, Pt “coated wire” electrodes with the same membrane composition, previously used for the potentiometric determinations, both with and without ionophore 2, were examined for solutions of n-BuNH3+ (as primary ion) and t-BuNH3+ (as interfering ion), ranging from 10-3 to 10-1 M. The resulting diagrams were interpreted using the Randles equivalent circuit shown in Chart 1.23 (20) Buck, R. P. The impedance method applied to the investigation of Ion-selective electrodes, Presented at the 3rd Symposium on Ion-Selective Electrodes, Ma`trafu ¨ red, Hungary, 1980. (21) Xie, S. L.; Cammann, K. J. Electroanal. Chem. 1987, 229, 249-263. (22) Masuda, Y.; Liu, J.; Sekido, E. J. Electroanal. Chem. 1991, 313, 95-107. (23) Covington, A. K.; Zhou, D. M. J. Electroanal. Chem. 1992, 341, 77-84.
4634 Analytical Chemistry, Vol. 70, No. 21, November 1, 1998
Figure 6. FRA diagram obtained in 0.1 M n-BuNH3+, with and without ionophore 2 in the membrane.
Chart 1. Randles Equivalent Circuit
The impedance was measured by applying an alternate current of decreasing frequency and measuring the corresponding potential. The calculated impedance can then be represented by complex numbers, the imaginary component of which is reported, changed in sign, versus the real one. The resulting plot, called a Nyquist plot,24 is shown in Figures 5 and 6. The high-frequency (low Z′ values) semicircle is relevant to the “bulk” characteristics of the membrane (Rb||Cg circuit in Chart 1). The low-frequency (high Z′ values) region typically responds to processes taking place at the sample-membrane phase boundary. In particular, the charge-transfer resistance (Rct) due to the interaction of the ionophore with the ions in solution and the capacity of the double layer (Cdl) cause the formation of the second (24) Bard, A. J.; Faulkner, L. R. Electrochemical Methods: Fundamentals and Applications; John Wiley & Sons: New York, 1980; Chapter 9.
semicircle, while the impedance W, the “Warburg impedance”, arises at low frequencies when diffusion and ion-coupling phenomena are prevailing over interaction due to complexation by neutral carriers. The second semicircle in Figure 5, trace relative to the interfering ion, is relevant to the interaction of t-BuNH3+ with the membrane (Rct; Cdl from 300 to 400 kΩ). The high-frequency region of the diagrams strongly depends on the “bulk” characteristics of each electrode. Since the thickness of the membrane is not easily reproducible with “coated wire” electrodes, a direct comparison between responses to primary and interfering ions on the same electrode was carried out. Figure 5 shows that, at a given concentration (10-2 M) of t-BuNH3+, the second semicircle decreases as the concentration of n-BuNH3+ is increased. This can be interpreted in terms of the n-BuNH3+ Rct being negligible with respect to Rb and shows that the potentiometric selectivity for the linear isomer is also due to a favorable kinetic contribution. Finally, Figure 6 shows a comparison between the diagrams obtained with a 10-1 M n-BuNH3+ solution and electrodes with and without ionophore in the membranes. In the spectrum of the electrode without ionophore, a Warburg impedance can be observed. This is a further confirmation of the efficiency of calix[5]arene derivative 2 as a selective ionophore for n-BuNH3+ ions. CONCLUSIONS From a screening on calix[5]arene derivatives 1-5, it emerges that pentaisohexyl ether 2 has a useful application as an ionophore in n-BuNH3+ ISEs, in terms of both detection limit and selectivity, due to the high degree of preorganization of its hydrophobic cavity and to the weak interaction of phenolic oxygens at the lower rim
with sodium ions present in the membrane. A kinetic contribution has also been demonstrated by FRA. The potentiometric selectivities found for electrode 2, based on 2, are 1 order of magnitude lower than the thermodynamic selectivities determined by UV in a previous work, probably due to the difference in solvents and anions in solution and also possibly to ion-pairing between butylammonium and tetraphenylborate ions in the membrane. Conversely, the strong affinity of ionophores 3-5 for the sodium ion in the lipophilic salts used in membranes causes a lowering of their efficiency and selectivity as ISE sensing agents. These observations demonstrate that calix[5]arenes bearing long alkyl chains at the lower rim (e.g., 2) are the molecules of choice for the development of highly efficient ISEs for the recognition of linear alkylammonium ions in the presence of inorganic ions (e.g., Na+). Selectivities of electrodes based on structurally related calixarenes, containing carbonyl or ethereal oxygen moieties (e.g., 3-5), on the other hand, are strongly depressed by the presence of sodium ions. The findings of this model study, confined to isomeric butylammonium 6a-d, pave the way for the development of calix[5]arene-based ISEs for the on-line detection of molecules of biological interest, incorporating the n-BuNH3+ structural motif (e.g., γ-aminobutyric acid, biogenic amines, lysine-containing peptides, etc.). From a practical point of view, it is of considerable value to have now available potentiometric sensing agents which are able to recognize an n-alkylammonium subunit but at the same time do not suffer from interference by the inorganic ions omnipresent in physiological fluids.
Received for review April 7, 1998. 1998.
Accepted July 27,
AC9803840
Analytical Chemistry, Vol. 70, No. 21, November 1, 1998
4635