A single molecule multi analyte chemosensor differentiates among Zn²⁺, Pb²⁺ and Hg²⁺: modulation of selectivity by tuning of solvents

Abstract

A chemosensor detects and differentiates among Hg²⁺, Pb²⁺ and Zn²⁺ with fluorescence enhancement at different wavelengths under different solvent conditions. In methanol, Hg²⁺ and Pb²⁺ displace Zn²⁺ from the [Zn²⁺-sensor] complex displaying ratiometric behaviour. In water, the chemosensor recognises only Hg²⁺ with fluorescence enhancement, and offers intracellular detection of Hg²⁺.

---

Fluoroscence sensing is a rapidly evolving field of research.¹ Sensing multiple analytes with a single chemosensor is a challenging task. It involves eliciting different responses by the molecule in the presence of different analytes. Such a molecule would involve either multiple binding sites giving rise to different binding modes, or multiple signal transducing units (e.g., chromophores/fluorophores/redox active moieties), multiple read-out modes or any combination of the above. Chemosensors with selectivity for two analytes have received much attention in recent years. There are reports of two-ion sensors that detect Ba²⁺–Hg²⁺,^2a Hg²⁺–Fe²⁺,^2b Al³⁺–Zn²⁺,^2c Zn²⁺–Cd²⁺,^2d Cu²⁺–Zn²⁺,^2e Hg²⁺–Pb²⁺,^2f and Hg²⁺–Cr³⁺.^2g The above examples, for brevity, are non-exclusive. Selective detection of three analytes with a chemosensor having only one kind of fluorophore is rare in the literature. Such a chemosensor is highly coveted in terms of convenience and economy. Three analyte molecular receptors have been used for the construction of molecular logic gates where an emission output rather than the differentiation of response by multiple analyte is important.³ The Zn²⁺ ion is present in several metalloenzymes and is indispensable for proper cellular functions. It plays a crucial role in normal functioning of the central nervous system.⁴ Zinc deficiency causes impaired liver, kidney and neural functions,⁵ whereas an excess intake suppresses absorption of copper and iron, leading to other secondary problems.⁶ Pb²⁺ and Hg²⁺, on the other hand, are toxic even at a very low levels and cause serious damage to the skin, nervous and renal systems.⁷ The US-EPA guidelines allow a maximum acceptable concentration of 0.015 and 0.002 ppm of inorganic lead and mercury respectively in drinking water.⁷

In this communication, we report a coumarin based fluorescent chemosensor 1 on a thiocarbamate scaffold that displays different selectivity towards Hg²⁺, Pb²⁺ and Zn²⁺ ions under different conditions. In non-aqueous media, chemosensor 1 detects Hg²⁺, and Pb²⁺ with enhancement of the ligand fluorescence band at 410 nm. With Zn²⁺, the chemosensor displays ratiometric fluorescence behaviour at 410 and 495 nm. Both Hg²⁺ and Pb²⁺ can displace Zn²⁺ from the [Zn²⁺·1] complex and demonstrate ratiometric behaviour. In aqueous media, the chemosensor selectively recognises only Hg²⁺ with a remarkable fluorescence enhancement at nmol L⁻¹ concentrations. The S donor atom in the photo-induced electron transfer (PET) process ensures its activity even at low pH.⁸ In aqueous media, the sensor also detects intracellular Hg²⁺.

The bipodal chemosensor 1 consists of two thiocarbamate moieties attached to coumarin units via ethylenediamine spacers (Fig. 1). It was synthesized from bis-bromomethyl salicylaldehyde and ethylenediamine-coumarin conjugate in the presence of CS₂ under aqueous conditions (Scheme S1, SI‡). The compound was characterised by 1D and 2D NMR and other spectroscopic methods. The sensing behaviour of 1 was investigated by UV-vis and fluorescence measurements with several metal ions i.e., Na⁺, K⁺, Al³⁺, Ca²⁺, Cd²⁺, Cr³⁺, Mn²⁺, Fe²⁺, Fe³⁺, Co²⁺, Ni²⁺, Cu²⁺, Zn²⁺, Pb²⁺, Li⁺, Ba²⁺, Ag⁺ and Hg²⁺ in CH₃OH and H₂O with 1% DMSO as a co-solvent in the presence of HEPES buffered at pH 7.0. The absorption spectrum of 1 (10 μmol L⁻¹) in MeOH showed absorption bands at 280 nm and 335 nm. Systematic titration of sensor 1 with increasing concentrations of Zn²⁺ revealed a continuous intensity increase of the absorption band at 280 nm with a new charge transfer transition band at 400 nm (Fig. S1, SI‡). In contrast, titration of 1 with one equivalent of Hg²⁺ resulted in a small decrease in the absorption at 280 nm with a 12 nm bathochromic shift whereas the absorption at 335 nm increased in intensity (Fig. S2, SI‡). Under the same conditions with increasing concentrations of Pb²⁺ a decrease in the absorption intensity with a bathochromic shift to 290 nm (Δλ = 10 nm) was observed (Fig. S3, SI‡). Interestingly, in aqueous media, 1 showed ratiometric behaviour with Hg²⁺. Upon titration of 1 (10 μmol L⁻¹) with Hg²⁺ (0–30 μM) the absorption at 400 nm gradually decreased in intensity, whereas the absorption at 280 nm went up with an isosbestic point at 350 nm (Fig. S4, SI‡).

Fig. 1 Structure of chemosensor 1.

The emission spectrum of 1 (λ_ex = 320 nm, slit 5/5 nm) in MeOH showed a characteristic fluorescence band (quantum yield, Φ_f = 0.116) at 410 nm (Fig. 2A). Upon addition of the above perchlorate salts, 1 with Zn²⁺ showed a significant decrease in the 410 nm emission band. A new red shifted (85 nm) intense emission band centered at 495 nm (Φ_f = 0.50) was also observed with an isoemission point at 450 nm, indicating Zn²⁺ selective ratiometric behaviour (Fig. 2B). The new band at 495 nm in the presence of Zn²⁺ is most likely an ICT based optical response.^2c

Fig. 2 (A) Change in fluorescence intensity of 1 (10 μmol L⁻¹) in MeOH/DMSO (99 : 1, v/v) with different cations (20 μM) in the presence of HEPES buffer at pH 7.0. (B) Fluorescence spectra of 1 (10 μmol L⁻¹) upon addition of Zn²⁺ (0 to 40 μmol L⁻¹) in MeOH/DMSO (99 : 1, v/v). Inset: Zn²⁺ titration profile at 410 and 492 nm.

Furthermore, under the same conditions, Hg²⁺ and Pb²⁺ only enhanced the intensity (Φ_f = 0.58 and 0.39) of the 410 nm fluorescence band, suggesting 1 as an efficient Hg²⁺, Pb²⁺ selective fluorescence turn-ON sensor based on the PET mechanism (Fig. 2A). With other metal ions, a small fluorescence enhancement (Cd²⁺, Ag⁺, Cu²⁺) or quenching (Ni²⁺, Mn²⁺, Co²⁺, Fe²⁺, Fe³⁺) was observed (Fig. 2A and S5, SI‡).

The stoichiometry of the Zn²⁺·1 association was determined by a Job's plot from the UV-vis absorption data (Fig. S7, SI‡).⁹ The plot revealed a 2 : 1 ratio of the Zn²⁺ ions associating with each molecule of 1. A peak at m/z 907 in ESI-MS further supported the 2 : 1 binding stoichiometry (Fig. S20, SI‡). The binding constant for the complex of Zn²⁺ with 1 was calculated using the method of Lehrer and Chipman from the fluorescence data (Fig. S11, SI‡).¹⁰ The stoichiometry of the Zn²⁺·1 association thus obtained was 2.08, which is again in perfect agreement with the 2 : 1 ratio obtained earlier. From the intercept of the best fitting straight line, an association constant of 7.07 × 10⁴ mol² L⁻² was obtained. The analytical detection limit of Zn²⁺ determined following the reported method¹¹ (Fig. S15, SI‡) was 0.58 μmol L⁻¹.

Interestingly, the addition of Hg²⁺ and Pb²⁺ to a solution of the Zn²⁺·1 complex results in a shift in the fluorescence emission from green to violet. Thus, the Zn²⁺·1 ensemble can be used to detect Hg²⁺ and Pb²⁺via the displacement of the Zn²⁺ ion based on the higher affinity of 1 for these two ions. Upon addition of an increasing concentration of Hg²⁺ and Pb²⁺, the emission maximum of the Zn²⁺·1 complex at 495 nm gradually decreased with an increase of the band at 410 nm (Fig. 3A and 3B) with the formation of clear isoemission points (at 480 nm for Hg²⁺, and 465 nm for Pb²⁺). Moreover, the replacement of Zn²⁺ by Hg²⁺ and Pb²⁺ has been confirmed by ¹H NMR experiments (Fig. 4B). Thus, the Zn²⁺·1 chemosensing ensemble serves as a ratiometric sensor for Hg²⁺ and Pb²⁺ ions via the competitive metal ion displacement approach. The analytical detection limit¹¹ for Pb²⁺ under these conditions was 1.6 × 10⁻⁸ mol L⁻¹ (Fig. S16 and S17, SI‡). The effect of Zn²⁺, Hg²⁺ and Pb²⁺ with 1 was independently studied with ¹H NMR spectroscopy in CD₃OD (Fig. 4B). In the presence of Zn²⁺, small downfield shifts of the ethylenediamine and the coumarin H_a protons were observed indicating their proximity to the binding site. With 1, Hg²⁺ and Pb²⁺ induced a significant upfield shift of the methylene protons next to the amide linkage (H_b), and a downfield shift of methylene protons next to the thiocarbamate moiety (H_c) in addition to the downfield shift of H_a. Besides, when 1 eqv. of Hg²⁺ (or Pb²⁺) was added to the Zn²⁺·1 ensemble, the chemical shifts were almost identical to those of the Hg²⁺·1 (or Pb²⁺) complex (Fig. 4B). This is an added proof for the higher affinity of 1 towards Hg²⁺ and Pb²⁺ compared to Zn²⁺. Based on the above results the proposed binding mode of 1 with Zn²⁺, Hg²⁺ and Pb²⁺ is shown in Fig. 4A.

Fig. 3 Change in fluorescence spectrum of Zn²⁺·1 (10 μmol L⁻¹ of 1 mixed with 20 μmol L⁻¹ of Zn²⁺) in MeOH/DMSO (99 : 1, v/v) upon addition of (A) Hg²⁺ and (B) Pb²⁺ in HEPES buffered at pH 7.0. Inset: (A) (i) Ratio of emission intensity [I₄₁₅/I₄₉₅] as a function of eqv. of Hg²⁺ (ii) and Job's plot between 1 and Hg²⁺. (B) Ratiometric emission intensity [I₄₁₀/I₄₉₅] as a function of eqv. of Pb²⁺.

Fig. 4 (A) Probable binding model; (B) ¹H NMR titration of sensor 1 with Zn²⁺, Hg²⁺ and Pb²⁺.

To determine the stoichiometry, a Job's plot obtained from the emission data showed a 1 : 1 association of 1 with both Hg²⁺ and Pb²⁺ (Inset, Fig. 3(ii) and Fig. S8, SI‡). Furthermore, the result is consistent with the ESI-MS experiment of 1 with excess Pb²⁺ and Hg²⁺ (Fig. S21 and S22, SI‡). Under these conditions, the binding constant for Hg²⁺ and Pb²⁺ was found to be 3.08 × 10⁵ M⁻¹ and 3.90 × 10⁵ M⁻¹ respectively (Fig. S12 and S13, SI‡).¹²

The fluorescence behaviour of 1 was then studied in the presence of various metal ions (Na⁺, K⁺, Al³⁺, Ca²⁺, Cd²⁺, Cr³⁺, Mn²⁺, Fe²⁺, Fe³⁺, Co²⁺, Ni²⁺, Cu²⁺, Zn²⁺, Pb²⁺, Li⁺, Ba²⁺, Ag⁺ and Hg²⁺) in aqueous medium (H₂O/DMSO, 99 : 1, v/v) (Fig. 5A). The emission spectrum (λ_ex = 320 nm) of 1 in H₂O showed a weak fluorescence band (Φ_f = 0.028) at 410 nm. Interestingly, among the above metal ions, only the Hg²⁺ ion enhanced the fluorescence signal remarkably (Φ_f = 0.48, 17-fold) with a 5 nm red shift in the fluorescence peak (Fig. 5B). The fluorescence intensity increased linearly with the addition up to one equivalent of Hg²⁺ ions (Fig. 5B, inset). There was no significant change in the fluorescence spectra with any of the other metal ions except for Ag⁺, which showed a small enhancement in the emission spectra (Fig. 5A).

Fig. 5 (A) Emission spectra of 1 (10 μM) in aqueous medium upon addition of metal ions (20 μmol L⁻¹). (B) Changes in the emission spectra of 1 (10 μmol L⁻¹) in aqueous medium with Hg²⁺ (0 to 40 μmol L⁻¹) in the presence of HEPES buffer at pH = 7.0. Inset: Hg²⁺ titration profile at 415 nm.

Therefore, in aqueous medium, sensor 1 is highly specific towards Hg²⁺. Under aqueous conditions, Pb²⁺ does not display any fluorescence enhancement, and thus can clearly be distinguished from Hg²⁺. Competitive experiments were also carried out to investigate the effects of other coexisting cations along with Hg²⁺ in the presence of the chemosensor (Fig. S6, SI‡). The results clearly demonstrate that the fluorescence response of 1 is not affected in the presence of other metal ions in aqueous medium. The analytical detection limit for Hg²⁺ with 1 in aqueous medium was found to have a remarkably low value of 9 nmol L⁻¹ (Fig. S17, SI‡).¹¹ Under aqueous conditions, the 1 : 1 association of Hg²⁺ with 1 was confirmed by a Job's plot (Fig. S9, SI‡) and a mass spectrum (Fig. S22, SI‡). The NMR spectrum of 1 with Hg²⁺ in D₂O/DMSO-d₆ showed a larger downfield shift of several protons compared to the ones in CD₃OD indicating a stronger association between 1 and Hg²⁺ in aqueous medium (Fig. S10, SI‡). Indeed, the binding constant in water was found to be 1.2 × 10⁶ M⁻¹ following a Benesi–Hildebrand analysis (Fig. S14, SI‡).¹²

The pH dependence of the chemosensor was investigated. Both 1 as well as the Hg²⁺ complex displayed their fluorescence behaviour from pH 2 to 8 (Fig. S18, SI‡). This indicates that the PET process perhaps involves an electron transfer from the thiocarbamate sulphur to the fluorophore. We envisaged that compound 1 might cross the cell membrane efficiently at physiological pH since the molecule is uncharged and less polar, and after complexation with Hg²⁺, owing to its charged nature it might stay inside the cell. Thus, 1 was tested for the recognition of Hg²⁺ ions in cells using fluorescence imaging experiments. HeLa S3 cell lines were exposed to Hg²⁺ (500 μmol L⁻¹) in DMEM for 2 h at 37 °C (see SI‡). Subsequently, the cells were washed twice to remove the remaining Hg²⁺ ions, and incubated with 1 (10 μM) in PBS for 10 min at 25 °C. The cells were washed again with PBS to remove any residual dye from the cells. The epifluorescence image of HeLa S3 loaded with 1 (10 μmol L⁻¹) under the same conditions showed negligible intracellular fluorescence (Fig. 6B). In contrast, the cells treated with Hg²⁺ and 1 revealed highly enhanced intracellular fluorescence (Fig. 6C). These results demonstrate that 1 can be used to detect Hg²⁺ in living cells.

Fig. 6 Fluorescence images of HeLa S3: (A) only cells, (B) cells with 1 (10 μM), (C) 1- stained cells exposed to Hg²⁺.

In conclusion, we have synthesized a coumarin based new fluorescence probe 1 that detects multiple metal ions in organic and aqueous media. In methanol sensor 1 shows a ratiometric response to Zn²⁺. Differentiation of the metal ions under different conditions is summarized in Table 1. Furthermore, Hg²⁺ and Pb²⁺ displace Zn²⁺ from the Zn²⁺·1 ensemble showing a second ratiometric fluorescence response with isoemission points at 480 and 465 nm. In water, chemosensor 1 exclusively detects Hg²⁺ demonstrating its excellent selectivity for Hg²⁺. The sulphur donor in the PET ensures the sensor activity even at low pH. The detection limit of Hg²⁺ in water of 9 nmol L⁻¹ allows it to detect the minimum level as set by the US-EPA for drinking water. The highly enhanced intracellular fluorescence of 1 in the presence of mercury ions demonstrates that 1 can detect Hg²⁺ in living cells.

Table 1 Differentiation of the metal ions from their fluorescence behaviour

Metal salt in	Enhancement at^b	Inference^c
a In addition, Zn^2+·1 displays ratiometric behaviour with Pb^2+/Hg^2+. b The values in parentheses are the Φf values; λex = 320 nm, slit 5/5 nm. c The values in parentheses are the Φf values in presence of the respective metal ions. See text.
1 in MeOH^a	(a) 495 nm (—)	Zn^2+ (0.50)
	(b) 410 nm (0.12)	Pb^2+/Hg^2+ (0.058/0.039)
1 in water	(c) 415 nm (0.03)	Hg^2+ (0.48)
	(d) but not (c)	Pb^2+

---

The authors thank DST, India for financial support, IISER for facilities, and CSIR for a SRF to JH.

References

1. (a) A. W. Czarnik, Acc. Chem. Res., 1994, 27, 302; (b) B. Wang and E. J. Anslyn, Chemosensors: Principles, Strategies, and Applications, John Wiley and Sons, New York, 2011..
2. (a) X. Liu, M. Zhu, J. Li, X. Yin, H. Zheng, Z. Zuo, C. Ouyang, H. Liu, Y. Li and D. Zhu, J. Org. Chem., 2008, 73, 5008; (b) Y. Chen, L. Wan, X. Yu, W. Li, Y. Bian and J. Jiang, Org. Lett., 2011, 13, 5774; (c) D. Maity and T. Govindaraju, Chem. Commun., 2012, 48, 1039; (d) L. Xue, Q. Liu and H. Jiang, Org. Lett., 2009, 11, 3454; (e) M.-M. Yu, Z.-X. Li, L.-H. Wei, D.-H. Wei and M.-S. Tang, Org. Lett., 2008, 10, 5115; (f) R. Pandey, R. K. Gupta, M. Shahid, B. Maiti, A. Misra and D. S. Pandey, Inorg. Chem., 2012, 51, 298; (g) P. Das, A. Ghosh, H. Bhatt and A. Das, RSC Adv., 2012, 2, 3714.
3. A. P. de Silva, Chem.–Asian J., 2011, 6, 750.
4. B. K. Bitanihirwe and M. G. Cunningham, Synapse, 2009, 63, 1029.
5. A. S. Prasad, Br. Med. J., 2003, 326, 409.
6. G. J. Fosmire, Am. J. Clin. Nutr., 1990, 51, 225.
7. http://water.epa.gov/drink/contaminants/index.cfm; last accessed 26 Mar 2012.
8. Y. Lu, S. Huang, Y. Liu, S. He, L. Zhao and X. Zeng, Org. Lett., 2011, 13, 5274.
9. P. Job, Ann. Chim., 1928, 9, 113.
10. (a) S. Lehrer and G. D. Fashman, Biochem. Biophys. Res. Commun., 1966, 23, 133; (b) D. M. Chipman, V. Grisaro and N. Shanon, J. Biol. Chem., 1967, 242, 4388; (c) S. Saha, A. Ghosh, P. Mahato, S. Mishra, S. K. Mishra, E. Suresh, S. Das and A. Das, Org. Lett., 2010, 12, 3406.
11. A. Caballero, R. Martinez, V. Lloveras, I. Ratera, J. Vidal-Gancedo, K. Wurst, A. Tarranga, P. Molina and J. Veciana, J. Am. Chem. Soc., 2005, 127, 15666.
12. (a) H. A. Benesi and J. H. Hildebrand, J. Am. Chem. Soc., 1949, 71, 2703; (b) M. Barra, C. Bohne and J. C. Scaiano, J. Am. Chem. Soc., 1990, 112, 8075.

---

Footnotes

† Dedicated to Professor H. Junjappa.

‡ Electronic supplementary information (ESI) available: details of experimental procedures and characterization data for the products and spectroscopic methods and data. See DOI: 10.1039/c2ra20822a

---