Anal. Chem. 2002, 74, 1327-1332
Perbrominated closo-Dodecacarborane Anion, 1-HCB11Br11-, as an Ion Exchanger in Cation-Selective Chemical Sensors Shane Peper, Martin Telting-Diaz, Philip Almond, Thomas Albrecht-Schmitt, and Eric Bakker*
Department of Chemistry, Auburn University, Auburn, Alabama 36849
The 2,3,4,5,6,7,8,9,10,11,12-undecabromocarborane anion, 1-HCB11Br11- (UBC-) has been evaluated for its suitability as an ion exchanger in solvent polymeric membrane electrodes and bulk optodes. Experiments comparing the chemical stability of the perhalogenated carborane anion to that of the best lipophilic tetraphenylborate, 3,5-[bis(trifluoromethyl)phenyl]borate (TFPB-), demonstrated that in the presence of 0.2 M acetic acid TFPB- was completely lost within 6 h, while the concentration of UBC- decreased by less than 10% in the same time period. Thin-film bulk optodes containing BME-44 as potassium-selective ionophore, ETH 5294 as chromoionophore, and UBC- as ionic sites exhibited a K+ response similar to analogous optodes containing TFPB-, with comparable selectivities over Na+ and Ca2+. Potentiometric measurements evaluating the selectivity behavior of UBC- in both ionophore-free and ionophorecontaining electrodes were performed. Ionophore-free PVC membranes containing UBC- as ion exchanger and either DOS or NPOE as plasticizer also demonstrated selectivity similar to TFPB--containing membranes. Sodium-selective membranes containing the ionophore 4-tertbutylcalix[4]arenetetraacetic acid tetraethyl ester (sodium ionophore X) and UBC- as ionic sites showed a Nernstian response for sodium and selectivity comparable to that found in analogous electrodes containing TFPB-. Tetraphenylborate derivatives have been used as ion exchangers in cation-selective solvent polymer membrane electrodes and bulk optodes for many years. They were initially introduced into ISE membranes to facilitate Donnan exclusion, which is the electrostatic repulsion of lipophilic anions trying to extract into the membrane.1 In addition to reducing anion interference, tetraphenylborates also decrease membrane resistance.2 It was later found that the presence of such borates in optimized concentrations improves ionophore selectivity by stabilizing the concentration of ion-ionophore complexes.1 The delocalized monoanionic charge that these compounds possess, in combination with their sterically hindered molecular structure (see Figure 1), make them very weakly coordinating. This is a characteristic that leads to weak, nonspecific ion pair formation and maximum ionophore-mediated selectivity of the membrane.3 (1) Bakker, E.; Bu ¨ hlmann, P.; Pretsch, E. Chem. Rev. 1997, 97, 3083. (2) Bakker, E.; Pretsch, E. Anal. Chim. Acta 1995, 309, 7. 10.1021/ac0156262 CCC: $22.00 Published on Web 02/14/2002
© 2002 American Chemical Society
Because the unsubstituted tetraphenylborate (TPB-) is susceptible to decomposition via acid hydrolysis, oxidants, and light, the search for more chemically stable derivatives began many years ago.1,2,4 The limited lipophilicity of TPB- was recently addressed by Kimura by the introduction of a derivative that was covalently attached to a sol-gel matrix.5 Clearly, however, the most successful derivative for general use has been the highly substituted 3,5-[bis(trifluoromethyl)phenyl]borate (NaTFPB).6 The presence of strong electron-withdrawing groups decreases the tendency for cleavage of the boron-phenyl bond, because the amount of localized charge at the ipso carbons is significantly reduced.4 Furthermore, the presence of electron-withdrawing groups suppresses the π-coordination of the phenyl groups, thus making the compound more inert and improving its electrochemical stability by increasing the reduction potential.4 Even though halogenated derivatives, such as NaTFPB, are more lipophilic and more resistant to phenyl cleavage, acid-hydrolyzed decomposition still occurs albeit at a much slower rate.3,6 This shortcoming limits the use of tetraphenylborates in systems that require an acidic sample pH, as in the case of heavy metals, such as Pb2+.7 Compounds that may be suitable alternatives to tetraphenylborates are carboranes, specifically closo-dodecacarboranes. The chemical structure of the perbromo-substituted derivative appears in Figure 1. These compounds possess many characteristics that may make them suitable ion exchangers. Very weak ion pair formation is observed due to the lack of electron lone pairs and π-electrons, a property rarely found in anions.8 Moreover, as stated by Reed, the dodecacarborane anion, CB11H12, behaves “like a 3D analogue of benzene” because of the versatile functionalization chemistry that these compounds possess.8 The desired lipophilicity of this class of carboranes can easily be tailored both at the boron vertexes4,8-10 and at the carbon vertex.11,12 The most (3) Rosatzin, T.; Bakker, E.; Suzuki, K.; Simon, W. Anal. Chim. Acta 1993, 280, 197. (4) Strauss, S. Chem. Rev. 1993, 93, 927. (5) Kimura, K.; Sunagawa, T.; Yajima, S.; Miyake, S.; Yokoyama, M. Anal. Chem. 1998, 70, 4309. (6) Nishida, H.; Takada, N.; Yoshimura, M.; Sonoda, T.; Kobayashi, H. Bull. Chem. Soc. Jpn. 1984, 57, 2600. (7) Ceresa, A.; Bakker, E.; Hattendorf, B.; Gu ¨ nther, D.; Pretsch, E. Anal. Chem. 2001, 72, 343. (8) Reed, C. Acc. Chem. Res. 1998, 31, 133. (9) Xie, Z.; Tsang, C.-W.; Sze, E. T.-P.; Yang, Q.; Chan, D. T. W.; Mak, T. C. W. Inorg. Chem. 1998, 37, 6444. (10) Tsang, C.-W.; Xie, Z. Chem. Commun. 2000, 1839. (11) Jelinek, T.; Baldwin, P.; Scheidt, W. R.; Reed, C. A. Inorg. Chem. 1993, 32, 1982.
Analytical Chemistry, Vol. 74, No. 6, March 15, 2002 1327
Figure 1. Three-dimensional chemical structures of the ion exchangers discussed in this work. (Left) 3,5-[bis(trifluoromethyl)phenyl]borate (TFPB-); (right) perbrominated closo-dodecacarborane anion, 1-HCB11Br11- (UBC-).
lipophilic derivatives that have been synthesized are those of the perhalogenated9 and peralkylated dodecacarborane anion.10 In addition to potentially unparalleled lipophilicity, the carboranes possess many other characteristics that make them suitable for electrochemical applications. For example, they are not susceptible to acid and base hydrolysis and they are relatively inert to electrochemical oxidation (∼2.0 V vs ferrocene/ferrocenium at Pt in dichloromethane).8 High Ih symmetry and tangentially delocalized σ-bonding make the carboranes one of the most chemically stable classes of compounds in chemistry. Furthermore, their bulky size (nearly 1 nm in diameter) and sufficient charge delocalization meet the criteria imposed for sufficient ion exchanging. Another advantage, important for bulk optode studies, is their lack of absorption in the UV-visible spectrum. In this paper, we present the trimethylammonium salt of the 2,3,4,5,6,7,8,9,10,11,12-undecabromocarborane anion (TMAUBC) as a more robust alternative to TFPB- as an ion exchanger in solvent polymeric membrane ion-selective electrodes and bulk optodes. Chemical stability in acidic media and lipophilicity of the perbrominated carborane are evaluated and compared to the traditionally used anionic site, NaTFPB. Bulk optode films based on ion-exchange equilibria are used to evaluate function and influence on selectivity of TMAUBC. In addition, potentiometric evaluation of the selectivity behavior of the perbrominated dodecacarborane is compared to that of NaTFPB in ISE membranes with and without a neutral ionophore. EXPERIMENTAL SECTION Reagents. For membrane preparation, poly(vinyl chloride) (PVC), bis(2-ethylhexyl sebacate) (DOS), 2-nitrophenyl octyl ether (NPOE), sodium tetrakis[3,5-bis(trifluoromethyl)phenyl]borate (NaTFPB), 4-tert-butylcalix[4]arenetetraacetic acid tetraethyl ester (sodium ionophore X), 2-dodecyl-2-methyl-1,3-propanediyl bis[N(12) Xie, Z.; Tsang, C.-W.; Xue, F.; Mak, T. C. W. J. Organomet. Chem. 1999, 577, 197.
1328 Analytical Chemistry, Vol. 74, No. 6, March 15, 2002
[5′-nitro(benzo-15-crown-5)-4′-yl]carbamate] (potassium ionophore III, BME-44), and tetrahydrofuran (THF) were of Selectophore quality from Fluka (Milwaukee, WI). Cesium carborane (CsCB11H12) and silver 7,8,9,10,11,12-hexabromocarborane (AgCB11Br6H6) were of the highest quality available from Strem Chemicals (Newburyport, MA). Tris(hydroxymethyl)aminomethane (TRIS) was of ACS grade from Aldrich (Milwaukee, WI). Chloride salts of lithium, sodium, potassium, calcium, and magnesium were Puriss quality from Fluka. All salt solutions were made with deionized Nanopure water (18 MΩ cm). For syntheses, triflic acid (99%) and bromine (99.8%) were purchased from Alfa Aesar (Ward Hill, MA). Trimethylammonium chloride (98%), octadecanoic acid chloride, 4-aminoazobenzene, triethylamine (99.5%), and lithium aluminum hydride were acquired from Aldrich. Diethyl ether, dichloromethane, toluene, and hexane were ACS grade from Fisher Scientific (Norcross, GA). Syntheses. Synthesis of [Me3NH][1-H-CB11Br11]. The synthesis of trimethylammonium-2,3,4,5,6,7,8,9,10,11,12-undecabromocarborane ([Me3NH][1-H-CB11Br11], TMAUBC) was performed according to the procedure previously reported by Xie et al.9 but at half scale. To summarize, a thick-walled Pyrex tube was charged with CsCB11H12 (0.05 g, 18 mmol), bromine (0.5 mL, 9.7 mmol), and triflic acid (0.5 mL, 5.65 mmol). The tube was sealed under vacuum and placed in an oven, which was heated to 200 °C and maintained at that temperature for 4 days. After removal of excess bromine and triflic acid, the brown residue was washed with a 5% (w/v) solution of NaOH until a neutral pH was reached. The solution was extracted with diethyl ether (3 × 20 mL), and after concentrating, an aqueous solution of Me3NHCl was added dropwise until no more precipitate formed. The white precipitate was filtered off, washed with deionized water, and dried under vacuum, giving [Me3NH][1-H-CB11Br11] as a white solid (147 mg, 76% yield). The formula of [Me3NH][1-H-CB11Br11] was confirmed by ESI-MS, 1H NMR (250 MHz, acetone-d6), and IR. Elemental analysis calculated for C4H11NB11Br11 (1071.00): C, 4.48; H, 1.11;
N, 1.31. Found: C, 4.92; H, 1.15; N, 1.33. Caution: Lipophilic anion exchangers are potentially hazardous to your health if swallowed. While clinical tests based on this lipophilic derivative have not been performed, alkali metal salts of the simple dodecacarborane anion, CB11H12-, that were passed straight through living organisms exhibited the equivalent toxicity of NaCl.8 The procedure for the synthesis of ETH 5315 has previously been described13 and gave a yield of 163 mg (20%). The structure was confirmed by 1H NMR (250 MHz, CDCl3) and FAB-MS. Optode Leaching Experiments. Optode membranes for leaching experiments contained 1.8 mg (16.7 mmol/kg) of ETH 5315, 155 mg of DOS (66 wt %), 78 mg of PVC (33 wt %), and either 4.7 mg of NaTFPB (22 mmol/kg) or 5.7 mg (22 mmol/kg) of [Me3NH][1-H-CB11Br11]. The components were dissolved in 1.5 mL of THF. From each cocktail, two membranes of the same composition were cast onto two 35-mm quartz disks by means of a spin-coating device.14 After the films were air-dried for 1 h, they were placed in a flow-through cell,1 which was mounted into a Hewlett-Packard 8452A diode array UV-visible spectrophotometer and filled with 0.2 M HOAc. The solution was replaced after 4 h of stasis by flowing 0.2 M HOAc continuously through the flow cell at a rate of 1.2 mL/min. Absorption spectra were recorded between 300 and 800 nm at 1-min intervals. Optode K+ Response. Cocktails containing 78.16 mg (33 wt %) of PVC and 154.78 mg (66 wt %) of DOS, 4.72 mg (20 mmol/kg) of the potassium-selective ionophore BME-44 (potassium ionophore III), 1.37 mg (10 mmol/kg) of the chromoionophore ETH 5294 (chromoionophore I), and either 2.21 mg of NaTFPB (10 mmol/kg) or 2.75 mg of TMAUBC (10 mmol/kg) as anionic sites were dissolved in 1.5 mL of THF. Thin films ∼2 µm thick were cast onto quartz disks as mentioned above. Samples made from the chloride salts of potassium, sodium, and calcium were buffered with 1 mM TRIS, pH 7.46. The films were allowed to equilibrate for 2 h prior to measurement. A fluorescence microscope was used to assess chromoionophore protonation as previously described.15 Fluorescence was collected using a filter cube with a 510-560nm excitation filter, a 565-nm dichroic mirror, and a 595-nm longpass emission filter and spectrally resolved with a dispersing element. Neutral density filters 4 and 8 were used in conjunction with a 40× objective lens. An exposure time of 1000 ms was used for spectral acquisitions. Membrane Preparation and Potentiometric Measurements. ISE membranes of ∼200-µm thickness were prepared by pouring a solution of ∼140 mg of total components, dissolved in 1 mL of THF, into 22-mm glass rings affixed onto glass plates. After solvent evaporation, 6-mm-diameter membrane disks were cut from the parent membrane and assembled into Philips Electrode bodies. A four-electrode batch was fabricated for each composition examined which consisted of 10 mmol/kg ion exchanger, i.e., TMAUBC or NaTFPB (0.9-1.1 wt %), plasticizer DOS or o-NPOE (66 wt %), and PVC (33 wt %). For comparison of sodium ISE responses incorporating either TMAUBC or NaTFPB as ion exchanger, the membranes also contained 20 mmol/kg sodium ionophore X and plasticizer DOS at the same compositions as stated above. Electrodes containing the TMAUBC ion exchanger (13) Lerchi, M.; Bakker, E.; Rusterholz, B.; Simon, W. Anal. Chem. 1992, 64, 1534. (14) Seiler, K.; Simon, W. Anal. Chim. Acta 1992, 266, 73. (15) Tsagkatakis, I.; Peper, S.; Bakker, E. Anal. Chem. 2001, 73, 315.
were preconditioned in 10-2 M LiOH for 4-6 h in order to deprotonate and extract trimethylammonium into the aqueous phase. An 8-h conditioning time in a 10-4 M CaCl2 solution followed for all electrodes. The internal filling solutions used were 10-3 M CaCl2 for electrodes containing the ion exchangers only and 10-3 M NaCl for Na-X-based electrodes. Selectivity measurements were all determined from calibration curves obtained in the 10-4-10-1 M range of the ions tested in the order of the most to least discriminated ion. Calculations were performed based on readings taken at the highest concentration level measured. All EMF measurements were made against a Ag/AgCl reference electrode (Metrohm 6.0729.100) with a 1 M LiOAc bridge electrolyte. The instrumentation used to acquire potentiometric data has been described earlier.16 Measurement values were corrected for liquid junction potentials using the Henderson formalism, and ion activities were calculated according to the Debye-Hu¨ckel approximation.17 RESULTS AND DISCUSSION It has been known for many years that decomposition of tetraphenylborates occurs via hydrolysis in acidic media.1,3,4 Experiments comparing the rate of decomposition of various tetraphenylborate derivatives was previously reported by Simon and co-workers.3 They found that the derivative with the slowest decomposition rate was the highly substituted NaTFPB. The presence of strong electron-withdrawing groups on this compound decreases the tendency for cleavage of the boron-phenyl bond, because the amount of localized charge at the ipso carbons is significantly reduced4 as compared to the unsubstituted analogue. Decomposition experiments comparing the chemical stability of TFPB- in acid with that of various carborane anions were done using optode films. By using an optode film that contains a lipophilic anion of interest and a chromoionophore with an acidic pKa value (ETH 5315, pKa ) 5.2 18), the rate of anion decomposition, leaching, or both can be determined. In contact with a weakly acidic solution, the countercation of the lipophilic ion exchanger must be the protonated chromoionophore to satisfy electroneutrality. Consequently, the initial absorption spectrum of the film indicates that the chromoionophore is fully protonated. As a cation exchanger leaves the film, either by decomposition or insufficient lipophilicity, a hydrogen ion must be abstracted from the chromoionophore and coextract with the anion. As the chromoionophore becomes deprotonated, there is a decrease in the absorption peak corresponding to the protonated form of the chromoionophore (518 nm), which is indicative of the leaching that is occurring. Initial studies evaluated the stability and lipophilicity of the commercially available anions, CB11H12- and CB11Br6H6-. It is expected from the fragment method for the estimation of octanolwater partition coefficients19 that the substitution of hydrogens by halogens increases the lipophilicity of a compound. Indeed, the unsubstituted carborane anion possessed insufficient lipophilicity and was not able to fully protonate the chromoionophore, although thick ISE membranes for sodium, potassium, and lead (16) Bakker, E. J. Electrochem. Soc. 1996, 143, L83. (17) Meier, P. C. Anal. Chim. Acta 1982, 136, 363. (18) Bakker, E.; Lerchi, M.; Rosatzin, T.; Rusterholz, B.; Simon, W. Anal. Chim. Acta 1993, 278, 211. (19) Hansch, C.; Leo, A. Substituent Constants for Correlation Analysis in Chemistry and Biology; John Wiley & Sons: New York, 1979.
Analytical Chemistry, Vol. 74, No. 6, March 15, 2002
1329
Figure 2. Acid decomposition of anionic sites in DOS-PVC (2:1) optode membranes containing ETH 5315, visualized by a decrease in absorbance of the protonated form at 518 nm.
containing this anion were functional (data not shown). The hexabrominated carborane anion, however, was able to fully protonate the chromoionophore, but after only 2 h, was complete leaching from the optode film observed (data not shown). This led to the search and synthesis for an even more lipophilic carborane anion, the perbrominated 1-H-CB11Br11-. A comparison of the leaching/decomposition rates of NaTFPB and TMAUBC is shown in Figure 2. The tetraphenylborate derivative is completely lost in ∼6 h, while in the same time period, the perbrominated carborane shows less than a 10% signal decrease. It is clearly demonstrated that 1-H-CB11Br11- has sufficient lipophilicity and chemical stability, as compared to NaTFPB. The rate of leaching for both anions appears to be slightly increased by changing the flow rate of the contacting acetic acid solution. This is illustrated in Figure 2 by the small rate increase upon initiation of continuous flow with 0.2 M acetic acid. The rate of decomposition of the NaTFPB was obtained repeatedly and differs from that previously reported in the literature.3 Furthermore, leaching experiments performed with the chromoionophore ETH 5418 also support the leaching data reported here for TFPB- (data not shown). The results shown in Figure 2 demonstrate that TMAUBC has superior lipophilicity/stability under the conditions shown than the best tetraphenylborate reported to date. Nonetheless, even more lipophilic carborane derivatives are still desired because TMAUBC exhibited an approximate 10% decrease in absorbance of the protonation peak of ETH 5315, demonstrating that some loss does still occur. Fortunately, because of the versatile functionalization chemistry of carboranes, lipophilicity can easily be tailored by substitution with heavier halogen substituents such as iodine9,12 or with alkyl groups10 at the boron vertexes. Furthermore, substitution may also occur at the carbon vertex. Many substituted alkyl and phenyl derivatives have been described,11,12 and exploration of some of these compounds is in progress in our laboratory. In addition to their use in the leaching studies, optodes were used to evaluate the ion-exchanging capabilities of UBC- in potassium-selective optical thin films. The normalized response 1330 Analytical Chemistry, Vol. 74, No. 6, March 15, 2002
Figure 3. Normalized optical K+ response curves and selectivities over sodium and calcium for thin DOS-PVC (2:1) films containing the potassium ionophore BME-44, the H+-chromoionophore ETH 5294, and either TMAUBC (top) or NaTFPB (bottom) as lipophilic cation exchangers. The y-axis denotes the mole fraction of protonated chromoionophore (Ind).
curve of UBC--containing films for potassium appears as Figure 3 and is compared to the analogous behavior of TFPB--containing optode films. The potassium activity in a given sample was determined by using the fluorescence emission of the neutral chromoionophore, ETH 5294. Peaks at 655 and 693 nm, corresponding to the deprotonated and protonated forms of the chromoionophore, respectively, were used to monitor ion exchange. As potassium ions enter the film, the chromoionophore is deprotonated to maintain electroneutrality; this corresponds to a decrease in fluorescence intensity at 693 nm and, conversely, an increase in the deprotonated emission peak at 655 nm. By taking the ratio of these two emission peaks, a response curve may be generated. The use of ratiometric fluorescence measurements to normalize response curves in terms of the degree of protonation of the chromoionophore has been described previously.15 The theoretical response curve used to fit the data in Figure 3 has been reported repeatedly.15,20 The locations of the response curves and the 3 orders of magnitude measuring ranges shown in Figure 3 are nearly identical for films containing both cation exchangers. The determined logarithmic ion-exchange constants, according to eq 3 in ref 15, were -4.56 for TFPB- and -4.15 for UBC-. UBC-(20) Kurihara, K.; Ohtsu, M.; Yoshida, T.; Abe, T.; Hisamoto, H.; Suzuki, K. Anal. Chem. 1999, 71, 3558.
Table 1. Potentiometric Selectivity Coefficients and Electrode Slopes for Ion-Selective Membranes Containing the Ion Exchangers TFPB- and UBCa a,b for DOS-PVC (2:1) log Kpot Na,J and electrode slopes
ion Jz+ Li+ Na+ K+ Mg2+ Ca2+ a
TFPB-0.27 ( 0.02 0 0.55 ( 0.02 -2.37 ( 0.08 -2.0 ( 0.1
b a,b for NPOE-PVC (2:1) log Kpot Na,J and electrode slopes
UBC58.4 ( 0.5 58.9 ( 0.7 59.1 ( 0.6 24 ( 3 28 ( 1
-0.14 ( 0.01 0 0.45 ( 0.01 -2.91 ( 0.03 -2.57 ( 0.04
TFPB57.6 ( 0.2 56.5 ( 0.1 56.6 ( 0.3 22.7 ( 0.7 29 ( 2
-0.6 ( 0.1 0 1.3 ( 0.2 -1.50 ( 0.06 -0.99 ( 0.09
UBC47 ( 2 51.2 ( 0.7 57.5 ( 0.2 28.1 ( 0.7 29 ( 2
-0.61 ( 0.08 0 1.21 ( 0.01 -1.7 ( 0.1 -1.6 ( 0.1
31 ( 2 57.6 ( 0.3 49 ( 2 14 ( 1 20 ( 2
Average from three electrodes; standard deviation given. b Slope (second column for each data set): between log a ) -3.1 and -1.
Figure 4. Potentiometric selectivity comparison of ionophore-free DOS-PVC (2:1) membranes containing 10 mmol kg-1 of either TMAUBC or NaTFPB. Solid lines with theoretical slope.
containing films appear to enhance the competitive extraction of potassium relative to hydrogen ions. Moreover, the selectivity over sodium appears to be improved for the UBC--containing system by ∼0.2 logarithmic unit. These differences are perhaps too small to merit interpretation at this stage and should be followed by more thorough experiments that assess the ion pair formation constants of these structurally different cation exchangers. In addition to optical experiments, the selectivity behavior of the perbrominated carborane was evaluated potentiometrically using ionophore-free ISE membranes. As expected, all membranes show a selectivity pattern in order of increasing cation lipophilicity. In PVC-DOS membranes containing only UBC-, the selectivity observed was nearly identical to PVC-DOS membranes containing only NaTFPB (see Table 1 and Figure 4), and the electrode slopes were close to Nernstian (see Table 1). The selectivity over magnesium in PVC-DOS, where the largest discrepancies were found for both ion exchangers, may be an artifact due to the observed sub-Nernstian slopes. Overall, the data suggest that the UBC- anion behaves similarly to TFPB- in terms of ionexchanging ability. In addition to the selectivity evaluation of 1-H-CB11Br11- in simple ion-exchanger membranes, selectivity was also examined in membranes containing a sodium ionophore. Figure 5 compares the potentiometric responses of PVC-DOS membranes with sodium ionophore X and either UBC- or TFPB- as ion exchanger. As shown in Table 2, both electrodes exhibit near-Nernstian slopes
Table 2. Potentiometric Selectivity Coefficients and Electrode Slopes for DOS-PVC (2:1) Ion-Selective Membranes Containing the Sodium Ionophore Na-X and the Ion Exchangers TFPB- and UBCa a,b log Kpot Na,J and electrode slopes
ion Jz+ Li+ Na+ K+ Mg2+ Ca2+
TFPB-3.85 ( 0.05 0 -2.80 ( 0.01 -4.32 ( 0.05 -4.25 ( 0.03
UBC25 ( 1 57.8 ( 0.2 43 ( 2 24.9 ( 0.4 30 ( 2
-3.8 ( 0.2 0 -2.6 ( 0.2 -4.3 ( 0.2 -4.1 ( 0.2
30 ( 2 56 ( 3 48.9 ( 0.3 31 ( 5 33 ( 9
a Average from three electrodes; standard deviation given. b Slope (second column for each data set): between log a ) -3.1 and -1.
for Na+ and have nearly identical selectivity coefficients for the discriminating ions tested. Electrode slopes for a number of interfering ions are sub-Nernstian, which indicates that the values found here are conservative upper limits for those ions.16 CONCLUSIONS The perbrominated closo-dodecacarborane anion, 1-H-CB11Br11-, has been evaluated as an improved ion exchanger in ionselective chemical sensors. Superior chemical stability, high lipophilicity, and versatile functionalization chemistry make dodecacarboranes a promising alternative to even the best substiAnalytical Chemistry, Vol. 74, No. 6, March 15, 2002
1331
Figure 5. Potentiometric selectivity comparison of DOS-PVC (2:1) membranes containing sodium ionophore X and either TMAUBC or NaTFPB as ionic additives. Solid lines with theoretical slope.
tuted tetraphenylborates known, such as TFPB-. In addition, lack of absorption of ultraviolet and visible wavelengths make carboranes attractive for optical sensing applications. Selectivity comparable to TFPB- was achieved by the perbrominated carborane in ISE membranes both with and without a sodium-selective ionophore. In bulk optode films containing the potassium-selective ionophore BME-44 and UBC-, the normalized potassium response, the 3 orders of magnitude measuring range, and the selectivity of Na+ and Ca2+ were similar to that observed for films containing TFPB-. These results indicate that important sensing characteristics such as lifetime and signal stability can be significantly improved by introducing this new class of anionic additives. The synthesis of more electron-withdrawing dodeca-
1332
Analytical Chemistry, Vol. 74, No. 6, March 15, 2002
carborane derivatives with even higher lipophilicity, along with a more detailed characterization of the ion pairing and lipophilicity characteristics, is currently under investigation in our group. ACKNOWLEDGMENT The authors gratefully acknowledge financial support from the National Institutes of Health (Grants GM59716 and GM58589) and Beckman-Coulter, Inc., for this research.
Received for review September 14, 2001. Accepted January 16, 2002. AC0156262