Colorimetric Cyanide-Selective Sensor Based

Letter pubs.acs.org/ac

Ratiometric Fluorescent/Colorimetric Cyanide-Selective Sensor Based on Excited-State Intramolecular Charge Transfer−ExcitedState Intramolecular Proton Transfer Switching Wei-Chi Lin, Sin-Kai Fang, Jiun-Wei Hu, Hsing-Yang Tsai, and Kew-Yu Chen* Department of Chemical Engineering, Feng Chia University, 40724 Taichung, Taiwan ROC S Supporting Information *

ABSTRACT: A novel salicylideneaniline-based fluorescent sensor, SB1, with a unique excited-state intramolecular charge transfer−excited-state intramolecular proton transfer (ESICT−ESIPT) coupled system was synthesized and demonstrated to fluorescently sense CN− with specific selectivity and high sensitivity in aqueous media based on ESICT−ESIPT switching. A large blue shift (96 nm) was also observed in the absorption spectra in response to CN−. The bleaching of the color could be clearly observed by the naked eye. Moreover, SB1-based test strips were easily fabricated and low-cost, and could be used in practical and efficient CN− test kits. Density functional theory (DFT) and time-dependent DFT (TD-DFT) calculations further support the cyanide-induced ESICT−ESIPT switching mechanism. The results provide the proof of concept that the colorimetric and ratiometric fluorescent cyanide-selective chemodosimeter can be created based on an ESICT−ESIPT coupled system. he recognition and sensing of cyanide (CN−) has continuously received significant attention due to the element’s lethal effect on living organisms and the environment.1 Cyanide is widespread in industrial processes, such as gold mining, metallurgy, electroplating, and the synthesis of fibers and resins.2 Therefore, more and more research is focusing on the design and synthesis of new colorimetric and ratiometric fluorescent cyanide-selective chemosensors.3 Versatile mechanisms for the detection of cyanide have been developed, including those based on excited-state intramolecular charge transfer (ESICT),4 excited-state intramolecular proton transfer (ESIPT),5 hydrogen-bonding interactions,6 supramolecular self-assembly,7 complex formations with metal ions and boron derivatives,8 and attachment with quantum dots.9 In addition, reaction-based chemosensors for CN− have been used due to their high selectivity over other anions and the fact that they can effectively reduce the interference of hydrogen bonding and the acidity of the media. Effective sensors based on this theory have been prepared using squaraine,10 oxazines,11 acridium salts,12 benzyl derivatives,13 trifluoroacetophenone derivatives,14 coumarin derivatives,4,15 calix[4]pyrrole derivatives,16 and dicyanovinyl derivatives.16a,17 Fluorescence based detection usually depends on the intensity change at a single wavelength, which is regarded to be more sensitive compared with absorption spectroscopy due to its high sensitivity and easy implementation under diversified environmental conditions. Nevertheless, the signal output could be easily overshadowed by the background noise of the sample media. To overcome this drawback, ratiometric fluorescent

T

© 2014 American Chemical Society

sensing has been used and this permits the measurement of relative fluorescence intensities at two different wavelengths, which can serve as an internal reference for self-correction and therefore the reliability of the measurements is substantially enhanced.18 Recently, several potential ESICT−ESIPT coupled systems have been synthesized and investigated.19 Since the ESICT states are involved in the reactant or product side of the ESIPT reaction, the interplay between two emissions well separated on the wavelength domain can be modulated not only by molecular structural modifications but also by the change in its surrounding media. These features dramatically broaden the possibilities for the design of wavelength-ratiometric fluorescent sensors and probes. However, to the best of our knowledge, ESICT−ESIPT based fluorescent chemosensors for the detection of cyanide have not been investigated yet. Thus, this research used a unique ESICT−ESIPT coupled molecule, SB1, for the first time as a fluorescent sensor, which was able to sense CN− with specific selectivity and high sensitivity in aqueous media based on ESICT−ESIPT switching. The structure of SB1 was composed of a well-known ESIPT molecule, salicylideneaniline (Scheme 1),19a in which diethylamino and dicyanovinyl groups were attached to act as electron-donating and electron-withdrawing groups, respectively. The electrophilic nature of the dicyanovinyl group can be Received: March 21, 2014 Accepted: May 8, 2014 Published: May 8, 2014 4648

dx.doi.org/10.1021/ac501024d | Anal. Chem. 2014, 86, 4648−4652

Analytical Chemistry

Letter

Scheme 1. Synthetic Route of SB1 and SB2 and the Reaction Mechanism of SB1 with the Cyanide Anion for the Formation of SB1-CN

modulated by CN−, which interrupts the π-conjugation,17 and its role in the sensing mechanism can be clarified by a comparison with the spectrum of SB2. The high yield synthesis of SB1 was readily prepared through a condensation of 4-diethylaminosalicylaldehyde (3) with 2-(4aminobenzylidene)malononitrile (2a) in the presence of acetic acid (Scheme 1). Detailed synthetic procedures and product characterizations are provided in the Supporting Information. Figure 1 shows the absorption and emission spectra of SB1 in different solvents of varying polarity, and pertinent

Figure 2. Relaxation processes for ESICT−ESIPT coupled system SB1. CTeq* and PCTeq* denote the polarization equilibrium of CT* and PCT*, respectively. This is to show the origin of dual emission.

increase of solvent polarity, which results in a sharp decrease of the CTeq* energy and, hence, the significant increase of the CTeq* → PCTeq* barrier, prohibiting ESIPT (Figure S3, Supporting Information). The sensing properties of SB1 were examined in THF/H2O solutions (8:2 [v/v], containing 0.01 M HEPES, pH = 7.3) through the addition of tetrabutylammonium salt of various anions, including F−, Cl−, Br−, I−, SCN−, AcO−, NO3−, BzO−, H2PO4−, HP2O7−, HSO4−, and CN− (Figure 3). Upon the

Figure 1. Absorption (left) and emission (right) spectra of SB1 in cyclohexane (black line), diethyl ether (blue line), THF (cyan line), dichloromethane (green line), and acetonitrile (red line). Excitation at 460 nm.

photophysical data is given in Table S1, Supporting Information. Compound SB1 shows two absorption bands at ∼330 and 470 nm; the former is attributed to the π−π* transition and the latter to an intramolecular charge transfer (ICT) transition. This viewpoint can be further supported by a theoretical approach based on density functional theory (Table S2, Supporting Information). Unlike the small shift in absorption spectra, the fluorescence spectra of SB1 are largely red-shifted if there is any increase in the solvent polarity, which indicates strong ICT characteristics for the excited states of the compound (Figure S1, Supporting Information). More importantly, dual emission can be resolved in nonpolar cyclohexane, which upon deconvolution, is composed of a charge transfer emission (CTeq*) and a large Stokes shifted proton-transfer tautomer emission (PCTeq*) band maximized at 520 and 560 nm, respectively (Figure S2, Supporting Information). Therefore, a remarkable ESICT−ESIPT coupled system was evidently established (Figure 2). It is should be noted that charge transfer emission is dominant in polar solvents (THF, dichloromethane, and acetonitrile) due to the

Figure 3. (a) Absorption and (b) emission spectra (excitation at 418 nm) and (c) colorimetric and (d) fluorimetric (excitation at 365 nm) responses of SB1 (2.5 × 10−5 M) in a THF/H2O (8:2, v/v, containing 0.01 M HEPES, pH = 7.3) solution upon the addition of 20 equiv of various anions.

addition of 20 equiv of various anions (except for CN− ion), the absorption spectra of SB1 did not show any significant change (Figure 3a). Nevertheless, in the presence of CN− ion, both absorption bands at 337 nm (π−π*) and 472 nm (ICT) diminished with the concomitant appearance of a new band at 376 nm. The bleaching of the color can be clearly observed by the naked eye (Figure 3c). A gradual spectral change is also shown in Figure S4, Supporting Information, where two clear isosbestic points can be observed at 339 and 418 nm, indicating a clean transformation to a new species. This is in good 4649

dx.doi.org/10.1021/ac501024d | Anal. Chem. 2014, 86, 4648−4652

Analytical Chemistry

Letter

agreement with a 1:1 binding stoichiometry by Job’s plot of the UV−vis absorption (Figure S4, Supporting Information). Furthermore, the time-dependence changes in the absorption spectra of SB1 upon the addition of cyanide were almost completed within 15 min. Accordingly, the pseudo-first-order rate constant was calculated to be 0.355 min−1 (Figure S5, Supporting Information). Upon the addition of F−, Cl−, Br−, I−, SCN−, AcO−, NO3−, BzO−, H2PO4−, HP2O7−, and HSO4−, no substantial change was observed in the emission spectra of SB1. The only significant response appeared when CN− was added (Figure 3b), and a weak fluorescence (Φ = 0.002) was observed at about 523 nm and the band at 626 nm (Φ = 0.15, ESICT) diminished. The 523 nm emission of SB1-CN (Scheme 1) was quite similar to that of SB2,19a which was unambiguously assigned to the ESIPT emission. The blue shift induced by the addition of CN− was large with an isoemissive point at 531 nm (Figure S6, Supporting Information). After the addition of 20 equiv of CN−, the peak at 626 nm was completely quenched. The ratio of emission intensities (I523/I626) exhibited a 75-fold decrease for the cyanide anion (Figure 4), which was easily

Figure 5. Absorbance (A376/A472) and fluorescence (I626) responses (excitation at 418 nm) of SB1 (2.5 × 10−5 M) in the presence of CN− and additional anions in THF/H2O (8:2, v/v) at pH = 7.3 (red bar, SB1 + anion; black bar, SB1 + anion + CN−).

examined over an extensive range of pH values; the detection of CN− could be well performed in the pH range of 5.0−9.0 in HEPES buffered solutions (Figure 6). The results not only confirmed the marvelous selectivity of sensor SB1 for CN− but also revealed its high potential for practical applications.

Figure 6. pH Dependence of absorbance of SB1, measured at 472 nm, with and without CN− (20 equiv).

The product SB1-CN was analyzed by 1H NMR spectroscopy. Figure 7 shows the 1H NMR of SB1 upon the addition of tetrabutylammonium cyanide in a THF-d8 solution. The addition of cyanide resulted in a slow reduction of the vinlyic proton signal (Hf) at 8.10 ppm, which finally disappeared with the addition of 1 equiv of cyanide, while a new signal appeared at 4.25 ppm (Hf′), corresponding to the β-proton of dicyanovinyl. Meanwhile, the upfield shifts of the two sets of aromatic protons (Hd′ and He′) are observed at the ortho and meta positions relative to the ethylene group. The product SB1-CN was further confirmed by mass spectrometry analysis, where a peak at m/z 370.1664 corresponding to [SB1 + CN]− was clearly observed (Figure S8, Supporting Information). These results are consistent with the proposal that CN− attacks the β-conjugated position of dicyanovinyl moiety. The optimized geometries of SB1 and SB1-CN are shown in Figure S9, Supporting Information. Except for the ethyl substituents of the amine, SB1 was nearly planar, with an angle of 10.1° between the two phenyl rings. Such planar conformation provided efficient π-conjugation and hence favored the efficient ICT transition from the amino group to the dicyanovinyl moiety. The ground-state geometry underwent an apparent twist upon the addition of CN− (SB1-CN), while the π-conjugation between the dicyanovinyl and the salicylideneaniline was interrupted. The interruption of the πconjugation resulted in a significant blue shift in the absorption

Figure 4. Emission spectra of SB1 (2.5 × 10−5 M) in the presence of different concentrations of CN− (0−22 equiv at 2.0 equiv interval) in a THF/H2O (8:2, v/v, containing 0.01 M HEPES, pH = 7.3) solution (excitation at 418 nm). Inset: ratio of fluorescent intensities at 523 and 626 nm as a function of CN− concentration.

observed by the naked eye under the irradiation of a UV-lamp (Figure 3d). In addition, the fluorescence intensity (I626) decreased in a linear fashion with the concentration of CN− (Figure S7, Supporting Information) and thus can be potentially used for the quantification of CN−. The detection limit20 of the fluorescence spectrum changes calculated on the basis of 3σ/m is 2.4 μM for CN−, which is quite close to the maximum level permissible in drinking water (1.9 μM).21 To further explore the utility of SB1 as an ion-selective chemosensor for CN−, competitive experiments were performed with 20 equiv of CN− and 20 equiv of various other anions (F−, Cl−, Br−, I−, SCN−, AcO−, NO3−, BzO−, H2PO4−, HP2O7−, HSO4−) in a HEPES-buffered (pH = 7.3) solution of SB1 (Figure 5). The absorption spectrum of SB1 with CN− was not significantly influenced by the subsequent addition of competing anions. Similarly, the ESICT−ESIPT switching behavior was not affected even in the presence of all anions together. Moreover, the selectivity of SB1 to CN− was 4650

dx.doi.org/10.1021/ac501024d | Anal. Chem. 2014, 86, 4648−4652

Analytical Chemistry

Letter

immersed in the aqueous media of CN− with different concentrations, a clear color change from orange to colorless was observed (Figure 9) and potentially competitive ions

Figure 9. Photographs of test strips of SB1 at various concentrations of CN− (mM): (A) 0, (B) 20.0, (C) 40.0, (D) 60.0, and (E) 80.0.

exerted no considerable influence on the detection of CN− by the test strips. Therefore, the SB1-based test strips can conveniently detect CN− in solutions without any additional equipment. In conclusion, a novel salicylideneaniline-based chemodosimeter SB1 was demonstrated to sense CN− with a fluorescent reaction in aqueous media based on ESICT− ESIPT switching. A large blue shift (96 nm) was also observed in the absorption spectra in response to CN−, and hence, the bleaching of the color could be clearly observed by the naked eye. Furthermore, the SB1-based test strips were easily fabricated and low-cost, useful in practical and efficient CN− test kits. The results prove the colorimetric and ratiometric fluorescent cyanide-selective chemosensor can be created based on an ESICT−ESIPT coupled system.

Figure 7. 1H NMR spectra changes of SB1 in THF-d8 upon the addition of CN− anion.

spectrum and decreased the ICT character. Figure S10, Supporting Information gives the frontier molecular orbitals of SB1, SB1-CN, and SB2. In the case of SB1, the electron density in the HOMO was localized on the salicylideneaniline moiety, while that in the LUMO was mainly localized on the dicyanovinyl group. Consequently, a charge separation can be expected to be generated while an electron is promoted from the HOMO to the LUMO. However, the electron distribution in SB1-CN, particularly in the LUMO, was quite different from that of SB1. Because of the interruption of the π-conjugation, the electron population in the LUMO of SB1-CN was confined to the salicylideneaniline moiety, without involving the malononitrile. Additionally, the electronic configuration of SB1-CN was similar to that of SB2 (Figure S11, Supporting Information), as was its weak ESIPT fluorescence (oscilator strength, f = 0.0085; Figure 8). Motivated by the obvious color change of the system in solution, test strips were prepared by immersing filter papers (3 × 1 cm2) in the THF/H2O solution of SB1 (1.0 mM) and then drying them in air. When the SB1-based test strips were

■

ASSOCIATED CONTENT

S Supporting Information *

Experimental details, procedures for synthesis, NMR, UV−vis, and PL spectra, and computational results. This material is available free of charge via the Internet at http://pubs.acs.org.

■

AUTHOR INFORMATION

Corresponding Author

*Phone: +886 4 24517250 ext. 3683. Fax: +886 4 24510890. Email: [email protected]. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS This paper is dedicated to Professor Tahsin J. Chow on the occasion of his 65th birthday. We thank the National Science Council (Grant NSC 101-2113-M-035-001-MY2) for the financial support.

■

REFERENCES

(1) Kulig, K. W. Cyanide Toxicity; U.S. Department of Health and Human Services: Atlanta, GA, 1991. (2) Miller, G. C.; Pritsos, C. A. In Cyanide: Social, Industrial, and Economic Aspects; TMS Publishing: Warrendale, PA, 2001; pp 73−81. (3) Xu, Z.; Chen, X.; Kim, H. N.; Yoon, J. Chem. Soc. Rev. 2010, 39, 127−137. (4) Kim, H. J.; Ko, K. C.; Lee, J. H.; Lee, J. Y.; Kim, J. S. Chem. Commun. 2011, 47, 2886−2888. (5) (a) Zhang, P.; Shi, B.-B.; Wei, T.-B.; Zhang, Y.-M.; Lin, Q.; Yao, H.; You, X.-M. Dyes Pigm. 2013, 99, 857−862. (b) Lee, K.-S.; Kim, H.J.; Kim, G.-H.; Shin, I.; Hong, J.-I. Org. Lett. 2008, 10, 49−51. (6) Sun, S.-S.; Lees, A. J. Chem. Commun. 2000, 1687−1688. (7) Shi, B. B.; Zhang, P.; Wei, T. B.; Yao, H.; Lin, Q.; Zhang, Y. M. Chem. Commun. 2013, 49, 7812−7814.

Figure 8. Selected frontier molecular orbitals involved in the excitation and emission of SB1-CN. E (K) and E* (K*) denote the enol (keto) form in the ground and excited state, respectively. GSIPT stands for ground state intramolecular proton transfer. 4651

dx.doi.org/10.1021/ac501024d | Anal. Chem. 2014, 86, 4648−4652

Analytical Chemistry

Letter

(8) Kim, Y.-H.; Hong, J.-I. Chem. Commun. 2002, 512−513. (9) Touceda-Varela, A.; Stevenson, E. I.; Galve-Gasión, J. A.; Dryden, D. T. F.; Mareque-Rivas, J. C. Chem. Commun. 2008, 1998−2000. (10) Ros-Lis, J. V.; Martínez-Máñez, R.; Soto, J. Chem. Commun. 2002, 2248−2249. (11) (a) Tomasulo, M.; Raymo, F. M. Org. Lett. 2005, 7, 4633−4636. (b) Ren, J.; Zhu, W.; Tian, H. Talanta 2008, 75, 760−764. (12) Yang, Y.-K.; Tae, J. Org. Lett. 2006, 8, 5721−5723. (13) (a) Cho, D.-G.; Kim, J. H.; Sessler, J. L. J. Am. Chem. Soc. 2008, 130, 12163−12167. (b) Sessler, J. L.; Cho, D.-G. Org. Lett. 2008, 10, 73−75. (14) Chung, Y. M.; Raman, B.; Kim, D. S.; Ahn, K. H. Chem. Commun. 2006, 186−188. (15) (a) Zhou, X.; Lv, X.; Hao, J.; Liu, D.; Guo, W. Dyes Pigm. 2012, 95, 168−173. (b) Peng, M.-J.; Guo, Y.; Yang, X.-F.; Wang, L.-Y.; An, J. Dyes Pigm. 2013, 98, 327−332. (16) (a) Hong, S.-J.; Yoo, J.; Kim, S.-H.; Kim, J. S.; Yoon, J.; Lee, C.H. Chem. Commun. 2009, 189−191. (b) Kim, S.-H.; Hong, S.-J.; Yoo, J.; Kim, S. K.; Sessler, J. L.; Lee, C.-H. Org. Lett. 2009, 11, 3626−3629. (17) (a) Vallejos, S.; Estévez, P.; García, F. C.; Serna, F.; de la Peña, J. L.; García, J. M. Chem. Commun. 2010, 46, 7951−7953. (b) Lin, Y.-D.; Pen, Y.-S.; Su, W.; Liau, K.-L.; Wen, Y.-S.; Sun, C.-H.; Chow, T. J. Chem.Asian J. 2012, 7, 2864−2871. (18) Tremblay, M. S.; Halim, M.; Sames, D. J. Am. Chem. Soc. 2007, 129, 7570−7577. (19) (a) Fang, T.-C.; Tsai, H.-Y.; Luo, M.-H.; Chang, C.-W.; Chen, K.-Y. Chin. Chem. Lett. 2013, 24, 145−148. (b) Seo, J.; Kim, S.; Park, S. Y. J. Am. Chem. Soc. 2004, 126, 11154−11155. (20) Yang, M. H.; Thirupathi, P.; Lee, K. H. Org. Lett. 2011, 13, 5028−5031. (21) Guidelines for Drinking-Water Quality, 3rd ed.; World Health Organization: Geneva, Switzerland, 2004.

4652

dx.doi.org/10.1021/ac501024d | Anal. Chem. 2014, 86, 4648−4652