An efficient multichannel probe to detect anions in different media and its real application in human blood plasma

Abstract

A simple coumarin derived chromo and fluorogenic chemosensor has been synthesized and tested to detect anions in different medium. The designed chemosensor has shown a high colorimetric response for F⁻ and AcO⁻ ions in acetonitrile in which the color of solution changed from a yellow-green to red and orange respectively whereas, in aqueous medium the probe showed high fluorogenic response and selectively reacted with cyanide through a Michael type nucleophilic addition reaction in which a nonfluorescent (switched-off) color of the solution, “turn-on” to became fluorescent blue (switched-on). The multichannel mode of interaction between the probe and anions has been confirmed by absorption, emission, FT-IR, NMR, ESI-MS spectral data analysis. Moreover, the probe has shown high sensitivity to detect cyanide in a real sample, human blood plasma as well as on paper strips.

---

Introduction

Recently, the recognition and sensing of anions, particularly, F⁻, AcO⁻ and CN⁻ have attracted considerable interest owing to their involvement in various biological and industrial processes as well as their detrimental effect on organisms and the environment.^1–4 For example, fluoride is useful in prevention of dental caries and treatment of osteoporosis however, a high intake of fluoride is responsible for fluorosis, nephrotoxic changes and even urolithiasis in organisms.^3,5 Similarly, carboxylate anion exhibit many biochemical functions in the enzymes and antibodies and plays important roles in different metabolic processes.^1,3 Moreover, besides high toxicity, cyanide has wide applications in various industrial processes.^5,6 The contamination of cyanide in water occurs from the discharge of various chemical, metal mining industries and waste water treatment facilities.⁵ A small amount of cyanide is lethal to organism because of its strong affinity with cytochrome a₃.⁷ Cyanide has also been used as a chemical warfare agent and terror material.⁸ World Health Organization (WHO) has set permissible level of cyanide in drinking water as 0.07 mg l⁻¹ = 2.7 μmol l⁻¹.^9b,c

The recognition of anions has gained much importance in the area of supramolecular chemistry. For effective and successful anion recognition event the choice of potential ionophore is crucial because specific geometry, high electronegativity, basicity and nucleophilicity of different class of anions interact differently with an ionophore either by hydrogen bonding interaction, deprotonation and possible nucleophilic addition reactions. Additionally, the thermodynamic feasibility of a non-covalent recognition event between an anion and receptor depends upon structural specificity of the molecular system and microenvironment. Consequently, the typical structural and electronic changes encountered due to interaction with a specific anion modulate the optoelectronic behavior of a receptor system, which can be easily followed by different spectroscopic techniques like UV-vis absorption, fluorescence and NMR. Therefore, the serious ongoing environment and biological concern it is necessary to develop selective chromo and fluorogenic chemosensors, based on small organic molecular systems to recognize anions selectively, which is sensitive to the naked eye, without resorting any expensive instrumentation.^10,11

Recently, some good receptor systems consisting of potential ionophores such as urea, thiourea, amide, thiazole, phenylhydrazone, and aldimine have been developed to detect F⁻ and AcO⁻ ions either by hydrogen bonding interaction or deprotonation.^1,12–18 Moreover, the intrinsic high nucleophilic property of cyanide can be utilized to induce a selective nucleophilic addition reaction with a suitably deigned receptor system in aqueous medium. Some boron derivatives,²² quantum dots²³ strategies involving H-bonding interaction,²⁴ coordination,²⁵ covalent⁸ bonding and ensemble based displacement approach²⁶ have been successfully utilized to detect CN⁻ under different conditions. However, chemical reaction based chemodosimeter approach are more sensitive, reliable and selective to detect cyanide in the environment of competitive and/or interfering anions like, F⁻ and AcO⁻.^19–21

Keeping above facts in mind we herein report a very simple multichannel chemosensor based on a coumarin-derived intramolecular charge transfer (ICT) probe to recognize F⁻, AcO⁻ and CN⁻ ions with good optoelectronic responses in different medium. In the present structural motif fluorescent coumarin unit has been utilized as a signaling unit, and nitro (–NO₂) substituent containing phenyl ring (DNP) to regulate the flow of charge along with hydrazone and enone fragments, as potential ionophore sites. Interestingly, the designed chemosensor has shown high colorimetric sensitivity for F⁻ and AcO⁻ in acetonitrile, whereas in aqueous medium it has shown fluorogenic response to recognize cyanide selectively through a Michael type nucleophilic addition reaction, in which nonfluorescent color of solution, “turn-on” to a fluorescent blue. The potential analytical applicability of the probe has been examined to sense cyanide in real contaminated water, in biological medium, and on paper strips.

Result and discussion

Synthesis and photophysical behavior of probes in acetonitrile

The synthetic route to obtain coumarin-derived probes 3 and model compound 4 is shown in Scheme 1. The salicylaldehyde and ethylacetoacetate were refluxed in the presence of catalytic amount of piperidine in ethanol to obtain compound, 2 which was subsequently refluxed with 2,4-dinitrophenylhydrazine (DNP) and phenylhydrazine in ethanol to afford desired molecules 3 and 4 in good yields. Compounds were well characterized by ¹H, ¹³C NMR, FTIR, and mass (ESI-MS) spectroscopy (Fig. S1–9†). Structurally, probe 3 and 4 is different in the presence and absence of nitro substituent. Due to the presence of electron withdrawing nitro substituent the hydrogen attached to hydrazine fragment of 3 become more acidic and labile in comparison to 4.

Scheme 1 (i) Ethylacetoacetate/piperidine/ethanol/reflux, (ii) 2,4-dinitrophenylhydrazine/ethanol/reflux and (iii) phenylhydrazine/ethanol/reflux.

The optoelectronic behavior of probe has been examined through the absorption and emission spectroscopy in solvents of different polarity and was found good in acetonitrile (MeCN). The polar protic solvents such as, water, ethanol and methanol use to establish strong hydrogen bonding interaction with probes having potential ionophores like, –NH/OH function. On the other hand, it is important to mention that anions like, F⁻/AcO⁻ prefer to interact with receptors having –NH/–OH fragments through hydrogen bonding interaction or deprotonation and the extent of H-bonding interaction depend on the basicity and structural specificity of anions.²⁷ Therefore, the solvent systems that itself is capable to establish H-bond with potential receptor site should be avoided in a sensitive anion recognition studies to understand the real mode of interaction between probe and tested anions.²⁸ To understand the possibility of H-bonding interaction both absorption and emission spectra of probe were acquired in different solvents like benzene, dichloromethane (DCM) acetonitrile (ACN), tetrahydrofuran (THF), methanol (MeOH) and dimethylsulfoxide (DMSO) (Fig. S10†). The photophysical data presented in Table S1† has suggested about the possibility of intermolecular H-bonding in a polar aprotic solvent like DMSO. Moreover, the polarity of different class of solvents has also shown their impact on the emission behavior of probe. Acetonitrile is a less viscous, medium polar aprotic solvent and is miscible with water and a variety of ionic and nonpolar compounds are easily soluble in this homogenous system. Therefore, we intended to see the behavior of probe toward different anions in acetonitrile and water.

The absorption spectrum of 3 (10 μM) shows a typical low energy band (π → π*) at 380 nm (ε = 2.9 × 10⁴ M⁻¹ cm⁻¹). Upon interaction with different anions (10 equiv.) such as, F⁻, Cl⁻, Br⁻, I⁻, S²⁻, N₃⁻, SCN⁻, AcO⁻, CN⁻, H₂PO₄⁻ (as their tetrabutylammonium salts) significant change occurred only in the presence of F⁻ and AcO⁻ in which the band, centered at 380 nm diminished and a new band appeared at ∼527 nm, and the color of solution was changed from a yellow-green to red-orange which was sensitive to the naked eye (Fig. 1). The considerable bathochromic shift observed in the presence of F⁻ and AcO⁻ (ε = 3.3 × 10⁴ and 2.1 × 10⁴ M⁻¹ cm⁻¹) is attributed to an intramolecular charge transfer (ICT) mechanism in which charge propagation occurred probably due to interaction of anions with potential ionophore unit of probe 3 either by deprotonation and/or hydrogen bonding interaction, respectively. To confirm, that absorption spectra were followed by the addition of TBAOH to a solution of probe 3 (10 μM). Interestingly, the typical low energy band (π → π*) absorption band at 380 nm (ε = 2.9 × 10⁴ M⁻¹ cm⁻¹) upon addition of TBAOH (10 equiv.) diminished and a new band appeared at ∼528 nm, which was almost close to the bands in the presence of F⁻ and AcO⁻. The color of solution was also changed from a yellow-green to red which was sensitive to the naked eye (Fig. S10c†). Therefore, suggesting the possibility of deprotonation. The other tested anions failed to exhibit any significant change in the absorption spectrum of 3.

Fig. 1 Change in UV-Vis absorption spectra of 3 (10 μM) upon addition of various anions in ACN. Change in color of 3 (10 μM) in the presence of different anions (10 equiv.).

The binding affinities of 3 with F⁻ and AcO⁻ were realized through the absorption titration studies (Fig. 2 and S11†). Upon a gradual addition of F⁻ (0–5 equiv.) to a solution of 3 (10 μM) the absorption band centered at 380 nm decreased almost completely and the absorbance of a new band enhanced ratiometrically at 527 nm and color of the solution changed from a light yellow-green to red. The formation of a sharp isosbestic point at 433 nm suggested the presence of more than one species in the medium (Fig. 2a). Similarly, upon addition of AcO⁻ (0–5 equiv.) to a solution of 3 (10 μM) the molar absorptivity of the band centered at 380 nm also decreased considerably and a new band appeared at 526 nm (Fig. S11a†) and color of solution changed from a yellow-green to orange. The formation of isosbestic point at 436 nm also supported the existence of more than one species in the medium.

Fig. 2 (a) Absorption titration spectra of 3 (10 μM) with F⁻. Inset: Job's plot and (b) Benesi–Hildebrand plot with F⁻.

Job's plot analysis revealed a 1 : 1 stoichiometry for an interaction between 3 and F⁻/AcO⁻ (Fig. 2a and S11a,† inset) for which the association constant was determined²⁹ by fitting the absorption titration data in eqn (1) and was found to be K_{assoc (F⁻)} = 2.0 × 10⁶ M⁻¹ (Fig. 2b) and K_{assoc (AcO⁻)} = 3.4 × 10³ M⁻¹ (Fig. S11b†). The change in free energy (ΔG) for a probable complex, 3–F⁻ and 3–AcO⁻ were found to be −23.96 kJ M⁻¹ and −12.03 kJ M⁻¹ at 298 K, respectively. Thus, suggesting high affinity for fluoride anion.

The high selectivity of probe 3 for F⁻ and AcO⁻ in acetonitrile was further confirmed by examine the absorption spectra of model probe 4. In the absence of tested anions probe 4 (10 μM) showed absorption bands at ∼377 and ∼274 nm. However, upon interaction with tested anions almost negligible change was observed in the absorption spectra of 4 (Fig. S12†). Therefore, it is interesting to mention that presence of nitro substituent in 3 make the –NH proton of hydrazine unit relatively more acidic and susceptible for deprotonation and/or hydrogen bonding interaction only with F⁻ and AcO⁻. Consequently, the charge propagation incurred due to rise in electron density on the probe led to a significant bathochromic shift and subsequent color response in the medium which was sensitive to the naked eye.

Studies in aqueous medium

The α–β unsaturated lactone/enone unit of coumarin is prone to undergo nucleophilic addition reaction to form a Michael adduct in a polar protic medium. In a polar protic medium the nucleophilicity and reactivity of anions, in general, govern by hydration factor,²⁶ and among the anions cyanide is known for its high inherent nucleophilic character²⁶ and is expected to react with probes to form adducts. Therefore, we intended to carry out anion interaction studies with probes in a pure aqueous medium.

pH studies

The optical behavior of probes (10 mM) have been examined at different pHs in a B/BH⁺ buffer (0.5 M HCl and 1 M NaOH) by performing acid–alkali spectrophotometric titration experiment, as reported previously.^18a At pH 7 probe 3 shows low energy absorption bands at 380 nm (ε = 1.85 × 10⁴ M⁻¹ cm⁻¹) and a high energy band at ∼265 nm (ε = 2.0 × 10⁴ M⁻¹ cm⁻¹) respectively. Under acidic conditions, pH 1 to 5 the absorption band centered, at 380 nm decrease along with a blue shift of ∼20 nm to appear at 360 nm whereas, in alkaline condition, pH 8–14 the absorptivity of band increased along with a red shift of ∼38 nm to appear at 418 nm. Similarly, probe 4 (10 μM) showed distinct low energy charge transfer and high energy, π → π* electronic transition bands at 372 nm (ε = 3.0 × 10⁴ M⁻¹ cm⁻¹) and 278 nm (ε = 6.8 × 10⁴ M⁻¹ cm⁻¹) respectively. Under acidic conditions, pH 1 to 5 absorption, at 372 nm decreased with a blue shifted of ∼14 nm to appear at 358 nm whereas, in between pH 8–14 hyperchromic shift occurred corresponding to low energy band, at 372 nm while absorptivity of high energy band decreased (Fig. 3). The absorption pattern at different pHs suggested about the possibilities of H-bonding interactions between enone and hydrazone functions of the probes. Further, it is interesting to mention that the probes 3 and 4 do not show any significant change in the emission behavior at different pHs (Fig. S13†).

Fig. 3 pH-absorption spectra of 3 and 4 in B/BH⁺ buffer.

Anion interaction studies

The absorption spectrum of 3 (10 μM) in aqueous medium (10 mM HEPES buffer, pH = 7.4) exhibited broad low and high energy bands at 375–380 nm and ∼265 nm respectively (Fig. 4a) while upon excitation at 380 nm a very weak broad emission band observed at ∼450–500 nm (Fig. 4b). Upon interaction with different anions (20 equiv.) such as, F⁻, Cl⁻, Br⁻, I⁻, SO₄²⁻, CO₃²⁻, SCN⁻, AcO⁻, CN⁻, N₃⁻, HS⁻, S²⁻, H₂PO₄⁻ (as their sodium salt) probe 3 displayed relatively high affinity for CN⁻ in which molar absorptivity corresponding to the band, at 380 nm decreased with a blue shift of ∼10–15 nm (Fig. 4a). It is important to mention that probe 3 at 10 μM concentrations does not revealed any colorimetric response however, when the colorimetric observation was taken at 10 mM concentration probe 3 appeared as a light yellow color which upon interaction with different anions showed naked-eye sensitive red-brown color selectively with cyanide anion (Fig. S12† inset images). Additionally, when the interference studies were performed the absorption spectra remained unchanged by the addition of tested anions to a solution of 3 + CN⁻ (Fig. S14†). Thus, suggested high selectivity of 3 for cyanide in aqueous medium. Moreover, when the emission spectrum of 3 (10 μM) was acquired in the presence of tested anions, at 380 nm excitation, fluorescence enhancement occurred only with CN⁻ in which the relative fluorescence intensity enhanced, ∼10 times and a new emission band appeared at 495 nm and the nonfluorescent color of the probe solution ‘turn-on’ to a fluorescent blue-green (Fig. 4b inset). While other tested anions (200 equiv.) failed to exhibit any considerable change in the emission spectra.

Fig. 4 (a) Change in absorption and (b) emission spectra of 3 (10 μM, λ_ex = 380 nm) upon addition of different anions (0–200 equiv.). Inset: change in color of 3 (10 μM) upon addition of CN⁻ (0–200 equiv.) in aqueous medium (10 mM HEPES buffer, pH = 7.4).

The absorption and emission titration experiments were performed to understand the binding affinity of 3 with CN⁻. Upon addition of 0–200 equiv. of CN⁻ to a solution of 3 the absorption band (at 380 nm) decreased gradually along with hypsochromic shift and the intensity of high energy band, centered at 253 nm increased concomitantly. The isosbestic point appeared at 250 nm suggested the existence of more than one species in the medium. Similarly, fluorescence titration studies showed significant fluorescence enhancement at 495 nm upon a gradual addition of 0–200 equiv. CN⁻ (Fig. 5b) and the non-fluorescent (switched-off) color of the solution “turn-on” to appear as a bright blue-green. Job's plots analysis consistently revealed a 1 : 1 stoichiometry for an interaction between 3 and CN⁻ (Fig. 5a and 6b, images).

Fig. 5 (a) Change in absorption and (b) emission titration spectra of 3 (10 μM, λ_ex = 380 nm) upon addition of CN⁻ (0–200 equiv.) in aqueous medium (10 mM HEPES buffer, pH = 7.4). Inset: shows Job's plot.

Fig. 6 (a) Emission and absorption interaction spectra of 4 (10 μM) with different anions (0.1 M) such as, F⁻, Cl⁻, Br⁻, I⁻, SO₄²⁻, CO₃²⁻, SCN⁻, AcO⁻, CN⁻, N₃⁻, S²⁻, H₂PO₄⁻ and HS⁻ (as their sodium salt) and (b) absorption and emission titration spectra of 4 (10 μM, λ_ex = 380 nm) upon addition of CN⁻ (200 equiv.) in aqueous medium. Inset: change in color of 4 upon interaction with CN⁻ (UV = 365 nm) and Job's plot.

Furthermore, in aqueous medium (10 mM HEPES buffer, pH = 7.4) model probe 4 (10 μM) upon interaction with different anions (0.1 M) showed selectivity for cyanide in which the low energy band, at 372 nm exhibited hypochromic shift while the high energy band at 278 nm blue shifted to appear at 247 nm (Fig. 6a, inset). Similarly, the emission spectrum of 4 (10 μM) revealed enhanced emission, “turn-on” (∼10 times) at 473 nm and the nonfluorescent color of solution changed to a fluorescent blue-green (switched-on). The sensitivity of probe for CN⁻ was further ascertained by interference studies in which, the absorption spectra remained unchanged upon addition of excess of tested anions to a solution of 4–CN⁻ (Fig. S12† and 6 inset, images). A 1 : 1 stoichiometry for an interaction between 4 and CN⁻ has been realized further by titration experiments in which the absorption band (at 372 nm) decreased with formation of a new band at 247 nm. The isosbestic point appeared at 251 nm suggested about the existence of a new species in the medium. Similarly, the emission spectra displayed a gradual fluorescence enhancement upon addition of same equivalents of CN⁻ to a solution of 4 (Fig. 6a). Thus, in aqueous medium probes have shown affinity for cyanide through a chemodosimeter approach, wherein cyanide reacted with enone function of probe through a Michael type addition reaction. The typical change in the absorption pattern of probe 3 and 4 upon interaction with cyanide revealed different pattern which clearly suggest about the nucleophilic addition reaction to form Michael adduct at pH 7–8.

Nature of interaction between probe 3 and anions

The prominent change observed in the absorption spectra of 3 with F⁻ and AcO⁻ are possible either due to H-bonding interaction or deprotonation of labile acidic proton available on the potential receptor site. To substantiate, ¹H NMR titration studies have been performed in CDCl₃ and DMSO-d₆ (c = 2 × 10⁻¹ mol l⁻¹). The ¹H NMR spectrum of 3 (Fig. 7 and S38†) has shown singlet at δ 11.369, 9.158 and 8.046 ppm attributable to resonances of –NH, H3′ and H4 protons. Similarly, resonances appeared at δ 8.369 (d, J = 9.3 Hz), 8.094 (d, J = 4.8 Hz), 7.627–7.571 (t, J₁ = 8.1, J₂ = 8.7 Hz), 7.393–7.314 (m) ppm are assignable to H5′, H6′, H5, H7, and H6, H8 protons, respectively.

Fig. 7 Stacked ¹H NMR titration spectra of 3 upon addition of F⁻ (0, 0.1, 0.2, 0.3, 0.4, 0.6, 0.8 and 1.2 equiv.) in CDCl₃.

Upon first addition of F⁻ (0.1 equiv.) ion to a solution of 3 almost all resonances shifted downfield, marginally (Fig. S15†). Further, increasing the concentration of F⁻ (0.2–0.3 equiv.) the –NH signal shifted downfield, Δδ = 0.028 to appear at δ 11.397 ppm while H3′ signal shifted upfield to appear at δ 9.144 ppm (Fig. S16 and 17†). Importantly, the H4 resonances, after addition of 0.3 equiv. of F⁻ shifted downfield, Δδ = 0.014 to appear at 8.060 ppm while resonances corresponding to H5′, H6′, H5–H7 and H6–H8 protons shifted upfield respectively. Thus, suggested about the possibility of H-bonding interaction between 3 and F⁻ involving H4 and –NH protons. However, upon increasing the concentration of F⁻ (0.4 to 0.6 equiv.) the –NH signal shifted downfield and H4 shifted toward upfield, δ = 7.999 (Δδ = 0.047) ppm respectively (Fig. S18 and 19†) and after addition of 0.8 to 1.2 equiv. of F⁻ the –NH signal become broadened along with upfield shift in resonances of other protons (Fig. S20 and 21†). Thus, the observed de-shielding and shielding effect clearly suggested about the interaction of F⁻ through H-bonding interaction as shown in Scheme 2. Consequently, the net charge propagation induced by F⁻ corroborate the significant change in the absorption spectra of probe.^14,18

Scheme 2 The possible mode of binding for 3, F⁻ and AcO⁻.

Similarly, the ¹H NMR titration experiment between 3 and AcO⁻ illustrated almost similar mode of interaction (Fig. S22†). Upon addition of AcO⁻ (0.1 equiv.) ion to a solution of 3 the –NH signal shifted downfield to appear at δ 11.391 ppm along with downfield shift in resonances of H3′, H5′, H6′, H4, H5, H7 and H6, H8 protons (appear at δ 9.177, 8.388, 8.110, 8.064, 7.643 and 7.411 ppm) respectively (Fig. S23†). After addition of 0.2 equiv. of AcO⁻ the –NH proton further shifted downfield, Δδ = 0.030 whereas H3′, H5′, H6′, H4, H5–H7 and H6–H8 resonances slightly shifted upfield to appear at δ 9.175, 8.385, 8.109, 8.062, 7.642, and 7.409 ppm respectively (Fig. S24†). Moreover, upon increasing the concentration of AcO⁻ (0.3 to 0.4 equiv.) the chemical shift corresponding to –NH signal further shift downfield to become broadened while H4 signal shifted upfield to appear as singlet at δ 8.056 ppm (Δδ = 0.01 ppm) along with upfield shift in the resonances of rest of protons, respectively (Fig. S25 and 26†). Thus, the typical shifts in ¹H NMR titration spectra of 3 in the presence of AcO⁻ clearly suggested about the H-bonding interactions involving protons of lactone ring and potential ionophore site as shown in Scheme 2.

To have an insight about the interaction of cyanide with 3 and 4 through the possible Michael type addition reaction probe-cyanide adducts were prepared by treating both the probes with NaCN in methanol for 1–1.5 h. The desired compounds were isolated and characterized by ¹H, ¹³C NMR, FTIR and ESI-MS data analysis. The ¹³C NMR spectra of 3 and 4 revealed C4 carbon of coumarin unit at δ 150.48 and 153.68 ppm respectively, which upon reaction with CN⁻ disappeared and new resonances corresponding to methine 30.46 and 25.45 whereas cyanide carbons generated at 118.14 and 118.09 ppm (Fig. S5, S9, S27 and S28†) respectively. The FTIR spectra of adducts, 3–CN⁻ and 4–CN⁻ showed characteristic vibration bands corresponding to CN function at 2219 cm⁻¹ and at 2254 cm⁻¹ (Fig. S4, S8, S29 and S30†) respectively. Similarly, the mass spectral (ESI-MS) data analysis illustrated molecular ion peak for 3 and its adduct [3 + CN + H]⁺ at m/z 368.9 and at m/z 396.9 respectively (Fig. S6 and S31†).

Moreover, upon formation of 3 + CN⁻ adduct the H4 proton of lactone ring (s, δ 8.04 ppm) disappeared completely and a new resonance appeared in upfield region at δ 4.09 ppm along with upfield shift in the phenyl ring protons (Fig. 8, S2 and S32†). Similarly, the ¹H NMR (CDCl₃) spectrum of 4 showed resonances at δ 7.97 (s) and 6.83 (s) ppm attributable to H4 and –NH protons. Similarly, multiplet appeared at δ 7.47 and δ 7.27 ppm are assignable to H3′, H5′, H5, H8 and H4′, H6 and H7 resonances while a doublet appeared at δ 7.09 ppm (d, J = 18.3 Hz) is attributable to H2' and H6′ protons respectively (Fig. S7 and S33†). When the ¹H NMR spectrum of an adduct 4 + CN⁻ was acquired the H4 proton resonances of 4 decreased and a new signal appeared in upfield region at δ 3.86 ppm while the phenyl ring protons appeared in the range of δ 7.61 to 7.03 ppm, respectively (Fig. S33†). Additionally, a new signal appeared at δ 8.05 ppm could be assigned to the –OH function proton of adduct, 4–CN⁻ because upon addition of D₂O to a solution of adduct, 4–CN⁻ the probable –OH/–NH function resonances disappeared completely (Fig. S34†). We could rationalize these observations by assuming the possibility of existence of distinct tautomeric forms in which adduct, 4–CN⁻ acquired sufficient energy to stay in its individual forms. However, no such distinct resonances corresponding to –OH/NH functions were observed in case of adduct, 3–CN⁻ because of a rapid proton transfer from hydrazone to enol form of coumarin unit. As expected the presence of electron withdrawing nitro substituent on probe 3 makes the hydrazine unit proton relatively more acidic as well as labile. Since, probe 4 does not have nitro substituent, the extent of rapid proton transfer and/or equilibrium between –NH/–OH protons relatively slow and distinct resonances corresponding to both –OH and –NH function observed in the NMR spectra. Thus, the ¹H NMR spectra clearly supported about the formation of Michael adducts as a result of nucleophilic addition reaction between probes and nucleophilic cyanide anion. A plausible mechanism of interaction has been shown in Scheme 3.

Fig. 8 Stacked ¹H NMR of 3 and 3–CN adduct in CDCl₃.

Scheme 3 A plausible mode of reaction between 3 and cyanide.

Studies based on DFT calculation methods. The minimum energy structures of 3 and its probable complexes, 3–F⁻; 3–AcO⁻ and adduct 3–CN⁻ (Fig. 9) were optimized by Density Functional Theory (DFT) calculation method employing Gaussian 03 suite and basis sets B3LYP/6-31G and B3LYP/6-31+G for probe 3 and their possible complexes respectively.³⁰ The probable complex, 3–F⁻ indicated about the structural changes around the potential ionophore site, in which bond angles corresponding to –C–C=N–; –C=N–N–; =N–N–C–; =N–N–H– units changed from 129.48°, 122.22°, 118.91°, 116.81° to 127.18°, 122.69°, 118.50°, 121.31° and the N–N and N–H unit bond distances changed from 1.35 Å to 1.39 Å and 1.02 Å to 1.10 Å respectively. Moreover, the significant change in an angle of ∼66.28° between the coumarin and 2,4-dinitrophenylhydrazine units may be accounted for an optimum –H⋯F⋯H type of H-bonding interaction (1.4 Å) between 3 and F⁻ involving H4 hydrogen atom of coumarin unit. Similarly, for a probable complex, 3–AcO⁻ bond angles corresponding to –C–C=N–; –C=N–N–; =N–N–C–; =N–N–H– units changed from 129.48°, 122.22°, 118.91°, 116.81° to 114.02°, 115.83°, 31.03°, 111.98° along with change in bond distances between N–N and N–H units from 1.35 Å to 1.40 Å and 1.02 Å to 1.08 respectively, along with change in an angle of ∼122.76° possibly due to –H⋯O⋯O⋯H type H-bonding interaction (1.53 Å) with AcO⁻ anion. Furthermore, the optimized minimum energy structure of an adduct 3–CN⁻ also suggested about the change in bond angles; –C–C=N–; –C=N–N–; =N–N–C–; =N–N–H– from 129.48°, 122.22°, 118.91°, 116.81° to 127.18°, 122.69°, 118.50°, 121.31° along with change in bond distances between the N–N and N–H units from 1.35 Å to 1.39 Å and 1.02 to 1.10 Å respectively. Thus, the theoretical and experimental observations corroborate the possible multichannel mode of interaction between probe and anions in different medium.

Fig. 9 B3LYP/6-31G and 631+G optimized minimum energy structure of 3, 3–F⁻, 3–AcO⁻ and 3–CN⁻.

Kinetics. In order to understand optimum response time for an interaction between probes and cyanide anion through a chemodosimeter approach reaction kinetics has been done by the addition of cyanide anions (200 equiv.) to the solution of 3 and 4 (10 μM) and emission spectra were acquired at the interval of 5 min. Kinetic plots have revealed the saturation in the emission spectral pattern of probe 3 and 4 within ∼50–60 min (Fig. 10).

Fig. 10 Reaction kinetic plots for 3 and 4 with cyanide.

Sensitivity and limit of detection (LOD)

Further, limit of detections (LOD) for F⁻, AcO⁻ and CN⁻ were estimated in acetonitrile and aqueous medium (10 mM HEPES buffer, pH = 7.4) respectively, as reported previously from author laboratory.³¹ A stock solution of 3 was prepared in aqueous medium (10 mM HEPES buffer, pH = 7.4) and diluted serially from 10 to 1 μM. The fluorescence spectra of different concentration solutions were acquired. The emission intensities vs. concentrations plot outfitted a linear calibration curve with standard deviation (σ) as 0.1343 (Fig. S35a†). The calibration sensitivity (m) 1.93 for CN⁻ was obtained from the slope of fluorescence curve where I₀ and I indicate the emission intensities of 3 in the presence and absence of CN⁻, respectively (Fig. S35b†). The LOD for cyanide was estimated using eqn (2) and was found to be 0.21 μM (5.2 ppb) which was comparable to the reported methods and close to allowed cyanide contamination in drinking water (7.7 μM; 0.2 ppm; as described by EPA, USA).^9,19 Similarly, the calibration curve for F⁻ and AcO⁻ were obtained by acquiring absorption spectra of 3 (10 to 1 μM) in acetonitrile (standard deviation, σ = 0.00283) (Fig. S36a†) and from the slop of absorption curves estimated calibration sensitivities were found to be 0.009 and 0.0049 (Fig. S36b and c†) which has given LOD as 0.94 and 1.7 μM respectively.

Analytical applications

Detection of CN⁻ in real contaminated water

Keeping in mind the high toxicity of cyanide it is vital to detect it in natural environment by a visual detection method. The real water sample have been collected from the river Ganga and filtered (0.2 μm filter paper, Whatman™). The stock solutions (10 ml) containing around 0, 50, 100 and 200 equiv. of CN⁻ were prepared by the addition of 0, 20, 40, 80 μl of aqueous NaCN solution (0.1 M) in each vials, respectively. After proper mixing the pH of the solutions were optimized and found to be in the range of 7.1 to 7.4. Now the real sample solutions (3 ml, each) were mixed with a fixed concentration of 3 (1 ml; 1 × 10⁻⁵ M) in each vials (4 ml; A–D) and vortex it for 30–45 min. Fig. 11 shows that a nonfluorescent probe solution turn-on to a naked-eye sensitive fluorescent blue-green under UV light. The intensity of color was brighter when the concentration of cyanide was around 200 equiv. (vial D) however, one can also easily observe the color change in the vials B and C which contain around 50 and 100 equiv. of CN⁻, respectively. Moreover, upon addition of CN⁻ (0.1 M) to a solution of 3 (0.1 M) a naked eye sensitive colorimetric change were also observed in which a yellow color solution of probe immediately changed to a dark red at room temperature.

Fig. 11 Images show (i) color changes under UV light (365 nm) upon addition of probe 3 (1 ml; 1 × 10⁻⁵ M) to real aqueous solutions (3 ml) having CN⁻ (a) 0.0 equiv., (b) 50 equiv., (c) 100 equiv. and (d) 200 equiv. (ii) Colorimetric change for 3 (0.1 M) upon addition of CN⁻ (0.1 M) in aqueous medium. (iii) Change in color of paper strips (a) 3, (b) 3 + 25 equiv., (c) 3 + 50 equiv., (d) 3 + 100 equiv., (e) 3 + 150 equiv. and (f) 3 + 200 equiv. of CN⁻.

Detection of CN⁻ on cellulose paper strip

Further to demonstrate the practical application of 3 paper strip test was performed. Small cellulose paper strips (Whatman™) were dipped in aqueous solution of 3 (1 mM) and dried in air. Now the paper strips were dipped in aqueous solution of sodium cyanide containing different equivalents of CN⁻ (25, 50, 100, 150 and 200 equiv.) and left them for 30 min. After that paper strips were dried in air and visualized under UV light (at 365 nm). The observed light to dark blue colors clearly, demonstrated the potential application of 3 to detect cyanide anion on the strips (Fig. 11).

Detection of cyanide in human blood plasma

In order to understand the potential analytical application of present chemodosimeter 3 to detect cyanide in biological medium we first analyze behavior of 3 in human blood plasma. Probe 3 (10 μM) was mixed with human blood plasma in HEPES buffer (9.8 ml, 0.2 ml human blood plasma; pH 7.4) as done previously,³² and emission spectra were acquired at 378 nm excitation. In human blood plasma probe 3 displayed almost similar emission spectral behavior that indicates about no undesirable interaction of 3 with blood protein and other electrolytes present in the blood plasma. Moreover, when the cyanide (10 mM) was added to the probe solution containing blood plasma (10 μM) significant fluorescence enhancement occurred in which relative fluorescence intensity enhanced approximately ∼7.7 times and an emission band appeared at 450 nm (Fig. 12). Notably, the readily detectable naked-eye sensitive nonfluorescent color of solution changed to blue (Fig. 12, inset). Further, the usability of 3 to detect cyanide in biological samples has been demonstrated by performing emission titration experiment in blood plasma. It was interesting to mention that upon a gradual addition of CN⁻ (0–100 equiv.) to a solution of 3 (10 μM) in blood plasma, relative emission intensity enhanced successively and emission band appeared at 450 nm (Fig. 12). The change in emission spectral behavior of 3 in biological sample was almost similar to those observed in aqueous medium with cyanide. Furthermore, the LOD to detect cyanide in blood plasma has been estimated, as mentioned above and was found to be 0.37 μM (9.62 ppb) (Fig. S37†).

Fig. 12 (a) Interaction and (b) titration emission spectra of 3 (10 μM, HEPES buffer–human blood plasma; 9.8 : 0.2, v/v; pH 7.4) with CN⁻ ions (10 mM) in aqueous medium. Inset: color change after addition of CN⁻ in 3 containing human blood plasma.

Conclusions

In summary, we have designed and synthesized a multichannel coumarin derived chemo and fluorogenic probes as a potential chemosensor to detect fluoride, acetate and cyanide anions in different medium. The probe has shown a naked eye detectable colorimetric response to recognize fluoride and acetate anions in acetonitrile. However, in aqueous medium probe has shown high fluorogenic affinity to detect cyanide anion through a reaction based chemodosimeter approach in which considerable enhanced emission, “turn-on” occurred and the color of solution switched-on to a fluorescent blue-green. Moreover, the present sensor motifs has shown high sensitivity to detect cyanide in real water sample, on test paper strips, and in human blood plasma through a reaction based chemodosimeter approach.

Experimental

General

Salicylaldehyde, aniline (nitro, methoxy) and sodium nitrite were purchased from Sigma Aldrich Pvt. Ltd. Ethanol and metals as their nitrate salts were purchased from Merck India Pvt. Ltd. Spectroscopic grade solvents were used in spectroscopic studies. The absorption spectra was recorded at room temperature on a Shimadzu 1700 and Perkin-Elmer Lambda 35 spectrophotometer using a quartz cuvette (path length = 1 cm). FTIR spectra (KBr pellets) were recorded on a Varian-3100 spectrometer. NMR spectra (chemical shifts in δ ppm) were recorded on a JEOL AL 300 FT-NMR (300 MHz) spectrometer using tetramethylsilane (TMS) as an internal standard.

The absorption and fluorescence titration data were utilized to calculate association constants by Benesi–Hildebrand method (B–H method) employing eqn (1) for a 1 : 1 stoichiometry.

1/(I − I₀) = 1/(I − I_f) + 1/K(I − I_f)[M] (1)

where K is the association constant, I is the absorbance/fluorescence intensity of the free probe 3/4, I₀ is the observed absorbance/fluorescence intensity of the 3–CN⁻/F⁻/AcO⁻, and I_f is the absorbance/fluorescence intensity at saturation level.

Limit of detection (LOD) has been estimated using eqn (2)

LOD = 3σ/m (2)

where, σ stands for the standard deviation of blank solution of 3/4 and m stands for calibration sensitivity toward F⁻, AcO⁻ and CN⁻ in MeCN and aqueous solution of 3/4.

Synthesis of compound 2

To a solution of salicylaldehyde (1.22 g, 10.0 mmol) in ethanol (20 ml) ethylacetoacetate (1.95 g, 1.5 mmol) and catalytic amount of piperidine were added drop wise for 30 min with stirring. The reaction mixture was refluxed with constant stirring for 6 h. The completion of reaction was judged with the consumption of staring material (monitor on TLC). Now the reaction mixture was allowed to stand at room temperature for 10 h. A yellowish white color precipitate so obtained was filtered and washed with ethanol (2 × 5 ml). The product was recrystallized in ethanol (8 ml) to afford off-white color compound. Yield 90%. R_f = 0.6 (hexane–ethyl acetate: 3/2, v/v). ¹H NMR (CDCl₃) δ (ppm): 8.50 (s, 1H, benzyl), 7.66 (t, 2H, J = 9 Hz, benzyl), 7.39 (m, 2H, benzyl), 2.73 (s, 3H, methoxy); IR (KBr) ν_max (cm⁻¹); 3102, 2840, 1664, 1608, 1577, 1526, 1478, 1286, 1171, 1143, 1102, 906, 855, 751, 719, 687, 578, 452.

Synthesis of compound 3

Compound 2 (188 mg, 1 mmol) and 2,4-dinitrophenylhydrazine (198 mg, 1 mmol) were constituted in warm ethanol (10 ml) and refluxed the reaction mixture for 4 h. After complete chemical reaction (monitored on TLC) solvent was evaporated in vacuum and washed with ethanol (3 × 1 ml) and recrystallized with ethanol (5 ml) to afford orange red color crystalline compound 3. Yield 70%. R_f = 0.5 (hexane–ethyl acetate: 3/2, v/v). ¹H NMR (CDCl₃) δ (ppm): 11.36 (s, 1H, NH), 9.15 (s, 1H, benzyl), 8.36 (d, 2H, J = 9.6 Hz, benzyl), 8.09 (d, 2H, J = 4.2 Hz, benzyl) 8.04 (s, 1H, benzyl), 7.62 (t, 2H, benzyl), 7.39 (m, 2H, benzyl), 2.48 (s, 3H, methyl). IR (KBr) ν_max (cm⁻¹); 3318, 3117, 1727, 1614, 1594, 1447, 1422, 1370, 1330, 1281, 1133, 1105, 1058, 967, 833, 770, 741, 680, 637, 586, 511, 461. ¹³C NMR (CDCl₃): 225.6, 192.1, 185.9, 150.4, 144.5, 128.7, 124.8, 116.7, 108.5, 77.4, 62.9, 54.9, 30.2. m/z⁺ (ESI-MS) = 368.9.

Synthesis of compound 4

Similarly, compound 2 (188 mg, 1 mmol) and phenylhydrazine (108 mg, 1 mmol) were reacted in ethanol (8 ml) to afford compound 4. Yield 70%. R_f = 0.6 (hexane–ethyl acetate: 3/2, v/v). ¹H NMR (CDCl₃) δ (ppm): 7.97 (s, 1H, benzyl), 7.47 (d, 4H, J = 12.9 Hz, benzyl), 7.27 (m, 4H, benzyl) 7.09 (d, 2H, J = 7.8 Hz, benzyl), 6.83 (s, 1H, NH), 2.22 (s, 3H, methyl); IR (KBr) ν_max (cm⁻¹); 3301, 3054, 1720, 1702, 1601, 1577, 1550, 1494, 1445, 1375, 1357, 1257, 1162, 1119, 1023, 964, 876, 781, 688, 639, 579, 506, 454. ¹³C NMR (CDCl₃): 190.7, 160.4, 153.6, 144.4, 139.5, 131.3, 129.2, 124.4, 120.7, 116.3, 113.3, 77.4, 13.8.

General synthesis for 3/4–CN adduct

Compound 3 (92 mg, 0.25 mmol) and 4 (69 mg, 0.25 mmol) were treated with NaCN (13 mg, 0.25 mmol) in methanol (4 ml) for 2 h at room temperature. The precipitate so obtained was washed with water, to get respective dark reddish brown and dark yellow color precipitate of adducts in good yield.

Adduct 3–CN

Yield 72%. ¹H NMR (CDCl₃) δ (ppm): 11.38 (s, 1H, NH), 9.17 (s, 1H, benzyl), 8.39 (d, 1H, J = 11.4 Hz, benzyl), 8.11 (d, 2H, J = 4.2 Hz, benzyl), 7.64 (t, 2H, J = 16.8 Hz, benzyl), 7.41 (m, 2H, benzyl), 4.09 (s, 1H, CH), 2.50 (s, 3H, CH₃); IR (KBr) ν_max (cm⁻¹); 3436, 2924, 2219, 1734, 1597, 1524, 1489, 1456, 1257, 1208, 1048, 1026, 969, 755, 702. ¹³C NMR (CDCl₃): 173.5, 160.9, 137.8, 132.0, 129.4, 128.3, 124.2, 118.1, 115.7, 77.4, 40.2. ESI-MS: m/z [3 + CN + 2H]⁺ = 396.9.

Adduct 4–CN

Yield 74%. ¹H NMR (CDCl₃) δ (ppm): 8.05 (s, 1H, OH), 7.61 (d, 1H, J = 3.3 Hz, benzyl), 7.56 (d, 1H, J = 4.8 Hz, benzyl), 7.49 (d, 1H, J = 16.5 Hz, benzyl), 7.35 (m, 4H, benzyl), 6.92 (s, 1H, NH), 3.86 (s, 1H, CH), 2.30 (s, 2H, methyl); IR (KBr) ν_max (cm⁻¹); 3304, 2254, 1735, 1602, 1488, 1445, 1393, 1257, 1155, 1120, 1026, 966, 864, 755, 688, 622, 508. ¹³C NMR (CDCl₃): 140.0, 131.0, 129.9, 129.3, 126.8, 123.8, 122.4, 118.0, 77.4, 25.4.

Acknowledgements

Authors are thankful to the Council of Scientific and Industrial Research (CSIR), New Delhi for financial support and senior research fellowships (to SSR, RA, PS and MS).

References

1. (a) P. A. Gale, Coord. Chem. Rev., 2001, 213, 79–128; (b) J. L. Sessler and J. M. Davis, Acc. Chem. Res., 2001, 34, 989–997; (c) C. Suksai and T. Tuntulani, Chem. Soc. Rev., 2003, 32, 192–202.
2. M. E. Moragues, R. Martinez-Manez and F. Sancenon, Chem. Soc. Rev., 2011, 40, 2593–2643.
3. (a) S. Y. Kim and J.-I. Hong, Org. Lett., 2007, 9, 3109–3112; (b) T. Gunnlaugsson, A. P. Davis, J. E. O'Brien and M. Glynn, Org. Lett., 2002, 4, 2449–2452; (c) E. J. Jun, K. M. K. Swamy, H. Bang, S.-J. Kim and J. Yoon, Tetrahedron Lett., 2006, 47, 3103–3106; (d) J. Yoon, S. K. Kim, N. J. Singh and K. S. Kim, Chem. Soc. Rev., 2006, 35, 355–360.
4. P. A. Gale and R. Quesada, Coord. Chem. Rev., 2006, 250, 3219–3244.
5. (a) K. W. Kulig, Cyanide Toxicity, U.S. Department of health and human services, Atlanta, GA, 1991; (b) Z. Xu, X. Chen, H. N. Kim and J. Yoon, Chem. Soc. Rev., 2010, 39, 127–137; (c) S. K. Kwona, S. Kou, H. N. Kim, X. Chen, H. Hwang, S.-W. Nama, S. H. Kim, K. M. K. Swamya, S. Park and J. Yoon, Tetrahedron Lett., 2008, 49, 4102–4105.
6. G. C. Miller and A. Pritsos, Cyanide: Social, Industrial and Economic Aspects Proceeding of the TMS Annual Meeting, 2001, p. 73.
7. B. Vennesland, E. F. Comm, J. Knownles, J. Westly and F. Wissing, Cyanide in Biology, Academic, London, 1981.
8. (a) C. Young, L. Tidwell and C. Anderson, Cyanide: Social, Industrial, and Economic Aspects, Minerals, Metals, and Materials Society, Warrendale, 2001; (b) J. D. Johnson, T. L. Meisenheimer and G. E. Isom, Toxicol. Appl. Pharmacol., 1986, 84, 464; (c) R. Koenig, Science, 2000, 287, 1737; (d) M. Shahid and A. Misra, Anal. Methods, 2013, 5, 434–437.
9. (a) G. C. Miller and A. Pritsos, Proceeding of the TMS annual meeting, 2001, p. 73; (b) M. Sheffer, Guidelines for drinking-water quality, 3rd edn, vol. 1, World Health Organization, Geneva, Switzerland, 2008; (c) EPA National Primary Drinking Water Standards 2003, see http://www.epa.gov/safewater/contaminants/index.html for more information.
10. R. Sheng, P. Wang, Y. Gao, Y. Wu, W. Liu, J. Ma, H. Li and S. Wu, Org. Lett., 2008, 10, 5015–5018.
11. H. J. Kim, S. Bhuniya, R. K. Mahajan, R. Puri, H. Liu, K. C. Ko, J. Y. Lee and J. S. Kim, Chem. Commun., 2009, 7128–7130.
12. I. J. Shenderovich, P. M. Tolstoy, N. S. Golubev, S. N. Smirnov, G. S. Denisov and H. H. Limbach, J. Am. Chem. Soc., 2003, 125, 11710–11720.
13. E. B. Veale, G. M. Tocci, F. M. Pfeffer, P. E. Kruger and T. Gunnlaugsson, Org. Biomol. Chem., 2009, 7, 3447–3454.
14. A. Misra, M. Shahid and P. Dwivedi, Talanta, 2009, 80, 532–538.
15. C. R. Bondy and S. J. Loeb, Coord. Chem. Rev., 2003, 240, 77–99.
16. A. Helal and H. S. Kim, Tetrahedron, 2010, 66, 7097–7103.
17. J. Shao, H. Lin, M. Yu, Z. Cai and H. Lin, Talanta, 2008, 75, 551–555.
18. (a) A. Misra and M. Shahid, J. Phys. Chem. C, 2010, 114, 16726–16739; (b) M. Shahid, P. Srivastava and A. Misra, New J. Chem., 2011, 35, 1690–1700.
19. J. Isaad and A. E. Achari, Tetrahedron, 2011, 67, 4196–4201.
20. Y. K. Tsui, S. Devaraj and Y. P. Yen, Sens. Actuators, B, 2012, 161, 510–519.
21. R. Martinez-Manez and F. Sancenon, Chem. Soc. Rev., 2003, 103, 4419–4476.
22. (a) R. Badugu, J. R. Lakowicz and C. D. Geddes, J. Am. Chem. Soc., 2005, 127, 3635–3641; (b) Y. Kim, H. Zhao and F. P. Gabbai, Angew. Chem., 2009, 48, 4957–4960.
23. A. Touceda-Varela, E. I. Stevenson, J. A. Galve-Gasion, D. T. F. Dryden and J. C. Mareque-Rivas, Chem. Commun., 2008, 1998–2000.
24. N. Gimeno, X. Li, J. R. Durrant and R. Vilar, Chem.–Eur. J., 2008, 14, 3006–3012.
25. Q. Zou, X. Li, J. Zhang, J. Zhou, B. Sun and H. Tian, Chem. Commun., 2012, 48, 2095–2097.
26. (a) M. Shahid, S. S. Razi, P. Srivastava, R. Ali, B. Maiti and A. Misra, Tetrahedron, 2012, 68, 9076–9084; (b) M. Shahid and A. Misra, Anal. Methods, 2013, 5, 434–437.
27. Y. M. Hijji, B. Barare, A. P. Kennedy and R. Butcher, Sens. Actuators, B, 2009, 136, 297–302.
28. M. Boiocchi, L. Del Boca, D. E. Gomez, L. Fabbrizzi, M. Licchelli and E. Monzani, J. Am. Chem. Soc., 2004, 126, 16507–16514.
29. H. A. Benesi and J. H. Hildebrand, J. Am. Chem. Soc., 1949, 71, 2703–2707.
30. M. J. Frisch, G. W. Truncks and H. B. Schlegel, GAUSSIAN 03, Revision E.01, Gaussian Inc., Wallingford, CT, 2007.
31. P. Srivastava, R. Ali, S. S. Razi, M. Shahid and A. Misra, Sens. Actuators, B, 2013, 181, 584–595.
32. S. Kumar, P. Singh, G. Hundal, M. Singh and S. Kumar, Chem. Commun., 2013, 49, 2667–2669.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c4ra023388a

---