Supramolecular analyte recognition: experiment and theory interplay

Abstract

Due to the large impact of analyte recognition on biological and environmental studies, the development of supramolecular recognition and sensing strategies has attracted considerable interest in recent years. Although researchers employ a variety of spectroscopic (optical and or fluorescence) and/or the analytical techniques based on changes in mass, electrochemical behaviour and other changes to assess the perturbation of physico-chemical changes associated with a supramolecular sensing event, theoretical calculations have proven to be a pivotal tool to substantiate experimental results. Such calculations assume additional significance as the information gained in such calculations can help designing new, even more efficient chemosensors and the time and resources consumed in the synthesis of sensors and associated trials, errors and operational difficulties are considerably reduced. This review presents some supramolecular sensing events wherein the sensing mechanism is supported by theoretical predictions of the binding modes based on the perturbations in the energies of the frontier molecular orbitals upon recognition of the guest species.

---

Paramjit Kaur

Paramjit Kaur received her Ph.D. degree from Guru Nanak Dev University in 1990. She did a post-doctorate at the Departmento de Quimica Inorganica at Universidad Complutense de Madrid before joining as a Lecturer in the Department of Chemistry at Guru Nanak Dev University in 1997, where she is now a full Professor. Her research interest focuses on donor–acceptor chemistry and chemosensors.

Kamaljit Singh

Kamaljit Singh received his Ph.D. degree from Guru Nanak Dev University (GNDU) in 1989. After doing post-doctorate work at Instituto de Quimica Organica General (CSIC), Madrid, he became a Reader in Applied Chemistry Department of GNDU in 1995. He has been a full Professor in the Department of Chemistry at Guru Nanak Dev University since 2003. His research interests focus on synthetic organic chemistry, molecular electronics and recognition. He has been an INSA-RSC and The British Council Visiting Scientist at the University of Manchester, Institute of Science and Technology. He was the recipient of the Bronze Medal of the Chemical Research Society of India (2009).

---

1. Introduction

Owing to the influence of various anions and cations in our daily life, the design and development of molecular receptors possessing the ability to selectively recognize analytes and signal a quantifiable response has emerged as an active area of research. The molecular sensors are generally composed of a selective receptor and a signaling subunit. The function of the receptor is to recognize an analyte mainly through supramolecular interactions; the signalling subunit transduces the response by a signal output in terms of color change, emission change (enhancement or quenching of emission intensity), redox shifts etc.¹ As a consequence of this, supramolecular molecular sensors have attained a special status owing to their implication in the fabrication of sensing devices. For instance, the receptors which exhibit color and emission changes in the presence of a guest species have emerged as ideal candidates for use in optical fibre devices, whereas the redox active molecular sensors signalling redox changes upon the addition of analytes, can find applications in the development of amperometric sensing devices etc. In general, the binding/interaction of the receptor with the analyte can modulate (i) the UV-visible absorption properties of the former, which is manifested in terms of bathochromic or hypsochromic shifts in the absorption bands responsible for “naked-eye” color changes during a sensing event, (ii) the fluorescence properties of the receptor manifested as enhanced/quenched/shifted or the appearance of new emission band(s), and (iii) the redox properties of the receptor in terms of shifts (cathodic or anodic) in the redox potentials of the receptor and/or the appearance of new peaks. Such changes are measurable and quantifiable leading to a wealth of information regarding binding constants, stoichiometry and formation receptor-analyte species, limits of detection, sensitivity, as well as selectivity of the sensing process. Theoretical calculations have been frequently used as a robust tool to substantiate the proposed binding hypothesis and have thus gained significance. Optimization of the molecular structure and geometry, evaluation of the energies of the participating frontier molecular orbitals (FMOs), highest occupied and lowest unoccupied molecular orbitals (HOMO and LUMO, respectively) of the receptor, assignment and correlation of the molecular electronic transitions, atomic charges etc. provide an in-depth insight into the understanding of the sensing mechanism and these studies can also help to design more efficient sensors. The current review provides an overview of the representative work during the recent past mainly 2009 to 2012, wherein the experimental results of supramolecular sensing events have been supported by such calculations and the binding mechanisms have been unequivocally suggested.

2. Cation recognition

2.1 Charge-transfer (CT) based systems

A coumarin-based fluorogenic probe bearing a 2-picolyl unit 1 was developed² as a chemosensor for Cu²⁺ (Fig. 1), wherein the quenching of fluorescence of the coumarin moiety after complexation with Cu²⁺ via ligand to metal d-orbital and/or ligand to metal charge-transfer (LMCT) was proposed to be the operative mechanism. In order to gain deeper insight into understanding what type of energy changes the frontier orbitals of 1 undergo upon interaction with the guest, ab initio calculations suggested the mechanism wherein LMCT having the main contribution from HOMO → LUMO+4 (18%) and HOMO → LUMO+6 (11%) excitations from the coumarin moiety to Cu²⁺ centre, provides a pathway for the non-radiative deactivation of the excited state which in fact, is one of the causes of the observed quenching of the fluorescence emission.

Fig. 1 Coumarin based fluorogenic probe for Cu²⁺ detection.

A semisquarylium dye 2 was found to be highly selective for Hg²⁺ ions.³ The intensity of both the absorption and emission bands of the dye decreased in the presence of Hg²⁺ ions, and was attributed to the formation of a 2 : 1 (2 : Hg²⁺) coordination complex during the sensing event (Fig. 2). The formation of the complex is supported by the calculated geometry indicating bridging of the Hg²⁺ ion between the sulfur atom and the carbonyl oxygen atom. Calculation of the electron distribution in the HOMO and LUMO orbitals of the dye revealed that HOMO–LUMO excitation displaced the electron density from the thiazole moiety to the cyclobutene moiety reflecting the strong intramolecular charge-transfer (ICT) character of the dye (Fig. 3).

Fig. 2 Complexation of Hg²⁺ by semisquarylium dye.

Fig. 3 Electron density distribution of the HOMO and LUMO energy levels of 2 [reprinted with permission from ref. 3. Copyright 2009 Elsevier].

Complexation with Hg²⁺ via the sulfur atom of thiazole reduces the electron donating ability of the thiazole moiety, restricting the ICT, and thus causes the observed spectral changes.

A donor (D)-π-acceptor (A) system 3 has been developed as a chemosensor for Ni²⁺.⁴ The D–A dyad 3 exhibits ICT, supported by DFT calculations. The comparison of the electron distribution in the FMOs revealed that the HOMO–LUMO excitation displaced the electron density from the donor unit 4-[bis(pyridin-2-ylmethyl)amino] to the isophrone moiety (Fig. 4a). The observed hypsochromic shift in the absorption band of 3 in the presence of Ni²⁺ has been attributed to the formation of a complex via N atoms of the donor unit and thereby restricting the ICT (Fig. 4b).

Fig. 4 (a) Electron density distribution of the HOMO and LUMO energy levels of 3 and (b) the energy minimised structure of the 3 : Ni²⁺ complex [reproduced from ref. 4 by permission of the publisher (Taylor and Francis Ltd.) http://www.tandf.co.uk/journals].

Recognition of the guests by a D–A type dyad is generally manifested by either a hypochromic or bathochromic shift in the ICT band arising from an electronic transition between the appropriate HOMO and LUMO, or no shift is observed at all depending upon the participation of the coordination sites. If the guest coordinates to the donor site, it would lead to a hypsochromic shift, whereas binding to the acceptor site facilitates a bathochromic shift. A weak interaction with either of these two may not shift the ICT band at all and show only a hypochromic effect (reduction in the intensity). In line with this, very convincingly, Martinez-Manez et al.⁵ have evaluated the cation binding behaviour of 4–9 possessing the ICT. The DFT calculations performed on the model compound 4 predicted the main absorption band due to electron transfer from the HOMO orbitals of the N,N-dimethylaniline donor unit to the LUMO orbitals of the thiazole unit (Fig. 5). The hypsochromic, bathochromic and hypochromic effects exhibited by the derivatives in the presence of Hg²⁺, Pb²⁺, Fe³⁺ etc. have been attributed to the coordination site preference by the metal ions.

Fig. 5 The electron density distribution of the HOMO and LUMO energy levels of 4 [reproduced with permission from (ref. 5). Copyright (2011) Wiley-VCH].

The perturbation in the absorption spectral pattern upon complexation was well supported by DFT calculations. Thus, the interaction of 4 with Hg²⁺ by the aniline and thiazole units was proposed. When Hg²⁺ was coordinated by the nitrogen atoms of the azo-thiazole moiety, the TD-DFT calculations predicted the intense ICT (responsible for the bathochromic shift), while, the coordination at the aniline centre predicted the hypsochromic shift giving the band at a lower wavelength (Fig. 6).

Fig. 6 Coordination modes and colorimetric behaviour of 4 with Hg²⁺.

Taking advantage of the coordinative perturbations in the ICT, our research group developed a series of hetarylazo derivatives 10–13 as chemosensors for the detection of Zn²⁺, Hg²⁺, Fe³⁺, Cu²⁺, Pb²⁺ and Co²⁺ based upon the visual as well as the absorption spectral changes.^6–9 The “push–pull” chromophore 10 possessing an electron donating N,N-di(β-ethoxycarbonylethyl)aniline group and electron withdrawing thiadiazole moiety manifests its preference for Hg²⁺ and Fe³⁺ over other metal ions in the form of a bathochromically shifted twin absorption band.⁶ The bathochromic shift has been attributed to the facilitated ICT prevailing in 10, suggested to be as a result of the interaction of metal ions with the sulfur atom of the thiadiazole and the nitrogen atom of the azo linkage. The proposed binding mode was well corroborated by DFT calculations. The energy minimized structure of 10 with Hg²⁺ depicted the interaction of Hg²⁺ with the sulfur atoms of the thiadiazole rings and the azo nitrogen atoms of two molecules of 10, resulting in a 2 : 1 stoichiometry of the resulting complex, which was also supported by the predicted 2 : 1 stoichiometry from the Job plot. However with Fe³⁺, the best optimized structure predicted the same binding mode, but with 1 : 1 stoichiometry, as supported by the Job plot (Fig. 7). TD-DFT calculations further established the HOMO−1 → LUMO+1 as the main contributing transition to the bathochromically shifted band in the case of recognition of Hg²⁺ and Fe³⁺ with a small contribution from the HOMO → LUMO transition (corresponding to one with a longer wavelength in the twin band) in the case of Hg²⁺, and HOMO−3 → LUMO (corresponding to the one with a shorter wavelength in the twin band) in the case of Fe³⁺ detection (Fig. 8).

Fig. 7 The energy minimised structures of (a) (10)₂ : Hg²⁺ and (b) 10 : Fe³⁺ complexes.

Fig. 8 Comparative energy level diagrams of the main transitions (absorption) along with the participating FMOs (isovalue = 0.03).

A structurally similar “push–pull” system 11, with a triazole moiety as the electron withdrawing group, also exhibited a bathochromically shifted absorption band in the presence of Cu²⁺.⁷ Like 10, the bathochromically shifted low energy ICT band is suggestive of the binding of Cu²⁺ with the triazole moiety, as supported by the DFT calculations (Fig. 9).

Fig. 9 Energy minimised structure of the f 11 : Cu²⁺ complex.

The energy minimized structure supports the 1 : 1 stoichiometry of the complex, as established by the Job plot. The TD-DFT calculations predict the HOMO−3 → LUMO and HOMO−1 → LUMO transitions as the main contributors to the bathochromically shifted low energy absorption band responsible for the color change in the 11 : Cu²⁺ complex (Fig. 10). Furthermore the receptor 12 behaved differently in the sense that the inclusion of the aza crown moiety induced a change in selectivity when compared to receptor 10.⁸ The significant feature was that the low energy ICT band of 12 exhibited contrasting behaviour in the presence of Hg²⁺ and Pb²⁺.

Fig. 10 The energy level diagram of the main transitions (absorption) observed in the 11 : Cu²⁺ complex, along with the participating FMOS (isovalue 0.03).

In the presence of Hg²⁺, a bathochromic shift in the ICT was noted, whereas in the presence of Pb²⁺, a hypsochromic shift was observed. The shift in the case of Hg²⁺ could be ascribed to the interaction of Hg²⁺ with the 1,3,4-thiadiazole ring and azo nitrogen atom, similar to that found for the receptor 10. On the other hand, the hypsochromic shift in the case of Pb²⁺ suggested the interaction with the lone pair of the nitrogen and oxygen atoms of the aza crown group. The proposed binding modes and the resultant effects on the absorption spectrum were also confirmed by performing DFT calculations on the metal complexes (Fig. 11).

Fig. 11 The energy minimised structures of the (a) 12 : Hg²⁺ and (b) 12 : Pb²⁺ complexes.

The experimental results were in good agreement with the TD-DFT studies. While for 12 : Hg²⁺, the twin absorption band is seen to be contributed to largely by the HOMO → LUMO transition (64%) and to a lesser extent by the HOMO → LUMO + 1 transition (36%), in 12 : Pb²⁺, the HOMO → LUMO transition is the sole contributor to its main ICT band. In comparison to that with free 12, it could be seen that the predominant characteristic band in all three cases, i.e., 12, 12 : Hg²⁺ and 12 : Pb²⁺, originates from the HOMO → LUMO transition, albeit having different energies (Fig. 12). The receptor 13 responded to Cu²⁺, Co²⁺ and Hg²⁺ in terms of exhibiting a hypsochromic shift in its low energy ICT band (Cu²⁺ > Co²⁺ > Hg²⁺) hinting at the coordination of metal ions involving the two pyridines, as well as the aniline nitrogen atoms of the donor site.⁹ The proposed binding mode, as well as the order of hypsochromic shift, were well validated with the help of ab initio calculations. The predicted bond lengths indicate the weak interaction of Hg²⁺ with the ligating atoms in the binding pocket, as compared to Cu²⁺ and Co²⁺. The order of the binding constants determined experimentally was in good agreement with the calculated ΔG values (Fig. 13).

Fig. 12 Comparative energy level diagram of the main transitions along with the participating FMOs (isovalue = 0.030).

Fig. 13 The energy minimised structures of 13, 13 : Cu²⁺, 13 : Co²⁺, and 13 : Hg²⁺, along with the respective ΔG (KJ mol⁻¹) values.

Furthermore, the TD-DFT calculations performed on the 13 : M²⁺ complexes provided an insight into the nature of the transitions corresponding to the main ICT absorption bands obtained experimentally and the data is presented in Table 1. In the case of 13 : Cu²⁺, as evident from the FMOs (Fig. 14a), the electron density for the HOMO, HOMO−3 and HOMO−7 was also dispersed onto Cu²⁺. It was thus proposed that this dispersal could be a cause for the observed hypsochromic shift in the low energy band, and also the band was regarded as a combination of ICT/MLCT (metal to ligand charge transfer). Since a similar dispersal of one of the contributing orbitals, HOMO−1, was also observed in Co²⁺ (Fig. 14b), the hypsochromically shifted band was regarded as a combination of ICT/MLCT, like in the case of Cu²⁺ recognition. However, no such dispersal was predicted in the case of 13 : Hg²⁺ (Fig. 14c), and the band was assigned to be of ICT nature only. This difference in the nature of the main absorption band was attributed to the open shell nature of Cu²⁺ and Co²⁺ in comparison to the close shell nature of Hg²⁺.

Table 1 Assignment of the main absorption transitions on the complexes on the basis of TD-DFT calculations

S. no.	Complex	Hypsochromically shifted band at λ (nm)	Main contributory transitions
1.	13 : Cu^2+	352	HOMO−3 → LUMO(30%), HOMO−11 → LUMO(10%), HOMO−7 → LUMO(9%), HOMO → LUMO(5%)
2.	13 : Co^2+	370	HOMO−3 → LUMO(32%), HOMO−1 → LUMO+1(24%), HOMO−5 → LUMO(11%)
3.	13 : Hg^2+	400	HOMO−1 → LUMO(100%)

---

Fig. 14 The electron density distribution of various energy levels of (a) 13 : Cu²⁺, (b) 13 : Co²⁺ and (c) 13 : Hg²⁺ complexes.

Furthermore, the resulted hypsochromic shift due to the dispersal of the electron density of the HOMOs was supported by the natural bond orbital (NBO) calculations, which indicated the decreased negative charge on the donor part on complexation with the metal ions, unequivocally suggesting the presence of M²⁺ ions in the binding pocket created by the pyridyl and the aniline nitrogens, hence preventing the ICT.

Similarly, in another investigation,¹⁰ we reported the perturbations in the strong ICT exhibited by 14 as a result of the electrophilic substitution reaction of Hg²⁺ resulting in the formation of a covalently linked Hg²⁺ complex. The formation of the product was well corroborated by the DFT calculations predicting the changes in the respective C–C bond lengths expected to be affected by covalent bonding, as shown in Fig. 15. The NBO calculations were performed to confirm the expected change in the electronic density upon the respective unit atoms after complexation.

Fig. 15 The energy minimised structures of (a) 14 and (b) 14 : Hg(ClO₄)₂ [reprinted with permission from ref. 10. Copyright 2010 Elsevier].

X-shaped fully π-conjugated cruciforms where molecules intersect at a central core also exhibit interesting molecular recognition properties. Such properties are engineered through modulation of the FMOs by structural changes leading to spatial separation between the donor and acceptor groups. Miljanic et al.¹¹ explored the influence of electronic effects on the separation of the FMOs and the associated optical changes in a benzobisoxazole based cruciform 15 (Fig. 16).

Fig. 16 A representative example of spatially separated FMOs [reprinted with permission from (ref. 11). Copyright (2011) American Chemical Society].

The hypsochromic and bathochromic shifts in the absorption and emission spectra experienced by the cruciform on protonation, were found to be in good agreement with the preferred stabilisation of the FMOs.

A crown ether based derivative 16 responded to Hg²⁺ by exhibiting visible ratiometric chromogenic behavior (a gradual decreasing absorbance at 354 nm and increasing absorbance at 408 nm).¹² The bathochromic shift has been attributed to the binding of Hg²⁺ with oxygen atoms in the crown ether instead of nitrogen atoms of the azo (–N=N) bonds. A binding mode (Fig. 17), expected to enhance the stability as well as the conjugation of the complex as a consequence of the formation of six rigid five membered rings, which could be responsible for the observed bathochromic shift, has been proposed on the basis of DFT calculations.

Fig. 17 The energy minimised structure of the 16 : Hg²⁺ complex [reprinted with permission from ref. 12. Copyright 2011 Elsevier].

Coumarin based derivatives 17 and 18 have been reported¹³ to exhibit two photon absorption (TPA) activity in the longer wavelength region, 760–860 nm in the presence of Mg²⁺ and Zn²⁺, respectively, which has been attributed to increased charge delocalization on metal binding and was well supported by theoretical calculations. The DFT studies predicted decreased HOMO–LUMO gaps on complexation, facilitating the charge-transfer transition, and the high TPA activity.

2.2 Photo electron transfer (PET) based systems

A coumarin-based compound 19 containing an ionophore with O and N heteroatoms exhibits “turn-on” behaviour in the presence of Hg²⁺ ions through the formation of the complex 20 (19 : Hg²⁺).¹⁴ A PET process operative from the ionophore to the excited chromophore 19 quenches the emission intensity close to zero. The enhancement of the emission intensity in the presence of Hg²⁺ is attributed to the binding of Hg²⁺ to the ionophore and thereby restricts the PET. The proposed interaction of Hg²⁺ with the ionophore was found to be in good agreement with the DFT calculations which predicted the coordination of Hg²⁺, not only with the heteroatoms of the macrocycle, but also with the sulfur atom of the thioureido linker (Fig. 18).

Fig. 18 The energy minimised structure of complex 20 [reproduced with permission from ref. 14. Copyright (2010) Royal Society of Chemistry].

The azadiene 21 has been developed as a chemosensor for Cu²⁺ and Hg²⁺ exhibiting absorption as well as emission based changes.¹⁵ The non-fluorescent nature of 21 despite having strong fluorophores like pyrene, has been attributed to the result of the PET quenching of the excited state of pyrene by the lone pair of electrons on the nitrogen atom of the bridging group.

Involvement of this lone pair of electron in the complexation with the metal ions, imposes restriction on the PET, resulting in enhanced emission intensity. DFT studies supported the formation of a 1 : 1 stoichiometric complex of 21 with Cu²⁺.

The best optimized structure of the 21 : Cu(OTf)₂ complex exhibited hexacoordination around the Cu²⁺ ion (Fig. 19), one site being occupied by an azadiene N atom (d_N–Cu = 2.037), four sites by OTf oxygen atoms (two equatorial d_O–Cu = 2.024 and 2.115 Å, two axial d_O–Cu = 2.534 and 2.619 Å) and a weak contact with one of the carbon atoms of the pyrenyl group (d_C–Cu = 2.350 Å).

Fig. 19 The energy minimised structure of 21 : Cu²⁺ [reprinted with permission from ref. 15. Copyright 2010 Elsevier].

Similar fluorescence enhancement and the bathochromic shift in the absorption peak were exhibited by the pyrene–thiourea based derivative 22 in the presence of Cu²⁺ and Hg²⁺.¹⁶ Through the mechanism of PET quenching as proposed for 21, a geometry 22 : Cu²⁺ (Fig. 20) has been proposed in support of the observed fluorescence “on–off” behaviour (Fig. 21).

Fig. 20 Fluorescence “on–off” behaviour of 22 upon recognition of Cu²⁺.

Fig. 21 The energy minimised structures of (a) 22 and (b) the 22 : Cu²⁺ complex [reprinted with permission from ref. 16. Copyright 2010 Elsevier].

A sugar-aza-crown ether based fluorescent compound 23 recognizes Cu²⁺ and Hg²⁺, selectively.¹⁷ The observed fluorescence intensity enhancement has been attributed to the PET that occurs upon complexation of the nitrogen atoms by the metal ions. The suggested complexation mode was further supported by the DFT studies. The most stable optimized structure obtained for Hg²⁺ indicates the presence of Hg²⁺ in the coordination centre of 23, surrounded by two nitrogen and two oxygen atoms from the linker and the ribosyl unit, respectively, with the average bond lengths of 2.52 Å (Hg–N) and 2.53 Å (Hg–O) (Fig. 22).

Fig. 22 Energy minimised structure of 23 : Hg²⁺ [reprinted with permission from ref. 17. Copyright 2011 Elsevier].

A benzoimidazole based derivative 24 was found to be selective for Zn²⁺.¹⁸ The bathochromically shifted absorption and fluorescence enhanced behaviour exhibited by 24 in the presence of Zn²⁺ has been assigned to the change in dihedral angle between the aryl and benzoimidazole plane of 24, influencing PET from the tertiary amine nitrogen atom to the fluorophore in the excited state of 24, on the basis of the DFT calculations. The dihedral angle (30.63°) between the above said two planes, noted in the best optimized structure of free 24 (Fig. 23), has been suggested to cause photo-induced electron as well as photo-induced energy transfer, leading to weak fluorescence. However, the coordination of Zn²⁺ to the bipicolyl amine site reduced the dihedral angle (19.63°) owing to the coplanarity, thus causing a red shift in the absorption and emission bands. As a result of the change in dihedral angle, the PET process gets restricted which results in the observed emission enhancement.

Fig. 23 Energy minimised structure of the 24 : Zn²⁺ complex depicting the dihedral angle [reprinted with permission from ref. 18. Copyright 2012 Elsevier].

An otherwise non-fluorescent rosamine derivative 25 having [15]aneNO₂S₂ as a binding unit exhibits fluorescence enhancement in the presence of Ag⁺ and Hg²⁺.¹⁹ The non-fluorescent nature of 25 has been assigned to the electron transfer from the HOMO of [15]aneNO₂S₂ to the HOMO of the excited fluorophore 25 which is lower in energy (Fig. 24a). However, upon binding with metal ions, the HOMO of [15]aneNO₂S₂ gets more stabilized, thereby restricting the electron transfer to the HOMO of the fluorophore at a higher energy, thus leading to the enhanced fluorescence intensity (Fig. 24b).

Fig. 24 The comparative energy levels of (a) 25 and (b) 25 : Ag⁺, depicting the mechanism of fluorescence enhancement in the 25 : Ag⁺ complex.

Interestingly, the fluorescence intensity enhancement effect of both the metal ions was entirely opposite in acetonitrile and water. This contrasting behaviour was also well supported by the DFT calculations. In acetonitrile, the HOMO energy gap between [15]aneNO₂S₂-Ag⁺ and the fluorophore was predicted to be much less (−0.48 eV) compared to [15]aneNO₂S₂-Hg²⁺ (−0.71 eV). Thus, the extensive PET inhibition by Hg²⁺ induces much more fluorescence enhancement than Ag⁺. On the other hand, in water, the HOMO energy gap between [15]aneNO₂S₂-Ag⁺ and the fluorophore (−0.43 eV) was predicted to be much more than that between [15]aneNO₂S₂-Hg²⁺ and the fluorophore (−0.01 eV). Thus, Ag⁺ could induce more fluorescence enhancement in water (Fig. 25).

Fig. 25 Comparative energy levels of 25, 25 : Ag⁺ and 25 : Hg²⁺ complexes in acetonitrile and water.

2.3 Binding mode based systems

Imidazole-annelated ferrocene derivatives 26 and 27 have been developed as multichannel chemical probes for Pb²⁺ ions.^20a The sensing event is accomplished via chromogenic (bathochromically shifted absorption band), fluorescent (chelation enhanced fluorescence, CHEF) and redox (anodically shifted redox peak) changes, in both 26 and 27 in the presence of Pb²⁺ ions. The derivatives 28 and 29 also exhibited a similar behaviour. The associated color changes are shown in Fig. 26. The reversibility of the sensing event upon the addition of ethylenediamine suggested binding of the metal ion to the receptor through weak interactions. This hypothesis was, in fact, supported by Truhlar's hybrid meta functional MPW1B 95,^20b–d DFT calculations, which are useful for studying systems with non-covalent interactions.

Fig. 26 Color changes observed for 26–29 in the presence of Pb²⁺ (top) absorption and (bottom) emission [reproduced with permission from (ref. 20a). Copyright (2009) American Chemical Society].

The calculations predicted binding of both the pyridine as well as imidazole N atoms of 26 with Pb²⁺ (d_Pb–N = 2.565 Å and d_Pb–N = 3.037 Å, respectively) (Fig. 27), leading to the formation of the 26 : Pb(ClO₄)₂(CH₃CN)₂ complex. However, the receptor 27, which lacks an additional pair of nitrogen atoms, could not presumably act as a bidentate ligand and led to a more favourable (27)₂ : Pb²⁺ stoichiometric complex. Furthermore, the significant bathochromic shift observed in the absorption spectrum, as a consequence of binding led to the color change detectable by the naked eye. The proposed binding model was in good agreement with the TD-DFT calculations performed for 26 : Pb²⁺. The main low energy absorption band in 26 is contributed to by a HOMO/HOMO−1 → LUMO transition (Fig. 28).

Fig. 27 Calculated structure for the 26 : Pb(ClO₄)₂(CH₃CN)₂ complex. Only the nitrogen atom of CH₃CN is shown [reprinted with permission from (ref. 20a). Copyright (2009) American Chemical Society].

Fig. 28 Calculated molecular orbital diagram for (a) 26 and (b) the 26 : Pb(ClO₄)₂(CH₃CN)₂ complex [reprinted with permission from (ref. 20a). Copyright (2009) American Chemical Society].

Although, both the HOMO and LUMO of the free 26 gets stabilized upon complexation with Pb²⁺, the LUMO experiences a significant stabilization with respect to the HOMOs, which is responsible for the observed bathochromic shift.

A ferrocene-azaquinoxaline based dyad 30 detects Hg²⁺, Pb²⁺ and Zn²⁺ ions via three different channels exhibiting identical sensing properties towards each of the three metal ions as deduced from the anodically shifted redox peaks, as well as high fluorescence enhancement and visible colorimetric change.²¹ However, a more pronounced color change of the ligand was observed in the case of complexation with Hg²⁺ ions. This has been well supported by the binding behaviour of 30 towards different metal ions predicted on the basis of DFT calculations (Fig. 29). In the case of Zn²⁺, the most stable complex stoichiometry (30)₂ : Zn(ClO₄)₂ indicates ^4,5N coordination in one of the ligands (d_N–Zn⁵ = 2.198 Å, d_N–Zn⁴ = 2.414 Å, but an almost exclusive ⁵N binding mode for the other ligands (d_N–Zn⁵ = 2.127 Å, d_N–Zn⁴ = 2.877 Å), presumably due to steric crowding around the relatively small (Zn²⁺) metal ion, thus indicating the presence of two different heterocyclic ligands (Fig. 29a). In contrast to this, the larger Pb²⁺ cation in (30)₂ : Pb(ClO₄)₂ allows approximation of both the ligands exhibiting the ^4,5N binding mechanism (d_N–Pb⁵ = 2.546 and 2.864 Å, d_N–Pb⁴ = 2.769 and 2.747 Å), thus indicating the presence of two very similar heterocyclic ligands (Fig. 29b). In the case of Hg²⁺, the structure of the most stable complex is 30 : Hg(ClO₄)₂ exhibiting typical N–Hg bond distances (d_N–Hg⁵ = 2.491 Å, d_N–Hg⁴ = 2.518 Å) (Fig. 29c). However, when the terminal ¹N atom in the 1 : 1 optimised structure was also made to involve binding with an identical molecule, formation of a 2 : 2 stoichiometric (30)₂ : ((Hg(ClO₄)₂)₂ as a stable complex was revealed (Fig. 29d).

Fig. 29 Calculated structures for (a) (30)₂ : Zn(ClO₄)₂, (b) (30)₂ : Pb(ClO₄)₂, (c) 30 : Hg(ClO₄)₂ and (d) (30)₂ : [Hg(ClO₄)₂]₂ [reprinted with permission from (ref. 21). Copyright (2009) American Chemical Society].

A prominent feature of this binding model is the interaction of the internal Hg²⁺ cation with the adjacent ferrocenyl iron atom, which was not observed in the complexes of the other two metal ions, and thus this additional binding mode has been held energetically responsible for the pronounced color change. The predicted MO level diagram for 30 shows that the low energy optical transition should involve the HOMO → LUMO transition, the HOMO being located on ferrocene and the π*-type LUMO on the heterocyclic moieties and that this energy difference decreases subsequent to complexation. This decrease has been predicted to be more pronounced for the 2 : 2 complex (Table 2).

Table 2 Predicted optical transitions for 30 and its complexes with Zn²⁺, Pb²⁺ and Hg²⁺

S. no.	Compound	Transition	λmax
1.	30	HOMO–LUMO π*	399 nm
2.	(30)2 : Zn(ClO4)2	HOMO–LUMO π1*, π2*	523 nm
3.	(30)2 : Pb(ClO4)2	HOMO–LUMO π1*/π2*, π2*/π1*	460 nm
4.	30 : Hg(ClO4)2	HOMO–LUMO π*	499 nm
5.	(30)2 : [Hg(ClO4)2]2	HOMO–LUMO π1*	568 nm

---

A series of pyrene derivatives 31–33 was synthesized and evaluated for their behaviour towards the recognition of cations.²² Out of these, only 31 could yield significant results.

The molecule exhibited a strong static excimer emission in the presence of Cu²⁺ indicating the formation of a dimer of 31. The recognition of Cu²⁺ ion led to self-assembled pyrenyl excimer formation which has additionally been rationalized by the DFT calculations. The best optimized structures of dimeric 31–33 suggested the contribution of hydrogen bonding in the self-assembling of all the three pyrene derivatives. The binding energies have been calculated as 12.0, 19.4 and 22.4 kcal mol⁻¹ for 31, 32 and 33, respectively (Fig. 30). Due to the high binding energies, sometimes it is thermodynamically less favorable for a guest species to disturb the stable electrostatic interactions between the two monomer units. In line with this hypothesis, only 31 having the lowest binding energy of the three, led to the recognition of Cu²⁺. Furthermore, the lowest energy structure of the 31 : Cu²⁺ complex (Fig. 31) indicated the participation of the amide oxygen next to the pyrene in the recognition event which was also supported by natural population analysis (NPA) suggestive of the distinct electron density change in oxygen upon Cu²⁺ ion complexation. The strong fluorescence excimer band for the 31 : Cu²⁺ complex was rationalized by TD-DFT calculations suggesting the HOMO to LUMO excitations from one pyrene to the other facing pyrene (Py–Py* interaction) responsible for the observed behaviour.

Fig. 30 Global minima structures for the dimers of (a) 31, (b) 32 and (c) 33. Hydrogen bonds are shown as dotted lines [reprinted with permission from (ref. 22). Copyright (2009) American Chemical Society].

Fig. 31 (a) Lowest energy structure for 31 : Cu²⁺, (b) HOMO of 31 : Cu²⁺ and (c) LUMO of 31 : Cu²⁺ complex [reprinted with permission from (ref. 22). Copyright (2009) American Chemical Society].

The coumarin based fluorescent compound 34 exhibits a “turn-off” response upon recognition of Cu²⁺.²³ The proposed complexation mode of 34 with Cu²⁺ (Fig. 32) was well supported by DFT calculations. The energy minimized structure indicates that Cu²⁺ ions occupy the central coordination site of 34 and the whole molecule attains a planar structure with a Cu–N bond length of 2.00 Å, and a Cu–O bond length of 2.03 and 1.95 Å (for Cu–coumarin and Cu–benzoylhydrazine, respectively).

Fig. 32 Formation and energy minimized structure of 34 : Cu(NO₃)₂ [reproduced with permission from ref. 23. Copyright (2011) Royal Society of Chemistry].

The effects of structural modifications on the optical properties of π-conjugated systems have been demonstrated in α-monoacylated, α,α′- and α,β-diacylated dipyrrins (35–37), respectively, which are fluorescence “turn-on” probes for Zn²⁺ ions.²⁴ The observed bathochromic shifts in emission wavelengths in the series 35 to 37, upon interaction with Zn²⁺ were found to be in good agreement with the predicted HOMO–LUMO energy gaps (Fig. 33).

Fig. 33 Comparative energy level diagram of (a) Zn(35)₂, (b) Zn(36)₂ and (c) Zn(37)₂ [reprinted with permission from (ref. 24). Copyright (2013) American Chemical Society].

The triad 38 detects Hg²⁺ via electrochemical, spectral and optical transduction channels.²⁵ The corresponding experimental responses, such as a positive shift in potential and a bathochromic shift of the absorption bands accompanied by a color change detectable by the naked-eye, exhibited by 38 in the presence of Hg²⁺ have been attributed to the grabbing of Hg²⁺ by both side-arms of the receptor by the action of both azadiene N atoms, as predicted by DFT-based quantum chemical calculations (d_N–Hg = 2.201 Å, N–Hg–N 151.5°). The energy minimised structure of the complex 38 : Hg²⁺ exhibits C₂-symmetry, as shown in Fig. 34.

Fig. 34 The energy minimised structure of 38 : Hg²⁺ [reproduced with permission from ref. 25. Copyright (2009) Royal Society of Chemistry].

3. Anion recognition

3.1 Reaction based systems

A spiropyran derivative 39 behaves as a selective and sensitive CN⁻ receptor in aqueous media under UV irradiation utilizing the nucleophilicity of the CN⁻.²⁶ The observed hypsochromic shift in the absorption band of 39 upon cyanide addition is attributed to the addition of CN⁻ to the spirocarbon of the merocyanine (MC) form of 39 (formed upon UV irradiation) producing a spirocyclic-opened species (39 : CN⁻) (Fig. 35). The observed hypsochromic shift is in good agreement with the ab initio calculations, predicting that the lowest singlet electronic transitions of both 39 (MC) and 39 : CN⁻ are mainly contributed by a HOMO–LUMO transition having π → π* character. The predicted lower HOMO–LUMO energy difference (Fig. 35, inset) in 39 (MC) (2.67 eV) as compared to 39 : CN⁻ (3.29 eV), has been assigned as the probable reason for the observed chromogenic response.

Fig. 35 Recognition of CN⁻ through the photo-induced nucleophilic substitution to spiropyran 39 (inset: HOMO–LUMO energy changes).

The receptors bearing salicylaldehyde hydrazone functionality 40 and 41 detect CN⁻ from aqueous solution via fluorogenic and chromogenic transduction channels, respectively, utilizing the nucleophilicity of CN⁻.²⁷ The nucleophilic attack of CN⁻ at the imine functionality of 40 and 41 leads to the formation of 1 : 1 adduct of CN⁻ with both the receptors, although with revived fluorescence in the case of non-fluorescent 40, and a visual color change in 41.

The DFT calculations performed on 40 predicted a dispersed HOMO electron density over the whole molecule which gets partially dispersed from the 4-(N,N-dimethylamino)benzamide fluorophore to the salicylaldehyde hydrazone moiety upon excitation leading to the weak fluorescence. However, in the CN⁻ adduct, HOMO–LUMO excitation completely moves the electron density distribution from phenoxide anion to the fluorophore (Fig. 36) resulting in the decay of the excited state, thus reviving the fluorescence.

Fig. 36 Comparative electron density distribution in the HOMO and LUMO energy levels of 40 and 40 : CN⁻ [reprinted with permission from ref. 27. Copyright 2009 Elsevier].

The cationic boron compounds 42 and 43 detect F⁻ in a mixture of H₂O–CHCl₃.²⁸ The sensing process is accomplished by the formation of zwitterion fluoroborates (Fig. 37), which are characterized by bathochromically shifted absorption band (ICT in nature) in comparison to 42 and 43. The bathochromic shift is attributed to the decreased HOMO–LUMO gap, localized on the mesityl (Mes) and the pyridinium units, respectively.

Fig. 37 Detection of F⁻ by cationic boronic derivatives.

The N-benzamido-bisthiourea derivative 44 undergoes interesting changes in its optical behaviour upon detecting F⁻ ions.²⁹ A concentration dependent hypsochromic shift followed by a bathochromic shift in the absorption band of 44 was observed upon the gradual addition of F⁻ indicating the presence of three different species: 44, 44 : F⁻ and 44 : (F⁻)₂, in equilibrium. Similarly, emission spectral studies exhibit initial fluorescence quenching without any spectral shift followed by the appearance of a new bathochromically shifted emission band. On the basis of these observations, it was proposed that at a lower concentration, the interaction of the F⁻ is through hydrogen-bonding with the bisthiourea moiety (responsible for the hypsochromic shift) corresponding to the formation of 44 : F⁻ species, which is then followed by deprotonation of thiourea NH at higher concentration of F⁻ (responsible for the bathochromically shifted absorption along with emission quenching), corresponding to the formation of 44 : (F⁻)₂ species (Fig. 38).

Fig. 38 Energy minimised structures of 44 : F⁻ and 44 : (F⁻)₂ [reproduced with permission from ref. 29. Copyright (2010) Royal Society of Chemistry].

The DFT calculations predicted for all three species that while the HOMO is localized on the sulfur atom of thiourea and is nonbonding in nature, the LUMO is π* in nature and is localized on the second thiourea group of 44. The contribution to the absorption band of 44 is mainly from the HOMO → LUMO transition (S₁ state). In 44 : F⁻, the thiourea attains a LUMO+1 orbital, thus the hypsochromically shifted band is the result of a HOMO → LUMO+1 transition (S₁ state). In 44 : (F⁻)₂, the energy of HOMO → LUMO (S₁ state), becomes lower, thus accounting for the new bathochromically shifted absorption band (Fig. 39). The formation of this new state is responsible for the new emission band. The ICT band of benzoxadiazole derivative 45, resulting from the ICT from N,N-dimethylamine group acts as a donor to the dicyanovinyl group acting as an acceptor, getting hypsochromically shifted in the presence of CN⁻.³⁰

Fig. 39 Comparative energy levels of 44, 44 : F⁻ and 44 : (F⁻)₂.

This shift has been attributed to the inhibited ICT, as a result of the addition of CN⁻ to the α-position of the dicyanovinyl group reducing its electron accepting ability. The optimized structure of 45 and 45 : CN⁻ indicates a significant difference in the π-conjunction with respect to a change of conjugated bridge (–C=C–) and dihedral angle of 114° between the benzoxadiazole group and dicyano group in 45 to a saturated bridge (–C–C–) with a dihedral angle of 137° in 45 : CN⁻ (Fig. 40). This was further corroborated by the calculated HOMOs and LUMOs of 45 and 45 : CN⁻. From the locations of LUMO in 45 (Fig. 41), it is clear that the ICT takes place through the conjugated bridge between the benzoxadiazole and dicyano groups, whereas in 45 : CN⁻, it is restricted.

Fig. 40 Energy minimised structures of (a) 45 and (b) 45 : CN⁻ [reprinted with permission from (ref. 30). Copyright (2011) American Chemical Society].

Fig. 41 The electron density distribution of the various energy levels of 45 and 45 : CN⁻ [reprinted with permission from (ref. 30). Copyright (2011) American Chemical Society].

The “turn-on” fluorescence chemosensing behaviour of 8-formyl-7-hydroxycoumarin 46 in the presence of CN⁻ was investigated and authenticated to a large extent by TD-DFT studies in the ground and the first excited states of both 46 and its CN⁻ addition product 47.³¹ On the basis of these studies, it was proposed that in the first singlet excited state (S₁), the phenolic proton of 46 transfers to the neighbouring formyl group along with the intramolecular hydrogen bond, whereas 47 observes a proton transfer process in the ground state, and consequently has a similar structure in the first excited state. This accounts for the experimentally observed strong fluorescence during the sensing event. Furthermore, the molecular orbital analysis also predicts an S₀–S₁ transition at a lower energy in 47 as compared to 46. From the shapes of the FMOs involved in 46 and 47, it can be seen that while the HOMOs of both 46 and 47 are located on the coumarin ring, the LUMO orbitals are additionally spread over the 8-formyl substituent also in 46, whereas, in 47 these are located mainly on the coumarin unit (Fig. 42). On the basis of this, it was concluded that the S₁ states of both are ICT states and are operative in reverse directions. Prior to this, similar types of studies were reported for the sensing mechanism of a F⁻ chemosensor, phenyl-1H-anthra(1,2-d)imidazole-6,11-dione.³²

Fig. 42 Electron density distribution in the HOMO and LUMO energy levels of 46 and 47 [reproduced with permission from (ref. 31). Copyright (2010) Wiley-VCH].

A unique anion to cation relay recognition concept has been established in the sensing of F⁻ followed by Cu²⁺ ions (Fig. 43).³³ The proposed fluorescence “turn-on” behaviour of 48 in the presence of F⁻, based upon formation of a specific cyclisation product 49 and “turn-off” behaviour of the in situ formed 49 in the presence of Cu²⁺ via formation of a 1 : 1 complex 50 (Fig 43), has been well supported by DFT calculations. The decrease in dihedral angle in 49 in comparison to 48 suggested the enhanced planarity and subsequent delocalised π bonds which enhanced the fluorescence emission. Also, the identical energy gaps between the HOMOs and LUMOs of 48 and 49 are in good agreement with the observed unperturbed absorption spectral pattern. On the other hand, the best optimised structure of the Cu²⁺ complex 50 of the in situ formed complex 49, suggested the binding of Cu²⁺ to nitrogen atoms of the imino and benzothiazole groups. A LMCT, as predicted by the localisation of HOMO−2 and LUMO in 50, may lead to the observed “turn-off” fluorescent behaviour and bathochromic shift in the absorption spectrum of 3 (Fig. 44).

Fig. 43 Anion to cation relay recognition mechanism.

Fig. 44 HOMO and LUMO energy levels of (a) 48, (b) 49 and (c) 50 [reprinted with permission from (ref. 33). Copyright (2011) American Chemical Society].

3.2 Displacement/elimination reaction based systems

A supramolecular ensemble 51 resulted from the complexation of chromenolate anion with bis-pyridinium calix[4] pyrrole, detects pyrophosphate over other competing anions including phosphate anions.³⁴ The sensing event takes advantage of the fluorescence dye displacement mechanism releasing the fluorescent chromenolate anion and trapping pyrophosphate anion by 51. The supramolecular trapping is facilitated by the combined benefit of electrostatic, anion-π and hydrogen bond interactions.

The pyrene derivative 52 exhibits a hypsochromic shift in its absorption as well as emission spectra upon detecting F⁻ ions over other anions.³⁵ The mechanism involving elimination of the trimethylsilyl substituent by the F⁻ ion has been proposed to be responsible for the sensing event. DFT calculations have suggested that upon elimination of the trimethylsilyl group, the HOMO–LUMO energy difference increases, which is responsible for the observed spectral as well as visual changes (Fig. 45).

Fig. 45 Comparative energy difference between the HOMOs and LUMOs of 52 and 53 [reproduced with permission from ref. 35. Copyright (2011) Royal Society of Chemistry].

4. Neutral analyte recognition

4.1 Reaction/redox reaction based systems

The absorption as well as the emission bands of a diketopyrrolopyrrole 54 undergo hypsochromic shifts in the presence of cysteine.³⁶ The absorption band at 659 nm was suggested to be an ICT band on the basis of the localization of the HOMO and LUMO on the diketopyrrolopyrrole (fluorophore) core and malononitrile moiety, respectively, in the ground state as predicted by DFT calculations. On the other hand, emission transition at 666 nm has been assigned as a conformation transformation process with the excited state LUMO energy level lower than that of the ground state level and the excited state HOMO energy level higher than that of the ground state, thereby decreasing the energy required for the HOMO–LUMO transition in the excited state, as depicted in Fig. 46. However, in 55, the adduct of 54 and cysteine, the magnitude of the ICT transition was predicted to be decreased in the ground state as a result of the increased localization of the ICT transition (S₀ to S₁) on the diketopyrrolopyrrole chromophore. Like 54, the emission transition in 55 has also been assigned as a conformation transformation process, but with the HOMO–LUMO energy difference greater than the free 54 in the excited state. These results are in good agreement with the observed hypsochromic shift in the absorption and emission bands of 54 in the presence of cysteine.

Fig. 46 The predicted energies of the HOMOs and LUMOs of 54 and its cysteine adduct 55. GS-ground state, ES-excited state, CT-conformational transition.

A pyrene based sensor 56 (Fig. 47) exhibits a significant colorimetric change from light yellow to pink as well as enhanced fluorescence in the presence of lysine over a number of amino acids.³⁷ The sensing protocol has been ascribed to the formation of a Schiff base 57. The DFT calculations revealed the formation of a stacked dimer of 57 (Fig. 47b), proposed to be stabilized by water mediated intramolecular hydrogen bonding between C=N, OH and lysine COOH groups, as depicted in Fig. 47c.

Fig. 47 Energy minimised structures of (a) 56, (b) dimeric form of 57 and (c) dimeric form of 57–(H₂O)₂ [reproduced with permission from ref. 37. Copyright (2011) Royal Society of Chemistry].

The hybrid material Z30-SQ 58 obtained by the inclusion of squaraine dye 59 in the supercages of zeolite Y with a SiO₂/Al₂O₃ ratio of 30 (Z30) (Fig. 48), was found to be sensitive towards the presence of volatile propylamine and propylthiol over other volatile organic compounds, like acetone, ethanol, dichloromethane, hexane, acetonitrile, ethyl acetate, ethyl ether and acetic acid.³⁸ The sensing mechanism has been ascribed to the reaction of 59 with thiol and the amine compounds resulting in the spectral and visual changes. Interestingly, the reaction with thiol was partially reversible, whereas with amine it was irreversible, pointing towards a weaker thiol adduct 60 as compared to the amine adduct 61. This was further corroborated by the optimized geometries of these two adducts, predicting the S–C and N–C bond lengths of 1.90 and 1.50 Å, respectively (Fig. 49a). However, when geometry optimization was performed with acetone or propanol, no reaction was found (the distance between the oxygen atom of acetone–propanol and 59 was found to increase to 4 Å) (Fig. 49b).

Fig. 48 Recognition of n-propylamine and n-propylthiol by supercaged squarine 59 [reproduced with permission from ref. 38. Copyright (2011) Royal Society of Chemistry].

Fig. 49 Energy minimised structures of 59 (a) with acetone, (b) with n-propanol, (c) 60 and (d) 61 [reproduced with permission from ref. 38. Copyright (2011) Royal Society of Chemistry].

A dimethylamino-1,4-benzoxazin-2-one based derivative 62 detects cysteine and homocysteine via chemodosimetric action forming the product 63, with enhanced fluorescence intensity and a hypsochromically shifted absorption band.³⁹ The main absorption band of 62, possessing an electron donor amino group and electron withdrawing carbonyl group has been assigned as an ICT with contribution from the HOMO–LUMO transition (3.35 eV). Subsequent to the formation of 63, the ICT gets perturbed leading to the enhanced energy difference between the HOMO and LUMO (3.39 eV) which is responsible for the experimentally observed hypsochromic shift.

The dialdehyde substituted borondipyromethane (BODIPY) dye 64, being highly electron deficient and prone to facile reductions, exhibited pH dependent “on–off” fluorescence and thus could be labelled as a pH sensor.⁴⁰ The electron deficiency was also predicted by DFT calculations performed on both the di-substituted and unsubstituted derivatives. The electron distribution patterns in the HOMO and LUMO states of 64 and 65 predicted that while the HOMO states of both the molecules are superimposable in the BODIPY core, the LUMO states indicate more delocalization towards the 3,5-diformyl groups in 64 (Fig. 50), leading to an electron deficient BODIPY core facilitating its reduction, thus pointing towards the important role of formyl groups in the pH sensitivity of 64.

Fig. 50 Electron density distribution patterns in the HOMO and LUMO states of (a) 64 and (b) 65 [reprinted with permission from (ref. 40). Copyright (2011) American Chemical Society].

4.2 CT based system

The phosphorescent transition metal complexes, equipped with significant Stokes shifts, luminescence in the visible region and long lifetimes have attracted great interest as the basic building blocks for the development of phosphorescence based chemosensors. In this context, the iridium(III) complexes have attained an important place in the family of phosphorescent dyes.⁴¹ Working in this line, a cyclometallated iridium(III) complex 66 was employed for the detection of thiols.⁴² The complex manifested its selectivity towards thiols in the form of a hypsochromic shift in the low energy absorption band and enhanced emission. The DFT calculations have rendered the conversion of the mainly intraligand (ILCT) transition in 66 to the MLCT/LLCT states in its adduct with cysteine, as a primary cause for the observed experimental changes.

In addition to the above examples, we have recently reported^43,44 sensing properties of a quinolone based derivative 67, which detected Hg²⁺, Cu²⁺, Fe³⁺, as well as HSO₄⁻. A unique feature was that it discriminated between cations and anions through binding at two different sites, thus constituting one of the rare examples of receptors of this type. The proposed binding modes on the basis of various experimental studies have been well corroborated by DFT calculations. These calculations were suggestive of the involvement of the quinolone nitrogen in binding to the metal ions, and the response was observed at different energies, leading to the distinctive recognition, whereas trapping of the HSO₄⁻ ion is accomplished through intermolecular hydrogen bonding interactions between the C-4 NH substituent [N–H⋯O (d_H⋯O = 2.21 Å, angle NHO = 154.99°)] and C–5H [d_O⋯H = 2.77 Å, angle CHO = 104.48°], of the quinolone moiety (Fig. 51).

Fig. 51 Energy minimised structures of (a) 67 : Cu²⁺ and (b) 67 : HSO₄⁻ [reproduced with permission from ref. 44. Copyright (2013) Royal Society of Chemistry].

5. Computational methods

The computational methods used in these investigations include: Hartree–Fock (HF),²⁷ Density Functional Theory (DFT),²⁷ DMol3³ approach, atoms-in-molecules (AIM) calculations using the Gaussian 98/03/09 suite of programmes,^2,23,27 MOPAC 2007 software,¹⁵ AIM2000 software,²¹ Materials Studio 4.2 or 4.4 package,^3,4 Turbomole 5.9.0 programme package,¹⁴ Win MOPAC 3.0,²⁶ AMBER programme²² etc. The various functionals used for the calculations include the Becke's three parameterised Lee–Yang–Par (B3LYP) exchange function,²⁷ Becke-Perdew 86 (BP86) function,¹⁴ Perdew–Burke–Ernzerhof (PBE) function of generalised gradient approximation (GGA) level,³ meta-hybrid functional MO6-2X³⁴ and Truhlar's hybrid meta functional mPW1B95.²⁰ For metals, such as Zn, Hg, Pb, Cu, Ag, Cs etc., LanL2DZ,²¹ Stuttgart–Dresden (SDD),¹⁸ def2-SVP,¹⁴ Stuttgart relativistic small core (STRSC)¹⁵ basis sets with effective core potential were employed, whereas for all other atoms, 3-21G,²² 6-31G,²¹ TZVP³¹ basis sets were used with diffuse and polarization functions added in some cases. In the case of the donor atoms (N, O, F), a basis set with added diffuse functions denoted as aug-cc-pVTZ¹² was used. The solvent effects were studied by using either the Cossi or Barone' CPCM (Conductor-like polarisable continuum model) modification of the Tomasi's PCM formalism²¹ or the counterpoise approach or the COSMO method¹⁵ by semiempirical level. Bond orders were characterised by the Wiberg bond index¹⁵ (WBI) and calculated with the Natural Bond Orbital (NBO) population analysis,²⁵ as the sum of squares of the off-diagonal density matrix elements between atoms. Atomic charge calculations were performed using Natural Population analysis (NPA)²² and Natural Bond Orbital (NBO) population analysis.¹⁵ The non-relativistic gauge-including atomic orbital (GIAO)²⁵ approach was employed to obtain the magnetic shielding tensor. The TD-DFT calculations were performed using the same method as for the respective geometry optimisation. In order to search for significant bond critical points (BCPs) around the spatial region where weak host–guest interactions are taking place, Bader's AIM (atoms-in-molecules) methodology¹⁵ was used.

6. Summary and perspective

This review presents a representative compilation of sensing events where the mechanism of the sensing event has been explained by a rational corroboration by theoretical support. A comparable sketch of the sensing profiles of the chemosensing events discussed in the review is presented in Table 3. Theoretical calculations have become indispensable tools to corroborate the experimental data related to the electronic changes that accompany a sensing event for different analytes and explain the fundamental aspects concerning the modulation of different electronic states and energies, which in turn help in the search for more efficient sensors in terms of sensitivity, selectivity as well as improved detection limit, the important features for the success of the action of the chemosensor. Additionally these predictions can not only be helpful in reducing the amount of trial and error in the development of efficient sensors, but can also lead to significant potential savings in time and resources to be consumed in the synthesis of sensors. This has been well demonstrated by L. Wang et al.⁴⁵ who have successfully employed this approach to develop fluorescent sensors for Zn²⁺. Recently, R. Jin,⁴⁶ in an interesting investigation has studied the interactions between diketopyrrole based derivatives and different halides using theoretical calculations and has established that the substituted derivatives can turn out to be efficient sensing materials.

Table 3 Absorption, emission and/or redox changes and limit of detection upon recognition of analyte by the chemosensors reported in this review

Compound	Analyte (medium, pH)	Absorption^a Δλmax (nm)	Color change	Emission change^b	Redox change mΔE1/2 (mV)^c	LOD^d	Ref.
a R = red shift, B = blue shift, LEB = low energy band, HEB = high energy band.b CHEF – chelation enhanced fluorescence, PET – photoinduced electron transfer.c AS = anodic shift.d LOD = limit of detection.e LOD based on emission studies.
1	Cu^2+ (H2O, 4–10)	15 R	Yellow to dark yellow	Quenched	—	0.5 μM	2
2	Hg^2+ (DMSO : H2O, 9 : 1 v/v)	Intensity decreases	Yellow to colorless	Quenched	—	—	3
3	Ni^2+ (CH3CN)	83 B	—	Quenched (inhibited ICT)	—	—	4
4	Hg^2+ (CH3CN)	52 R	Violet to blue	—	—	—	5
	Fe^3+(CH3CN)	12 B	Violet to red magenta	—	—	—
10	Hg^2+ (CH3CN : H2O, 9 : 1, v/v)	51–84 R	Yellow to purple	—	—	—	6
	Fe^3+ (CH3CN : H2O, 9 : 1, v/v)	51–84 R	Yellow to red	—	—	—
11	Cu^2+ (CH3CN : H2O, 9 : 1, v/v)	119 R	Yellow to purple	—	—	1.36 × 10^−5 M	7
12	Hg^2+ (CH3CN : H2O, 9 : 1, v/v)	45–83 R	Orange to purple	—	0.08 AS	5.0 × 10^−6 M	8
	Pb^2+ (CH3CN : H2O, 9 : 1, v/v)	113 B	Orange to light yellow	—	0.05 AS
13	Cu^2+ (CH3CN : H2O, 9 : 1, v/v)	125 B	Orange to light yellow	—	—	9.0 × 10^−6 M	9
	Co^2+ (CH3CN : H2O, 9 : 1, v/v)	107 B	Orange to light green	—	—	10^−5 M
	Hg^2+ (CH3CN : H2O, 9 : 1, v/v)	77 B	Orange to light green	—	—	10^−5 M
14	Hg^2+ (CH3CN : H2O, 1 : 1, v/v, 7.0)	LEB vanishes HEB 34 R	Blue to colorless	—	—	—	10
16	Hg^2+ (H2O, 4.0–9.0)	LEB 54 R	—	—		2.9 × 10^−8 M	12
17	Mg^2+ [C2H5OH (abs.), CH3CN (em.)]	107 R	—	Enhanced	—	—	13
18	Zn^2+ [(C2H5OH (abs.), CH3CN (em.)]	60 R	—	Δλem. = 15 nm, R, enhanced	—	—	13
19	Hg^2+ (H2O, 7.0)	—	—	Enhanced, due to PET	—	—	14
21	Hg^2+ (CH3CN)	LEB at 528 nm and disappearance of HEB at 233–393	Yellowish to deep pink	CHEF = 12	—	2.5 × 10^−6 M	15
	Cu^2+ (CH3CN)			CHEF = 27	—	4.5 × 10^−6 M
22	Hg^2+ (DMSO : H2O, 4 : 1, v/v, 7.8)	LEB 25 R HEB 3 B	Pale yellow to brown	Enhanced	—	10^−4 M (0.09 ppm)^e	16
	Cu^2+ (DMSO : H2O, 4 : 1, v/v, 7.8)	Intensity of absorption peaks enhanced	Pale yellow to green yellow	Enhanced	—	10^−4 (0.10 ppm)^e
23	Cu^2+ (CH3OH)	—	—	Enhanced	—	1.3 × 10^−4 M	17
	Hg^2+ (CH3OH)	—	—	Enhanced	—	1.26 × 10^−5 M
24	Zn^2+ (CH3CN)	36 R	—	Δλem. = 51 nm, R, enhanced	—	—	18
25	Ag^+ (H2O, 7.4)	—	—	Enhanced	—	—	19
	Hg^2+ (H2O, 7.4)	—	—
26	Pb^2+ (CH3CN : H2O)	LEB 44 R HEB 11 R	Colorless to orange	CHEF	150 AS	1.32 × 10^−8 M	20a
27	Pb^2+ (CH3CN : H2O)	LEB 23 R	Pale orange to red	CHEF	120 AS	2.5 × 10^−6 M	20a
28	Pb^2+ (CH3CN : H2O)	LEB 23 R	Colorless to yellow	Δλem. = 39 nm, R, CHEF	180 AS	1.32 × 10^−8 M	20a
29	Pb^2+ (CH3CN : H2O)	LEB 7 B HEB 160 B	—	Δλem. = 10 nm, R, CHEF	110 AS	4.5 × 10^−6 M	20a
30	Zn^2+ (CH3CN)	LEB 42 R	Orange to purple	CHEF = 184	10 AS	10^−6 M	21
	Hg^2+ (H2O)	LEB 112 R	Orange to green	CHEF = 204	90 AS	10^−6 M
	Pb^2+ (CH3CN)	LEB 40 R	Orange to purple	CHEF = 90	10 AS	10^−6 M
31	Cu^2+ (CH3CN)	—	Colorless to orange	Formation of a new excimer band	—	10^−6 M	22
34	Cu^2+ (H2O : DMSO, 9 : 1, v/v, 5.0–8.0)	39 R	—	Quenched	—	0.1 μM	23
38	Hg^2+ (CH3CN)	LEB 120 R HEB 70 R	Orange to deep red	—	210 mV AS	—	25
39	CN^− (H2O : CH3CN, 1 : 1, 9.3)	98 B	Pink to yellow	—	—	1.7 μM	26
40	CN^− (DMSO : H2O, 1 : 1, v/v)	40 R	No change	Δλem = 14 nm, B, enhanced	—	5.6 × 10^−8 M	27
41	CN^− (DMSO : H2O, 1 : 1, v/v)	New band at 525 nm	Pale yellow to dark red	—	—	1.5 × 10^−6 M	27
42	F^− (H2O : CHCl3)	LEB 49 R	Colorless to yellow	—	—	—	28
43	F^− (H2O : CHCl3)	LEB 75 R	Colorless to yellow	—	—	—	28
44	F^− (CH3CN)	8 B followed by 17 R	—	Δλem = 120 nm, R, enhanced	—	—	29
45	CN^− (CH3CN : H2O, 95 : 5, v/v)	LEB 72 B	—	Δλem = 55 nm, B, enhanced	—	1.47 × 10^−6 M	30
51	HP2O7^3− (CH3CN)	—	—	Enhanced	—	2 ppb	34
52	F^− (THF)	LEB 19 B HEB 18 B	Light green to colorless	λem B	—	10^−6	35
54	Cysteine (CH3CN : H2O, 4 : 1, v/v)	44 B	Purple to yellow	Δλem = 126 nm, B	—	—	36
56	Lysine [CH3CN : HEPES buffer (pH 7.4, 0.01 M), 1 : 9, v/v)]	Enhanced broad band at 500 nm	Light yellow to pink	Enhanced	—	—	37
58	Propylamine and propylthio	Vanishes	Bleaching	—	—	—	38
62	Cys/HCys (CH3CN : HEPES 3 : 7,v/v, 7.4)	70 B	Orange to yellow	Enhanced	—	6.8 × 10^−7 M	39
64	pH (acetate buffer soln, over a pH range of 4–9)	Intensity increases with increasing H^+ ion concentration	—	“on–off”	—	—	40
66	Thiol (cysteine) [DMF : HEPES buffer (pH 7.2), 4 : 1, v/v)]	100 B	Yellow to colorless	Enhanced	—	—	42
67	Fe^3+ (CH3OH : H2O, 9 : 1, 7.4)	LEB 0–13 R	—	94% quenched	—	9.24 × 10^−5 M	43
	Cu^2+ (CH3OH : H2O, 9 : 1, 7.4)	LEB 0–13 R	—	69% quenched	—	4.17 × 10^−4 M
	Hg^2+ (CH3OH : H2O, 9 : 1, 7.4)		—	87% quenched	—	2.94 × 10^−4 M
	HSO4^− (CH3OH : H2O, 9 : 1)	LEB 12 R	—	Quenched	—	—	44

---

Acknowledgements

The authors thank CSIR New Delhi for financial assistance (Projects: 01/2687/12-EMR-II) and UGC, New Delhi for the SAP project.

Notes and references

1. A. P. de Siva, H. Q. Nimal Gunaratne, T. Gunnlaugsson, A. J. M. Huxley, C. P. McCoy, J. T. Rademacher and T. E. Rice, Chem. Rev., 1997, 97, 1515–1566.
2. H. S. Jung, P. S. Kwon, J. W. Lee, J. I. 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.
3. J. S. Bae, S.-Y. Gwon, Y.-A. Son and S.-H. Kim, Dyes Pigm., 2009, 83, 324–327.
4. Y.-A. Son, S.-Y. Gwon and S.-H. Kim, Mol. Cryst. Liq. Cryst., 2011, 538, 327–332.
5. T. Abalos, M. Moragues, S. Royo, D. Jimenez, R. Martinez-Manez, J. Soto, F. Sancenon, S. Gil and J. Cano, Eur. J. Inorg. Chem., 2012, 76–84.
6. P. Kaur and D. Sareen, Dyes Pigm., 2011, 88, 296–300.
7. P. Kaur, D. Sareen and K. Singh, Talanta, 2011, 83, 1695–1700.
8. P. Kaur, D. Sareen and K. Singh, Dalton Trans., 2012, 41(29), 8767–8769.
9. P. Kaur, D. Sareen and K. Singh, Anal. Chim. Acta, 2013, 778, 79–86.
10. P. Kaur, S. Kaur, Y. Kasetti, P. V. Bharatam and K. Singh, Talanta, 2010, 83, 644–650.
11. J. Lim, T. A. Albright, B. R. Martin and O. S. Miljanic, J. Org. Chem., 2011, 76, 10207–10219.
12. Z. Yan, L. Hu, L. Nie and H. Lv, Spectrochim. Acta, Part A, 2011, 79, 661–665.
13. D. Ray, A. Nag, A. Jana, D. Goswami and P. K. Bharadwaj, Inorg. Chim. Acta, 2010, 363, 2824–2832.
14. S. Voutsadaki, G. K. Tsikalas, E. Klontzas, G. E. Froudakis and H. E. Katerinopoulos, Chem. Commun., 2010, 46, 3292–3294.
15. R. Martinez, A. Espinosa, A. Tarraga and P. Molina, Tetrahedron, 2010, 66, 3662–3667.
16. W.-C. Lin, C.-Y. Wu, Z.-H. Liu, C.-Y. Lin and Y.-P. Yen, Talanta, 2010, 81, 1209–1215.
17. Y.-C. Hsieh, J.-L. Chir, S.-T. Yang, S.-J. Chen, C.-H. Hu and A.-T. Wu, Carbohydr. Res., 2011, 346, 978–981.
18. X. Wang, Z. Liu, F. Qian and W. He, Inorg. Chem. Commun., 2012, 15, 176–179.
19. M. Hu, J. Fan, J. Cao, K. Song, H. Zhang, S. Sun and X. Peng, Analyst, 2012, 137, 2107–2111.
20. (a) F. Zapata, A. Caballero, A. Espinosa, A. Tarraga and P. Molina, J. Org. Chem., 2009, 74, 4787–4796; (b) Y. Zhao and D. G. Truhlar, J. Phys. Chem. A, 2004, 108, 6908–6918; (c) Y. Zhao and D. G. Truhlar, J. Phys. Chem. A, 2005, 109, 5656–5667; (d) J. Muniz, L. E. Sansores, A. Martinez and R. Salcedo, THEOCHEM, 2007, 820, 141–147.
21. F. Zapata, A. Caballero, A. Espinosa, A. Tarraga and P. Molina, Inorg. Chem., 2009, 48, 11566–11576.
22. H. S. Jung, M. Park, D. Y. Han, E. Kim, C. Lee, S. Ham and J. S. Kim, Org. Lett., 2009, 11, 3378–3381.
23. L. Huang, J. Cheng, K. Xie, P. Xi, F. Hou, Z. Li, G. Xie, Y. Shi, H. Liu, D. Bai and Z. Zeng, Dalton Trans., 2011, 40, 10815–10817.
24. Y. Ding, X. Li, T. Li, W. Zhu and Y. Xie, J. Org. Chem., 2013, 78, 5328–5338.
25. T. Romero, A. Caballero, A. Espinosa, A. Tarraga and P. Molina, Dalton Trans., 2009, 2121–2129.
26. Y. Shiraishi, K. Adachi, M. Itoh and T. Hirai, Org. Lett., 2009, 11, 3482–3485.
27. Y. Sun, Y. Liu and W. Guo, Sens. Actuators, B, 2009, 143, 171–176.
28. C. R. Wade and F. P. Gabbai, Dalton Trans., 2009, 9169–9175.
29. R. Koteeswari, P. Ashokkumar, V. T. Ramakrishnan, E. J. Padma Malar and P. Ramamurthy, Chem. Commun., 2010, 46, 3268–3270.
30. Z. Liu, X. Wang, Z. Yang and W. He, J. Org. Chem., 2011, 76, 10286–10290.
31. G.-Y. Li, G.-J. Zhao, K.-L. Han and G.-Z. He, J. Comput. Chem., 2011, 32, 668–674.
32. G.-Y. Li, G.-J. Zhao, Y.-H. Liu, K.-L. Han and G.-Z. He, J. Comput. Chem., 2010, 31, 1759–1765.
33. Y. Peng, Y.-M. Dong, M. Dong and Y.-W. Wang, J. Org. Chem., 2011, 77, 9072–9080.
34. P. Sokkalingam, D. S. Kim, H. Hwang, J. L. Sessler and C.-H. Lee, Chem. Sci., 2012, 3, 1819–1824.
35. H. Lu, Q. Wang, Z. Li, G. Lai, J. Jiang and Z. Shen, Org. Biomol. Chem., 2011, 9, 4558–4562.
36. L. Deng, W. Wu, H. Guo, J. Zhao, S. Ji, X. Zhang, X. Yuan and C. Zhang, J. Org. Chem., 2011, 76, 9294–9304.
37. Y. Zhou, J. Won, J. Y. Lee and J. Yoon, Chem. Commun., 2011, 47, 1997–1999.
38. J. V. Ros-Lis, R. Martinez-Manez, J. Soto, L. A. Villaescusa and K. Rurack, J. Mater. Chem., 2011, 21, 5004–5010.
39. M. Hu, J. Fan, H. Li, K. Song, S. Wang, G. Cheng and X. Peng, Org. Biomol. Chem., 2011, 9, 980–983.
40. S. Madhu, M. R. Rao, M. S. Shaikh and M. Ravikanth, Inorg. Chem., 2011, 50, 4392–4400.
41. (a) K. M. C. Wong and V. W. W. Yam, Coord. Chem. Rev., 2007, 251, 2477–2488; (b) D. L. Ma, W. L. Wong, W. H. Chung, F. Y. Chan, P. K. So, T. S. Lai, Z. Y. Zhou, Y. C. Leung and K. Y. Wong, Angew. Chem., Int. Ed., 2008, 47, 3735–3739; (c) K. Huang, H. Yang, Z. Zhou, H. Chen, F. Li, T. Yi and C. Huang, Inorg. Chim. Acta, 2009, 362, 2577–2580.
42. N. Zhao, Y.-H. Wu, L.-X. Shi, Q.-P. Lin and Z.-N. Chen, Dalton Trans., 2010, 39, 8288–8295.
43. P. Kaur, H. Kaur and K. Singh, RSC Adv., 2013, 3, 64–67.
44. P. Kaur, H. Kaur and K. Singh, Analyst, 2013, 138, 425–428.
45. (a) G. A. Hudson, L. Chang, J. Yu, Y. Yan, D. J. Dyer, M. E. McCarroll and L. Wang, J. Phys. Chem. B, 2010, 114, 870–876; (b) M. E. McCarroll, Y. Shi, S. Harris, S. Puli, I. Kimaru, R. Xu, L. Wang and D. Dyer, J. Phys. Chem. B, 2006, 110, 22991–22994.
46. R. Jin, Theor. Chem. Acc., 2012, 131, 1260.

---