|Home | About | Journals | Submit | Contact Us | Français|
This chapter covers recent advances in the development of polymeric materials containing discrete heterocyclic anion receptors, and focuses on advances in anion binding and chemosensor chemistry. The development of polymers specific for anionic species is a relatively new and flourishing area of materials chemistry. The incorporation of heterocyclic receptors capable of complexing anions through non-covalent interactions (e.g., hydrogen bonding and electrostatic interactions) provides a route to not only sensitive but also selective polymer materials. Furthermore, these systems have been utilized in the development of polymers capable of extracting anionic species from aqueous environments. These latter materials may lead to advances in water purification and treatment of diseases resulting from surplus ions.
The field of polymer chemistry has made tremendous progress since Staudinger first pioneered various polymerization methods in the early 1920’s [1, 2]. Staudinger proposed structural formulae for natural rubber, polystyrene, and polyoxymethylene, and is also credited with coining the term “macromolecule” which is still commonly used today. The field of synthetic polymers was advanced further in the 1930’s when Carothers developed two widely used synthetic polymers, neoprene and nylon [3, 4]. The importance of these materials cannot be overstated. For example, nylon and other polyamides are excellent fiber forming materials, and were initially used to make ropes, parachutes, and tents, but applications have since greatly expanded and these and other synthetic polymers can now be seen in almost all aspects of everyday life. This ubiquity has spawned wide-spread efforts to develop new and improved materials. A few areas of advanced focus within this general paradigm include creating synthetic replacements for biological tissues [5, 6] and developing macromolecular materials suitable for use in diagnostic and array technologies [7, 8].
Another major thrust in polymer chemistry has come from the field of “supramolecular chemistry”, a term coined by Lehn to mean chemistry beyond the molecule [9, 10]. As chemists began to exploit the weak non-covalent interactions (which represent the foundation of supramolecular chemistry) to create new molecular receptors and chemosensors, it soon became apparent that the same principles could lead to advances in the field of polymer chemistry [11–14]. A major subset of this latter effort has been devoted to the synthesis of so-called main-chain supramolecular polymers, where reversible/dynamic non-covalent interactions within the polymer backbone are used to assemble monomers into larger polymeric structures . Separate from this, a number of researchers began to realize the potential benefits of incorporating small molecule supramolecular receptors into polymeric structures. The vast majority of these latter systems consist of conjugated polymers containing receptors for positively charged or neutral species. One highly utilized approach involves the incorporation of crown ether species for metal ion recognition and sensing . For example, Swager reported how an appropriately designed, conjugated polymer chemosensor containing crown ethers would undergo recognizable changes as the result of potassium induced aggregation .
In spite of the above advances, it was only recently that the number of polymers that incorporate receptors for anionic species began to increase. This tardiness in terms of materials development is somewhat surprising given the role anionic species play in the environment and in biological processes. It is now well recognized that many anionic species, such as fluoride, nitrate, phosphate, and cyanide, constitute major environmental pollutants; there is thus a clear, widely appreciated need for materials capable of recognizing, capturing, and detecting these and other anionic species. Furthermore, excesses in phosphate anion levels in patients suffering from end-stage renal failure is a frequent complication that can lead to a disease state known as hyperphosphatemia [18, 19]. In most cases dialysis in combination with ingestible anion exchange resins is used to control the level of phosphate in the blood. However, neither of these methods is fully satisfactory, in part because the underlying methodologies are not phosphate selective. The authors thus believe that new materials, such as anion-specific dialysis membranes, could be used to control phosphate anion concentrations for patients on dialysis and lead to clinical advances. The incorporation of supramolecular receptors specific for anionic species into polymeric frameworks represents a first step towards achieving this long-range goal. Related anion-binding macromolecules could also see application in the recognition and sensing of anions of environmental concerns and in the areas of separation and purification. Taken in concert, these various potential benefits provide an incentive to prepare anion receptor-modified polymers. The goal of this chapter is to provide a summary of recent work in the area.
To the best of our knowledge, no reviews have appeared that cover the chemistry of macromolecular systems targeting the problem of anion recognition. The focus of this chapter will be on polymeric systems that incorporate heterocycle-based anion receptors into polymeric structures. Heterocyclic receptors as additives to polymers (Fig. 1A) for applications such as polyelectrolyte development, will be covered, as will be polymers that contain anion receptors either as pendant side-chains (Fig. 1B) or incorporated directly into the polymer main-chain (Fig. 1C). As will become apparent from the present review, the majority of the systems that exist today are conjugated polymers that have been applied as so-called chemosensors. However, it is our hope and expectation that a detailed description of these and related systems will inspire the synthesis of new and improved systems for anion detection, sequestration, extraction, and separations, among other conceivable applications.
According to the principle of Occam’s Razor, “entities must not be multiplied beyond necessity,” which one can interpret to mean that often the simplest explanation is the correct one. Analogously, in engineering and materials science, one attempts to get the biggest bang for one’s buck by using cheap and simple materials to build up complexity and solve pressing problems. Over-complicated syntheses or expensive starting materials can lead to greatly diminished returns and harpoon the chances that the molecule or material will ever see wide-spread application. Such appreciations may well have inspired early researchers working at the interface of anion recognition and polymer research. This is because most initial efforts in the area were focused on integrating heterocyclic anion receptors into polymeric systems by simply blending various anion receptors into the polymers. Such approaches, which are akin to blending chocolate chips into cookie dough, are markedly different than synthetic strategies that involve attaching recognition units to the polymer backbone or creating receptors that are functionalized in such a way that they may act as precursors for polymer synthesis. These approaches are thus treated separately in this chapter, with the focus of this particular section being on polymeric materials whose properties have been modified via the blending in of anion receptors.
Arguably the most important area where the blending strategy has been successfully applied involves the development of ion-selective electrodes (ISEs). Such analytical tools derive their selectivities from the specificities of the constituent ionophores and their abilities to form stable complexes with analytes that are subsequently transported into electrode membranes. Ideally, an ISE should respond to a specific ion and should not be affected by the presence of other ions in solution. A number of ISEs utilizing polymer matrix liquid membranes have been developed and investigated. As detailed below, the incorporation of appropriately chosen receptors for a target ion into the electrode membrane represents a particularly valuable approach to anion-specific ISE development.
Early work in the development of ISEs for anionic species utilized polyamines. One such example was reported by Carey and Riggan who incorporated cyclic aza-ethers (1 – 4) into polyvinyl chloride (PVC) membranes for construction of phosphate selective electrodes (Fig. 2) . The chemical inertness of PVC, combined with its adsorptivity and stability features, has traditionally made it an ideal choice for electrode development. In the particular case of the Carey and Riggan study, ISEs were constructed by dipping an electrode into a membrane-forming “cocktail” which consisted of 20 wt % ionophore (i.e., 1 – 4), 45 wt % PVC, and 35 wt % dibutyl sebacate in tetrahydrofuran (THF). This resulted in the formation of a PVC matrix with a thickness of 15–30 μm over the open end of the electrode.
The electrode produced using 1 as the enmeshed ionophore showed excellent selectivity for dibasic phosphate (HPO42−), exhibiting a near-Nernstian slope of ~ 29 mV per activity decade and a linear range of 10−7 – 10−1 M at pH = 7.2. Ionophores 1–4 demonstrated an inverse relationship between ring size and the linear range of the electrodes made with each ionophore. The electrode developed with 1 was also shown to respond to varying pHs, a finding that led the authors to suggest that this electrode responds specifically to dibasic phosphate anions. Furthermore, the electrode demonstrated selectivity over other interfering anions (e.g., HPO42− > SCN− ≈ Cl− > NO3− > SO42−). The selectivity of the electrode is consistent with the so-called Hofmeister bias , with the exception that Cl− gives rise to a greater response than NO3−. The authors postulate that possible interactions between chloride and membrane components, such as PVC, could account for this observation. Carey and Riggan reasoned that the inherent selectivity of 1 could be attributed to preferential conformational changes upon binding, according to preliminary modeling results. Gratifyingly, the electrode showed great durability, retaining selectivity and usability over a period of 9 months while being repeatedly used in tests of human saliva and phosphate standards.
Umezawa et al. developed a similar ISE based on a modified aza-crown ether, 8 , which was obtained by treating a C16H33-substituted malonate (5) with tetraethylenepentamine (6) in absolute methanol for three weeks under conditions of reflux. This produced macrocycle 7 (Scheme 1). The amide linkages were then reduced to amines using BH3·THF. This procedure allowed receptor 8 to be isolated in 40% yield after purification involving recrystallization from a mixture of MeCN and MeOH. The receptor was then used to create a PVC-based ISE. As above, good anion selectivity was seen with this system, in this case towards adenosine triphosphate (ATP4−) in comparison to phosphate and other biologically ubiquitous anions. In HEPES buffer at pH 6.7, a linear response to ATP4− was found from 10−3 – 10−7 M with a response slope of −14.5 mV. Additionally, the electrode did not appear to exhibit any memory effect (hysteresis) corresponding to changes in concentration.
At physiological pH, the aza-crown ether derivative 8 was believed to be triply protonated (i.e., existing as a trication). The use of such a charged receptor, as opposed to a neutral receptor, was thought to account for the observed selectivity for ATP4− and/or its monoprotonated form, HATP3−. This finding was considered to be unique at the time since it meant that the neutral form of the receptor (as used to prepare the modified electrode) undergoes protonation at the surface of the ISE, thus producing a more effective, charged receptor. This finding inspired the construction of yet additional ISEs.
In 1998 Umezawa, Sessler, and coworkers conducted a comprehensive study of a variety of aliphatic and heteroaromatic lipophilic amines (Fig. 3) and their potentiometric anionic response to neutral phenols . The amines used in this study included the aza-crown ether derivative discussed above (8), tri(decyl)amine (9), 4,7-diphenyl-1,10-phenanthroline (12), terpyridine (13), and sapphyrin (14) .
The electrodes developed by Umezawa and Sessler utilized PVC matrix membranes. Membranes that incorporated aliphatic amines (9–11) exhibited potentiometric selectivities that correspond to the acidity (hydrogen donor activity) as well as the lipophilicity (extractability) of the phenols. The authors explained the anionic response on the basis of a decrease in the amount of charge separated by the protonated amines and counteranions across the membrane. The pH profiles of heteroaromatic amines 12, 13, and 14 led the authors to suggest that protonation of these latter species is less facile, but still leads to charge separation across the membrane interface generating a membrane potential. Interestingly, the membrane developed using sapphyrin (14) as the ionophore showed selectivity for catechol, reflecting an innate geometrical discrimination.
In a separate study, Umezawa and Sessler studied the potentiometric response of membranes containing expanded porphyrins towards carboxylate anions and inorganic anions . The expanded porphyrins used in this study included sapphyrin (14), rubyrin (15), and triphenylrosarin (16). Prior to using these particular species as ISE elements, Sessler and coworkers had shown that various expanded porphyrins are capable of binding anionic species in their protonated forms through a combination of hydrogen bonding and electrostatic interactions . Specifically, sapphyrin was demonstrated to be an efficient receptor for fluoride, phosphate and phenylphosphate [27, 28]. Rubyrin , which is a 26 π-electron species, was shown to demonstrate an affinity for chloride and GMP, whereas triphenylrosarin , a fully conjugated 24 π-electron macrocycle, was likewise shown to recognize the chloride anion. In both cases, prior protonation of the macrocycle was deemed necessary to achieve efficient anion binding.
Umezawa and coworkers prepared PVC membranes containing expanded porphyrins 14 – 16 using a procedure similar to that used to obtain the previously described systems. In this case, pH titrations involving the resulting electrodes served to demonstrate that protonation of the expanded porphyrins at the surface of the electrode is required for uptake of the targeted anions and to obtain a potentiometric response. This set of electrodes demonstrated a strong response to the benzoate anion, but did not respond well to inorganic anions or saturated aliphatic carboxylate anions. Interestingly, the electrode developed from sapphyrin (14) deviated from the Hofmeister series, and showed selectivity for fluoride, a much more hydrophilic anion, over chloride and bromide. The sapphyrin-based electrode in this study appeared to discriminate for fluoride based on size, the specificity of the charge-charge interactions, and via hydrogen bond formation. At the time of the work, a selective potentiometric response based on size represented a new approach to developing analyte-specific ISEs.
In 1999, Král, Sessler, and Gale developed PVC-derived ISEs from neutral anion receptors . This study utilized meso-octamethylcalixpyrrole (17) and its pyridine containing analogues dichlorocalixpyrrolepyridine (18) as well as tetrachlorocalixpyridine (19). This series of macrocyclic receptors represents a matched set, wherein it was considered that if the pyridine moieties remain unprotonated they should act as strong (17), intermediate (18), and weak anion receptors (19), respectively. Calixpyrrole is a known neutral macrocycle that binds anionic species through hydrogen bonding interactions, showing selectivity for F− > Cl− > H2PO4− when studied as the TBA salts in dichloromethane .
ISEs developed from 17 displayed a strong anionic response (negative slope) at lower pH values (i.e., 3.5 and 5.5) in the case of Br−, Cl−, and H2PO4−. A correspondingly lower response was observed with F−. However, at higher pH (i.e., 9.0) electrodes derived from 17 displayed a cationic response (positive slope) toward Cl− and Br−. Non-Hofmeister selectivity was also seen in the case of an ISE produced from 17 (i.e., Br− < Cl− < OH− ≈ F− < HPO42−). The authors rationalized this finding by suggesting that 17 acts as a direct anion binding agent at low pH, whereas at higher pH this same receptor acts, at least in part, as a hydroxide anion-complexing receptor. ISEs based on receptors 18 and 19 displayed selectivities in accordance with the Hofmeister series at pH 9.0. The authors also reported that at lower pH values ISEs derived from 18 and 19 displayed increased anionic responses and improved selectivities for hydrophilic anions (e.g., F− and H2PO4−); this observation was rationalized in terms of protonation of the pyridine containing receptors, which led to the higher anion binding effects. These results, when considered in concert, provide initial support for the notion that calixpyrroles, as well as other neutral receptors, may be used in the development of anion specific ISEs and, more broadly, polymeric systems capable of interacting specifically with negatively charged analytes.
The incorporation of heterocyclic receptors into polymeric materials has not been limited to the development of selective electrodes. For example, Anzenbacher and Nishiyabu designed chromogenic anion chemosensors based on calixpyrrole (17) that were produced by attaching non-chromophoric dye precursors to the macrocyclic framework . The ease and high yield of the underlying preparation offer obvious advantages over attaching pre-existing chromophores to calixpyrroles. The calixpyrrole derivatives 20 – 22 produced by these researchers were found to display dramatic color changes upon the addition of the fluoride, acetate, pyrophosphate, and phosphate anions.
In an effort to demonstrate the relevance of chemosensors 20 – 22 to health care applications, studies were performed using carboxylates of medical interest (salicylate, ibuprofen, naproxen). Toward this end, Anzenbacher and his group developed a new assay utilizing polyurethane films that contained receptors 20 – 22 embedded non-covalently within macromolecular matrices. According to the authors, the polyurethane support served two purposes. First, the polymer physically acted to screen off various blood plasma protein carboxylates from the active sites. Second, its hydrophobic nature prevented hydrophilic anions (e.g., HCO3−) from penetrating into the matrix and biasing the embedded sensor. The actual test experiments involved the use of a multi-well assay that contained the functionalized polyurethane. It was found that relatively lipophilic aromatic carboxylates were able to penetrate films containing 20, bind with the chemosensor, and produce a response. This response takes the form of a characteristic change in the color of the polymer film (Fig. 7). The affinity for the carboxylate anions was as follows: naproxen ≈ ibuprofen ≥ salicylate > laurate > acetate (prepared in plasma-like aqueous solution (PLAS)).
A follow up article in 2007 by Anzenbacher et al. was concerned with the development of easy to use chemosensors and arrays for anions in water . The low charge to radius ratio and high energy of solvation of various anions (e.g., F− has an ionic radius of 1.33 Å and ΔGhydration = −465 kJ/mol ) makes sensing in water a particularly challenging task. Anzenbacher utilized eight pyrrole hydrogen bond donor motifs as the chemosensor elements (Fig. 8). These motifs included N-confused calixpyrrole (23), calixpyrrole 21 and its derivatives 24 – 28, as well as the anthraquinone based dipyrrolylquinoxaline (DPQ) 29.
As in the previous study, Anzenbacher blended sensing elements 21 and 23 –29 with a polyurethane hydrogel, and cast the resulting modified polymer into a microwell array. Individual wells were filled with 400 nl of polymer-sensor mixture (approximately 0.08% sensor in polyurethane w/w) in a tecophilic THF solution (5% w/w) and dried to form polymer films (10 μm thick). The use of a hydrogel in these studies lends mechanical support to the sensor ensemble and presumably helps draw the bulk analyte into the matrix and partially remove the hydrate from the anion (all anions were applied in aqueous solutions). Wells prepared in this way produced a naked eye detectable color change upon exposure to the fluoride, pyrophosphate, and acetate anions. Anions were typically added as aqueous solutions (200 nl) of their corresponding TBA salts. The authors observed that the magnitude and dynamic range of the colorimetric response corresponded well with the affinity displayed by the chemosensor molecule in organic solution. Furthermore, multivariate analysis, specifically principal component analysis (PCA), was used to distinguish between 10 different brands of fluoride-containing toothpaste.
The final study considered in this section involves work reported by Ebdon and coworkers in 2004. They were able to improve upon the findings of Carey and Riggan (vide supra) by developing an improved phosphate-selective electrode containing receptors as both additives to polymers and as covalently linked side-chain functional elements . This study is important in that it allowed the relative benefits of receptors incorporated as additives into preformed polymers to be compared with those where the receptors are covalently attached to polymers. In prior work, Ebdon et al. demonstrated immobilization of quaternary ammonium species containing allyl groups via a free radical cross-linking reaction . The resulting membranes were used as receptors for nitrates. They thus employed similar methods as were used by Carey and Riggan to immobilize derivatives of receptor 1 (vide supra); this was done by substituting a terminal allyl-bearing alkyl group to the macrocyclic core. However, in contrast to these earlier researchers, Ebdon et al. utilized monomers 30 and 31 whose syntheses were based on previously reported procedures [20, 36]. These monomers were added to membranes composed of the relatively heat resistant polystyrene-block-polybutadiene-block-polystyrene (SBS); subsequent cross-linking was achieved via a hot-pressing process using a Bytec industrial heated press (obtained from Bytec, London, UK) .
To test the efficacy of a trapped ionophore in a polymer blend versus an ionophore covalently bound to a polymer, Ebdon and coworkers constructed two electrodes. One of these was a PVC-based membrane akin to that used by Carey and Riggan; the second electrode consisted of the previously mentioned SBS-based membrane containing covalently bound ionophores. Under the conditions of analysis, receptor 30 was found to be too hydrophilic for use as an embedded (i.e., not cross-linked) probe. Specifically, it was found to leach out of the PVC-based membrane during processing, such that no response to phosphate could be detected. On the other hand, the SBS-based electrode incorporating 30 worked well, as long as a hot-pressing process was used to ensure covalent cross-linking between 30 and the SBS backbone. This electrode gave a response slope of −16 ± 3 mV, had a linear range of 2.2 × 10−3 to 5 × 10−6 mol dm−3 H2PO4− and a limit of detection of 1.0 × 10−6 mol dm−3 H2PO4−. While these results compared favorably to those obtained by Carey and Riggan the nature of the comparison was not identical. Therefore to permit a more accurate assessment of the trapped versus covalently bound ionophore approaches, Ebdon and coworkers synthesized a more hydrophobic monomer, 31, which showed improved performance in the PVC matrix electrode. The PVC-31 electrode showed detection abilities comparable to those of electrodes created using 1 embedded in PVC. Specifically, a response slope of −34.0 ± 2.0 mV, a linear range of 2.2 × 10−3 – 5 × 10−6 mol dm−3 H2PO4− and a limit of detection of 5 × 10−6 mol dm−3 H2PO4− was seen. Due to 31 being simply embedded in the PVC matrix, the electrode had a response lifetime of merely 4 days, compared to the SBS-30 electrode. In the latter case, where the ionophore was covalently bound, a lifetime of 20 days was observed.
With these results in hand, Ebdon and coworkers sought to extend further the lifetime and improve the response of their phosphate-selective electrodes. To do this, they turned to clay composites, an approach known for improving stability, as well as altering the backfilling of electrolytes which resulted in significant improvement in the limit of detection in electrode systems [38, 39]. By blending in 23.8% (m/m) PoleStar™ 200R (clay-based filler), Ebdon and coworkers found that the lifetime of the electrode could be extended from 20 days to 40 days for the bound electrode. That comparable response behavior was seen for the ISEs based on mobile and bound ionophores is consistent with the notion that ionophore mobility is not necessary to get a working ISE. Nevertheless, it is important to appreciate that the ISEs based on covalently bound ionophores exhibit significantly increased robustness and stability as compared to analogous PVC-membranes containing “trapped” ionophores.
The work of Ebdon and coworkers thus provides an important demonstration of the potential advantages that can accrue from appending receptors covalently to polymeric materials. Taking this a step further, the formation of well-defined polymers with precisely placed receptors can be expected to enhance the performance of the aforementioned devices (and other applications) even further. Along these lines, the following sections will highlight two classes of such receptor functionalized polymers, specifically those where receptors are 1) attached as side-chain appendages to polymeric frameworks and 2) incorporated directly into polymer backbones.
The incorporation of anionic receptors into well-defined polymers and other macromolecules is expected to produce materials that offer a number of advantages as compared to their small molecule analogues. For example, polymers often have significantly different solubilities than small molecules, a feature that is likely to be particularly useful in the areas of anion extraction and separation. Attachment to a polymer provides a method for immobilizing anionic receptors and chemosensors and can prevent “leaching”, or loss of the receptor, which can be advantageous in applications involving the use of mixed phases. The use of polymers can also lead to beneficial matrix effects that can serve to decrease the solvation of a targeted anion or its countercation. Again, this can be useful in any application involving transfer of an anionic species from an aqueous to hydrophobic environment. Finally, polymeric materials can often be used to create mechanically-robust thin films; where the films contain anion receptors, new and practical applications in the areas of inter alia membrane technology, filter devices, and sensors can be easily envisioned.
In this section the focus will be on polymeric materials wherein anion receptors are incorporated as pendant side chains. As will be detailed below, several basic strategies have been used to prepare such materials. The use of differing synthetic methods, and the choice of the receptor system in question, has allowed a degree of affinity and selectivity to be designed into polymeric systems. However, the hope and expectation is that yet-improved systems can be obtained. Indeed, as implied in the introductory remarks, one objective of this review is to stimulate additional effort along these lines.
Early functionalized polymer systems targeting the problem of selective anion recognition relied on the use of secondary electrostatic interactions for the detection of anionic species. Indeed, rather than incorporating an anion receptor per se, these systems relied on cationic species that would then attract anions as the result, presumably to maintain charge balance. This strategy is elegantly embodied in the work of Leclerc and coworkers who in 2002 reported a polythiophene derivative containing a pendant imidazolium salt . This polymer was prepared by an oxidative polymerization procedure using FeCl3 as the oxidizing agent (Scheme 2) . The polymer was soluble in aqueous solution and at 55 °C (0.1 M NaCl or 10 mM tris(hydroxymethyl)aminomethane (Tris) buffer/0.1 M NaCl) the solutions were yellow in color (λmax = 397 nm). The red-shifted absorption maxima was attributed to the random-coil conformation and subsequent decrease in effective conjugation .
Polymer 34 was initially used for the colorimetric and fluorometric detection of nucleic acids. The cationic polymer displayed an ability to transduce oligonucleotide hybridization easily, with the specific capture of a 20-mer probe translating into a clear optical output. In a separate study, Ho and Leclerc used the related polymers 35 and 36 as colorimetric and fluorometric chemosensors for the specific detection of the iodide anion . The impetus for this latter study derives in part from the fact that iodide is a biologically relevant anion, mostly known for its role in thyroid functions. In the study itself, electrostatic interactions were thought to result in a conformational change of the polythiophene derivatives causing a change in the effective conjugation of the polymer chain. Iodide anions promote the aggregation and planarization of polymer 35 causing a red-shift in absorbance as well as fluorescence quenching. The optical effects, monitored using absorption and emission spectroscopies were shown to be independent of the corresponding counter cation (e.g., Na+ vs. K+). The authors also demonstrated that these systems were dependant on the length and nature of the side chain; for instance, they found that polymer 36 displayed reduced sensitivity under conditions similar to those used to analyze polymer 35.
In a more recent study Tang et al. employed a different approach towards creating polymers with positively charged appendages . Specifically, with the goal of exploiting the so-called “molecular wire effect” noted for conjugated polymers, these researchers synthesized polyphenylacetylenes (PPAs) containing imidazole moieties attached to the backbone (i.e., as pendant side chains). These materials were prepared via post-polymerization functionalization, wherein substitution with imidazole was used to generate polymer 37 (Fig. 10). Polymer 37 displayed poor solubility in THF, chloroform and dimethylformamide (DMF) but was readily soluble in ethanol. This material demonstrated both the strong luminescence (i.e., blue fluorescence) characteristic of disubstituted polyacetylenes, as well as the metal ion-coordinating ability of imidazoles. As such, it was proposed that polymer 37 would provide a new kind of potentially efficient chemosensor.
It was shown that Cu2+ could quench the fluorescence of polymer 37 (1.06 × 10−4 M solution in ethanol) completely at very low concentration (1.48 ppm) with the corresponding Stern-Volmer constant determined to be 3.7 × 105 M. Tang and coworkers were then able to take advantage of a “turn on” method of anion sensing by employing species that competitively bound Cu2+. In particular, low concentrations of CN− (i.e., 7.0 × 10−5 M−1 of sodium cyanide) served to “turn on” the fluorescence of the polymer 37·Cu2+ complex. The reactivation of the fluorescence signal was not observed in the presence of other anionic species. The detection of CN− is highly important given its inherent toxicity to mammals and its common use in industrial applications such as gold mining, electroplating, and metallurgy [45, 46].
In 2009, Tang and coworkers utilized polymer 37 to sense α-amino acids via a similar “turn on” fluorescence approach . The use of a normal UV lamp allowed histidine to be differentiated from other α-amino acids; in particular, histidine allowed for the visual observation of a restored blue fluorescence at concentrations as low as 4.0 × 10−5 M. Using fluorescence spectroscopy to monitor the “turned on” fluorescence signal, after exposure of the polymeric complex 37·Cu2+, allowed histidine to be detected at concentrations as low as 2.1 ppm.
Recognizing that the human eye is not generally as sensitive to the blue fluorescence produced by 37 as it is to green light, Tang et al. generated a second generation imidazole-pendant PPA (e.g., polymer 38) . In this polymer the residual halogen groups were completely converted to imidazole substituents (effected synthetically by converting 37 to the corresponding bromide); this yielded a polymer with strong green fluorescence (38). The chemosensing behavior of 38 could be observed visually or with the help of a normal UV lamp. This material was capable of detecting copper and cyanide at concentrations as low as 0.85 and 1.38 ppm, respectively.
A novel PPA containing naphthalimide subunits in the side-chain was developed by Tian and coworkers in 2009 . Previous work with naphthalimides demonstrated that these systems act as chemosensors for fluoride ions based on a unique intermolecular proton-transfer (IPT) signaling mechanism (Fig. 11) [50, 51]. Polymer 39 was synthesized using [Rh(nbd)Cl]2 as a catalyst, resulting in a material that was soluble in most common organic solvents.
Upon addition of tetrabutylammonium fluoride (TBAF) to an acetonitrile solution of 39, a drastic color change from colorless to yellow was observed. The change in the absorption and emission spectra of 39 allowed for a ratiometric detection of fluoride anions in solution. Titrations with other anionic species lead to no spectral changes. Ratiometric fluorescent probes allow for improved reliability and a greater effective dynamic range by providing a built-in correction for environmental effects [52, 53]. While there have been a number of ratiometric fluorescent probes developed for cationic species [54, 55], there are few reports of ratiometric fluorescent polymers for F− anions . Interest in the detection and recognition of the fluoride anion is of growing interest given its association with nerve gas (it is, for instance, a hydrolysis product of Sarin) and the role that UF6 plays in the enrichment of uranium-235.
Taking a more materials based approach to anion sensing, Li and coworkers developed a novel photonic ionic liquid (IL) system for the detection of various anions (Fig. 12) . Utilizing an imidazolium-based IL monomer (40), these investigators performed a photopolymerization of this monomer in 1:1 feed ratio with methylmethacrylate in the presence of the cross-linker ethylene glycol dimethacrylate, and the initiator 2,2′-azoisobutyronitrile (AIBN). The polymerization itself was effected in the presence of silica colloidal crystals in a mixed solvent system consisting of a 1:1 mixture of methanol and chloroform. The silica colloidal crystals were then etched away with 1% HF solution to yield a photonic IL film with an inverse opal structure. The resulting 3D ordered photonic IL was utilized in the naked eye detection of anions (vide infra).
It is well known that the hydrophobicity and hydrophilicity of ILs can be tuned by counteranion exchange . In the case of macromolecules containing such species, such tuning was expected to be characteristic of the material. The use of appropriate solvents leads to swelling or shrinking of the films, which should result in a change in the stop gap of the photonic ILs. Similar effects were expected from counteranion exchange. As can be seen from an inspection of Fig. 13, such effects were observed for exposure of the photonic IL films to nitrate (NO3−), tetrafluoroborate (BF4−), perchlorate (ClO4−), phosphorous hexafluoride (PF6−), bis(trifluoromethylsulfonyl)imide (Tf2N−) and bromide (Br−). Subsequent monitoring by FTIR served to confirm adsorption of the anions to the films. The most hydrophobic anion, Tf2N−, results in the largest shift in absorbance, 76 nm, while the most hydrophilic, NO3−, led to a relatively modest shift of only 18 nm.
Boron-based heterocyclic receptors have also been appended to polymers. Fabre and coworkers in 1999 reported one of the first examples of an immobilized boronate anion receptor . In particular, Fabre et al. detail a polypyrrole functionalized with pendant boronate pinacol ester derivatives (41) that can function as an electrochemical chemosensor for anions when electrodeposited onto a platinum electrode via anodic oxidation performed in acetonitrile. The concentration of anionic analyte was then monitored via shifts in the cyclic voltammagrams. To synthesize the pyrrolic monomer, Fabre and coworkers began with 1-(phenylsulfonyl)-3-vinylpyrrole. Hydroboration by diisopinocampheylborane yielded the intermediate diethyl boronate derivative. Reaction with pinacol then gave the pinacol boronate derivative. Subsequent deprotection of the phenylsulfonyl group by electrochemical reduction gave monomer 41 in 13% overall yield. Electropolymerization of 41 was achieved in anhydrous MeCN using 10−1 M Bu4NPF6 as the supporting electrolyte.
In water/acetonitrile mixtures, the polymer-coated platinum electrode based on 41 exhibited a reversible redox process at −0.11 V versus a standard calomel electrode (SCE). Upon addition of a 2 mM solution of potassium fluoride (KF) to the system, a cathodic shift to −0.37 V versus SCE was observed, a result that is consistent with the expectation that the polymer would be easier to oxidize when fluoride is bound. Fabre, et al. also reasoned that the negative charge of the fluoride anion might be stabilized by the positive charge created on the polymer backbone as the result of oxidation. Presumably, this adds to the signal response. In contrast to fluoride, very little response was seen when the boronate functionalized polypyrrole was exposed to the chloride or bromide anions, requiring a minimum concentration of 20 mM before a recognizable response was seen in the cyclic voltammetry (CV) traces.
A different set of electrochemical sensors for anion detection was developed by Royal and coworkers, who relied on ferrocene-viologen (4,4′-bipyridinium) based redox active receptors, as well as related polymer films . The stated long-range goal of this research was to develop a sensor for ATP2− that would function in biological milieus. This is a particularly challenging objective given the ubiquity and relatively high concentrations of other anionic species in the human body, including in particular chloride and inorganic phosphate. Cognizant of this challenge, Royal et al. first synthesized the small molecule receptors 42 and 43. These compounds were made so as to test the sensing abilities of the ferrocene-viologen receptors per se. Here, it is to be noted that receptor 42 represents a ferrocene with viologens bound at the 1 and 1′ positions, while 43 is an ansa system wherein two viologens are connected through two ferrocenes to form a macrocycle that is formally analogous to the cyclobis(paraquat-p-phenylene) “blue box” made famous by Stoddart . Films of polymer-44 were deposited onto an electrode surface via oxidative electrochemical polymerization of the pyrrole fragment (0.9 V vs. Ag/Ag+ 10−2 M in MeCN + TBA salts of perchlorate (TBAP).
These investigators chose CV and differential pulse voltammetry (DPV) as the probe methods used to investigate the binding affinities of 42, 43 and polymer-44. In the case of the redox active receptors 42 and 43, exposure to H2PO4−, SO42−, HPO42−, CF3COO−, F− and Cl− produced no significant shifts in the CV or DPV curves. However, the Fc0/+ couple in 42 was found to undergo a shift of −10 mV when exposed to HSO4−, S2O42− and ATP2−, while a shift of −10 mV in this wave was seen when 43 was exposed to PhPO42− and S2O42−. Of particular note is that the ferrocene linked system 43 experienced a shift in the Fc0/+ wave of −25 mV when titrated with 2 two-molar equivalents of ATP2−, thus revealing a selectivity for this all-important biological analyte. Royal, et al. argued that this selectivity can be attributed to 43 forming a hexacationic receptor upon oxidation, increasing the electrostatic interactions with ATP2−. Charge-transfer interactions between the electron rich adenosine ring system and the viologens are also thought to contribute to the anion binding process.
Upon oxidative electropolymerization of monomer 44 in MeCN, an ATP2− selective electrode was produced. When the polymer-coated electrode was titrated against H2PO4−, PhPO42− and halide anions, which had previously caused no shifts or only modest shifts in the CV peaks of receptors 42 and 43, little in the way of response was likewise observed. On the other hand, exposure to ATP2− over a concentration range of 10−5 to 10−3 M produced a cathodic shift in the Fc0/+ peak that increased to a maximum ΔEp of −35 mV (i.e., from Ep = 460 mV to Ep = 425 mV vs. SCE). This was considered an excellent result that highlights the possible detection of ATP in aqueous environments, which could in due course allow for the use of these electrodes in more biologically-oriented environments.
In this section we have covered polymeric materials that incorporate anion receptors as pendant side chains. These systems have been applied as chemosensors for species such as CN−, F−, and phosphate derivatives. This approach toward material development is attractive because it allows the beneficial features of various monomeric receptors (e.g., specificity and signal output) to be combined with those of polymers (e.g., solubility and stability). In this section, the receptors discussed derived their selectivity from electrostatic interactions or subsequent de-protonation upon exposure to anionic species. In contrast, the ensuing section will focus on incorporation of neutral heterocyclic receptors with a strong affinity for anionic species.
The previous section provides an introduction into polymeric materials containing receptors capable of recognizing anionic species. The aforementioned systems incorporate anion responsive or sensing moieties as side-chain appendages. However, the innate functionality of these materials was typically derived from electrostatic interactions or deprotonation after exposure to anionic species. This often resulted in a colorimetric or fluorometric response. In this section, we will look at neutral receptors capable of binding anions with a high affinity, often through multiple hydrogen bonding interactions. In particular, the structural incorporation of calixpyrrole into polymeric systems will be covered, as well as other molecular receptors (e.g., crown ether, DPQ, and diindolylquinoxaline (DIQ)). Earlier in this chapter, polymer matrices containing calix[n]pyrrole, and derivatives thereof, were described. These materials not only represent a creative approach to ISE development, but also are exemplary of initial progress in the development of polymers containing neutral anion receptors. One advantage to the strategies described below is the ability to tune material properties based on choice of receptor as well as polymer design (e.g., alteration of polymer type, cross-linking, or monomer loading levels).
Early progress in this field is represented by two sets of resins prepared by Kaledkowski [62, 63]. In the first of these, calixpyrrole (45) and calixpyrrolethiophene (47) were attached to cross-linked vinylbenzene chloride/divinylbenzene copolymer beads. A condensation of phenol-substituted calixpyrrole with formaldehyde (46) was used to synthesize the second set of resins. Both these materials displayed an ability to “capture” halides and cyanide anions from MeCN (i.e., non-aqueous environments). Furthermore, resins that contained the “expanded” calixpyrrolethiophene unit demonstrated an enhanced affinity for larger anionic species, such as iodide. This result correlates well with the selectivities observed for the stand-alone receptor .
In a more recent study, Sessler, Bielawski, and coworkers reported poly(methyl methacrylate)s (PMMAs) containing pendant calixpyrroles . As mentioned in the chapter on calix[n]pyrroles by Gale and Lee, these new polymeric materials were utilized in the extraction of TBAF and TBACl from aqueous environments. The calixpyrrole methacrylate monomer (48) was prepared in 84% yield from treatment of a hydroxymethyl calixpyrrole derivative with methacryloyl chloride in the presence of triethylamine (TEA). Homopolymer (49) was prepared via treatment of monomer 48 with 1 mol% AIBN in THF solution. The reaction was then stirred at 70 °C for 17 h under a nitrogen atmosphere. The resulting viscous polymer solution was precipitated by dropwise addition into excess cold methanol, and recovered in 66% yield. Analysis using gel permeation chromatography (GPC) revealed a number-average molecular weight (Mn) of 23,600 Da (relative to PMMA standards) and a polydispersity index (PDI) of 2.3. Furthermore, this conventional free radical polymerization technique was also used to generate a calix-pyrrole containing PMMA copolymer (50). Using GPC analysis, copolymer 50 was found to possess a Mn of 85,500 Da and a PDI of 2.1. Sessler and Bielawski concluded from the high molecular weight of 50, compared to 49, that the steric bulk of the calixpyrrole negatively impacted the growth of the polymer (48).
In an effort to explore the anion binding ability of copolymer 50 under interfacial conditions, Sessler and Bielawski utilized NMR spectroscopic techniques. Towards this end, a D2O solution of TBAF (90 mM) was exposed to a CD2Cl2 solution of polymer 50 (effective concentration of calixpyrrole repeat unit = 6.5 mM). The layered mixture was shaken for 20 min and then centrifuged to separate the organic and aqueous layers. Analysis of the organic layer (CD2Cl2) via 1H NMR spectroscopy revealed a substantial downfield shift in the pyrrole NH protons (Δppm = 0.32 ppm), which is typically observed upon anion binding. Furthermore, peaks corresponding to the TBA+ counter cation (at δ = 3.2 ppm) were observed, confirming that both the anion (F−) and the cation (TBA+) were present in the organic phase. A greater downfield shift in the NH pyrrole proton signal was observed upon exposure to TBACl under analogous extraction conditions and concentrations. These results are consistent with the ability of polymer 50 to extract chloride over fluoride. This selectivity runs counter to the anion affinities displayed by calixpyrrole in dichloromethane . However, the observed affinities are in agreement with the so-called Hofmeister bias, as chloride is a more hydrophobic anion (ΔGh = −340 KJ mol−1) than fluoride anion (ΔGh = −465 KJ mol−1) [21, 66]; chloride anion was thus expected to be extracted more readily than this latter, more hydrophilic species. Supporting this rationalization came from the finding that octamethylcalixpyrrole and PMMA were both capable of extracting TBACl, but 35% less efficiently than polymer 50. In the case of TBAF only polymer 50 was capable of removing fluoride from the aqueous layer.
The results described above represent one of the first examples of a molecular receptor functionalized polymer capable of extracting anionic species from aqueous environments. However, polymer 50 displayed a low affinity for “hard” salts containing hydrophilic cations (e.g., K+ and Na+). In 2008, Sessler, Bielawski, and coworkers addressed this issue by appending benzo--crown-5-ether, a subunit known for its ability to complex cationic species , as a pendant group to the calixpyrrole functionalized PMMA polymer backbone . The resulting copolymer (i.e., 51) was prepared by the previously described free radical polymerization. Initial quantitative evidence that copolymer 51 could extract chloride salts came from a visual extraction test utilizing water-soluble dye 52 (Fig. 18). An aqueous solution of 52 was extracted with a CH2Cl2 solution of copolymer 51, as well as control solutions of octamethylcalixpyrrole (Fig. 18B), benzo--crown-5-ether (Fig. 18C), and a mixture of both receptors (Fig. 18D). Only the organic layer containing copolymer 51 displayed evidence of successful extraction of dye 52, as observed by a light blue colored organic phase.
The amount of dye (52) removed from the aqueous layer, upon exposure to copolymer 51 and control compounds, was quantified with the use of UV-Vis spectroscopy. Analysis of the aqueous layer, post-extraction, confirmed that copolymer 51 was able to extract dye 52 into the organic phase 54% more effectively than octamethylcalixpyrrole, benzo--crown-5-ether, or the mixture of receptors.
The promise of the results described above, led Sessler and Bielawski to explore whether or not copolymer 51 was capable of extracting two hard ions, namely potassium fluoride (KF). Thus, a 3.4 M D2O solution of KF was exposed to a CD2Cl2 solution of copolymer 51 (effective concentration of the calixpyrrole and crown ether repeat units 6.25 and 4.86 mM, respectively) and extracted via the previously describe procedure. This resulted in the appearance of a signal at δ = −121.7 ppm in the 19F NMR spectrum of the organic phase. The fluoride anion concentration in this phase was quantified via addition of fluorobenzene (final concentration: 14.21 mM) as an internal standard (δ = −114.3 ppm) to the aforementioned extraction experiment. Comparative integration showed copolymer 51 was capable of extracting KF (7.55 ± 0.04 mM) more efficiently than copolymer 50 (5.71 ± 0.03 mM). The extraction efficiencies determined by 19F NMR were further supported by flame emission spectroscopy (FES), which was used to quantify the amount of potassium extracted into the organic phase (6.84 ± 0.05 mM).
In more recent collaborative work involving the Sessler and Bielawski groups, the above strategy was applied to the development of polymeric chemosensors. In this work, quinoxaline derivatives known as dipyrrolylquinoxaline (DPQ) and di-indolylquinoxaline (DIQ) were utilized as the anion chemosensor elements. DPQs contain pyrroles appended through an α-pyrrolic position to the 5 and 6 positions of the quinoxaline. First developed by Oddo  in the early 20th century, DPQs were later “rediscovered” by Sessler and coworkers in 1999 as an efficient fluorometric and colorimetric small molecule chemosensors for anions, e.g., fluoride (F−), dihydrogen phosphate (H2PO4−), etc. . The 6-nitro derivative (nitro-DPQ) showed remarkable affinity and selectivity for the fluoride anion, exhibiting a binding constant of 118,000 M−1 in dichloromethane. It also produced an easily discernible yellow-to-red color change. On the other hand, DIQ demonstrated a high affinity and selectivity for dihydrogen phosphate, with the nitro derivative exhibiting a binding constant of 20,000 M−1 in dichloromethane.
A two step modification was used to synthesize methylacrylamide monomers of DPQ (53) and DIQ (55). Subsequently, free radical polymerization, using the previously described synthetic method, resulted in methylmethacrylate copolymers of DPQ (54) and DIQ (56). The resulting copolymers showed relatively high molecular weights (ca. 40,000 Da) and polydispersities on the order of 2.1 to 2.5, similar to previously observed systems.
Thin films of the DPQ polymer 54 (ratio of DPQ to MMA repeat units = 1:10; unpublished results) were prepared using Langmuir-Blodgett techniques. Upon exposure to HF vapors (generated from a 48% aqueous solution of HF) a colorimetric response from bright yellow to red was observed (cf. Fig 19b). After two minutes the red color began to fade (Fig. 19c), and was no longer observable after 10 minutes.
Likewise, thin films of copolymer 54 were prepared and used as a “dip-stick” test (i.e., direct exposure to solutions of HF). Exposure to a 48% solution of aqueous HF resulted in a drastic colorimetric change from yellow to dark purple (Fig. 19e). Upon exposure to a 25% aq. solution of HF a change from yellow to red was observed. However, control studies using HCl as the acid solution showed similar colorimetric changes, suggesting that the observed color changes are a result of protonation of the imine in the DPQ, rather than binding of the F− anion. Incorporating DPQ and DIQ sensing moieties into polymeric systems allows for potential application in material-based devices, and demonstrates the tunability of this approach based on monomer selection.
As previously mentioned, polymeric materials often provide advantages beyond those generally displayed by their constituent monomers. One such advantage, displayed by certain polymers (particularly conjugated polymers), is their ability to conduct charge. This type of conductive polymer can be used to transform a chemical signal into an easily detectable optical or electrical event . The incorporation of anion receptors into such materials should facilitate sensing applications via signal amplification. With this goal in mind, it was considered useful to incorporate discrete receptors into the main-chain of covalent polymers. An excellent example of this strategy was reported by Lee et al., who in 2007 succeeded in incorporating quinoxaline receptors into a set of derivatized polyfluorene copolymers (Fig. 20) . The polymer in question, poly[ortho-diaminophenylene-fluorene)-co-(quinoxaline-fluorene)] (57), was prepared by effecting polymerization of 4,7-dibromo-2,1,3-benzothiadiazole, 9,9-dihexylfluorene-2,7-bis (trimethyleneborate), and 5, 8-dibromo-2,3-diphenylquinoxaline through Suzuki coupling, followed by reduction to the ortho-diamino group using lithium aluminum hydride. The authors suggested that a selective interaction between the ortho-diamino group and the fluoride anion results not only in a colorimetric change, observable by the naked-eye, but also in fluorescence quenching as the result of a photoinduced electron transfer process (PET).
In 2008, Lee and coworkers reported a new set of azomethine-containing conjugated polymers containing fluorene and/or quinoxaline units that are closely attached to their main chains; as above, these systems, represented by canonical structures 58 and 59, were synthesized by Suzuki-coupling reactions followed by hydrogenation and condensation with cyclohexanone (Fig. 21) .
Polymers 58 and 59 were used in the naked-eye detection of acid vapors. Towards this end, the polymers were deposited onto filter paper and exposed to acid vapors. This led to a dramatic color change from bright red to bluish violet (Fig. 22). This color change, which is thought to result from protonation of the imine unit, as opposed to a structural alteration, was found to be reversible, with the color fading when the deposited polymer was left exposed to air for 5 minutes. Protonated versions of polymers 58 and 59, obtained by pre-exposure to TFA, were also used to prepare a set of anion-induced colorimetric sensors. Specifically, exposure of these TFA-treated polymers to iodide and acetate ions resulted in drastic chromatic changes, which allowed for naked eye detection of these particular negatively charged species.
As previously mention, DPQ is a highly effective receptor for the fluoride anion, thus several groups have developed polymeric systems that rely on this quinoxaline derivative as the anion chemosensor element. Sun and coworkers prepared a DPQ-containing poly(phenylene ethynylene) backbone via palladium-catalyzed Sonagashira cross-coupling reactions, producing polymer 60 (Fig. 23) . These systems were found to undergo a bathochromic shift in the UV-Vis absorption spectra when exposed to TBAF in CH2Cl2 solution. A quenching of the fluorescence intensity was also seen.
Based on careful 1H NMR spectroscopic titrations, Sun et al. attributed the bathochromic shift in the UV-Vis absorption spectra, as well as the fluorescence quenching, to deprotonation of one of the pyrrolic units, rather that hydrogen bonding of the pyrrolic N-H protons. From UV-Vis absorbance titrations, they were able to calculate an effective binding constant (per DPQ subunit) of 2.52 × 103 M−1 for I− and 1.44 × 103 M−1 for pyrophosphate (HP2O73−).
Anzenbacher et al., have also prepared conjugated polymers containing DPQs. These systems were prepared from monomers of DPQ functionalized with ethylenedioxythiophenes (EDOTs) which were then electropolymerized and subsequently doped (Fig. 24) . The polymeric systems prepared in this way (61 and 62) allow for two forms of anion sensing, namely by following the anion-induced color changes via UV-Vis absorption spectroscopy, or using a bipotentiostat to monitor the changes in current produced by the addition of anions. Polymer 61 demonstrated good affinities for the fluoride and pyrophosphate anions, yielding effective binding constants of 48,000 and 61,100 M−1, respectively. Polymer 62 is similar to 61 but contains chlorides in the α pyrrolic position of the pyrrolic subunits. The chloride modified polymer showed selectivity toward dihydrogen phosphate, exhibiting an effective per DPQ binding constant of 90,000 M−1. This polymer also displayed an affinity for fluoride and pyrophosphate, although reduced compared to that displayed by the dihydrogen phosphate anion (the effective affinity constants were 24,000 and 11,000 M−1 for fluoride and pyrophosphate, respectively).
A somewhat different class of conjugated polymers containing neutral heterocycles as anion chemosensors was reported by Wang et al.; these researchers produced polyphenylenes copolymerized with phenol-substituted oxadiazoles (systems 65 and 66) as efficient fluorescent sensors for fluoride anions . Wang and coworkers began by first synthesizing the small molecule model compounds 2,5-bis(2-hydroxyphenyl)-1,3,4-oxadiazole (63) and 2-(2-hydroxyphenyl)-5-phenyl-1,3,4-oxadiazole (64), systems that they found to act as colorimetric and fluorescent chemosensors of F− and H2PO4− (Fig. 25) .
It was demonstrated that 63 exhibited a high affinity for dihydrogen phosphate and fluoride anions (Ka= 7.9 × 105 and 8.6 × 104 M−1, respectively) in DMF, while 64 exhibited a slightly higher affinity for dihydrogen phosphate (Ka= 1.8 × 106 and 4.1 × 104 M−1 for H2PO4− and F−; studied as the respective TBA salts in DMF). The ability of the small molecule model systems to bind dihydrogen phosphate in organic media led to the assumption that polymeric systems would follow suit. However, contrary to this expectation, Wang et al. reported that polymers 65 and 66 (Fig. 26) exhibited higher affinities for fluoride over dihydrogen phosphate when studied in chloroform solutions (5 μM concentrations). The emission intensity of polymer 65 decreased up-to 380-fold upon exposure to F−, which was accompanied by an emission red-shift of up to 15 nm. However, the addition of H2PO4− led to only a 6-fold reduction in emission intensity (both anions studied as the respective TBA salts). The anion binding properties of 66 were similar to 65 but a significant reduction in sensitivity for F− was observed. This latter finding was attributed to the different number of hydroxyl binding sites in each repeat unit. The more global change in affinity, namely selectivity for fluoride over dihydrogen phosphate, was ascribed to the fact that the polymers exist in coil conformations. Such structures were expected to limit access to the hydroxyl sites in the case of bulky anions such as dihydrogen phosphate, while still permitting smaller, more charge dense anions, such as fluoride, to bind. Based in part on further studies (vide infra) from the Wang group, it is likely that this “binding” involves at least partial deprotonation of the hydroxyl protons.
In a separate study involving conjugated fluorescent polymers as anion sensors, Wang et al. prepared hydroxyl-functionalized polyquinoline-derived macromolecules (PQOH, 68; cf. Fig 27) . Polymer 68 was prepared via a nickel-catalyzed coupling of a methoxy-protected precursor followed by demethylation (BBr3 followed by water) . A hydroxyl-containing quinoline, 67, was synthesized as a small molecule control.
Both the model compound 67 and the polymer 68 (Fig. 22) were found to undergo naked-eye detectable color changes, as well as fluorescence “turn-on” in the presence of TBAF in DMSO. Upon addition of this fluoride anion salt, solutions of 67 were found to change from colorless to yellow, with even larger changes, from colorless to red, being seen in the case of polymer 68. Other anionic species, namely chloride, bromide and dihydrogen phosphate (all studied as their TBA salts), did not produce color changes in the case of polymer 68. Exposure of polymer 68 to TBAF, but not to other anions, was also found to engender a new fluorescent emission peak at 620 nm. This observation is most easily rationalized in terms of the hydroxyl groups undergoing deprotonation when exposed to F− (an anion recognized for its basicity in organic milieus) a chemical change that was expected to give rise to an intramolecular charge transfer interaction.
Deprotonation also likely accounts for the efficacy of a set of polyhydroxybenzoxazole-based colorimetric and fluorometric chemosensors for fluoride (cf. structure 71) introduced by the Lee group, whose other contributions are featured elsewhere in this chapter . In this case, a Suzuki cross-coupling polymerization strategy was employed (Scheme 4). Specifically, palladium-catalyzed polymeric coupling of the dibromo-benzoxazole monomer (69) with a dialkoxyphenylene di-boronic acid precursor (70), followed by deprotection of the benzyl groups, gave rise to the desired polymer, namely poly[2-(2′-hydroxyphenyl)benzoxazole] (71). This polymer, a grayish-white powder with Mn = 5090 Da and a polydispersity of 1.56, proved soluble in most organic solvents.
Polymer 71 exhibited a UV-Vis absorbance maximum at 331 nm, and upon excitation at 330 nm, was characterized by a fluorescence emission maxima at 414 nm and 518 nm (both spectra being observed in chloroform). The unique fluorometric results are attributed to excited-state intramolecular proton transfer (ESIPT) of the enol form and the excited state of a polar tautomeric keto form [82, 83]. Furthermore, naked eye and spectroscopic analysis of polymer 71 in DMF in the presence of various anionic salts demonstrated the use of this macromolecular system as a fluoride chemosensor. Specifically, upon the addition of TBAF, DMF-polymer solutions were found to undergo a naked-eye detectable shift from colorless to yellow. This process, when followed by UV-Vis spectroscopy, is characterized by the emergence of a new maximum at 420 nm. Unfortunately, Lee et al. did not quantify these results to determine binding selectivities or limits of detection.
A different class of polymeric anion chemosensors containing receptors within the main-chain are those that rely on the use of bipyridyl moieties as the anion recognition motif. An excellent example of this type of material was reported by Lee and coworkers . These researchers carried out a Knoevenagel condensation of 5,5′-bis(cyanomethyl)-2,2′-bipyridine (72) and 2′,5′-didecyloxy-p-terphenyl-4,4”-dialdehyde (73); this gave rise to a bipyridine polymer linked by cyanostyryl groups (74, cf. Scheme 5). This polymer proved to be a dark yellow solid with an Mw = 12,500 Da and a polydispersity of 1.47.
Lee et al. tested a number of anions against polymer 74, including chloride, sulfate and dihydrogen phosphate (all anions tested as their corresponding TBA salts using DMF as the solvent). However, only the hydroxide anion appeared to induce any significant colorimetric shift in the UV-Vis spectrum or fluorescent amplification. Initial fluorescence spectral analyses of polymer 74 in DMF revealed a broad emission peak at 535 nm when excitation was effected at 404 nm. Upon addition of TBAOH, the photoluminescence intensity increased with a strong new emission peak at 480 nm and bright, blue-green fluorescence visible by the naked-eye also being observed. In the case of the UV-Vis absorption spectrum, the initial absorption maximum at 404 nm seen for 74 was found to disappear as TBAOH was titrated into the DMF solution. A blue-shift in the absorption maximum (to 354 nm) takes place, resulting in a color shift from orange to colorless. These results were rationalized in terms of changes in the electronics of the system that occur as the hydroxide anion binds to the bipyridine subunits. Interestingly, the authors did not report having tested fluoride or cyanide; both of these anions are relatively basic and may be capable of being bound by the bipyridine recognition sites present in polymer 74 as well.
Valijaveettil and coworkers developed a series of poly(p-phenylene carbazole) copolymers that enabled the colorimeteric and fluorometric detection of iodide anions (Fig. 28) . Carbazole is an interesting functional group for incorporating into polymers given its high thermal stability and moderately high oxidation potential [86, 87]. The polymerization reactions used to produce polymers 75 – 79 relied on the use of a Suzuki polycondensation process carried out in the presence of potassium carbonate, a palladium catalyst, and cetyltriethylammonium bromide (CTAB) used as the chain transfer reagent (40 mol%). Thermal gravimetric analyses of 75 – 79 revealed that these polymers were stable up to 390 C.
Polymers 75 – 79 proved to be highly fluorescent displaying quantum yields in the range of 0.58 – 0.78 when studied in THF; these values are significantly higher than those of typical polycarbazoles , and led to the consideration that these polymers could be used as fluorometric anion chemosensors. In fact, spectral changes were only observed when these systems were exposed to iodide (e.g., TBAI, LiI, NaI, and KI) among other common anions, including fluoride, chloride, bromide, nitrate, perchlorate, dihydrogen phosphate, and hydrogen sulfate (all studied as the TBA salts in THF solution). The polymers were colorless in solution; however, upon addition of iodide salts a change to yellow was observed. A red shift in absorbance and emission was also observed upon addition of iodide salts. The red shift in absorbance is attributed to the lower level of carbazole loading present in polymers 78 and 79. In all cases, the nature of the iodide counter cation was found to have no influence on the observed color change.
This chapter provides a summary of polymeric systems that contain heterocyclic anion receptors. The polymeric materials covered within this chapter are diverse and cover a broad range of applications. The ability to incorporate neutral receptors (and/or charged species) into different varieties of polymers is beginning to be translated into a corresponding ability to tune the anion recognition properties of macromolecules to meet specific needs. A number of these needs were discussed, including applications involving the use of receptor-polymer combinations as additives for ion-selective membrane development, in the extraction of ions from aqueous environments, and the selective sensing of anionic species. Likewise, a broad range of anion receptors, including aza-crown ethers, calixpyrrole, carbazole, and quinoxalines were reviewed in the context of highlighting these applications. The anion specific receptors can be directly blended with existing polymers as so-called “additives” to enhance material properties. Furthermore, synthetic design also allows for the incorporation of anion receptors into the polymers themselves, either as pendant side-chains or directly included as part of the main-chain polymer backbone. The examples given in this chapter are meant to provide a general overview of the field of anion specific polymers and provide a summary of the current state-of-the art. As such, it is expected that the present review will make it clear to the reader that there is much left to do within this fast-evolving area of materials chemistry.
This work was supported by the National Institute of Health (grant GM 58907 to J.L.S.) and the Robert A. Welch Foundation (grants F-1018 and F-1621 to J.L.S. and C.W.B., respectively).