Cyanostar: C–H Hydrogen Bonding Neutral Carrier Scaffold for Anion

Jan 22, 2018 - (22, 23) Various functional groups can be incorporated into the ionophore for anion binding, such as transition metal Lewis acids, dipo...
0 downloads 5 Views 2MB Size
Download PDF
Article Cite This: Anal. Chem. XXXX, XXX, XXX−XXX

pubs.acs.org/ac

Cyanostar: C−H Hydrogen Bonding Neutral Carrier Scaffold for Anion-Selective Sensors Elsayed M. Zahran,†,§ Elisabeth M. Fatila,‡ Chun-Hsing Chen,‡ Amar H. Flood,‡ and Leonidas G. Bachas*,† †

Department of Chemistry, University of Miami, Coral Gables, Florida 33126, United States Department of Chemistry, Indiana University, 800 East Kirkwood Avenue, Bloomington, Indiana 47405, United States § Applied Organic Chemistry Department, National Research Centre, Cairo, 12622, Egypt ‡

S Supporting Information *

ABSTRACT: Cyanostar, a pentagonal macrocyclic compound with an electropositive cavity, binds anions with CH-based hydrogen bonding. The large size of the cyanostar’s cavity along with its planarity favor formation of 2:1 sandwich complexes with larger anions, like perchlorate, ClO4−, relative to the smaller chloride. We also show that cyanostar is selective for ClO4− over the bulky salicylate anions by using NMR titration studies to measure affinity. The performance of this novel macrocycle as an anion ionophore in membrane ion sensors was evaluated. The cyanostar-based electrodes demonstrated a Nernstian response toward perchlorate with selectivity patterns distinctly different from the normal Hofmeister series. Different membrane compositions were explored to identify the optimum concentrations of the ionophore, plasticizer, and lipophilic additive that give rise to the best perchlorate selectivity. Changing the concentration of the lipophilic additive tridodecylmethylammonium chloride was found to impact the selectivity pattern and the analytical dynamic range of the electrodes. The high selectivity of the cyanostar sensors and their detection limit could enable the determination of ClO4− in contaminated environmental samples. This novel class of macrocycle provides a suitable scaffold for designing various anion-selective ionophores by altering the size of the central cavity and its functionalization.

M

ionophores,20 the design of ionophores with high degrees of anion selectivity remains a formidable challenge. Anions have broader structural diversity, larger sizes with relatively low charge densities, and often higher hydration energies than cations.21 The design of anion-selective ionophores commonly follows standard recognition principles of complementarity.22,23 Various functional groups can be incorporated into the ionophore for anion binding, such as transition metal Lewis acids, dipoles, Coulombic contacts, halogen bonding, and hydrogen bonding.22,24,25 Receptors bearing hydrogen-bonding sites retain much interest on account of their versatility. Generally, the NH hydrogen-bond donors employed in anion receptors are based on amines, guanidinium cations, pyrroles, ureas, and amides. Although routinely exploited in the fabrication of anion-selective sensors,26,27 some of the NH donors can become protonated, which might affect the selectivity of the electrode at various pHs.28−30 Furthermore,

embrane-based ion-selective sensors are well established for routine analysis of several analytes in clinical, industrial, and environmental fields. In recent years, attention has been directed to nonequilibrium approaches to improve the detection limits of such ion sensors.1−10 Additional methodologies have been investigated in the last two decades to lower detection limits and enhance selectivity. These include controlling the ion flux across the membrane by adjusting the inner filling solution, optimizing the matrix of the sensing membrane, attaching the ionophore to the polymer backbone, and designing novel, highly selective ionophores.11−18 Significant progress in these areas has produced sensors with nanomolar and picomolar detection limits.13,19 These advances lay foundations for simple analytical tools to replace highly sophisticated techniques, such as, inductively coupled plasmamass spectroscopy, atomic absorption spectroscopy, and X-ray fluorescence spectroscopy. Underpinning both the conventional and nonequilibrium approaches to membrane sensors are ionophores designed to confer optimized binding strengths, structural selectivities, and enhanced compatibility with the membrane medium. As such, ionophores remain a cornerstone of advanced membrane ion sensors. Despite the maturity of the design of cation-selective © XXXX American Chemical Society

Received: September 29, 2017 Accepted: December 29, 2017

A

DOI: 10.1021/acs.analchem.7b04008 Anal. Chem. XXXX, XXX, XXX−XXX

Article

Analytical Chemistry

Figure 1. (a) The cyanostar (CNStar) macrocycle and (b) its preferred binding as a 2:1 sandwich complex with anions (X−). (c) The Hofmeister series. (d) Selectivity preferences of cyanostar in 60/40 v/v % dichloromethane/methanol (part of this figure has been adapted from ref 39. Copyright 2013 Flood and co-workers).

macrocycle to form π-stacked 2:1 “sandwich” complexes (Figure 1b). The central cavity of the two cyanostars is lined with 20 CH subunits that hydrogen bond to the bound anion. The macrocycle displayed high perchlorate affinities that were, at the time, unprecedented. Perchlorate affinities are not routinely measured (presumably a result of being too low). However, since the discovery that the large cavities in cyanostar can host such large anions, two other classes of charge-neutral ionophores have shown strong solution affinities for perchlorate, the bambusurils44 from Sindelar and co-workers and the tricarbazole triazolophanes45 from Flood and co-workers. Here, we introduce cyanostar as a new hydrogen-bond donating macrocyclic ionophore (CNStar, Figure 1a) incorporated into a simple ISE to confer enhanced detection limits and selectivity toward perchlorate. The performance of the cyanostar ionophore was evaluated in the ISE platform and found to display enhanced limits of detection and selectivity for perchlorate. Excellent selectivity can be produced relative to one of the most lipophilic anions, salicylate, based on the ionophore’s extremely poor binding of this competing anion. Depending on membrane composition, the intrinsic characteristics of the cyanostar can be combined with those of the membrane to produce an ISE with optimal performance for perchlorate sensing in aqueous solutions that simulate waters contaminated with perchlorate. The latter can be found in rivers, lakes, underground water, and sediments close to industrial sites and military training facilities.46 Thus, these ISEs offer a simple alternative for detecting perchlorate in realworld situations.

several NH-based ionophores experience aggregation and competition for hydrogen bond donor sites.31 Use of CH hydrogen bonding circumvents some of these problems; e.g., they are less sensitive to pH, but their use has been limited to only a few anion receptors.32,33 The introduction of macrocyclic triazolophane receptors18,34,35 helped spur interest in building anion-selective ionophores based solely on CH functionalities. The key success of the design of these receptors depends on preorganizing multiple, extrinsically activated CH donors inside the central cavity of a macrocycle to provide affinity and define size complementarity for the target anion.36 The parent triazolophane demonstrated high selectivity for chloride and bromide over common interfering anions in organic solutions, and this functionality transferred well to a membrane selective electrode platform for sensing chloride in aqueous solutions.35,37 Designing ionophores for bulky and weakly coordinating anions has been challenging because of the complexity of synthesizing larger size receptors that have complementary binding functionalities. As a result, ion-selective electrodes for bulky anions have mostly been derived from an ion-exchange mechanism, where a lipophilic anion exchanger (e.g., a quaternary ammonium salt) enables transfer of the anion into the membrane phase. In such electrodes, the selectivity is determined by the relative lipophilicity of the anion and follows what is commonly known as the Hofmeister series (Figure 1c). Herein, we explore a new ionophore design principle that circumvents the ion-exchange mechanism and improves on the Hofmeister selectivity. We chose perchlorate, a weakly coordinating anion,38 to demonstrate this principle. To achieve that, we designed receptors that have appropriately positioned hydrogen-bonding sites and are able to form “sandwich-type” complexes with anions. We avoid positively charged functionalities in order to reduce ion-exchange mechanisms. Cyanostar (CNStar), a new star-shaped macrocycle39 (Figure 1), was prepared and demonstrated to have size-selective binding (Figure 1) toward ClO4−.40−42 The electron deficient nature of the cyanostilbenes and the macrocycle’s shape persistence43 enhance the binding to anions that have relatively low electron density, like ClO4−. Furthermore, the toroid (shallow bowl) shape of the cyanostar rigid structure along with the π−π interactions between the two molecules allow the

■

EXPERIMENTAL SECTION Reagents. The cyanostar ionophore, C5-symmetric penta-tbutylpentacyanopentabenzo[25]annulene macrocycle, was synthesized and purified according to published procedures.39,47 All chemicals used for the preparation of ion-selective membranes, including 2-nitrophenyl octyl ether (NPOE), bis(2-ethylhexyl) sebacate (DOS), tridodecylmethylammonium chloride (TDMACl), and poly(vinyl chloride) (PVC), were purchased in Selectophore grade from Fluka (Ronkonkoma, NY). High purity sodium perchlorate, sodium chloride, sodium bromide, sodium nitrate, sodium salicylate, sodium iodide, sodium thiocyanate, sodium bicarbonate, and sodium acetate B

DOI: 10.1021/acs.analchem.7b04008 Anal. Chem. XXXX, XXX, XXX−XXX

Article

Analytical Chemistry

Figure 2. (a) Top view of the crystal structure showing the salicylate threading inside [(TBA)CNStar(salicylate)(H2O)2(salicylate)CNStar(TBA)] and (b) side view (tert-butyl substituents removed for clarity); CCDC #1573051. (c) Structural views showing H2O and salicylate bonding residing in the core of the cyanostar dimer.

Membranes were conditioned in 1 × 10−2 M solution of the anion overnight before evaluating the potentiometric performance. Potentiometric Measurements. All potentiometric measurements were performed at 22 ± 0.5 °C using the following cell assembly: Ag/AgCl|3.0 M KCl∥1.0 M CH3COOLi∥sample solution|ISE membrane|1 × 10−3 M NaCl|Ag/AgCl. The cell potential was measured using an 8-channel SC-2345 interface (SCC-LP01, National Instruments, Austin, TX). This interface was connected to a PC computer using a PCI-6036E multifunction I/O data acquisition device. A custom-designed data acquisition script written in Lab VIEW 7.0 (National Instruments) software was used to record the EMF measurements. A Fisher Scientific Accumet 915 pH/Ion meter, which was equipped with an Orion pH glass electrode (Beverly, MA), was used throughout for pH studies and buffer preparations. The effect of different buffers was evaluated by monitoring the change in the electrode’s potential with incremental additions of the buffer solution to deionized water. The selectivity coefficients were determined by the separate solutions method according to the conditioning procedure described by Bakker et al.49 The cyanostar-based electrode was used for determination of perchlorate concentrations in horse serum, tap water, and natural water adjusted to pH 7.4 using 0.5 M HEPES buffer.

were obtained from Sigma-Aldrich (St. Louis, MO). The tetrabutylammonium salicylate used for anion binding studies was purchased from TCI America and used as received. The buffers N-(2-hydroxyethyl)piperazine-N-(2-ethanesulfonic acid) (HEPES) along with tetrahydrofuran (THF) were also purchased from Sigma-Aldrich. Serum from a donor New Zealand horse that was tested (negative) by enzyme immunoassay (EIA) was purchased from Invitrogen (Carlsbad, CA). Natural water samples were collected from Lake Osceola, a local reservoir at the University of Miami, Coral Gables, Florida. Tap water was collected from the laboratory at the University of Miami. Deionized water of 18.2 MΩ (E-Pure water purification system; Thermo Scientific, Marietta, OH) was used for all solution preparations and measurements unless mentioned otherwise. Solution Binding Studies of Cyanostar with the Salicylate Anion. Binding studies of cyanostar with tetrabutylammonium salicylate were conducted using 1H NMR spectroscopy. 1H NMR and diffusion studies were recorded on Varian INOVA (500 and 600 MHz) spectrometers. Deuterated solvents (CD2Cl2, CD3OD) were purchased from Cambridge Isotope Laboratories (Tewksbury, MA) and used without further purification. High-resolution electrospray mass spectrometry (ESI-MS) analyses were performed on a Waters Synapt Time-of-Flight mass spectrometer. Calibration was performed using sodium trifluoroacetate. Spectroscopic grade dichloromethane and methanol (Omnisolv, EMD Millipore, Billerica, MA) were used for ESI-MS studies. Membranes and Electrode Preparation. The PVCbased ion-selective electrodes were prepared according to established procedures reported elsewhere.48 Briefly, specific amounts of cyanostar, TDMACl, and either NPOE or DOS plasticizers were dissolved in 1 mL of THF to prepare membrane cocktails. After sonicating the solution for 30 min to dissolve all the material, the cocktail was left at rest to release any air bubbles. The PVC membrane was then prepared by pouring the solution into a 22 mm-i.d. glass ring located on a glass slide and allowing the THF to evaporate slowly overnight at room temperature. Subsequently, the as-formed 150−200 μm thick membrane was cut into small disks of 7 mm diameter using a cork borer to install onto Philips IS-561 (Gläsblaserei Möller, Zurich, Switzerland) electrode bodies. A solution of 1 × 10−3 M NaCl was used as the internal filling solution.

■

RESULTS AND DISCUSSION Designing ionophores for large anions is challenging because of the need to synthesize larger size receptors with preorganized functionalities that can bind the anion. We report here a strategy to accomplish this that involves the use of cyanostar, a toroid-shaped macrocycle with CH hydrogen bonding sites pointing to the interior of the macrocycle. The design allows for the formation of π-stacked 2:1 “sandwich” complexes with anions through π−π interactions between two cyanostar molecules. We evaluate this methodology using perchlorate as a model anion. We hypothesize that the strong binding affinity of cyanostar toward perchlorate in addition to its compatibility with the membrane environment make it a good neutral carrier ionophore for membrane-based ion sensors. The selectivity of the 2:1 sandwich complexes seen in a solvent mixture of methanol and dichloromethane (Figure 1d) suggests a selectivity series for the anions of interest running from highest to lowest: ClO4− ≫ I− > SCN− ≫ NO3− > Br− ≫ Cl−. Given C

DOI: 10.1021/acs.analchem.7b04008 Anal. Chem. XXXX, XXX, XXX−XXX

Article

Analytical Chemistry

Figure 3. (a) 1H NMR titration data of CNStar (1 mM) with tetrabutylammonium salicylate in 60/40 v/v CD2Cl2/CD3OD (298 K, 600 MHz). (b) The structure of the cyanostar and (c) saturation plot demonstrating the effect of salicylate binding on cyanostar 1H NMR chemical shifts in (a).

quantify binding for other anions (Figure 1d).39 A plot of the peak shifts of the cyanostar protons (Ha, Hb, Hc) shows that saturation requires ∼20 equiv. of the added anion. For proton Hd, we see an inflection point with 4 equiv. of added anion. For comparison, the ClO4− anion saturates at 0.5 equiv. as expected for direct and highly cooperative formation of a 2:1 sandwich complex39 with strong binding affinity (β2 ∼ 1011 M−2). Diffusion data (Figure S2) using all aromatic protons were found to corroborate the 1H NMR data. That is, after the addition of 0.5 equiv. of the salicylate salt, we only observed a small decrease in diffusion for the cyanostar species (diffusion coefficient shifts from 5.9 to 5.8 × 10−10 m2 s−1). This finding is indicative of a 1:1 complex dominating in solution rather than formation of larger 2:1 sandwich complexes that normally show a sizable decrease in the diffusion coefficient down to ∼5 × 10−10 m2 s−1 (298 K; in 60/40 v/v CD2Cl2/CD3OD, with 0.5 equiv. of anion).55 Analysis of the solution species by electrospray ionization mass spectrometry (2 mM cyanostar, Figure S1) showed a complex spectrum when 1 equiv. or less of salicylate was added. The two major species observed were the 2:1 and 1:1 complexes. With 4 equiv. of salicylate used, the spectrum simplifies to show the 1:1 species dominating. Overall, the data demonstrate successful fitting using HypNMR to a model showing a weak 2:1 and 1:1 complex.

that thiocyanate and nitrate typically exhibit strong responses in conventional ion-selective electrodes based on ion exchangers,50 the intrinsic selectivity of cyanostar is promising. The salicylate anion was not among the anions originally investigated for binding to cyanostar,39 though it is a model drug that commonly interferes with anion measurements in biological samples. For this reason, we undertook the first characterization of the binding of salicylate to cyanostar in solution and in the solid state. Weak Binding of Salicylate to Cyanostar. The solidstate structure of cyanostar with salicylate (Figure 2, Table S1), present as the tetrabutylammonium (TBA+) salt, offers clues to its solution behavior. The X-ray quality crystals of [(CNStar)(salicylate)2(H2O)2(TBA)2] were grown by slow diffusion of diethyl ether into a saturated solution of cyanostar with 1 equiv. of TBAalicylate in chloroform. The crystal structure of cyanostar with the salicylate anion shows the inclusion of two water molecules that bridge the two salicylate anions inside a pair of π-stacked cyanostars. The crystal structure differs significantly from the 2:1 structure of cyanostar with perchlorate40 and all other anions.39,51−54 One of the major differences to the crystal structure of [(CNStar)2(ClO4)(TBA)], which shows perchlorate held inside the central cavity of cyanostar dimers, is that the salicylate anions are not held inside the cyanostar cavity. Rather, the salicylate anions perch outside of the cyanostar’s mean plane, and the carboxylate functional groups are almost flush with the cyanostar’s mean plane (Figure 2b). Also unique is the stoichiometry. We have previously seen 2:2 stoichiometries for two cyanostars with either two bisulfate anions42,55 or two organophosphates.54 This is the first structure, however, with bridging waters inside the stacked pair of cyanostars, although the complexation of water is not without precedence with other receptors.56−58 On the basis of this solid-state structure, the solution-phase affinity of the salicylate anion for the cyanostar was expected to be weak. To evaluate cyanostar’s salicylate affinity, we conducted a 1H NMR titration (Figure 3, 1 mM cyanostar) in 60/40 v/v CD2Cl2/CD3OD to match the original conditions used to

K1

CNStar + sal− ⇌ CNStar·sal−

(1)

K2

CNStar·sal− + CNStar ⇌ (CNStar)2 ·sal−

(2)

Our estimate of the binding constant, for formation of the 1:1 complex, is 200 M−1 (log K1:1 = 2.3 ± 0.1) and, for the 2:1 complex, is 200 M−1 (log K2:1 = 2.3 ± 0.2). Thus, even though salicylate is a lipophilic anion that might prefer to partition out of water, its cyanostar affinity is many orders of magnitude lower than the one for ClO4−. Cyanostar-Based Ion-Selective Electrodes. To study the performance of the cyanostar macrocycle as an anion-selective ionophore, PVC membranes containing various concentrations of cyanostar and the lipophilic salt tridodecylmethylammonium D

DOI: 10.1021/acs.analchem.7b04008 Anal. Chem. XXXX, XXX, XXX−XXX

Article

Analytical Chemistry Table 1. Composition and Potentiometric Response of Cyanostar-Based Membranes CNStar (wt %) I II III IV V VI VII VIII IX X XI XII a

2.0 2.0 2.0 2.0 1.0 1.0 1.0 2.0 2.0 1.0 1.0 0.0

TDMACl (mol %) 0.0 25.0 50.0 75.0 25.0 50.0 75.0 25.0 50.0 50.0 75.0 1.0

plasticizer (wt %)

slope ClO4− (mV/decade) −38.2 −56.7 −54.1 −56.0 −50.6 −57.0 −55.6 −56.5 −53.6 −51.1 −53.2 −55.0

NPOE, 66.0 NPOE, 65.1 NPOE, 64.9 NPOE, 65.7 NPOE, 65.9 NPOE, 65.8 NPOE, 65.7 DOS, 65.2 DOS, 64.9 DOS, 65.9 DOS, 65.7 NPOE, 66.0

detection limita (M)

linear range (M) 6.3 6.3 2.0 5.0 6.3 1.0 1.0 1.6 6.3 1.0 4.0 5.0

× × × × × × × × × × × ×

−6

10 −1.6 10−7−1.0 10−7−1.6 10−7−1.6 10−6−1.0 10−7−1.0 10−6−1.0 10−6−6.3 10−7−1.0 10−5−1.0 10−6−1.0 10−6−1.0

× × × × × × × × × × × ×

−3

10 10−2 10−1 10−1 10−2 10−1 10−1 10−3 10−1 10−1 10−1 10−1

3.2 2.0 5.0 2.0 6.3 5.0 5.0 8.3 2.0 1.0 8.0 1.0

× × × × × × × × × × × ×

10−6 10−7 10−8 10−7 10−7 10−8 10−7 10−7 10−7 10−6 10−7 10−6

Detection limits were determined in unbuffered solution.

the ISE using membrane I, which has no lipophilic positive sites (i.e., TDMACl) added. Lipophilic cationic sites, like quaternary ammonium salts, are typically added to membranes composed of neutral carriers to confer multiple benefits. These include initiating permselectivity toward anions, improving sensitivity, and dissipating the charges of the anion−ionophore complex at high analyte concentrations.63,64 Therefore, to bring the slope of the cyanostar-based electrodes to a Nernstian response and improve the detection limit, membranes bearing increasing amounts of TDMACl (Table 1) were prepared and their performance toward various anions was evaluated (Figure S4). The inclusion of just 25 mol % of TDMACl (relative to cyanostar) in membrane II enhanced the slope seen originally with electrode I to −56.7 mV per decade toward ClO4− over a linear range of 6.3 × 10−7−1.0 × 10−2 M with submicromolar detection limits. Although ClO4− is a highly lipophilic anion and generally situated at the top of the Hofmeister series, electrode II demonstrated an exceptional selectivity pattern (Figures 4 and

chloride (TDMAC1) were prepared using NPOE or DOS as plasticizers (Table 1). It should be noted that membranes prepared with only plasticizer and PVC demonstrate a weak cationic response because of residual negatively charged sites that are present in PVC due to the manufacturing process.59,60 Therefore, if the addition of an ionophore to such a membrane produces an anionic response, this would be an indication that the ionophore facilitates the transfer of anions from the solution phase to the membrane. Accordingly and for a preliminary evaluation of the potentiometric performance of the ionophore toward various anions, membrane I was prepared with 2 wt % of CNStar and NPOE plasticizer but with the ion exchanger TDMACl omitted. Generally, ion exchangers like TDMACl are added in equimolar ratios relative to the ionophore to initiate permselectivity toward anions.61 However, for strong binding ionophores, a weak anionic response can be achieved even in the absence of ion exchangers.62 Consistent with the large binding affinity of cyanostar, electrodes based on membrane I displayed an anionic response toward ClO4−, I−, and SCN− with higher selectivity toward ClO4− (Figure S3). Electrode I containing 0 mol % TDMACl demonstrated an anionic slope of −38.2 mV/decade toward ClO4− (Figure S3) over the linear range of 6.3 × 10−6−1.6 × 10−3 M. The electrode displayed an anionic response of −27.1 and −28.2 mV/decade toward iodide and thiocyanate, respectively. The anionic response of electrode I can be explained in light of the size, structure, and electron density of these anions. These results are consistent with the solution-phase titration data that showed I− and SCN− demonstrate the next highest binding affinity to cyanostar after ClO4− (Figure 1c).39 Although the high lipophilicity of these anions improves their extraction into the membrane phase, the anionic response comes as a result of the selective complexation by the ionophore. Electrode I demonstrated a cationic response toward all other common interfering anions (Figure S3). This observation rules out the role of the ClO4− anion’s lipophilicity as a governing driving force for the anionic response seen with electrode I. Consistently, the high discrimination ability of electrode I toward the lipophilic anion salicylate, log Kpot ClO4−,sal−= −6.80, also supports this conclusion. This performance is consistent with the NMR titrations and ESI-MS studies (vide supra). All these observations indicate that the response toward ClO4− derives from cyanostar’s high perchlorate affinity. In this case, the electron deficient cavity of cyanostar with its multiple CH units is responsible for the distinctive anionic response of

Figure 4. Response of electrode II (2 wt % CNStar, 25 mol % TDMACl) toward various anions.

5) that is distinctively different from ones that depend on the Gibbs free energy of hydration. It is evident from the comparison of the response of electrode II and the TDMACl-only membrane (electrode XII, Table 1) that the presence of cyanostar enhanced the selectivity toward ClO4− over other anions by at about 2 orders of magnitude or better. Furthermore, salicylate demonstrated an insignificant interference for electrode II with log Kpot ClO4−,sal− = −6.50 (Figure 6). Salicylate usually represents a strong interference for anionselective electrodes based solely on ion exchangers like E

DOI: 10.1021/acs.analchem.7b04008 Anal. Chem. XXXX, XXX, XXX−XXX

Article

Analytical Chemistry

improvement in electrode performance was observed by increasing the TDMACl content further to 75 mol % (electrode IV). Considering the use of strong ionophores in the phase boundary potential model, the response of the membranebased carriers in the electrode will be governed by the critical concentration ratio of the anion−ionophore complex and the lipophilic ionic sites.18,20 Therefore, decreasing the concentration of cyanostar in the membrane might lower the ClO4− detection limit by reducing the ion flux effect.66 To test this idea, we prepared membranes with 1 wt % of the cyanostar and various amounts of TDMACl (V, VI, and VII, Table 1) and evaluated their potentiometric performance (Figure S5). Membrane VI containing 1 wt % of cyanostar and 50 mol % of TDMACl demonstrated selectivity toward ClO4− with a slope of −57.0 mV per decade over the widest linear range of 1.0 × 10−7−1.0 × 10−1 M and a detection limit of 50 nM. This improved performance relative to membrane III containing 1 wt % can be attributed to increasingly balanced membrane electroneutrality and reduced coextraction of the counterion. The selectivity of membrane VI toward ClO4− over other anions is 1-to-2 orders of magnitude higher than the ion exchanger-based electrode XII. The selectivity of electrode VI is not the best among the membranes; rather, electrodes II (2% CNStar, Figure 5) and V (1% CNStar) containing 25 mol % of TDMACl demonstrated superior selectivity patterns. We note that, with lower amounts of TDMACl, 25 mol %, and more cyanostar (2%) (Figure 5) superior selectivity patterns are seen; pot pot e.g., log KClO = −6.50 and log KClO = −5.54. − − − − 4 ,sal 4 ,Cl Nevertheless, the improvement in the detection limit seen with electrode VI is very desirable for the determination of perchlorate in environmental samples. In this case, the compromise between selectivity and detection limit will ultimately depend on sample matrix. For neutral ionophores with pure CH hydrogen bonding, the dielectric constant of the plasticizer exerts a significant influence on the phase transfer of anions into the membrane.37,67 Therefore, to investigate the effect of the dielectric constant of the plasticizer on the performance of cyanostar-based electrodes, various membranes were prepared using DOS (ε = 4.8), which has significantly lower dielectric constant compared to NPOE (ε = 24.2).68,69 Electrodes plasticized with DOS (Figure S6) demonstrated lower slopes and shorter linear ranges than those plasticized with NPOE. For instance, electrode VIII containing 2 wt % CNStar and 25 mol % TDMACl with DOS exhibited strong Donnan failure at just 1.0 × 10−3 M ClO4−. This performance was improved by increasing the TDMACl concentration to 50 mol %, yet comparable electrodes plasticized with NPOE still showed an enhanced response. For instance, electrodes III and VI plasticized with NPOE demonstrated higher slopes over a wider linear range with improved detection limits (gray lines, Figure S6) in comparison to electrodes IX and X plasticized with DOS (red lines, Figure S6). However, the selectivity of electrodes plasticized with DOS toward the perchlorate anion over other lipophilic anions, like thiocyanate and salicylate, was improved by at least 2 orders of magnitude (compare electrodes III and IX, Figure 5). The type of plasticizer used might alter the ionophore/ion binding stoichiometry in the membrane. NPOE-plasticized membranes that have a dielectric constant that is closer to the methanol/ dichloromethane solvent used for the NMR titrations of CNStar demonstrated enhanced performance over the DOS-

Figure 5. Selectivity pattern of various CNStar-based electrodes toward ClO4− and common interfering anions compared to the selectivity pattern of a TDMACl-based electrode. Electrodes arranged in this order to demonstrate the comparison in the selectivity pattern between 1 wt % CNStar (V, VI, and VII) and 2 wt % CNStar (II, III, and IV). The effect of increasing the TDMACl concentration is manifested in each group. The selectivity performance of electrode IX plasticized with DOS is compared to all other electrodes based on NPOE.

Figure 6. Calibration of electrode VI toward ClO4− in different media: (⧫) 0.01 M HEPES buffer, pH 7.4, (■) tap water, (▲) lake water, and (●) horse serum.

TDMACl.65 This interference usually causes overestimation of clinically important anions for patients with high doses of aspirin because salicylate is a direct metabolite of aspirin. Despite its selectivity, electrode II displayed a slight memory effect and a Donnan exclusion failure at perchlorate concentrations higher than 1.0 × 10−2 M (Figure S4). Consequently, we increased the amount of TDMACl to 50 mol % (electrode III) to enhance the phase exchange and balance the charges of the anion−ionophore complex at high ClO4− concentrations. Consistently, electrode III demonstrated a wider linear range of 2.0 × 10−7−1.6 × 10−1 M with improved detection limit toward ClO4− of 50 nM (Table 1). Although this increase in TDMACl concentration also enhanced the response toward all the anions, the selectivity pattern was superior to a membrane based on the ion exchanger alone (electrode XII, Figure 5). For instance, the selectivity for ClO4− over Cl− seen in membrane XII was improved by 1 order of magnitude in membrane III to log Kpot ClO4−,Cl− = −5.54. No further F

DOI: 10.1021/acs.analchem.7b04008 Anal. Chem. XXXX, XXX, XXX−XXX

Article

Analytical Chemistry

that are very close to the pure 0.01 M HEPES buffer. Furthermore, the cyanostar-based electrodes in this study exhibited minimal elevation in the detection limit of less than 1 order of magnitude when compared to the measurements made in pure buffer. This outcome is attributed to the high selectivity of the electrode and suggests that these cyanostar-based electrodes can be used for the determination of ClO4− in various types of samples.

plasticized ones that have low dielectric constant. Taking these two effects together, the more lipophilic DOS and the effect of increasing the concentration of TDMACl, we conclude that the response of the cyanostar electrodes toward ClO4− results from the strong binding affinity of ClO4− to cyanostar. The response time was determined according to the IUPAC recommendation from the plot of electromotive force (EMF) versus time and establishing the time to reach a rate of voltage change of ΔE/Δt = −0.1 mV/s; the response time was found to be 19 ± 1 s for a 0.1 mM perchlorate solution. The cyanostar-based electrodes demonstrated consistent performance with less than 2 mV per decade slope reduction over a period of 4 months. This observation can be attributed to the stability of the cyanostar macrocycle39 and its compatibility with the PVC membrane phase. Furthermore, membranes stored dry were used after 6 months and displayed similar performance as when they were first fabricated.

■

CONCLUSIONS The design of new macrocyclic ionophores with preorganized central cavities based on CH hydrogen bonding is a versatile approach to improve the performance of membrane ion sensors. Cyanostar, a novel ionophore with electropositive cyanostilbene-based CH groups, allows perchlorate binding with 2:1 stoichiometry. Ion-selective electrodes derived from PVC membranes using cyanostar as a neutral carrier ionophore demonstrated selectivity toward perchlorate at substantially higher levels than the ones operating solely by ion-exchange mechanisms. Improved detection limits with enhanced selectivity could be achieved after tuning the concentration of TDMACl ion exchanger and the ionophore. The optimized electrodes demonstrated a response to ClO4− with a slope of −57.0 mV per decade over a linear range of 100 nM to 0.1 M and a detection limit of 50 nM. The sensing characteristics demonstrated by cyanostar electrodes motivate future modifications to the cyanostar toroid structure to design ionophores with unique selectivities toward other anions.

■

APPLICATIONS Perchlorate is widely used in explosives, rocket fuels, fireworks, and fertilizers. The United States Environmental Protection Agency (EPA) and other organizations have classified perchlorate as a persistent pollutant on account of its high solubility and stability in water. High levels of perchlorate have been detected in rivers, lakes, underground waters, and drinking water, especially in areas close to military practice and fireworks activities.70,71 Perchlorate has a negative impact on human health by interfering with iodide uptake, which then causes a decrease in thyroid hormone production,72,73 and is capable of causing human diseases like hypothyroidism, goiter, and mental retardation.74 Occasionally, up to 1000 mg/day of potassium perchlorate is recommended for patients with hyperthyroidism.75 Although there is no federal drinking water standard, EPA has defined an official reference daily exposure of 0.7 μg/ kg/day (24 μg/L drinking water) of perchlorate. Generally, ion chromatography (IC), IC-ESI-MS/MS, and HPLC-ESI-MS/ MS have been used in the determination of perchlorate in environmental and biological samples.76−78 The ISEs for perchlorate are typically based on ion exchangers that exhibit strong interference from other lipophilic anions, such as thiocyanate and nitrate. 79 The superior selectivity of cyanostar-based electrodes toward perchlorate, particularly over the common interfering anions like chloride log Kpot ClO4−,Cl− pot = −5.54 and salicylate log KClO4−,sal− = −6.50 enables the determination of ClO4− in various media. Generally, chloride is present at relatively high levels in environmental and biological samples. For example, in rivers, chloride can be present at 1 mM levels, though these can rise significantly during winter distribution of salt for deicing. In serum, chloride concentrations are typically 125 mM and are at 0.1−5 mM for intracellular levels. The presence of salicylate in blood samples acquired from patients taking aspirin for prescribed treatments leads to interferences that can cause erroneous estimation of other anions. To exploit the high selectivity of cyanostar-based electrodes in real world applications, calibration plots of perchlorate were constructed. Tap water and lake water were used to simulate environmental measurements, while horse serum made up to pH 7.3 using HEPES buffer simulated biological samples. CNStar-based electrodes demonstrated performances (Figure 6) for tap water, lake water, and horse serum with slopes of −54.9, − 55.8, and −54.2 mV per decade, respectively, values

■

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.analchem.7b04008. X-ray crystallography data, ESI-MS spectra of CNStar with salicylate, diffusion 1H NMR of CNStar with ClO4− and salicylate, and potentiometric response of various CNStar-based electrodes (PDF)

■

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Phone: +1(305) 284-4021. Fax: +1(305) 284-5637. ORCID

Amar H. Flood: 0000-0002-2764-9155 Leonidas G. Bachas: 0000-0002-3308-6264 Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS The authors acknowledge support from the University of Miami. E.M.Z. acknowledges the National Research Centre, Egypt, for granting an academic leave, and A.H.F. thanks the National Science Foundation (CHE 1412401) for support.

■

REFERENCES

(1) Bakker, E. TrAC, Trends Anal. Chem. 2014, 53, 98−105. (2) Makarychev-Mikhailov, S.; Shvarev, A.; Bakker, E. J. Am. Chem. Soc. 2004, 126, 10548−10549. (3) Gemene, K. L.; Meyerhoff, M. E. Electroanalysis 2012, 24, 643− 648. G

DOI: 10.1021/acs.analchem.7b04008 Anal. Chem. XXXX, XXX, XXX−XXX

Article

Analytical Chemistry (4) Greenawalt, P. J.; Garada, M. B.; Amemiya, S. Anal. Chem. 2015, 87, 8564−8572. (5) Bakker, E.; Bhakthavatsalam, V.; Gemene, K. L. Talanta 2008, 75, 629−635. (6) Zook, J. M.; Bodor, S.; Gyurcsanyi, R. E.; Lindner, E. J. Electroanal. Chem. 2010, 638, 254−261. (7) Bell-Vlasov, A. K.; Zajda, J.; Eldourghamy, A.; Malinowska, E.; Meyerhoff, M. E. Anal. Chem. 2014, 86, 4041−4046. (8) Bobacka, J.; Lewenstam, A.; Ivaska, A. J. Electroanal. Chem. 2001, 509, 27−30. (9) Shvarev, A.; Neel, B.; Bakker, E. Anal. Chem. 2012, 84, 8038− 8044. (10) Kim, Y.; Amemiya, S. Anal. Chem. 2008, 80, 6056−6065. (11) Jarolimova, Z.; Crespo, G. A.; Xie, X. J.; Afshar, M. G.; Pawlak, M.; Bakker, E. Anal. Chem. 2014, 86, 6307−6314. (12) Daunert, S.; Bachas, L. G. Anal. Chem. 1990, 62, 1428−1431. (13) Malon, A.; Vigassy, T.; Bakker, E.; Pretsch, E. J. Am. Chem. Soc. 2006, 128, 8154−8155. (14) Mi, Y. M.; Mathison, S.; Goines, R.; Logue, A.; Bakker, E. Anal. Chim. Acta 1999, 397, 103−111. (15) Boswell, P. G.; Buhlmann, P. J. Am. Chem. Soc. 2005, 127, 8958−8959. (16) Bereczki, R.; Takacs, B.; Langmaier, J.; Neely, M.; Gyurcsanyi, R. E.; Toth, K.; Nagy, G.; Lindner, E. Anal. Chem. 2006, 78, 942−950. (17) Guzinski, M.; Lisak, G.; Sokalski, T.; Bobacka, J.; Ivaska, A.; Bochenska, M.; Lewenstam, A. Anal. Chem. 2013, 85, 1555−1561. (18) Zahran, E. M.; Hua, Y. R.; Lee, S.; Flood, A. H.; Bachas, L. G. Anal. Chem. 2011, 83, 3455−3461. (19) Mendecki, L.; Chen, X. R.; Callan, N.; Thompson, D. F.; Schazmann, B.; Granados-Focil, S.; Radu, A. Anal. Chem. 2016, 88, 4311−4317. (20) Bakker, E.; Buhlmann, P.; Pretsch, E. Chem. Rev. 1997, 97, 3083−3132. (21) Flood, A. H. Beilstein J. Org. Chem. 2016, 12, 611−627. (22) Busschaert, N.; Caltagirone, C.; Van Rossom, W.; Gale, P. A. Chem. Rev. 2015, 115, 8038−8155. (23) Langton, M. J.; Serpell, C. J.; Beer, P. D. Angew. Chem., Int. Ed. 2016, 55, 1974−1987. (24) Brown, A.; Beer, P. D. Chem. Commun. 2016, 52, 8645−8658. (25) Hay, B. P.; Bryantsev, V. S. Chem. Commun. 2008, 0, 2417− 2428. (26) Berrocal, M. J.; Cruz, A.; Badr, I. H. A.; Bachas, L. G. Anal. Chem. 2000, 72, 5295−5299. (27) Liu, Y. L.; Qin, Y.; Jiang, D. C. Analyst 2015, 140, 5317−5323. (28) Shishkanova, T. V.; Sykora, D.; Sessler, J. L.; Kral, V. Anal. Chim. Acta 2007, 587, 247−253. (29) Piotrowski, T.; Radecka, H.; Radecki, J.; Depraetere, S.; Dehaen, W. Electroanalysis 2001, 13, 342−346. (30) Koseoglu, S. S.; Lai, C. Z.; Ferguson, C.; Buhlmann, P. Electroanalysis 2008, 20, 331−339. (31) Langton, M. J.; Serpell, C. J.; Beer, P. D. Angew. Chem., Int. Ed. 2016, 55, 1974−1987. (32) Hua, Y.; Flood, A. H. Chem. Soc. Rev. 2010, 39, 1262−1271. (33) Gale, P. A.; Hursthouse, M. B.; Light, M. E.; Sessler, J. L.; Warriner, C. N.; Zimmerman, R. S. Tetrahedron Lett. 2001, 42, 6759− 6762. (34) Li, Y.; Flood, A. H. J. Am. Chem. Soc. 2008, 130, 12111−12122. (35) Li, Y.; Flood, A. H. Angew. Chem., Int. Ed. 2008, 47, 2649−2652. (36) Busschaert, N.; Bradberry, S. J.; Wenzel, M.; Haynes, C. J. E.; Hiscock, J. R.; Kirby, I. L.; Karagiannidis, L. E.; Moore, S. J.; Wells, N. J.; Herniman, J.; Langley, G. J.; Horton, P. N.; Light, M. E.; Marques, I.; Costa, P. J.; Felix, V.; Frey, J. G.; Gale, P. A. Chem. Sci. 2013, 4, 3036−3045. (37) Zahran, E. M.; Hua, Y.; Li, Y.; Flood, A. H.; Bachas, L. G. Anal. Chem. 2010, 82, 368−375. (38) Wolsey, W. C. J. Chem. Educ. 1973, 50, A335. (39) Lee, S.; Chen, C. H.; Flood, A. H. Nat. Chem. 2013, 5, 704−710. (40) Hirsch, B. E.; Lee, S.; Qiao, B.; Chen, C.-H.; McDonald, K. P.; Tait, S. L.; Flood, A. H. Chem. Commun. 2014, 50, 9827−9830.

(41) Qiao, B.; Anderson, J. R.; Pink, M.; Flood, A. H. Chem. Commun. 2016, 52, 8683−8686. (42) Fatila, E. M.; Twum, E. B.; Sengupta, A.; Pink, M.; Karty, J. A.; Raghavachari, K.; Flood, A. H. Angew. Chem., Int. Ed. 2016, 55, 14057−14062. (43) Dobscha, J. R.; Liu, Y.; Flood, A. H. Shape-Persistent Anion Receptors. In Comprehensive Supramolecular Chemistry II; Elsevier: New York, 2017; pp 329−348. (44) Yawer, M. A.; Havel, V.; Sindelar, V. Angew. Chem., Int. Ed. 2015, 54, 276−279. (45) Lee, S.; Hirsch, B. E.; Liu, Y.; Dobscha, J. R.; Burke, D. W.; Tait, S. L.; Flood, A. H. Chem. - Eur. J. 2016, 22, 560−569. (46) Soldin, O. P.; Braverman, L. E.; Lamm, S. H. Ther. Drug Monit. 2001, 23, 316−331. (47) Singharoy, A.; Venkatakrishnan, B.; Liu, Y.; Mayne, C. G.; Lee, S.; Chen, C. H.; Zlotnick, A.; Schulten, K.; Flood, A. H. J. Am. Chem. Soc. 2015, 137, 8810−8818. (48) Zahran, E. M.; Gavalas, V.; Valiente, M.; Bachas, L. G. Anal. Chem. 2010, 82, 3622−3628. (49) Bakker, E.; Pretsch, E.; Buhlmann, P. Anal. Chem. 2000, 72, 1127−1133. (50) Crespo, G. A.; Afshar, M. G.; Barrabes, N.; Pawlak, M.; Bakker, E. Electrochim. Acta 2015, 179, 16−23. (51) Benson, C. R.; Fatila, E. M.; Lee, S.; Marzo, M. G.; Pink, M.; Mills, M. B.; Preuss, K. E.; Flood, A. H. J. Am. Chem. Soc. 2016, 138, 15057−15065. (52) Qiao, B.; Liu, Y.; Lee, S. M.; Pink, M.; Flood, A. H. Chem. Commun. 2016, 52, 13675−13678. (53) Qiao, B.; Hirsch, B. E.; Lee, S.; Pink, M.; Chen, C. H.; Laursen, B. W.; Flood, A. H. J. Am. Chem. Soc. 2017, 139, 6226−6233. (54) Zhao, W.; Qiao, B.; Chen, C. H.; Flood, A. H. Angew. Chem., Int. Ed. 2017, 56, 13083−13087. (55) Fatila, E. M.; Twum, E. B.; Karty, J. A.; Flood, A. H. Chem. - Eur. J. 2017, 23, 10652−10662. (56) He, Q.; Kelliher, M.; Bähring, S.; Lynch, V. M.; Sessler, J. L. J. Am. Chem. Soc. 2017, 139, 7140−7143. (57) Mason, S.; Llinares, J. M.; Morton, M.; Clifford, T.; BowmanJames, K. J. Am. Chem. Soc. 2000, 122, 1814−1815. (58) Custelcean, R.; Sellin, V.; Moyer, B. A. Chem. Commun. 2007, 1541−1543. (59) Schaller, U.; Bakker, E.; Spichiger, U. E.; Pretsch, E. Anal. Chem. 1994, 66, 391−398. (60) Lindner, E.; Graf, E.; Niegreisz, Z.; Toth, K.; Pungor, E.; Buck, R. P. Anal. Chem. 1988, 60, 295−301. (61) Bakker, E.; Malinowska, E.; Schiller, R. D.; Meyerhoff, M. E. Talanta 1994, 41, 881−890. (62) Tohda, K.; Higuchi, T.; Dragoe, D.; Umezawa, Y. Anal. Sci. 2001, 17, 833−839. (63) Tohda, K.; Umezawa, Y.; Yoshiyagawa, S.; Hashimoto, S.; Kawasaki, M. Anal. Chem. 1995, 67, 570−577. (64) Zubrowska, M.; Wroblewski, W.; Wojciechowski, K. Electrochim. Acta 2011, 56, 6114−6122. (65) Wright, D.; Sop, J. Am. J. Emerg. Med. 2015, 33, 1714.e3− 1714.e4. (66) Szigeti, Z.; Vigassy, T.; Bakker, E.; Pretsch, E. Electroanalysis 2006, 18, 1254−1265. (67) Gupta, K. C.; D’Arc, M. J. Anal. Chim. Acta 2001, 437, 199−216. (68) Jorge, M.; Gulaboski, R.; Pereira, C. M.; Cordeiro, M. N. D. S. J. Phys. Chem. B 2006, 110, 12530−12538. (69) Pabby, A. K.; Rizvi, S. S. H.; Requena, A. M. S. Handbook of Membrane Separations: Chemical, Pharmaceutical, Food, and Biotechnological Applications; CRC Press: Boca Raton, FL, 2008. (70) Sijimol, M. R.; Jyothy, S.; Pradeepkumar, A. P.; Chandran, M. S. S.; Ghouse, S. S.; Mohan, M. Environ. Forensics 2015, 16, 125−134. (71) Fram, M. S.; Belitz, K. Environ. Sci. Technol. 2011, 45, 1271− 1277. (72) Dasgupta, P. K.; Dyke, J. V.; Kirk, A. B.; Jackson, W. A. Environ. Sci. Technol. 2006, 40, 6608−6614. H

DOI: 10.1021/acs.analchem.7b04008 Anal. Chem. XXXX, XXX, XXX−XXX

Article

Analytical Chemistry (73) National Academies Press (U.S.); National Research Council (U.S.). Division on Earth and Life Studies. Health implications of perchlorate ingestion; National Academies Press: Washington, D.C., 2005; p xvi, 260 p. (74) Merrill, E. A.; Clewell, R. A.; Gearhart, J. M.; Robinson, P. J.; Sterner, T. R.; Yu, K. O.; Mattie, D. R.; Fisher, J. W. Toxicol. Sci. 2003, 73, 256−269. (75) Mattie, D. R.; Strawson, J.; Zhao, J. In Perchlorate: Environmental Occurrence, Interactions and Treatment; Gu, B., Coates, J. D., Eds.; Springer US: Boston, MA, 2006; pp 169−196. (76) Krynitsky, A. J.; Niemann, R. A.; Nortrup, D. A. Anal. Chem. 2004, 76, 5518−5522. (77) Oldi, J. F.; Kannan, K. Chemosphere 2009, 77, 43−47. (78) Jiang, S.; Li, Y.-S.; Sun, B. Chin. Chem. Lett. 2013, 24, 311−314. (79) Sabek, J.; Adriaenssens, L.; Guinovart, T.; Parra, E. J.; Rius, F. X.; Ballester, P.; Blondeau, P. Chem. - Eur. J. 2015, 21, 448−454.

I

DOI: 10.1021/acs.analchem.7b04008 Anal. Chem. XXXX, XXX, XXX−XXX