**3. "Ratiometric"sensors**

The third important sensor type is the "ratiometric" one. In materials showing these properties, the signal is estimated as the ratio between luminescence values at two different wavelengths, either affecting the effect of Ln3+ ions or affecting both the Ln3+ and the organic ligand.

$$\text{Sensory Resonse } (S) = \frac{I\_1}{I\_2} = \frac{\int\_{\lambda\_{1(\text{min})}}^{\lambda\_{1(\text{max})}} \text{Id}\lambda}{\int\_{\lambda\_{2(\text{min})}}^{\lambda\_{2(\text{max})}} \text{Id}\lambda},\tag{1}$$

where λ1,2(min/max) -are the initial and final coordinates of the bands in the spectrum involved in the integration.

"Ratiometric" sensors do not have the disadvantage to determine the presence of internal standards as in the case of "turn off" and "turn on" sensors. An unusual uniqueness of ratiometric sensory visibility materials is the naked eye color change of luminescent properties when they are in presence of an analyte.

The possible response mechanisms are similar to the "turn off" and "turn on" signals considered above. In addition, if two different Ln3+ ions are simultaneously present in a compound (**Ln-Ln' sensors**), energy transfer between them is possible, the efficiency of which depends on the Ln-Ln distances and other factors [41]; the analyte can affect the transfer efficiency constant. This type of response is defined as Metal-to-Metal Energy Transfer (MMET).

Ratiometric materials generally use the Eu3+/Tb3+ bimetallic pair, since these ions have the most efficient luminescence. Other lanthanide-based systems were used only in a small number of cases, for example, Ce3+/Tb3+ [42], Dy3+/Eu3+ [43], and Eu3+/Yb3+ [44] systems. To achieve greater accuracy, as a rule, it is better to employ the most intense lines in the emission spectra, for example, those corresponding to the europium <sup>5</sup> D0- 7 F2 (�612 nm) and terbium <sup>5</sup> D4- 7 F5 (�544 nm) transitions, also if this is not entirely correct, because the europium transition <sup>5</sup> D0- 7 F2 is partially superimposed on the low-intensity transition of terbium <sup>5</sup> D4- 7 F5 (�620 nm). The sensor response can be calculated using the europium <sup>5</sup> D0-7F4 (�700 nm) transition, which in many cases is also very intense and does not overlap with any terbium line, but lies in the region of reduced sensitivity of common photomultipliers [45]. The usual molar fractions ratio of lanthanides with a predominance of terbium ions is caused by the transfer of energy between lanthanide ions.

O2 in the triplet state can act as a luminescence quencher [46, 47]. The emission of Tb3+ ions is quenched by O2 molecules more efficiently than by Eu3+ due to the

*Luminescent Materials with* Turn-on *and* Ratiometric *Sensory Response Based… DOI: http://dx.doi.org/10.5772/intechopen.109189*

smaller difference in the energies of **<sup>3</sup> T** level of O2 and the excited state of corresponding Ln3+ ions. This principle is the basis for a sensor for gaseous oxygen, which is a mechanical mixture of terbium and europium complexes immobilized on a quartz surface [48].

The possibility to determine, using "ratiometric sensors", small H2O impurities against the background of organic solvents (for which Karl Fischer titration is usually used), which requires the use of toxic reagents and "capricious" equipment, is of great interest. In addition, "ratiometric" sensors can detect the mixture of light water in and D2O, which is impossible with Fischer titration and requires expensive mass spectrometers or precision IR spectrometers [49, 50]. Such materials are based on MOF structures containing intra-sphere water molecules. The sensor material is activated by heat treatment in vacuum, after which the obtained anhydrous powder is dispersed in the medium of the investigated solvent. This approach is not accidental: polymeric MOFs are insoluble in most organic solvents, which makes them easy to regenerate and reuse. OH quenching is the basis for the detection of methanol in ethanol and in the form of vapors in air [51], and a similar principle can be extended to CH oscillations in an elegant DMSO impurity sensor in deuterated DMSO-d6 [44]. Substitution of an OH group [45] or a water molecule [52] in the coordination sphere of lanthanide with F– ions also suppresses Eu3+ quenching more effectively than Tb3+, which was used in the development of fluoride-sensitive sensors.

Mixed-metal MOFs with a statistical distribution of Ln3+ ions are commonly used as material for "ratiometric" sensors. However, Tscelykh et al proposed in [53] to use solutions of pentafluorobenzoates or even Ln3+ chlorides, since the sensitivity of the sensor is directly dependent on the number of water molecules in the environment of Eu3+. The disadvantages of this approach include the complexity of material regeneration and contamination of the analyzed media, which makes flow analysis impossible and increases its cost.

Moisture sensors are closely related to pH sensor materials. "Ratiometric" pH sensors based on carboxylate MOFs have been also considered [54–56]. In the first two cases, as the pH increases, the color of the luminescence changes from green to red; the relative intensity of Eu3+ emission increases. This is explained by the strong pH dependence of the excitation transfer rate constant from Tb3+ to Eu3+, which was confirmed by measurements of τobs in monometallic and bimetallic complexes [55] (in the analysis of kinetic data, certain caution is required due to the non-monoexponential nature of the Eu3+ decay curve in the presence of luminescence sensitization by Ln ions [41, 50]). It has been described [56] that as the pH increases, the luminescence color, on the contrary, changes from red to green, that is, the relative intensity of Tb3+ luminescence increases. It was shown by the DFT method that the energy of the **<sup>3</sup> T\*** triplet level increases from 24.400 to 26.400 cm<sup>1</sup> upon transition from the protonated to the deprotonated form of the ligand. The resonant level of europium is much lower than both of these values, so the sensitization of europium luminescence (17.200 cm<sup>1</sup> ) does not change its efficiency (τobs(Eu) also changes slightly with pH). The resonant level of terbium (20.500 cm<sup>1</sup> ) lies higher, and an increase in the triplet level energy weakens the back transfer of excitation from Tb3+ to the ligand. This leads to a significant increase in τobs(Tb) and in the intensity of Tb3+ emission.

The MMET mechanism can be further confirmed by measuring the response in a bimetallic complex and in a mechanical mixture of complexes of two REEs. The decrease or disappearance of the response in the second case indicates the partial or complete participation of the MMET mechanism. A similar approach was used [57] where the MMET mechanism is related to the response to hydrosulfide anions.

Interestingly, the other two analyzed analytes (THF and Ag<sup>+</sup> cations) exhibit different response mechanisms.

The highly selective sensor for a potassium ions, as in the case of the "turn-on" material described above [13], contains diaza-18-crown-6 in the structure, which makes it selective to the tested s-metal cations [58]. It has been shown that the capture of the potassium ion by the crown fragment leads to an increase in the triplet level energy from 22.400 cm<sup>1</sup> to 23.400 cm<sup>1</sup> . This effect enhances the emission of Eu3+ to a greater extent than that of Tb3+. In addition, it was found that the efficiency of energy transfer from Tb3+ to a Eu3+ increases with increasing K<sup>+</sup> concentration, which leads to an increase in the contribution of europium bands in the spectrum.

For Ln-Ln' "ratiometric" sensors, in contrast to the "turn off" and "turn on" materials considered above, the response to conjugated organic analytes has been poorly studied. The sensory response of Eu-Tb bimetallic BTC-MOF film to a number of drugs [59] has been also investigated. The nature of the emission of this material changes significantly in the presence of a number of compounds, in particular, in the presence of coumarin and caffeine. The nature of the response to caffeine is not discussed, and for coumarin, an increase in *kLn ET* is assumed, which is confirmed by kinetic measurements, but the corresponding quantum chemical studies are not provided.

In the presence of only one lanthanide luminescent center, the "ratiometric" response can still be realized if the ligand is fluorescent or phosphorescent (Ln-L sensors). This is possible if the antenna sensitization efficiency is low, which can be caused by too large or too small energy gap between **<sup>3</sup> T\*** and the REE resonant level, as well as by a large metal-ligand distance. Possible mechanisms for the occurrence of this type of sensory response include binding of the lanthanide ion caused by decomposition of the parent complex, which weakens the efficiency of antenna sensitization (lanthanide ion ejection, LIE), and direct antenna function of the analyte.

The detection of Hg2+ ions was possible by using a composite material containing a luminescent terbium coordination polymer impregnated with a coumarin solution [60]. The response is selective to Hg2+ ions with respect to a wide range of s-, p-, and d-metal ions and is associated with the displacement of Tb3+ ions from the adenosine monophosphate environment, leading to a decrease in the Tb3+ luminescence intensity and to a weakening of its sensitization by coumarin. The driving force behind the displacement reaction of the lanthanide ion into a complex with a lower luminescence intensity can be not only the strength of the "cationic analyte-ligand" bond, but also the formation of a stable "analyte-lanthanide cation" complex. A similar strategy was implemented for phosphate ions [61] and for alkaline phosphatase [62]. In some cases, the response occurs when the ligand is destroyed, for example, in sensors for the determination of HClO [63] or formaldehyde [64]. The vulnerability of these two strategies is the impossibility of a sensor regeneration.

Radiative relaxation by the PET or FRET process can also lead to a response if the analyte has the appropriate LUMO energy. This approach was implemented in a sensor material for the detection of nitrofuranose and furazolidone against the background of other antibiotics in a dysprosium-containing sensor [65]. Trinitrophenol (TNP) similarly blocks the sensitization of terbium luminescence in specifically designed Tb-MOF via PET and FRET mechanisms [66].

As found for Ln-L sensors, the response manifests itself as a drop in the relative efficiency of lanthanide luminescence. The opposite is also possible if the ligand displaces quencher molecules (SQD) and exhibits antenna properties itself (DAAF). An example of this approach is in a study [67–69] describing the response to dipicolinic acid, an important biological marker associated with antrax disease.


### *Luminescent Materials with* Turn-on *and* Ratiometric *Sensory Response Based… DOI: http://dx.doi.org/10.5772/intechopen.109189*



### *Luminescent Materials with* Turn-on *and* Ratiometric *Sensory Response Based… DOI: http://dx.doi.org/10.5772/intechopen.109189*


 *sensors.*

**Table 3.** *"Ratiometric"*

*Luminescent Materials with* Turn-on *and* Ratiometric *Sensory Response Based… DOI: http://dx.doi.org/10.5772/intechopen.109189*

Bisphenol-A, an important industrial reagent in the production of plastics, can act as an effective antenna ligand for europium cations. This underlies the importance of the production of a highly sensitive sensor material based on a composite of carbon quantum dots and europium 5'-adenosine monophosphate [70].

Finally, a response can also be caused by an internal filter effect (IFE). This effect can be the main mechanism, as in the sensor for tetracyclines [71], or manifest itself simultaneously with quenching through ET mechanisms in sensors for Fe3+ [72] or for the aforementioned nitrofuranose and furazolidone (**Table 3**) [65].

As in the previous cases, the key stages in studying the mechanism of the "ratiometric" sensory response are the chemical analysis of the reaction product between the sensor and analyte; comparison of the excitation and emission spectra of the sensor and the absorption of the analyte; kinetic studies of excited state lifetimes and quantum chemical modeling. In general, the"ratiometric" response mechanisms usually coincide with those for"turn off" and"turn on" systems. An exception is the mechanism associated with changing the MMET constant (kLn ET). This mechanism must be reliably established by kinetic measurements, as well as by studying the response in a mechanical mixture of complexes of two lanthanides.

## **4. Conclusions**

The classification of sensory response mechanisms reported in this review, although perhaps not complete, allows us to successfully classify most of the cited works and make appropriate generalizations, moving from a descriptive style to a debatable one.

The reported examples show the significant progress in the field of lanthanidebased luminescent sensors achieved in recent years. A wide variety of analytes can be qualitatively or quantitatively determined using suitable lanthanide compounds, which requires rational system design. "Turn off" sensors can have a niche application in the analysis of nitroaromatic compounds, while more popular "Turn on" and "ratiometric" materials can be produced using fairly simple strategies: different quenching efficiency of various REE ions by bond vibrations, binding and removal of a quenching fragment by a suitable analyte (e.g., binding Cu2+ ions with sulfurcontaining ligands), etc.

Progress has touched not only the field of materials design, but also a reliable determination of the sensory response mechanism, which requires several spectroscopic and kinetic measurements, and in some cases quantum chemical calculations of orbital energies. The number of papers containing these studies has been increasing in recent years, and we hope that just such a systematic approach will become the standard in the future works.

## **Acknowledgements**

University of Camerino is gratefully acknowledged.

## **Conflict of interest**

The authors declare no conflict of interest.
