Promising ESIPT-based fluorescence sensor for Cu²⁺ and CN⁻ ions: investigation towards logic gate behaviour, anticancer activities and bioimaging application

Abstract

An excited state intramolecular proton transfer (ESIPT) process-based novel chromogenic and fluorogenic probe (2) was synthesized with the aim of sequential in situ detection of Cu²⁺ and CN⁻ ions under aqueous and biological conditions. The probe revealed chelating affinity for transition metal cations, amidst which only Cu²⁺ efficiently quenches the emission intensity. Further in situ addition of CN⁻ results in a metal displacement reaction and a turn on fluorescence response. This quencher displacement sensing strategy results in “ON–OFF–ON” type of fluorescence changes with high selectivity and great affinity in nanomolar detection of [CN⁻] and outlines the working principle of the IMPLICATION logic gate. The in vitro cytotoxic activity studied via the MTT assay revealed that in situ-formed Cu complexes worked as potential anticancer chemotherapeutic agents towards MCF-7 (human breast adenocarcinoma) cell lines (2 + Cu²⁺ IC₅₀ = 5.69 ± 0.26 μg mL⁻¹, 3 + Cu²⁺ IC₅₀ = 7.36 ± 0.29 μg mL⁻¹). The selective detection of Cu²⁺ and CN⁻ ions in biological systems was also explained by intracellular bioimaging studies in MCF-7 cell lines.

---

Introduction

By virtue of their spatio-temporal determination, real time chemical analysis and inherent high sensitivity, fluorescent sensory units are undeniably extremely efficient and potential tools and have been applied in significant biological molecular perception and biomedical diagnosis.^1,2 So, selecting precise fluorescent materials and alteration of the sensing activities to vary the fluorescent signals are two basic issues for fluorescent sensing. In the last decade, highly fluorescent dyes, noble metal nanoclusters, Schiff bases and iminium salts³ have attracted great attention as some of the most hopeful fluorescent sensors because of their potent ability to advance analytical sensitivity and specifically to be adopted for in vivo imaging applications. The sensing materials mentioned are based on various photophysical phenomena including fluorescence resonance energy transfer (FRET), photo-induced electron transfer (PET), aggregation-induced emission (AIE), intramolecular charge transfer (ICT), metal to ligand charge transfer (MLCT) and excited state intramolecular proton transfer (ESIPT).^4,5 Like fluorescent sensors, chromogenic sensing units have also received immense consideration in recent years, because ion recognition by the naked eye does not require expensive equipment. So designing and synthesis of discriminatory cation and anion sensing based on the above schemes have received enormous attention in supramolecular photochemistry, and pivotal roles continue to be performed by cations and anions in biological, chemical and environmental processes.^6,7

Copper is an essential trace element in humans and is involved in many biological and environmental processes.⁸ Deficiency and increased levels of Cu²⁺ cause serious problems to human health. Its homeostasis is key to severe diseases such as Menke's, Wilson's, Parkinson's and Alzheimer's diseases. High concentration levels cause the accumulation of Cu²⁺ in the liver and correspondingly are responsible for idiopathic copper toxicosis syndrome.⁹ In addition, Cu²⁺ takes part actively in the pathogenesis of cardiovascular disease, gastrointestinal disorders and diabetes.¹⁰ According to the US Environmental Protection Agency (EPA), [Cu²⁺] less than 1.3 ppm is acceptable in drinking water. In contrast with toxic metal ions, cyanide can cause death in a few minutes, even at low amounts, by suppressing the central nervous system. It directly affects all aerobic organisms by breaking the electron transport chain in the mitochondrial membrane and prevents respiration.¹¹ Chronic exposure to CN⁻ leads to many diseases, including those of the cardiac, vascular, endocrine, visual and metabolic systems.¹² Repeated exposure to low concentrations of cyanides over a long period causes nausea, muscle cramps, weakness, loss of appetite, paralysis of the arms and legs, and memory deficit.¹³ That is why it is one of the most lethal poisons known. Diverse detecting principles have been reported for cyanide recognition, using mechanisms such as nucleophilic addition of the anionic species onto activated carbonyl groups, hydrogen bonding interactions, Michael acceptor type activated C=C double bond, formation of cyanohydrin derivatives and demetallation of metal chelates.^14–16 Still, most of these sensory systems face interference from other anionic species, resulting in lack of selectivity and sensitivity for cyanide. This problem can effectively be avoided by utilizing copper–cyanide affinity. Use of high copper–cyanide affinity has become a valuable strategy to establish highly effective chemosensors for cyanide in aqueous media.

As the metal chelating entities were simply conjugated to the fluorophore units, the emission was not absolutely quenched after Cu²⁺ coordination, which is a hitch for quantitative investigations. So superior fluorophore units that bind Cu²⁺ directly to form strong, non-fluorescent Cu²⁺ complexes are desirable.¹⁷ Based on this prescription, we project and synthesize the excited state intramolecular proton transfer (ESIPT) process-based novel multifunctional chemosensors 3-((E)-((2-((E)-((1H-pyrrol-2-yl)methylene)amino)phenyl)imino)methyl)-2-methoxy-2H-chromen-4-ol (probe 2) and 3-((E)-((2-((E)-((1H-pyrrol-2-yl)methylene)amino)phenyl)imino)methyl)-2-ethoxy-2H-chromen-4-ol (probe 3) for finding both Cu²⁺ and CN⁻ ion concentrations [Scheme S1†]. The structures of the ligands were confirmed by elemental analysis, IR spectroscopy, UV-vis spectroscopy and ¹H NMR (Fig. S1 and S3†) and ¹³C NMR spectroscopy (Fig. S2 and S4†). All characterisation techniques confirmed the presence of a methoxy group in probe 2 and an ethoxy group in probe 3. This pointed towards in situ nucleophilic substitution via the solvent used. Mechanistic detail of the formation of probe 2 is shown in Fig. 1.

Fig. 1 Formation of probe 2 via in situ nucleophilic substitution of methoxide ion.

ESIPT has been developed as an emerging signalling mechanism; it is a rapid process with timescales ranging from fractions of picoseconds to tens of picoseconds, as the swiftness of proton transfer is far greater than that of electron transfer in the excited state.^18,19 One of the greatest advantages of ESIPT is large wavelength shifts that avoid overlap between absorption and emission spectra. ESIPT-based probe 2 behaves as a dual emitter as a result of the participation of the ground state and excited state of keto–enol tautomers [Fig. 2]. Absorption originates from the ground state of the stable enol form (E), potentially stabilised by intramolecular hydrogen bonding, and emission takes place from the preferred excited state of the keto form (K*), resulting in enhanced wavelength shift. Photoexcitation at 410 nm results in rapid conversion from the excited state of the enol (E*) to the excited state of the keto (K*) tautomer by ESIPT within a picosecond. Decay from K* to K (ground state of keto tautomer) gives rise to an emission band at 550 nm. At ground state level the less stable keto form (K) reverts to the more stable enol form (E) via reverse proton transfer. Molecules in the E* state that do not undergo ESIPT generate an emission band at 475 nm.

Fig. 2 The photophysical mechanism of probe 2 for illustration of ESIPT.

Results and discussion

Probes 2 and 3 were synthesized in 92% and 87% yields via a two-step process. As the presence of pyrrole, chromone and immine moieties provide the ligand with the capability to deliver multi-dentate metal coordination sites, we theorized that the probes could form stable coordination complexes with Cu²⁺ ions. Prior studies with various metal ions revealed that the probes had great affinity toward Cu²⁺ ions and displayed extremely sensitive and selective colorimetric sensing by the alteration in colour from yellow to brownish orange. When Cu²⁺ perchlorate salt was added to the ligands, the UV-vis absorptions were considerably red shifted to 381, 424 and 471 nm (Fig. S5a†) and fluorescence emission of the ligand at 550 nm was significantly reduced when excited at 410 nm. The fluorescence emission intensity of ligand displayed substantial reduction in the presence of Cu²⁺ ions. However, the addition of other metal ions, such as Mn³⁺, Pb²⁺, Cr³⁺, Zn²⁺, Hg⁺, Hg²⁺, Fe²⁺, Mg²⁺, Ni²⁺ and Na⁺, did not lead to substantial fluorescence changes, while Co²⁺ caused slight enhancement in fluorescence (Fig. S5b†).

Fig. 3 shows the absorbance and emission intensity changes of 2 (2 μM) in aqueous environment with the inclusion of Cu²⁺ ions. Remarkably, the fluorescence emission intensity around 550 nm was absolutely quenched when one equivalent of Cu²⁺ ions was added, presumably as a result of MLCT-based heavy metal ion effect, signalling gross formation of the 2–Cu complex.²⁰ Job's plot measurements and molar ratio analysis (Fig. S6†) of absorbance at 410 nm also supported the 1 : 1 stoichiometry between 2 and Cu²⁺ in situ. Fluorescence studies of the novel synthesized probe at different [Cu²⁺] revealed the lower detection limit to be 5.12 × 10⁻⁷ mol L⁻¹ as shown in Fig. S7.† The binding process of Cu²⁺ with probe were also investigated by optimized minimum energy structure (Density Functional Theory, DFT) calculation methods in the gas phase by applying the Gaussian 09W computational program with a B3LYP job using the basis set 6-31G (d, p) for probe and LANL2DZ for possible Cu²⁺ complexes respectively. The feasible Cu²⁺ complexes generated contained structural changes in the prospective site of probe. The bond lengths corresponding to –C=N (pyrrole), –C=N (benzene), –OH changed from 1.300 Å, 1.400 Å, 1.304 Å, 1.371 Å to 1.326 Å, 1.427 Å, 1.314 Å, 1.327 Å respectively. All the above bond length increases and –OH decreases are due to loss of ESIPT process upon binding with Cu²⁺. The energy band gap between HOMO–LUMO of the probable 2–Cu²⁺ complex was decreased (1.470 eV to 1.034 eV) owing to charge transfer between probe and Cu²⁺ ion, which also supports the red shift in the absorption spectra upon addition of Cu²⁺ (Fig. 4).

Fig. 3 Binding studies of probe 2 in aqueous solution (10 mM PBS buffer, 1.0% DMSO) with Cu²⁺: (a) UV-vis spectra with sequential addition of Cu²⁺. (b) Fluorescence emission titration with different concentrations of Cu²⁺ at 410 nm excitation wavelength with a slit width of 1/1. (Inset shows Benesi–Hilderbrand plot for calculating the binding constant.)

Fig. 4 Optimized structure with DFT and comparison of energy band gap between probe 2 and 2 + Cu²⁺.

Thus, the response of sensor, developed in situ by adding 1.0 equiv. of Cu²⁺ to ligand solution, to a variety of anions (sodium salts) including CN⁻, F⁻, Cl⁻, Br⁻, I⁻, NO³⁻, PO₄³⁻, SO₃²⁻, SO₄²⁻, S²⁻, HPO₄²⁻ was subsequently investigated. Specific fluorescence output of the probe in the presence of Cu²⁺ is not affected by the various counter anions with different shape and size, except CN⁻ ion as shown in Fig. 5(ii). CN⁻ coordinates with Cu²⁺ and forms an extremely stable [Cu(CN)_x]ⁿ⁻ complex having a lower solubility product constant (K = 1.27 × 10⁻²⁷). Addition of CN⁻ ion concomitantly enhances the fluorescence intensity (λ_em = 550) as a function of time and almost restores the original state of the probe, as shown in Fig. 6. Titration experiment at various [CN]⁻ revealed these ESIPT-based dual fluorescence sensors are capable of nanomolar detection of CN⁻ in aqueous and biological systems (Fig. S8†).

Fig. 5 Selectivity studies of probe 2 (2 μM). Probe 2 in aqueous solution (10 mM PBS buffer, 1.0% DMSO) with tested metal ions (i) and in situ formed 2 + Cu²⁺ with tested anions (ii).

Fig. 6 Time dependent fluorescence spectra of 2 + Cu²⁺ in aqueous solution (10 mM PBS buffer, pH 7.4, 1.0% DMSO) with 20 equivalents of CN⁻. Excitation at 410 nm and slit width of 1/1. Inset: plot of emission intensity at 550 nm vs. time.

As the quencher displacement strategy causes consequent changes in emission intensity at 550 nm, the resultant system works as a molecular switch at this fluorescence intensity, and can be used to perform Boolean logic operations. Molecular logic function was performed with the synthesized probe 2 along with Cu²⁺ (In-1) as well as CN⁻ (In-2) as inputs in the emission mode. In the existing system, the strong fluorescence at 550 nm is assigned as the ON state (output = 1) while the weak fluorescence is defined as the OFF state (output = 0). The threshold value of fluorescence intensity is specified as 15 000 CPS. Addition of Cu²⁺ to the probe 2 solution leads to fluorescence quenching to below the threshold level, whereas in situ addition of excess [CN⁻] almost restores the fluorescence intensity. The strong fluorescence intensity above the threshold label is recognized in the absence (0, 0) and presence (1, 1) of both inputs and also CN⁻ alone. These results simply establish the IMPLICATION logic gate, “→”, i.e., “p implies q” or “if p then q”, so the OFF state or 0 output is obtained only when the input is copper acetate. The working principle of the IMPLICATION logic gate is outlined in Fig. 7.

Fig. 7 Arrangement of the IMPLICATION logic gate. (a) Emission intensity changes of probe 2 in aqueous solution (10 mM PBS buffer, pH 7.4, 1.0% DMSO) with excitation at 410 nm and slit width of 1/1. (b) Changes in emission intensity under different input conditions. (c) Switch circuit diagram. (d) Truth table corresponding to logic gate.

Assessment of cytotoxicity of probes 2, 3, 2 + Cu²⁺ and 3 + Cu²⁺ complexes on human cell lines MCF-7 (breast cancer cell line), Hep G2 (human liver cancer cell line), HeLa (immortal cell line) and HEK-293 (Human Embryonic Kidney 293) was done via MTT assay. The cell lines were exposed for 24 h to medium containing the tested probes and complexes at 3 μg mL⁻¹. Screening for cytotoxicity showed that compounds have remarkable activity towards the MCF-7 cell line. In vitro anticancerous activity of the tested probes and respective complexes towards the MCF-7 cell line as well as HEK-293 cell lines (control experiment) was studied at various concentrations (0–60 μg mL⁻¹). In vitro anticancerous activity was calculated in terms of cell viability (%) via the following formula:

Cell viability (%) = [(mean OD of treated cells × 100)/mean OD of vehicle-treated cells]

A concomitant decrease in cell viability (%) was observed on subsequent increments of concentrations of the tested probes and their respective complexes. At higher concentrations, penetration of the tested probes and complexes inside the cells is increased, leading to more potent anticancerous activity.

Evaluation of the cytotoxicity of the tested compounds towards the MCF-7 cell line demonstrated that both the Cu²⁺ complexes exhibit potent cytotoxicity as compared with their respective parent probes. The 2 + Cu²⁺ complex showed best potency against the MCF-7 cell line with lowest IC₅₀ value (IC₅₀ = 5.69 ± 0.26 μg mL⁻¹) while the 3 + Cu²⁺ complex also displayed very pronounced potency (IC₅₀ = 7.36 ± 0.29 μg mL⁻¹). Probe 2 (IC₅₀ = 35.48 ± 0.34 μg mL⁻¹) and probe 3 (IC₅₀ = 48.99 ± 0.30 μg mL⁻¹) also exhibited moderate cytotoxicity against the MCF-7 cell line. However, the results clearly indicate a noncytotoxic effect on the HEK-293 cell line (normal cell line). The morphological changes discovered in the MCF-7 and HEK-293 cell lines exposed to the tested probes and complexes (30 μg mL⁻¹) for 24 h are presented in Fig. 8 and S9,† respectively. The MCF-7 cell line exhibited reduction of morphology and cell adhesion ability after this exposure.

Fig. 8 (a) Cytotoxic assessment. Morphological changes in MCF-7 cell line after 24 h of exposure with (i) control, (ii) probe 2, (iii) 2 + Cu²⁺, (iv) probe 3, and (v) 3 + Cu²⁺ complex. Images were taken by inverted phase contrast microscope at 200× magnification. (b and c) MTT assay of MCF-7 cell line: (b) probe 2 and 2 + Cu²⁺; (c) probe 3 and 3 + Cu²⁺ complex. *p < 0.05.

To establish the sensing ability of 2 towards Cu²⁺ and CN⁻ ions in living cells, bioimaging studies with the MCF-7 cell line were performed. When the cells preloaded with probe 2 (1 μM) were treated with equimolar cupric acetate, an immediate decrease in fluorescence intensity was observed inside the cells. When these cells were treated with various concentrations of NaCN (1–100 μM), fluorescence was marginally visible after 20 min of incubation prior to the NaCN treatment. The time dependent and concentration dependent increase of fluorescence intensity observed inside the cells apparently explains the chemodosimetric sensing of Cu²⁺ and CN⁻ ions, as shown in Fig. 9.

Fig. 9 (a) Confocal microscopic images of MCF-7 cells in the presence of probe 2 (1 μM), after the addition of cupric acetate (1 μM) and after the addition of NaCN (50 μM). (b) Confocal phase contrast images of cells shown in panel (a). (c) Overlay images of panels (a) and (b).

Conclusions

Owing to the great biological concern of Cu²⁺ and CN⁻, it is immensely desirable to establish colorimetric and fluorescent sensory units for these ions. Interestingly, the novel fluorescence chemosensor 2, containing pyrrole and chromone moieties and based on the ESIPT mechanism, offered highly selective and sequential sensing of Cu²⁺ and CN⁻ ions in 99% aqueous medium in the presence of other common interfering ions. The in situ complexation and quencher displacement mechanism were elucidated by fluorescence experiments and calculations. At 550 nm the emission intensity of the system worked as IMPLICATION molecular switch, which performed Boolean logic operations. The in situ formed probe–Cu²⁺ complexes exhibit high anticancer activity towards the MCF-7 cell line. Selective in situ sensing for biological applications was also studied in the MCF-7 cell line to express “ON–OFF–ON” fluorescence cellular images.

Acknowledgements

Ms Shubhrajyotsna Bhardwaj is thankful to CSIR for providing financial assistance.

References

1. (a) H. S. Jung, P. S. Kwon, J. W. Lee, J. Kim, C. S. Hong, J. W. Kim, S. Yan, J. Y. Lee, J. H. Lee, T. Joo and J. S. Kim, J. Am. Chem. Soc., 2009, 131, 2008–2012; (b) L. Peng, M. Wang, G. Zhang, D. Zhang and D. Zhu, Org. Lett., 2009, 11, 1943–1946; (c) J. Wu, W. Liu, J. Ge, H. Zhang and P. Wang, Chem. Soc. Rev., 2011, 40, 3483–3495.
2. (a) I. Yousuf, F. Arjmand, S. Tabassum, L. Toupet, R. A. Khan and M. A. Siddiqui, Dalton Trans., 2015, 44, 10330–10342; (b) O. M. Abdelhafez, H. I. Ali, K. M. Amin, M. M. Abdalla and E. Y. Ahmed, RSC Adv., 2015, 5, 25312–25324; (c) C. Zwergel, S. Valente, A. Salvato, Z. Xu, O. Talhi, A. Mai, A. Silva, L. Altucci and G. Kirsch, Med. Chem. Commun., 2013, 4, 1571–1579.
3. (a) R. K. Konidena and K. R. J. Thomas, RSC Adv., 2014, 4, 22902–22910; (b) Y. Liu, K. Ai, X. Cheng, L. Huo and L. Lu, Adv. Funct. Mater., 2010, 20, 951–956; (c) S. Kim, J. Y. Noh, K. Y. Kim, J. H. Kim, H. K. Kang, S. W. Nam, S. H. Kim, S. Park, C. Kim and J. Kim, Inorg. Chem., 2012, 51, 3597–3602; (d) Y. K. Yang and J. Tae, Org. Lett., 2006, 8, 5721–5723.
4. (a) Z. Zhou, M. Yu, H. Yang, K. Huang, F. Li, T. Yi and C. Huang, Chem. Commun., 2008, 3387–3389; (b) A. M. Hessels and M. Merkx, Metallomics, 2015, 7, 258–266; (c) J. H. Lee, A. R. Jeong, I. S. Shin, H. J. Kim and J. I. Hong, Org. Lett., 2010, 12, 764–767.
5. (a) P. Anzenbacher Jr, D. S. Tyson, K. Jursíková and F. N. Castellano, J. Am. Chem. Soc., 2002, 124, 6232–6233; (b) J. Wu, W. Liu, J. Ge, H. Zhang and P. Wang, Chem. Soc. Rev., 2011, 40, 3483–3495; (c) W. H. Chen, Y. Xing and Y. Pang, Org. Lett., 2011, 13, 1362–1365; (d) V. Luxami and S. Kumar, RSC Adv., 2012, 2, 8734–8740.
6. (a) C. Suksai and T. Tuntulani, Chem. Soc. Rev., 2003, 32, 192–202; (b) S. Banthia and A. Samanta, New J. Chem., 2005, 29, 1007–1010; (c) G. Sánchez, D. Curiel, I. Ratera, A. Tárraga, J. Veciana and P. Molina, Dalton Trans., 2013, 42, 6318–6326.
7. (a) T. Ábalos, S. Royo, R. M. Máñez, F. Sancenón, J. Soto, A. M. Costero, S. Gil and M. Parra, New J. Chem., 2009, 33, 1641–1645; (b) P. Calero, E. Aznar, J. M. Lloris, M. D. Marcos, R. M. Máñez, J. V. R. Lis, J. Soto and F. Sancenón, Chem. Commun., 2008, 1668–1670; (c) J. Du, M. Hu, J. Fan and X. Peng, Chem. Soc. Rev., 2012, 41, 4511–4535.
8. (a) Z. Wang, A. v. d. Bussche, P. K. Kabadi, A. B. Kane and R. H. Hurt, ACS Nano, 2013, 7, 8715–8727; (b) Y. Chen, D. Wang, X. Zhu, X. Zheng and L. Feng, Environ. Sci. Technol., 2012, 46, 12452–12458.
9. (a) Y. Nodaa, M. Asada, M. Kubota, M. Maesakoa, K. Watanabeb, M. Uemurab, T. Kihara, S. Shimohama, R. Takahashi, A. Kinoshita and K. Uemura, Neurosci. Lett., 2013, 547, 10–15; (b) Y. Takahashi, H. Miyajima, S. Shirabe, S. Nagataki, A. Suenaga and J. D. Gitlin, Hum. Mol. Genet., 1996, 5, 81–84.
10. (a) D. Kim, N. H. Kim and S. H. Kim, Angew. Chem., 2013, 52, 1139–1142; (b) E. Gaggelli, H. Kozlowski, D. Valensin and G. Valensin, Chem. Rev., 2006, 106, 1995–2044; (c) S. Dell’Acqua, V. Pirota, C. Anzani, M. M. Rocco, S. Nicolis, D. Valensin, E. Monzani and L. Casella, Metallomics, 2015, 7, 1091–1102; (d) R. A. Himes, G. Y. Park, G. S. Siluvai, N. J. Blackburn and K. D. Karlin, Angew. Chem., 2008, 47, 9084–9087.
11. B. Vennesland, E. E. Conn, C. J. Knowles, J. Westley and F. Wissing, Cyanide in Biology, Academic Press, London, 1981.
12. R. Koenig, Wildlife deaths are a grim wake-up call in Eastern Europe, Science, 2000, 10, 1737–1738.
13. S. Kimani, K. Sine, F. Bukachi, T. Katumbay and C. Maitai, Metab. Brain Dis., 2014, 29, 105–112.
14. (a) Y. Ding, T. Li, W. Zhu and Y. Xie, Org. Biomol. Chem., 2012, 10, 4201–4207; (b) Y. R. Zhu, H. Li, G. T. Yan, B. B. Shi, Y. M. Zhang, Q. Lin, H. Yao and T. B. Wei, RSC Adv., 2015, 5, 49953–49957.
15. (a) H. T. Niu, D. Su, X. Jiang, W. Yang, Z. Yin, J. He and J. P. Cheng, Org. Biomol. Chem., 2008, 6, 3038–3040; (b) M. E. Jun, B. Roy and K. H. Ahn, Chem. Commun., 2011, 47, 7583–7601.
16. (a) J. Jo and D. Lee, J. Am. Chem. Soc., 2009, 131, 16283–16291; (b) M. Dong, Y. Peng, Y. M. Dong, N. Tang and Y. W. Wang, Org. Lett., 2012, 14, 130–133; (c) C. L. Chen, Y. H. Chen, C. Y. Chen and S. S. Sun, Org. Lett., 2006, 8, 5053–5056; (d) H. C. Gee, C. H. Lee, Y. H. Jeong and W. D. Jang, Chem. Commun., 2011, 47, 11963–11965.
17. (a) G. R. You, G. J. Park, J. J. Lee and C. Kim, Dalton Trans., 2015, 44, 9120–9129; (b) L. Tang, P. Zhou, K. Zhong and S. Hou, Sens. Actuators, B, 2013, 182, 439–445.
18. (a) Z. Tu, M. Liu, Y. Qian, G. Yang, M. Cai, L. Wang and W. Huang, RSC Adv., 2015, 5, 7789–7793; (b) Z. Tu, Q. Zhang, M. Liu, Y. Qian, L. Wang and W. Huang, J. Mater. Sci., 2016, 51, 2972–2979; (c) G. Yang, S. Li, S. Wang, R. Hu, J. Feng, Y. Li and Y. Qian, Pure Appl. Chem., 2013, 85, 1257–1513.
19. (a) G. Yang, K. Zhang, F. Gong, M. Liu, Z. Yang, J. Ma and S. Li, Sens. Actuators, B, 2011, 155, 848–853; (b) R. Hu, J. Feng, D. Hu, S. Wang, S. Li, Y. Li and G. Yang, Angew. Chem., 2010, 49, 4915–4918.
20. H. S. Jung, J. H. Han, Z. H. Kim, C. Kang and J. S. Kim, Org. Lett., 2011, 13, 5056–5059.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c6ra22352d

---