Optimizing the Electron-Withdrawing Character on Benzenesulfonyl

Article pubs.acs.org/ac

Optimizing the Electron-Withdrawing Character on Benzenesulfonyl Moiety Attached to a Glyco-Conjugate to Impart Sensitive and Selective Sensing of Cyanide in HEPES Buffer and on Cellulose Paper and Silica Gel Strips Sivaiah Areti, Sateesh Bandaru, Deepthi S. Yarramala, and Chebrolu Pulla Rao* Bioinorganic Laboratory, Department of Chemistry, Indian Institute of Technology Bombay, Mumbai, Powai, Mumbai 400 076, India S Supporting Information *

ABSTRACT: Dansyl-derivatized, triazole-linked, glucopyranosyl conjugates, 5FLOH, 2FLOH, 1FLOH, and 0FLOH were synthesized and characterized. While the 5FLOH acts as a molecular probe for CN−, 2F LOH, 1FLOH, and 0FLOH acts as control molecules. The reactivity of CN− toward 5FLOH has been elicited through the changes observed in NMR, ESI MS, emission, and absorption spectroscopy. The conjugate 5FLOH releases a fluorescent product upon reaction by CN− in aqueous acetonitrile medium by exhibiting an ∼125-fold fluorescence enhancement even in the presence of other anions. Fluorescence switch-on behavior has been clearly demonstrated on the basis of the nucleophilic substitution reaction of CN− on 5FLOH. A minimum detection limit of (2.3 ± 0.3) × 10−7 M (6 ± 1 ppb) was shown by 5FLOH for CN− in solution. All the other anions studied showed no change in the fluorescence emission. The utility of 5FLOH has been demonstrated by showing its reactivity toward CN− on a thin layer of silica gel as well as on Whatman No. 1 cellulose filter paper strips. The role of glucose moiety and the penta-fluorobenzenesulfonyl reactive center present in 5FLOH in the selectivity of CN− over other anions has been demonstrated by fluorescence, absorption and thermodynamics study. Similar studies carried out with the control molecules showed no selectivity for CN−. The mechanistic aspects of the reactivity of CN− toward 5FLOH were supported by DFT computational study.

T

the small molecular probes reported in the literature for sensing CN− suffer from their low sensitivity, higher detection limit and nonselectivity, besides their turn-off nature.32−35 Therefore, there is a dire need for further research in addressing these short comings. All these were proposed to overcome by using a probe where a sensitive and selective electrophilic reaction center and a fluorescent probe were covalently connected to glucose (5FLOH) through a triazole moiety. The reaction of 5FLOH by cyanide ion is expected to result in generating fluorescent species due to its nucleophilic attack on the N-bound penta-fluorobenzenesulfonyl (PFBS) group. This has been demonstrated by various techniques as reported in this paper. The role of glucose moiety has been delineated by comparing the results obtained from its precursor molecular systems that possesses protected O-acetyl moiety. However, the reactivity of PFBS can be better understood when the data obtained in case of nonreactive analogues (2FLOH, 1F LOH, and 0FLOH) were compared. The reaction between 5FLOH and CN− was studied by NMR, ESI MS, emission, and absorption spectroscopy. Thus, for the first time, we report,

he detection of species, which are toxic to physiological systems, received considerable attention owing to its deleterious effect on human health.1−7 Among these, the cyanide ion stands as a crucial one since it is discharged into the environment due to increased utility of the compounds of cyanide in many chemical processes, such as electroplating, plastics manufacturing, tanning, and metallurgy.8−12 Cyanide and its other forms of species, such as, HCN, CNCl, NaCN, and KCN, will easily pass through the gastrointestinal tract and affects the hypoxia and lactate acidosis and thereby disturbs the central nervous system causing respiratory arrest that is responsible for death.13,14 Therefore, the development of small molecular probes having sensitive and selective detection of cyanide ions is still important. Among various strategies adopted in the literature, such as, hydrogen bonding,15−17 nucleophilic addition,18−22 and cyanide affinity to metal complexes,23−26 the one that is based on the chemical reaction of small molecular fluorescent probe attracts attention due to its high sensitivity and even the selectivity.27−30 To our knowledge, there have been no reports on the detection of CN− by a glycoconjugate. Though we find a copper complex of glycoconjugate that results in fluorescence enhancement for CN−, the same also shows fluorescence enhancement with H2PO4− and HSO4−, thus, turns out to be nonselective.31 Even © XXXX American Chemical Society

Received: October 26, 2015 Accepted: November 18, 2015

A

DOI: 10.1021/acs.analchem.5b04085 Anal. Chem. XXXX, XXX, XXX−XXX

Article

Analytical Chemistry Scheme 1. Synthesis of

nF

LOAc and

nF

LOH, Where n = 5, 2, 1, or 0a

(a) Et3N, CH2Cl2, rt, 1h; (b) catalytic amount of HClO4, acetic anhydride, 33 % HBr in CH3COOH, CH3CN, NaN3, reflux. 8h; (c) CuSO4·5H2O, sodium ascorbate, tert-BuOH : Water as 1:1 ratio, rt, 12 h; (d) different fluorobenzenesulfonyl chloride, THF, NaH, 0 °C, rt, 1 h; (e) anhydrous CH3OH, acetyl chloride, K2CO3, rt, 12 h. Inset: Intra- and intermolecular interactions noticed in the crystal structure of 5FLOAc. a

LOH as fluorescence turn-on as well as visually detectable receptor for CN− based on nucleophilic cleavage of the sulfonamide moiety. The real-time applicability of 5FLOH in detecting the CN− was demonstrated on silica gel sheet and on Whatman cellulose filter paper strips.

H NMR (400 MHz, D2O, δ ppm): 2.85 (s, 6H), 3.3 (s, 2H) 3.45−3.55 (m, 2H), 3.7−3.8 (m, 2H), 3.9 (d, J = 7.8 Hz, 1H), 4.15 (s, 2H), 5.3 (d, J = 7.8 Hz, 1H), 7.28 (d, J = 8.6 Hz, 1H), 7.56−7.50 (m, 3H), 7.85 (s, 1H), 8.25 (d, J = 8.4 Hz, 1H), 8.35 (d, J = 7.6 Hz, 1H), 8.55 (d, J = 8.6 Hz, 1H). 13C NMR (100 MHz, CD3OD, δ ppm): 30.8, 45.7, 62.4, 70.9, 74.0, 78.5, 81.2, 89.6, 116.8, 119.0, 121.2, 124.6, 125.6, 127.4, 130.6, 130.8, 133.0, 134.2, 134.3, 137.7, 143.8, 149.0, 151.6, 153.4. FTIR (KBr, cm−1): 1344 (ν−S=O), 2851 (ν−C−H), 2923 (νAr−C−H), and 3427 (ν−O−H). HRMS: chemical formula is C27H26F5N5O9S2 [M + Na+], calcd mass is 746.0984, found is 746.0981. Anal. Calcd for C27H26F5N5O9S2·CH3COOCH2CH3: C, 45.87%; H, 4.22%; N, 8.63%; S, 7.90%. Found: C, 45.50%; H, 3.85%; N, 8.18%; S, 7.67%.

5F

■

1

EXPERIMENTAL SECTION

Spectroscopy Studies. All the fluorescence and absorption titrations were carried out in 3:1 acetonitrile/(20 mM) aqueous HEPES buffer at pH 7.4 and this is simply referred as “aqueous acetonitrile”. The experiments were carried out by always using a 25 μL of bulk solution of 5FLOH (6 × 10−4 M) and the final solutions were maintained at 3 mL, resulting in a cuvette concentration of 5 μM 5FLOH, while the added anion solution will determine the mole ratio. For the absorption studies, the cuvette concentration of 5FLOH was kept constant at 10 μM, and all the other titration details are same as that used for fluorescence studies. The receptor 5FLOH (5 mM) was dissolved in 0.4 mL of CD3CN, and the 1H NMR spectra were recorded in absence as well as in the presence of different mole ratios of CN− by taking tetrabutylammonium cyanide (TBACN). Preparation of Samples on Silica Gel and Whatman Filter Paper Strips. TLC plate coated with silica gel and Whatman filter paper were cut into small strips of 3 × 1 cm2. The solution of the probe molecule was drop casted on the central portion of the strip and the solvent was dried by leaving it at room temperature. The dried strips were used as such for fluorescence measurements. Each of these experiments was repeated for three times in order to get error bars. Synthesis and Characterization. The synthesis and characterization details of the precursors and fluorinated derivatives were given in ESI 01−12 and that of the probe molecule (5FLOH) is given here. To a solution of 5FLOAc (890 mg, 0.01 mmol) in 3 mL of MeOH was added acetyl chloride (0.2 mL from a 0.02 M solution in MeOH) dropwise. The reaction mixture was stirred at room temperature until the starting material was consumed as checked by the TLC. The solution was neutralized by adding sodium bicarbonate and was filtered. The solvent was then removed in vacuum and obtained a solid product in 75% yield.

■

RESULTS AND DISCUSSION Synthesis and Characterization of 5FLOH, 2FLOH, 1FLOH, and 0FLOH. The probe and the control molecules, namely, 5F LOH, 2FLOH, 1FLOH, and 0FLOH have been synthesized by going through a number of steps starting from D-glucose on one hand and dansyl chloride on the other, as given in Scheme 1. The 1,3-dipolar addition reaction carried out between P2 and P4 resulted in a triazole precursor P5, which upon reaction with different fluorobenzenesulfonyl chlorides results in 5FLOAc, 2F LOAc, 1FLOAc, and 0FLOAc. These upon hydrolysis give the probe and control molecules, namely, 5FLOH, 2FLOH, 1FLOH, and 0F LOH with free −OH on carbohydrate moiety. The P2 is obtained from P1, and the P4 is obtained from D-glucose (P3) via its bromoderivative. All these were characterized by different techniques, such as, 1H and 13C NMR, FTIR, and ESI-MS (Experimental Section and Figures S01−S12, Supporting Information). Among these, the structure of 5FLOAc has been established by single-crystal XRD. Single Crystal XRD Structure of 5FLOAc. Among the derivatives reported in Scheme 1, single crystals of good diffraction quality were obtained only in case of 5FLOAc. The single crystals of 5FLOAc were obtained from slow evaporation of its solution in 1:2 vol/vol ratio of H2O/CH3OH. The 5FLOAc crystallizes in monoclinic system with P21 space group and the crystal structure was solved using the diffraction data and was B


Article

Analytical Chemistry refined (Figure S13, Supporting Information).36 The structure shows the glucose unit in its chair conformation with βanomeric form and the anomeric nature has been further deduced from 1H NMR spectra. The crystal structure shows that the molecule exists in an extended form. The distance between the oxygens of two different sulfonamide groups within the same molecule is 2.89 Å. The centroid to centroid distance between the two aromatic groups present within the same molecule of 5FLOAc (inset of Scheme 1) is 3.25 Å, indicating a strong intramolecular π···π interaction. Strong interaction has been possible, while one of the aromatic rings involved in this pair has electron withdrawing groups (PFB) and the other has electron-releasing groups (dansyl moiety). The centroid to centroid distance observed between the two neighbor molecules is 4.76 Å though the planar units are parallel suggesting that such π···π interactions are absent between the molecules in the lattice (inset of Scheme 1). In the lattice, the distance between the glucose C4−O (ether linked) and the triazole nitrogen of a neighbor molecule is 3.05 Å indicating the presence of some weak intermolecular interaction. The intermolecular F···π distance of 4.05 Å that is observed in the present case is shorter than that reported in the literature.37 Selective Reactivity of CN− on 5FLOH. The 5FLOH exhibits weak fluorescence emission at ∼520 nm when excited at 360 nm in aqueous acetonitrile owing to the intramolecular charge transfer (ICT).38,39 Due to the strong hydration effect of the cyanide ions,40,41 aqueous acetonitrile was used for all the studies reported in this paper. The weakly fluorescing 5FLOH has been used for the nucleophilic attack by anions at PFBS center, and this is expected to result in the formation of L1 (Figure 2d), which is fluorogenic as the cleaved product. To determine the effect of pH on the reaction of CN− with 5FLOH, fluorescence studies were performed in HEPES buffer solution and found that the emission intensity is high and unaltered during pH 7− 10 (Figure S14, Supporting Information). In order to study the reactivity of anions, 5FLOH (5 μM) was treated with these ions in HEPES buffer prepared in aqueous acetonitrile. Reaction of 5F LOH by cyanide results in the enhancement of fluorescence intensity as a function of added CN− and exhibits a highest fluorescence intensity of 125 ± 6-fold (Figure 1a). Similar reactions were carried out with other anions, viz., halides (F−, Cl−, Br−, I−), nitrogen-based anions such as SCN− and N3−, oxo-anions, such as, NCO−, CH3COO−, CO32−, NO2−, NO3−, SO42−, ClO4−, HSO4−, and HCO3−, and phosphate-based anions {H2PO4− (Pi) and disodium salts of nucleotides, namely, AMP 2− , ADP 2− , and ATP 2− } and the thiol nucleophiles, such as, cysteine, homocysteine, glutathione, thiophenol, and mercaptopropionic acid, but found no significant change in the fluorescence intensity of the reaction mixture (Figure S15, Supporting Information). Even the fluorescence titration studies carried out in the presence of metal ions showed no change in the original emission intensity (Figure S15, Supporting Information). The unaltered emission behavior in the presence of these ions is suggestive of their nonreactive nature toward 5FLOH (Figure 1b). The detection limit for CN− ion is (2.3 ± 0.3) × 10−7 M (6 ± 1 ppb) (Figure S16, Supporting Information). The probe 5FLOH showed lower detection limit as compared to the recently developed fluorescent probes in the literature for CN− (Figure S17, Supporting Information). The color changes from non fluorescent (in the absence of CN−) to greenish-yellow fluorescence only in the presence of CN− (under 365 nm

Figure 1. (a) Fluorescence spectra obtained during the titration of 5F LOH (5 μM) with CN− in aqueous acetonitrile HEPES buffer at pH = 7.4 at λex = 360 nm. Inset: Fluorescence intensity vs mole ratio of CN− at 520 nm. (b) Histogram showing the relative fluorescence intensity (I/I0) of 5FLOH at 520 nm in the presence of anions (100 μM). The top inset: Vials exhibiting the color upon reaction with the corresponding anions under 365 nm UV lamp. (c) Histogram showing the fluorescence intensity ratio (I/I0) at 520 nm with CN−. (d) Absorption spectra obtained during the titration of 5FLOH (10 μM) with CN− (100 μM). The inset shows plots of absorbance vs {[CN−]/ [5FLOH]} for three different bands.

UV light) and not in the presence of any other anion studied (inset of Figure 1b). This suggests that the CN− brings effective cleavage of the electron-withdrawing PFBS moiety from the weakly fluorescent probe 5FLOH to release L1 and L2 (Figure 2d). Thus, 5FLOH clearly distinguishes the CN− from all the other anions. Even the precursor tetra-acetylated molecule (5FLOAc) exhibits fluorescence enhancement upon reacting with CN−, however, the extent to which the enhancement takes place is low in case of 5FLOAc as compared to that observed for 5FLOH. Indeed the fold of enhancement is ∼125 in the case of 5FLOH, while it is only ∼30 in the case of 5FLOAc (Figure 1c). This is attributed to the difference observed in the original fluorescence intensity of these two molecules, namely, 5FLOAc and 5FLOH in the absence of CN−. Besides this, the 5FLOAc is less soluble in water as compared to 5FLOH, supporting that the 5F LOH is a better probe for CN− as compared to its acetylated precursor, 5FLOAc. In order to further support the reaction of CN− on 5FLOH, the absorption studies were carried out. The 5FLOH shows absorption bands at 360 and 220 nm with a shoulder at ∼263 nm. Upon the reaction of 5FLOH by CN−, the absorbance of the 360 and 220 nm bands increases, while the absorbance at ∼283 nm decreases (Figure 1d and inset). Under these conditions, the isosbestic point is observed at 325 nm. The reaction results in a bathochromic shift of the 360 nm band to appear at 395 nm at higher equiv of CN− and the same is associated with the change of its color from green to brown. However, when 5FLOH is treated with other anions, no significant spectral changes were observed (Figure S18, Supporting Information). Since the spectral changes were observed only with CN−, this is attributable to its unique reactivity toward 5FLOH. Reaction Products of CN− on 5FLOH by NMR and Mass Spectroscopy. In order to support the reaction of CN− on 5F LOH, 1H NMR spectra of 5FLOH (5 mM) were measured as a C


Article

Analytical Chemistry

Figure 2. (a) 1H NMR spectra measured during the titration of 5FLOH with CN− in CD3CN: (i) 5FLOH followed by CN−, where at room temperature (ii) 2.5, (iii) 5, and (iv) 10 equiv of CN−. The 1H NMR spectra shown in (v) and (vi) are the reaction of 5FLOH treated with 7.5 equiv of CN− at −10 and −20 °C, respectively. (b) 19F NMR spectra measured during the titration of (i) 5FLOH, and (ii) {5FLOH + 15 equiv CN−} in CD3CN. (c) ESI MS spectra obtained for the reaction mixture when 5FLOH is treated with 5 equiv of CN−. (d) Products of the reaction of 5FLOH with CN−.

Figure 3. Spectra obtained in the fluorescence titration of 5FLOH (λex = 360 nm) with CN−: (a) Whatman No. 1 cellulose filter paper strips and (b) silica gel strips. (c) Plot of emission intensity vs. concentration of CN− at 520 nm. The black one is on Whatman No. 1 cellulose filter paper strip, and the red one is from silica gel strips. (d) Photographs taken under UV light (365 nm) of 5FLOH-coated Whatman No. 1 cellulose filter paper (upper panel) and silica gel strips (lower panel) upon addition of increasing concentration of CN− from (i) to (viii): (i) 0, (ii) 0.010, (iii) 0.025, (iv) 0.1, (v) 0.25, (vi) 0.5, (vii) 0.75, and (viii) 1.0 mM. (e) The histogram showing the detection limit and relative fluorescence intensity (I/I0) at 520 nm of CN− by 5FLOH.

function of added CN − (0−15 mol equiv) and the corresponding spectra are shown in Figure 2a. Up to an addition of 5 equiv of CN−, the signals corresponding to 5FLOH, namely, 7.38, 7.74, 7.8, 8.1, 8.22, 8.52, and 8.74 ppm, diminishes in their intensity and disappear at higher mole ratio of CN−. This change is accompanied by the appearance of peaks corresponding to L1, namely, 5.4, 7.25, 7.85, 8.2, 8.31, and 8.56 ppm (Figure 2a). In order to find the ternary reaction species formed between CN− and 5FLOH, 1H NMR titrations were carried out at different temperatures starting from +30 down to −20 °C and found the spectra to be complex enough so as to possess the starting species, intermediate and the products at temperatures below −10 °C (Figure 2a). All this clearly support the release of L1 and L2 from 5FLOH due to the nucleophilic attack of CN− at PFBS center present in 5FLOH. Since the product L2 has fluorine bound aromatic moiety, 19F NMR spectrum of 5FLOH measured in CD3CN showed three

peaks at −134.4, −146.1, and −160.2 ppm corresponding to o-, m-, and p-F bound phenyl moiety. When 15 equiv of CN− is added to 5FLOH, the ortho- and para-F’s were shifted downfield by 42.4 and 7.9 ppm, respectively, and the meta-F is shifted to upfield by 0.7 ppm and all this supports the formation of L2 (Figure 2b). This spectrum matches well with the spectrum of pentafluoro-cyanobenzene reported in the literature42 and, hence, the product L2 is confirmed. The ESI MS showed peaks corresponding to these products, namely, L1 and L2 at m/z = 494.17 [L1 + H]+ and 232.28 [L2 + K]+, respectively (Figure 2c and Figure S19, Supporting Information). Thus, both the 1H NMR and mass spectra supported the nucleophilic attack of CN− on 5FLOH resulting in the release of L1 and L2, Figure 2d. 5F LOH as Reactive Probe for CN− when Coated on Silica Gel and Cellulose Paper Strip. In order to use 5FLOH as probe for the detection of CN− in various samples routinely, D


Article


Figure 4. (a) Schematic representation of the molecular structures of probe and control molecules. Fluorescence spectra obtained during the titration of (b) 5FLOH, (c) 2FLOH, (d) 1FLOH, and (e) 0FLOH (5 μM) with CN− in aqueous acetonitrile medium at λex = 360 nm). (f) Plot of fluorescence intensity vs [CN−]/[probe] mole ratio at 520 nm. (g) Histogram showing the relative fluorescence intensity (I/I0) of molecular probe at 520 nm in the presence of CN−. ITC thermograms for the reaction of CN− with 5FLOH, 2FLOH, 1FLOH, and 0FLOH with CN− in aqueous acetonitrile. (h) Baseline-corrected raw data for heats of reaction vs time of [CN−]/[probe]. (i) Heats of reaction vs the mole ratio of [CN−]/ [probe]. The solid line in the lower panel is a best fit, obtained upon using the one-site model. (j) Schematic representation of the molecular structures of control molecules. Fluorescence spectra obtained during the titration of (k) 2NO2LOH, (l) NO2LOH, and (m) MeLOH (5 μM) with CN− in aqueous acetonitrile medium at λex = 360 nm). (n) Plot of the relative fluorescence intensity (I/I0) vs [CN−]/[Probe] mole ratio at 520 nm.

eye (Figure 3d). Thus, the sensitivity of the reaction of 5FLOH toward CN− follows a trend, namely, solution >40× Whatman filter paper ≫100× silica gel, as can be seen from Figure 3e. Thus, the detection of CN− on Whatman filter paper is about 2.5× more sensitive than that on the silica gel. Reactivity Comparison among Different Fluorinated Derivatives toward Cyanide. The control molecular probes 2F LOH, 1FLOH, and 0FLOH (Figure 4a) were synthesized in order to understand the reactivity of the CN− ion toward 5FLOH, while keeping in mind the importance of the penta fluorobenzene group present in 5FLOH. However, in the case of 2FLOH, only 6 ± 1-fold of enhancement was observed with CN− ion because of the presence of semifluorinated moiety (2,4-difluorobenzene; Figure 4c), suggesting that the necessity of electron-withdrawing group and the importance of pentafluoro benzene moiety in the detection of CN−. Thus, the fold of enhancement with 5FLOH is 125 ± 12, while it is only 6 ± 1 in the case of 2FLOH. Similar titrations were carried out with the control molecules possessing 1F and 0F, namely,

an inexpensive and use-and-throw method has been developed by coating the silica gel sheets and Whatman No. 1 cellulose filter paper strips with 5FLOH (72.3 μg). Increasing concentration of CN− were added to such silica gel sheets to result in different mole ratios [CN−]/[5FLOH] in the range 0 to 50 and the sheets were allowed to dry. The corresponding fluorescence spectra and the color exhibited by these samples under UV light are shown in Figure 3. The probe 5FLOH exhibits weak fluorescence emission centered at 520 nm and shows increase in the emission intensity as the CN− concentration increases (Figure 3a,b). As the fluorescence intensity ratio plot is fairly linear in 10−250 μM of CN−, the detection of CN− by fluorescence is practically feasible using Whatman No. 1 cellulose filter paper strips and silica gel sheet in this concentration range (Figure 3c and Figure S20, Supporting Information). By using these 5FLOHcoated Whatman No. 1 cellulose filter paper and silica gel strips, concentrations as low as 10 (270 ppb) and 25 μM (670 ppb), respectively, of CN− is detectable under UV light by the naked E


Article


Figure 5. Reaction coordinate profile of addition of −CN¯ to 5FLOH via TS to form [Pr−H] at wB97xD/6-31G(d,p) level of theory and gas phase relative free energies and enthalpies are given in kcal/mol. The frontier molecular orbitals (MOs) of (b) 5FLOH (HOMO), (c) 5FLOH (LUMO), (d) L1 (HOMO), and (e) L1 (LUMO) for Pr (i.e., the cleaved product of L1 by −CN−). The energy levels of the MOs are shown (eV).

group, but has one well-known electron-releasing −CH3 group (MeLOH) was studied and found that this is not at all reactive toward CN− (Figure 4m). All this supports that strongly electron-withdrawing groups are essential to have a fruitful CN− attack at PFBS (Figure 4n). Nucleophilic Attack of CN− on 5FLOH by DFT Computational Study. In order to explain the reaction coordinate profile of 5FLOH by −CN−, the structure of the 5FLOH was built from the coordinates obtained from the crystal structure of its precursor (5FLOAc) and the computational studies were performed at wB97xD level of theory using 6-31G(d,p) basis set.44,45 The initial nucleophilic attack of CN− on PFBS present in 5FLOH led to the preproduct complex [Pr−H] by going through a substantially low energy transition state (TS) as can be seen from Figure 5. The activation free energy barrier of CN− addition to the probe molecule 5FLOH is −5.67 kcal/mol, the corresponding reaction free energy is −72.42 kcal/mol, and the corresponding enthalpy barrier and the reaction energies are also favorable for the CN− addition reaction (Figure 5). The nucleophilic attack of CN− on 5FLOH can be further supported by comparing the C···C bond distance in the transition state (2.059 Å) and the same in the corresponding product (in Pr−H: 3.989 Å; Figure 5a). The Figure 5 clearly reveals that there is no bonding between the pentafluoro cyanobenzene and the corresponding sulfur center (i.e., −SO2). This has been further confirmed by optimizing the pentafluoro cyanobenzene and SO2 independently in the gas phase as individual structures. From the free energy barriers and enthalpies, it can conclude that the cleavage of 5FLOH by CN− is a favorable process, and the reaction was more exothermic in nature. The exothermic reactivity was also supported by the ITC studies, as reported in this paper. Rationale for Fluorescence Sensing by Time-Dependent DFT. In order to rationalize the fluorescence sensing mechanism, the excited state calculations were carried out with the TDDFT. The ground state structures of the 5FLOH and the product (“Pr” is same as L1) were optimized in the gas phase with wB97xD using 6-31G(d,p) basis set in DFT (Figure 5b− e). The excited state calculations were carried out on the optimized geometries of the ground state (TDDFT//DFT optimized structures) and the corresponding results are given in Figure S22 of the Supporting Information. It is understood from this data that the S1 state is not a completely dark state, as supported by the oscillator strength of S0 → S1 (f = 0.1439), which conveys that the molecule 5FLOH is weakly fluorescent, and this is in agreement with the experimental results. For the CN− cleaved product (Pr), the oscillator strength of the S0 →

1F

LOH and 0FLOH and found no significant change in the fluorescence intensity during the addition of CN− ion. All this suggests that at least two F’s are required to be present on the aromatic moiety to expedite the CN− attack at PFBS (Figure 4d,e). All the data supports the necessity of both the pentafluoro benzene and the glucose moiety present in 5FLOH toward CN− detection (Figure 4f,g). In order to understand the thermodynamic aspects of binding of CN− by 5FLOH, 2FLOH, 1FLOH, and 0FLOH, isothermal titration calorimetry (ITC) studies were performed, and the corresponding thermograms were given in Figure 4h and also in the Supporting Information as Figure S21. While the interaction of CN− with 5FLOH and 2FLOH is exothermic, that of 1F LOH and 0FLOH showed no significant heat changes. The negative ΔH values obtained for 5FLOH (−110 kcal/mol) and 2F LOH (−35 kcal/mol) indicates the exothermic nature of the reaction and, thus, its feasibility (Figure 4i). On the other hand, the negative ΔS values 5FLOH (−354 cal/mol/deg) and 2FLOH (−97.2 cal/mol/deg) are indicative of the formation of the complex. In the case of 1FLOH and 0FLOH, no significant heat change was observed, supporting that no reaction takes place in this case. All these features can be comparatively understood from Figure 4i. Reactivity Comparison with Other Derivatives. Since a greater number of fluorines in the derivative led to higher reactivity by CN− and thereby higher sensitivity, the CN− reactivity studies were further extended to the derivatives containing nitro group(s), which has more electron-withdrawing power than that of fluorine (Figure 4j). Thus, two of the nitro derivatives, namely, 1NO2LOH and 2NO2LOH were studied for CN− reactivity and found enhancement in the fluorescence intensity by 15 ± 2 and 200 ± 25-fold (Figure 4k,l). Comparison of this data with that observed for 5FLOH (125 ± 15-fold) reveals that the sensitivity of 1NO2LOH is too low and that of the 2NO2LOH is too high, supporting that the 2NO2 LOH would be a better probe for CN− as compared to 5F LOH. However, when one looks at the reactivity of 2NO2LOH toward other anions, such as thiol-containing molecules, Cys shows high sensitivity, a result that we published recently.43 Though the 2NO2LOH shows greater fluorescence enhancement with CN− than that of 5FLOH, the former simply is nonselective as the same is sensitive toward both the Cys (other −SHcontaining molecules) as well as CN−. However, the 5FLOH does not undergo any reaction by Cys or other −SH molecules (Figure S14, Supporting Information), supporting that the 5F LOH is a better probe, possessing high selectivity. In addition, another derivative that has neither the fluorine nor the nitro F


Article

Analytical Chemistry Scheme 2. Schematic Representation of Different Features Noticed in Sensing CN− by

5F

LOH

and 0FLOH. Even at higher equivalents of CN−, the reaction does not contribute to the fluorescence intensity of 1FLOH and 0F LOH, however, 2FLOH shows fluorescence enhancement, but only to a 6-fold extent. Thus, the absence of PFBS does not hamper the detection of CN− by their reaction with 2FLOH, 1F LOH, and 0FLOH and hence, 5FLOH is a selective reactive probe. Thus, 5F is more selective to CN−, while the 2NO2 is more selective toward Cys, as reported by us recently,43 though both go through similar nucleophilic attack at PFBS center in the respective probe molecule. All these can be clearly noticed from Scheme 2. The feasibility has been supported by the negative ΔH values of CN− reaction with 5FLOH and 2FLOH and these are −110 and −35 kcal/mol, respectively, indicating the exothermic nature of the reaction. Further the CN− reaction with 1FLOH and 0FLOH showed no significant heat changes (Scheme 2). The DFT computations carried out for the CN− reaction with 5FLOH supported favorable process and the reaction was exothermic in nature. The TDDFT results showed that the oscillator strength of the S0 → S1 transition is 0.1919, supporting that the product should be more fluorescent than 5FLOH, and all these results were indeed observed experimentally. In order to carry out realtime applications, 5FLOH-coated filter paper and silica gel strips were studied at CN− concentrations as low as 10 and 25 μM, respectively. These were detected by observing the color under UV light. Thus, a weak fluorescent probe (5FLOH) reported in this paper is a highly promising molecular system for a real-time optical applications, mainly because of its specific and selective reactivity with CN− in the solution phase, on the silica gel, and on cellulose paper.

S1 transition is 0.1919, supporting that the product should be more fluorescent than 5FLOH, and this was indeed observed experimentally, as reported in this paper. In 5FLOH, the HOMO orbital is located on pentafluorobenzene and the LUMO is located on N,N-dimethylbenzene moiety (Figure 5b,c), whereas in the cleaved product, both the HOMO and LUMO are located on N,N-dimethylbenzene moiety supporting that the electronic transition is easy in the case of the product (Figure 5d,e).

■

CONCLUSIONS AND CORRELATIONS In summary, a reactive fluorescent probe (5FLOH) toward CN− has been synthesized by integrating the dansyl moiety linked through triazole and glucopyranosyl moiety. The conjugate 5F LOH is a reactive probe for CN− by switch-on fluorescence by ∼125-fold among the 20 anions and 5 thiols studied. The 5FLOH exhibited a minimum detection limit of (2.3 ± 0.3) × 10−7 M (6 ± 1 ppb) for CN− in aqueous acetonitrile. The fluorescence sensitivity is much higher with 5FLOH as compared to its acetylated precursor, 5FLOAc. The reactivity of CN− on 5FLOH was further supported by isosbestic point observed at 325 nm and a 35 nm bathochromic shift observed for the 360 nm band in the absorption study. The reactivity of CN− on 5FLOH has been demonstrated based on NMR, ESI-MS, emission, and absorption. The reactivity of CN− on 5FLOH has been demonstrated based on 1H NMR and ESI MS, where the signals corresponding to L1 and L2 appear, thereby supporting the release of these owing to the nucleophilic attack of CN− at the PFBS center. The role of glucose moiety as well as the PFBS reactive center in the selectivity of CN− over other anions has been proven by fluorescence study while the same studies show nonselectivity for CN− by the control molecules, 2FLOH, 1FLOH, G


Article


■

(20) Cheng, X. H.; Zhou, Y.; Qin, J. G.; Li, Z. ACS Appl. Mater. Interfaces 2012, 4, 2133−2138. (21) Park, S.; Kim, H. J. Chem. Commun. 2010, 46, 9197−9199. (22) Pramanik, S.; Bhalla, V.; Kumar, M. ACS Appl. Mater. Interfaces 2014, 6, 5930−5939. (23) Lee, J. H.; Jeong, A. R.; Shin, I. S.; Kim, H. J.; Hong, J. I. Org. Lett. 2010, 12, 764−767. (24) Zou, Q. Li; Zhang, J.; Zhou, J.; Sun, B. B.; Tian, H. Chem. Commun. 2012, 48, 2095−2097. (25) Liu, Y. L.; Lv, X.; Zhao, Y.; Liu, J.; Sun, Y. Q.; Wang, P.; Guo, W. J. Mater. Chem. 2012, 22, 1747−1750. (26) Lin, Q.; Sun, B.; Yang, Q. P.; Fu, Y. P.; Zhu, X.; Wei, T. B.; Zhang, Y. M. Chem. - Eur. J. 2014, 20, 11457−11462. (27) Cho, D. G.; Sessler, J. L. Chem. Soc. Rev. 2009, 38, 1647−1662. (28) Robbins, T. F.; Qian, H.; Su, X.; Hughes, R. P.; Aprahamian, I. Org. Lett. 2013, 15, 2386−2389. (29) Lin, Q.; Liu, X.; Wei, T. B.; Zhang, Y. M. Chem. - Asian J. 2013, 8, 3015−3021. (30) Lin, Q.; Lu, T. T.; Zhu, X.; Sun, B.; Yang, Q. P.; Wei, T. B.; Zhang, Y. M. Chem. Commun. 2015, 51, 1635−1638. (31) Tang, Y. H.; Qu, Y.; Song, Z.; He, X. P.; Xie, J.; Hua, J.; Chen, G. R. Org. Biomol. Chem. 2012, 10, 555−560. (32) Lin, W. C.; Fang, S. K.; Hu, J. W.; Tsai, H. Y.; Chen, K. Y. Anal. Chem. 2014, 86, 4648−4652. (33) Mahapatra, A. K.; Manna, S. K.; Pramanik, B.; Maiti, K.; Mondal, S.; Alia, S. S.; Mandal, D. RSC Adv. 2015, 5, 10716−10722. (34) Leea, M.; Moon, J. H.; Swamy, K. M. K.; Jeong, Y.; Kim, G.; Choia, J.; Lee, J. Y.; Yoon, J. Sens. Actuators, B 2014, 199, 369−376. (35) Wang, J.; Ha, S. C. Analyst 2011, 136, 1627−1631. (36) Crystallographic parameters for the structure of 5FLOAc: empirical formula, C35H34F5N5O13S2; molecular weight, 891.7; temperature 100 K; crystal system, monoclinic; space group, P21; a/ Ǻ , 7.486(2); b/Ǻ , 20.653(5); c/Ǻ , 12.537(3); β/°, 92.435(3); volume/ Ǻ 3 1936.8(8); Z, 2; absorption coefficient, 0.234; Dx (g cm−3), 1.529; number of reflections, 10454; reflections used, 9154; parameters, 541; final R, 0.0459 and wR2 0.1039. (37) Wheeler, S. E.; Houk, K. N. J. Am. Chem. Soc. 2008, 130, 10854−10855. (38) De Silva, A. P.; Gunaratne, H. N. Q.; Gunnlaugsson, T.; Huxley, A. J. M.; McCoy, C. P.; Rademacher, J. T.; Rice, T. E. Chem. Rev. 1997, 97, 1515−1566. (39) Bouffard, J.; Kim, Y.; Swager, T. M.; Weissleder, R.; Hilderbrand, S. A. Org. Lett. 2008, 10, 37−40. (40) Gotor, R.; Costero, A. M.; Gil, S.; Parra, M.; Martinez-Manez, R.; Sancenon, F.; Gavina, P. Chem. Commun. 2013, 49, 5669−5671. (41) Shiraishi, Y.; Sumiya, S.; Hirai, T. Chem. Commun. 2011, 47, 4953−4955. (42) Wei, Y.; Zhao, H.; Kan, J.; Su, W.; Hong, M. J. Am. Chem. Soc. 2010, 132, 2522−2523. (43) Areti, S.; Teotia, R.; Rao, C. P. Analyst 2015, 140, 7391−7398. (44) Wu, Q.; Yang, W. T. J. Chem. Phys. 2002, 116, 515−524. (45) Muxel, A. A.; Neves, A.; Camargo, M. A.; Bortoluzzi, A. J.; Szpoganicz, B.; Castellano, E. E.; Castilho, N.; Bortolotto, T.; Terenzi, H. Inorg. Chem. 2014, 53, 2943−2952.

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.analchem.5b04085. Experimental methods, synthesis and characterization, spectral data, crystallographic data, fluorescence, absorption data, and computational data. The CCDC number for 5FLOAc crystal data is 1432608 (PDF).

■

AUTHOR INFORMATION

Corresponding Author

*Phone: 91 22 2576 7162. Fax: 91 22 2572 3480. E-mail: [email protected]. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS C.P.R. acknowledges the financial support from DST (SERB & Nano mission), CSIR, and DAE-BRNS. S.A. and D.S.Y. acknowledges UGC for SRF fellowship. S.B. acknowledges IIT Bombay for the award of Institute Postdoctoral position. We thank Mr. Darshan S. Mhatre for some help with the crystal structure.

■

REFERENCES

(1) Wang, F.; Wang, L.; Chen, X.; Yoon, J. Chem. Soc. Rev. 2014, 43, 4312−4324. (2) Kaur, K.; Saini, R.; Kumar, A.; Luxami, V.; Kaur, N.; Singh, P.; Kumar, S. Coord. Chem. Rev. 2012, 256, 1992−2028. (3) Martínez-Máñez, R.; Sancenón, F. Coord. Chem. Rev. 2006, 250, 3081−3093. (4) Xu, Z.; Chen, X.; Kim, H.; Yoon, J. Chem. Soc. Rev. 2010, 39, 127−137. (5) Cho, D. G.; Kim, J. H.; Sessler, J. L. J. Am. Chem. Soc. 2008, 130, 12163−12167. (6) Sessler, J. L.; Cho, D. G. Org. Lett. 2008, 10, 73−75. (7) Kumari, N.; Jha, S.; Bhattacharya, S. J. Org. Chem. 2011, 76, 8215−8222. (8) Muir, G. D. Hazards in Chemical Laboratory; The Royal Chemical Society: London, 1977. (9) Kulig, K. W. Cyanide Toxicity; U.S. Department of Health and Human Services: Atlanta, GA, 1991. (10) Vennesland, B.; Comm, E. F.; Knownles, J.; Westly, J.; Wissing, F. Cyanide in Biology; Academic: London, 1981. (11) Young, C.; Tidwell, L.; Anderson, C. Cyanide: Social Industrial and Economic Aspects; Minerals, Metals, and Materials Society: Warrendale, 2001. (12) Guidelines for Drinking-Water Quality; World Health Organization: Geneva, 1996. (13) Takano, R. J. Exp. Med. 1916, 24, 207−211. (14) Baskin, S. I.; Brewer, T. G. In Medical Aspects of Chemical and Biological Warfare; Sidell, F., Takafuji, E. T., Franz, D. R., Eds.; TMM Publications: Washington, DC, 1997; Vol. 10, pp 271−286. (15) El-Ballouli, A. O.; Zhang, Y.; Barlow, S.; Marder, S. R.; Al-Sayah, M. H.; Kaafarani, B. R. Tetrahedron Lett. 2012, 53, 661−665. (16) Lee, D. Y.; Singh, N.; Satyender, A.; Jang, D. O. Tetrahedron Lett. 2011, 52, 6919−6922. (17) Shi, B. B.; Zhang, P.; Wei, T. B.; Yao, H.; Lin, Q.; Zhang, Y. M. Chem. Commun. 2013, 49, 7812−7814. (18) Jo, J. Y.; Olasz, A.; Chen, C. H.; Lee, D. J. Am. Chem. Soc. 2013, 135, 3620−3623. (19) Goswami, S.; Manna, A.; Paul, S.; Das, A. K.; Aich, K.; Nandi, P. K. Chem. Commun. 2013, 49, 2912−2914. H


Optimizing the Electron-Withdrawing Character on Benzenesulfonyl

Recommend Documents