A new colorimetric fluorescent sensor for ratiometric detection of cyanide in solution, test strips, and in cells

Abstract

A new high selective and sensitive fluorescent sensor for the detection of cyanide was developed based on the nucleophilic attack of CN⁻ with a color change from purple to colourless. The chemosensor was used for fabrication of test strips that can detect cyanide in aqueous samples. The living animal fluorescence experiment demonstrated the practical value of the sensor in tracing the CN⁻ in biological systems.

---

Anions play important roles in many areas such as medicinal biological, environmental chemistry and catalysis. Cyanide as one of the most useful anions, is extensively utilized in many fields such as gold mining, electroplating, metallurgy, synthetic fibers and resins industry.¹ But owing to its propensity of binding with iron species in cytochrome oxidase, which interferes with electron transport resulting in hypoxia,² cyanide does an extremely harm to the body. Minute quantity of cyanide could lead to deadly hazards.³ Therefore, easy and affordable detection methods are in great demand for various situations.

Up to now, numbers of sensors for CN⁻ have been invented.⁴ The general approaches used for the cyanide detecting are summarized as hydrogen bonding,⁵ nucleophilic addition on the carbonyl moiety,⁶ copper cyanide affinity (displacement approach),⁷ iminiumsalts,⁸ electron-deficient alkenes,⁹ and so on. But they themselves all have their own drawbacks such as poor solubility in aqueous media, low detection limits, complex synthesis process or poor application in the living cells. To date, colorimetric and ratiometric fluorescent chemosensors with high detection level and biocompatibility are still very limited.¹⁰

With these considerations in mind, we here developed a new chemodosimetric sensor (Phc) for the cyanide detection based on phenothiazine–hemicyanine conjugate as shown in Scheme 1. As its indolenium C-2 atom is an effective target for the nucleophilic analytes, the cyanide can easily combine with it, and the intramolecular conformational changes will block the intramolecular charge transfer (ICT), inducing a remarkable change in absorption and emission.

Scheme 1 The design concept of ratiometric fluorescent sensor-Phc for CN⁻.

Phc was condensed as a purple solid with a yield of 48%, the Phc–CN⁻ was also prepared by an addition reaction of Phc with Bu₄NCN in the solution of ethanol at room temperature with a 90% yield. Compound Phc and Phc–CN⁻ were well characterized by ¹H NMR, ¹³C NMR and MS.

The changes of the UV-vis spectra for the receptor in the presence of CN⁻ in a DMF–Tris·HCl buffer solution (pH = 7.4, 1 : 1, v/v) were shown in Fig. 1. The Phc showed an absorption band at 540 nm, which was attributed to the intramolecular charge transfer (ICT) transition of Phc. In UV-vis absorption spectrometry, with the increasing of the CN⁻, we found a decrease in absorbance at λ_max = 540 nm and λ_max = 375 nm. When we induced the CN⁻ to be 3.5 equiv., the absorbance intensities at 540 nm was found to be almost disappeared.

Fig. 1 UV spectra of Phc (100 μM) in the presence of increased concentration of CN⁻ (0–3.5 equiv.) in DMF–Tris·HCl buffer (10 mM, pH = 7.4, 1 : 1, v/v).

Next, we investigated the concentration dependent changes in the fluorescence spectra of Phc (10 μM) incubated with CN⁻. With the gradual addition of CN⁻, we found the fluorescence at 480 nm increasing sharply and the spectra had a slight redshift from 480 nm to 488 nm. What's more, the addition of 3.5 equiv. of CN⁻ to Phc induced a nearly 20-folds variation in the fluorescence ratio (I/I₀). Thus, The fluorescence intensities at 488 nm were plotted to obtain a calibration graph, which shown an excellent linear relationship, with the coefficient R = 0.99447 (Fig. 2).

Fig. 2 Fluorescence titration spectra of Phc (10 μM) in DMF–Tris·HCl buffer (10 mM, pH = 7.4, 1 : 1, v/v) upon gradual addition of CN⁻. The inset showed fluorescence intensity as function of CN⁻ on centration (DMF–Tris·HCl buffer, 1 : 1, v/v, pH = 7.4, excited at 350 nm, slits: 5 nm/5 nm).

As a new class of fluorescent dye based on hybrid phenothiazine–hemicyanine, Phc possesses a highly reactive indolium group for the cyanide detecting. The phenomenon indicated that the nucleophilic attacking of cyanide toward the indolium group interrupted the π-conjunction and blocked the molecular ICT process, after which the fluorescence of phenothiazine recovered. Meanwhile, an obvious color change from purple to colorless turned out (ESI, Fig. S5†). The test results are in good agreement with the aforementioned design concept.

The fluorescence intensities at 488 nm were plotted using the formula (I/I₀ − 1) as a function of the CN⁻ concentration to obtain a calibration graph, which showed an excellent linear relationship, with the coefficient R = 0.99909 (Fig. S6†). Interestingly, (I/I₀ − 1) varies almost linearly vs. the concentration of CN⁻ in the range of 5.0–30 μM. This phenomenon implied that Phc was potentially useful for the quantitative determination of CN⁻ concentrations. The detection limit DL of Phc for CN⁻ was determined from the following equation:

DL = KSb1/S

where K = 2 or 3 (we take 2 in this case); Sb1 is the standard deviation of the blank solution; S is the slope of the calibration curve. Using this equation, we calculated the detection limit of CN⁻ with Phc to be 6.67 × 10⁻⁸ M, which is superior to most of the sensor.¹¹ Noteworthy, this is about 30 times lower than the maximum level (1.9 μM) of CN⁻ in drinking water permitted by the World Health Organization (WHO).

Fig. 3 Variation of the relative fluorescence intensity at 488 nm of Phc (10 μM) in the presence of competitive anions (CN⁻, F⁻, Cl⁻, Br⁻, I⁻, N₃⁻, SCN⁻, NO₃⁻, EDTA, CH₃COO⁻, H₂PO₄⁻, HSO₃⁻ HPO₄²⁻, SO₄²⁻) and the fluorescence ratios of Phc (10 μM) to CN⁻ (3.5 equiv.) in the presence of 35 equiv. of other competitive analytes. Inset: black bar: 1 + anion, red bar: 1 + anion + CN⁻.

In addition, we studied the time-dependent fluorescence changes and found that the fluorescence of Phc increased to the top in only 1 min (shown in Fig. S7†). Most of the sensor invented now rarely studied the time-dependent fluorescence changes, so we believe this property is very excellent and is convenience for the practical use.

An important feature of Phc is its high selectivity toward the CN⁻ over the other competitive nucleophilic anions. Changes of fluorescence spectra of 1 (10 μM) caused by CN⁻ (35 μM) and miscellaneous competing species (350 μM) (including Cl⁻, Br⁻, I⁻, N₃⁻, SCN⁻, NO₃⁻, EDTA, CH₃COO⁻, H₂PO₄⁻, SO₄²⁻) in DMF–Tris·HCl buffer (10 mM, pH = 7.4, 1 : 1, v/v) were explored in Fig. 3. As could be seen from the diagram, these competitive species, which often showed strong interference in the cyanide detecting did not lead any significant fluorescence changes, and the fluorescence emission spectrum of Phc remained almost undisturbed. Competitive experiments had also proof the selectivity and stability of our sensor too. The absorption spectrum and emission spectrum were shown in Fig. S8.†

Fig. 4 Color changes of test strips containing Phc treated with various anions from left to right (none, Cl⁻, Br⁻, I⁻, N₃⁻, SCN⁻, NO₃⁻, EDTA, CH₃COO⁻, H₂PO₄⁻, SO₄²⁻).

Simultaneously, we studied the influence of the ions on the emission of Phc. As shown in the Fig. S4,† the metal cations induced negligible fluorescence changes for Phc. For Fe³⁺ and Cu²⁺ were found to have affinity towards cyanide, we collected the emission spectrum of Phc–CN⁻ in the present of these two cations. From the Fig. S9† we could figure out that when we induced the two cations into the Phc which reacted with CN⁻, the fluorescence changed little. If we first induced the cations and then added the CN⁻, we found that the fluorescence of solution with Cu²⁺ did not enhanced as high as others. We thought it was the complexation of CN⁻ and Cu²⁺ that influenced the reaction between the CN⁻ and Phc.

Next we calculated the binding constant according to the Benesi–Hildebrand equation (shown in Fig. S10†). K_a was calculated following the equation stated below:

1/(F − F₀) = 1/{K_a(F_max − F₀)[CN⁻]_n} +1/[F_max − F_o] (ref. 12)

The association constant (K_a) as determined by fluorescence titration method for sensor with CN⁻ was found to be 2.372 × 10⁴ M⁻¹.

We selected the filter paper to be the supporter of the sensor to make a test strips for the CN⁻ detecting. After dropping an aqueous solution of Phc on the neutral filter paper and drying it, a purple dipstick was formed. Exposed the modified dipstick to the aqueous solution, only CN⁻ induced the color fading from purple to colorless as shown in Fig. 4.

In order to further explore the potential biological application, we studied the capability of Phc to track the changes of CN⁻ in living GES cells (human breast cancer cells) and HeLa cells (human neuroblastoma cells). The living cells were first incubated with Phc (10 μM) for 15 min at 37 °C in 5% CO₂ atmosphere, washed with phosphate buffered saline (PBS, pH = 7.4) three times, and induced the CN⁻ (30 μM) into the solution for 15 min. From the phenomenon, extremely weak fluorescence could be observed (Fig. 5A and D) in the cells which only treated with Phc (10 μM), but after the inducing of CN⁻, we can clearly found bright green fluorescence in the GES cells (Fig. 5C) and the HeLa cells (Fig. 5F). So these facts implied that Phc had successfully immersed into the cells and our sensor was a membrane-permeable sensor. Moreover, owing to its quick response rate, it is possible for us to investigate the biological absorption process of involving CN⁻ in living cells.

Fig. 5 Fluorescence microscope imaging of live GES cells and HeLa cells with Phc (10 μM) before (A and D) and after(C and F) treated with CN⁻ (10 μM) for 15 min. Bright-field transmission images of GES cells and HeLa cells incubated with Phc (10 μM) are shown in (B and E).

Zebra fish is a tropical freshwater fish. We feed the fish in the water with Phc (10 μM) dissolved in DMSO for 1 h, then washed it with phosphate buffered saline (PBS, pH = 7.4) and observed it using the In vivo FX PRO Imaging system. As shown in Fig. 6, under the 390 nm light, we could find that the fish exposed only to Phc did not display apparent fluorescence (the image A), in contrast, we carried on to treat the fish with CN⁻ (30 μM) for another 1 h, a graceful fish emerged with a beautiful contour and strong green luminescence. From the intensity we can suspect that the accumulations of Phc in the gills and abdomen were higher compared to other parts of the fish body. So we can infer that our sensor had successfully immersed into the fish body, and induced the fluorescence enhancement.¹³ And as far as we know, this was the first time for the sensor of cyanide to be used in the living animal fluorescence experiment.

Fig. 6 Images of full grown zebra fish under 390 nm light: (A) fish incubated with only Phc (10 μM) for 1 h; (B) fish incubated with Phc (10 μM) and then incubated with CN⁻ (35 μM) for 1 h.

To get insight into the mechanism of the colorimetric and ratiometric fluorescence change in the presence of CN⁻, we carried out the DFT calculations with 6-31G* basis sets using a suite of Gaussian 09 programs in order to investigate the mechanism (Fig. 7).¹⁴ From the overhead structure we can see the significant difference in π-conjunction between Phc and Phc–CN⁻. Via a conjugated bridge (–C=C–), the benzothiolium groups and the phenothiazine groups were well planned with only a dihedral angle of 1.8°. Inducing the cyanide to indolenium C-2 atom blocked the plantation of phenothiazine and benzothiolium groups, then the dihedral angle converted to be 94.8°, the fluorescence of phenothiazine recovered.

Fig. 7 Rationalization of the photo-physical properties of Phc (A) and Phc–CN⁻ (B).

TD-DFT (time-dependent DFT) calculations indicated that the absorption band at the higher wavelength could be assigned to an ICT mechanism. The calculated transition at 480 nm with an oscillator strength of f = 0.06 correspond to a HOMO → LUMO transition (Fig. 8). In the case of Phc–CN⁻, the HOMO resides only on the phenothiazine group and the LUMO resides on the benzothiolium unit which clearly indicates the obstruction in the ICT process, so in the absorption spectrum of Phc–CN⁻ the band corresponding to the ICT is diminished, TD-DFT calculations showed only one strong transition at 336 nm with an oscillator strength of f = 0.05 which corresponds to a HOMO → LUMO transition as expected. So the calculations results are in good agreement with the experimental results.^15,16

Fig. 8 Frontier molecular orbitals involved in the vertical excitation (A and B) by TD-CAM-B3LYP/6-31G(d) method and emission (C) by TD-PBE0/6-31G(d) of Phc–CN⁻ method in DMF–H₂O solution.

In summary, we have successfully developed a new ratiometric fluorescent sensor for the detection of CN⁻. DFT and TD-DFT calculations suggested the sensing mechanism. Phc was highly reactive and selective to CN⁻, with a short responding time, low detection limit, huge color changing and not bad water-solubility. Moreover, Phc had an excellent biocompatible ability, and can indeed visualize the changes of intracellular cyanide in living cells and living fish and this was the first time for the sensor of cyanide to be used in the living animal fluorescence experiment. We envisage this new sensor an extensive application in tracing and adsorbing CN⁻ in the future work.

We acknowledge the Natural Science Foundation of China (no. 21174052) for the support of our work.

Notes and references

1. G. C. Miller and C. A. Pritsos, Cyanide: Soc., Ind. Econ. Aspects, Proc. Symp. Annu. Meet. TMS, 2001, pp. 73–81.
2. (a) K. Kulig, Cyanide Toxicity, U.S. Department of Health and Human Services, Atlanta, GA, 1991; (b) S. Baskin and T. Brewer, in Medical Aspects of Chemical and Biological Warfare, ed. F. Sidell, E. T. Takafuji and D. R. Franz, TMM Publication, Washington, DC, 1997, pp. 271–286; (c) H. N. Lee, Z. Xu, S. Kim, M. K. Swamy, Y. Kim, S.-J. Kim and J. Yoon, J. Am. Chem. Soc., 2007, 129, 3828.
3. Guidelines for Drinking-Water Quality, World Health Organization, Geneva, 1996.
4. (a) T. Gunnlaugsson, M. Glynn, G. Tocci, P. Kruger and F. Pfeffer, Coord. Chem. Rev., 2006, 250, 3094; (b) D. Cho and Y. Sessler, Chem. Soc. Rev., 2009, 38, 1647; (c) A. Li, J. Wang, F. Wang and Y. Jiang, Chem. Soc. Rev., 2010, 39, 3729; (d) H. Kim, Z. Guo, W. Zhu, J. Yoon and H. Tian, Chem. Soc. Rev., 2011, 40, 79.
5. (a) A. O. El-Ballouli, Y. Zhang, S. Barlow, S. R. Marder, M. H. Al-Sayah and B. R. Kaafarani, Tetrahedron Lett., 2012, 53, 661; (b) D. Y. Lee, N. Singh, A. Satyender and D. O. Jang, Tetrahedron Lett., 2012, 2, 6919.
6. (a) H. Li, Z. Wen, L. Jin, Y. Kan and B. Yin, Chem. Commun., 2012, 48, 11659–11661; (b) S. Goswami, S. Paul and A. Manna, Dalton Trans., 2013, 42, 10682–10686; (c) S. Madhu, S. K. Basu, S. Jadhavand and M. Ravikanth, Analyst, 2013, 138, 299.
7. (a) V. Bhalla, H. Singh and M. Kumar, Dalton Trans., 2012, 41, 11413; (b) G. Upendar Reddy, P. Das, S. Saha, M. Baidya, S. K. Ghosh and A. Das, Chem. Commun., 2013, 49, 255–257; (c) B. K. Kanungo, M. Baral, R. K. Bera and S. K. Sahoo, Monatsh. Chem., 2010, 141, 157–168.
8. (a) S. H. Mashraqui, R. Betkar, M. Chandiramani, C. Estarellas and A. Frontera, New J. Chem., 2011, 35, 57; (b) Z. Yang, Z. Liu, Y. Chen, X. Wang, W. He and Y. Lu, Org. Biomol. Chem., 2012, 10, 5073; (c) S. Kumar, P. Singh, G. Hundal, M. S. Hundala and S. Kumar, Chem. Commun., 2013, 49, 2667–2669.
9. (a) P. Kaur, D. Sareen and K. Singh, Dalton Trans., 2012, 41, 9607; (b) M. Dong, Y. Peng, Y.-M. Dong, N. Tang and Y.-W. Wang, Org. Lett., 2012, 14, 130; (c) Y.-D. Lin, Y.-S. Peng, W. Su, C.-H. Tu, C.-H. Sun and T. J. Chow, Tetrahedron, 2012, 68, 2523.
10. (a) K. Kaur, R. Saini, A. Kumar, V. Luxami, N. Kaur, P. Singh and S. Kumar, Coord. Chem. Rev., 2012, 256, 1992; (b) Z. Xu, X. Chen, H. N. Kim and J. Yoon, Chem. Soc. Rev., 2010, 39, 127; (c) Z. Yang, Z. Liu, Y. Chen, X. Wang, W. He and Y. Lu, Org. Biomol. Chem., 2012, 10, 5073.
11. (a) X. Huang, X. Gu, G. Zhang and D. Zhang, Chem. Commun., 2012, 48, 12195–12197; (b) S. Goswami, S. Paul and A. Manna, Dalton Trans., 2013, 42, 10682–10686; (c) L. Long, L. Zhou, L. Wang, S. Meng, A. Gong, F. Dub and C. Zhang, Anal. Methods, 2013, 5, 6605–6610.
12. S. Goswami, K. Aicha, S. Dasa, A. Kumar Dasa, A. Mannaa and S. Halderb, Analyst, 2013, 138, 1903–1907.
13. J. Hatai, S. Pal, G. P. Jose and S. Bandyopadhyay, Inorg. Chem., 2012, 51, 10129–10135.
14. M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria, M. A. Robb, J. R. Cheeseman, G. Scalmani, V. Barone, B. Mennucci, G. A. Petersson, H. Nakatsuji, M. Caricato, X. Li, H. P. Hratchian, A. F. Izmaylov, J. Bloino, G. Zheng, J. L. Sonnenberg, M. Hada, M. Ehara, K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O. Kitao, H. Nakai, T. Vreven, J. A. Montgomery, Jr, J. E. Peralta, F. Ogliaro, M. Bearpark, J. J. Heyd, E. Brothers, K. N. Kudin, V. N. Staroverov, R. Kobayashi, J. Normand, K. Raghavachari, A. Rendell, J. C. Burant, S. S. Iyengar, J. Tomasi, M. Cossi, N. Rega, J. M. Millam, M. Klene, J. E. Knox, J. B. Cross, V. Bakken, C. Adamo, J. Jaramillo, R. Gomperts, R. E. Stratmann, O. Yazyev, A. J. Austin, R. Cammi, C. Pomelli, J. W. Ochterski, R. L. Martin, K. Morokuma, V. G. Zakrzewski, G. A. Voth, P. Salvador, J. J. Dannenberg, S. Dapprich, A. D. Daniels, Ö. Farkas, J. B. Foresman, J. V. Ortiz, J. Cioslowski, and D. J. Fox, Gaussian 09, Revision D.01, Gaussian, Inc., Wallingford C. T., 2009.
15. H. Li, Z. Wen, L. Jin, Y. Kan and B. Yin, Chem. Commun., 2012, 48, 11659–11661.
16. P. B. Pati and S. S. Zade, RSC Adv., 2013, 3, 13457–13462.

---

Footnote

† Electronic supplementary information (ESI) available: Supplementary figures and methods. See DOI: 10.1039/c3ra46741d

---