**1. Introduction**

The use of coordination compounds, organometallic and inorganic species containing lanthanides (Ln) as luminescent materials is mainly due to electronic transitions inside f-shells [1–6]. The design of luminescent sensors based on lanthanides is currently catching up and the number of papers on emitting materials, in particular Organic Light-Emitting Diode (OLED), is increasing. Several lanthanide complexes and Metal Organic Frameworks (MOFs) were described as useful for ratiometric fluorescent sensing [2] or potentially employable as functional materials [5]. These compounds are often characterized by switchable and tunable properties allowing fine-tuned optical features and sensitive responses to small molecules and ions. Several reviews [7–12] have reported lanthanide-containing luminescent sensor materials, which can be used for detecting anions [9] or low molecular weight analytes [8], and also for detecting cations [12]. The rapid growth in the number of publications requires a systematization that could help in the choice of the right compounds for new devices. Here we report a general overview of the principles governing the lanthanide coordination compounds luminescence and the most important examples of Ln compounds for sensing both ions and low molecular weight compounds. We have decided to report here a classification of the species based on the type of *sensory signal being recorded*, in detail "turn off", "turn on" and "ratiometric" sensors.

## **2. Lanthanide complexes luminescence principles**

To understand how to control a sensory response by using a lanthanide complex as sensor material, it is necessary to consider the main processes occurring when Ln3+ compounds exhibit luminescence. In the absence of organic ligand environments, absorption and emission are due to f-f transitions in the electron shells of lanthanides which, at least theoretically, are possible for all ions on going from Ce3+ (4f15d1 ) to Yb3+ (4f13). At the same time, the structure of the levels first becomes progressively more complicated on going from Ce3+ to Eu3+, and then gradually becomes simpler on going from Tb3+ to Yb3+ ion. It is relevant that symmetric ions with respect to Gd3+, located in the center of the Ln series, have some similarity both in the structure of electronic levels and in luminescence features, such as the lifetime of the excited state (**Figure 1**).

The largest energy gap between the excited and ground levels is observed for the Gd3+ ion and corresponds to UV emission. Among those emitting in the visible region (**Table 1**), the most efficient phosphors are Tb3+ and Eu3+ ions; Sm3+ and Dy3+ ions are

#### **Figure 1.**

*Free ion energy levels of the trivalent lanthanide ions from Ce3+ (4f1 ) to Yb3+ (4f13). Levels are labeled by term symbols or, for some higher levels, capital letters. Reprinted from Dieke and Crosswhite, 39 copyright 1963, with permission from the Optical Society of America [13].*


*a λEM – emission wavelenght, τH2O and τD2O – observed excitation state lifetime of Ln3+ solvated with H2O or D2O respectively.*

#### **Table 1.**

*Ions emitting in the visible region.<sup>a</sup>*

somewhat lower emitters. Among the IR-emitting ions, the most effective is Yb3+, as well as Nd3+ and Er3+. Ce3+ luminescence is possible due to d-f electronic transitions. Other ions (Pr3+, Pm3+, Ho3+, Tm3+) can exhibit emission in the visible and/or IR regions, however, as a rule, their intensity is low and difficult to detect with a common equipment. These considerations explain why europium and terbium

compounds are used in most of the productions of lanthanide-based luminescent sensors.

With the exception of the d-f transitions for Ce3+ and Tb3+, the electronic transitions of lanthanides are characterized by low extinction coefficients (**Table 2**). The extinction coefficients are significantly lower not only with respect to those found for organic molecules but also with respect to d-cations which cause a low luminescence brightness of most inorganic REE compounds. In 1942, Weissman proposed a solution to this problem [16]. In chelate complexes containing organic ligands characterized by conjugated aromatic fragments, absorption and emission are spatially separated due to the "antenna" effect: the organic molecule effectively absorbs radiation, sequentially passing to singlet and triplet excited states, and then it can transfer



*Coordination Compounds of Lanthanides as Materials for Luminescent Turn Off Sensors DOI: http://dx.doi.org/10.5772/intechopen.109011*

**Table 2.**

*Absorption coefficients of Ln3+ ions corresponding to the most significant transitions [15].*

excitation energy to the Ln3+ ion, which in turn emits a typical narrow-band lanthanide-centered emission [17]. In **Figure 2**, a scheme showing the emission mechanism of the luminescence of lanthanide complexes is reported.

It is worth considering the Jabłoński diagram in **Figure 3** to better understand the principles governing luminescence for sensory materials.

Processes occurring during antenna sensitization of lanthanide cations are schematically presented in the Jabłoński diagram (**Figure 3**). Light absorption (A) occurs mainly due to the chromophore groups of the ligand, and the extinction coefficient of the ligand can exceed that of the lanthanide ion by several orders of magnitude [18]. The coordinated ligand is excited to one of the singlet excited states *S*\*, the transition

#### **Figure 2.**

*The emission process for a Ln3+ luminescent complex. Antenna, spacer, coordination site, and central ion are evidenced.*

#### **Figure 3.**

*Simplified Jabłoński diagram for luminescent Ln3+ complexes. The processes of absorption (A), radiative relaxation (abbreviated as r) – fluorescence (F), phosphorescence (P), and emission of lanthanides (Ln) are shown by straight lines; nonradiative relaxation processes (nr) are indicated by dotted lines; energy transfer (ET), internal conversion (ic), and intercombination conversion (isc) are indicated by dashed-dotted lines. The reverse energy transfer processes are not shown for clarity.*

to the most stable state being defined as internal conversion (*kic nr*). At this stage, the ligand can either emit a quantum of light (ligand fluorescence, rate constant *kF r* ), or non-radiatively relax (*kF nr*), or its excited state *S\** energy can be transferred to the triplet excited state *<sup>3</sup> T\** by inter-combination conversion (*kisc*). This triplet state is relatively long-lived due to the forbiddance of singlet-triplet transitions. The luminescent relaxation of the triplet excited state is called phosphorescence (*kP r* ) and is well observed when the energy of *<sup>3</sup> T\** is lower than the resonant level of lanthanide. This is usually the most common phenomenon for Gd3+ complexes [19]. The triplet excited state life is longer than that of the singlet one. Various relaxation pathways are possible for it: ligand phosphorescence (*kP r* ), nonradiative relaxation (*kP nr*) due to energy transfer into lattice vibrations (multiphonon relaxation), and energy transfer to the lanthanide ion (*k<sup>T</sup> ET*). In addition to this pathway, in rare cases, direct

excitation of the lanthanide ion through the singlet state of the ligand (*kET S* ) is also recorded. The transitions inside the f-shells of lanthanides are quite fast and are accompanied by the luminescence of the latters (*kLn<sup>r</sup>* ). Nonradiative relaxation of the excited states of lanthanides due to phonon lattice vibrations is also possible, which corresponds to the constant *kLnnr*.

Thus, the luminescence efficiency of monometallic lanthanide complexes depends on the rate of many processes, and minor changes in the system can dramatically affect it. To date, some rules have been developed that are necessary for the production of efficient luminescent materials based on REE compounds, primarily Eu3+ and Tb3+, but also Sm3+, Dy3+, IR-emitting Nd3+, Yb3+, and some other REE ions [17, 19–21].

## **3. Classification of sensory materials**

There are several ways to classify sensory materials:


In this manuscript we will consider examples of various materials, classifying them according to the type of recorded response, and evidencing also other relevant features.

From a phenomenological point of view, the most rational classification is on the basis of the type of *sensory signal being recorded*, the most common being "turn on", "turn off" and "ratiometric". These types of sensory response are unevenly represented in the literature (**Figure 4**). The most common type of response, i.e. the "turn off", is characterized by the least selectivity and unfortunately is not optimal for practical use.

It makes sense to distinguish the materials according to their structural features and typology. For example, it is useful to differentiate between soluble mononuclear

#### **Figure 4.**

*Percentage of different types of sensory response in lanthanide-based luminescent sensors.*

complexes, coordination polymers (CP's) and porous Metal–Organic Frameworks (MOF's) [22], or immobilized species on an inorganic support surface. Each of these materials has both advantages and disadvantages, and it is impossible to unambiguously define them as optimal without specifying the conditions for which they are optimal. For example, molecular complexes quickly react with an analyte and are convenient objects for studying the response mechanism; however, their regeneration is complicated, and they contaminate the sample. Complexes immobilized on carriers react more slowly and their preparation is time-consuming. Synthesis of insoluble MOF's is often even more difficult, but they offer the advantages of high chemical stability and increased selectivity due to limited pore and channel permeability.

A third possible classification of materials is according to the type of luminescent centers used:


The luminescence efficiency of Eu3+ and Tb3+ compounds usually turn out to be much higher, moreover, their emission bands lie in the visible region (**Table 1**), where the sensitivity of standard photomultipliers is high. As a result, most of the described sensor materials are built by using Eu3+ and Tb3+ compounds for "turn on", "turn off", and "ratiometric" sensors. The luminescence of IR-emitting ions Nd3+, Er3+, and Yb3+ are more sensitive to C–H bond vibrations, which makes them also promising to produce special sensor materials.

Finally, the fourth classification method allows one to distinguish between materials according to the mechanism of sensory response. The proposed classification is based on the effect that the analyte has on both the structure of the complex and on the processes reflected in the Jabłoński diagram.

The most obvious is the mechanism of action defined "**internal filter**"**effect** (**IFE**), in which the absorption spectrum of the analyte overlaps with the excitation or luminescence band of the sensor material. This mechanism (**Figure 5**) is not characterized by high selectivity and sensitivity but can be easily implemented for many analytes, for example, Fe3+ ions. For obvious reasons, only a "turn off" response can be achieved for materials based on this principle. The IFE does not require chemical interaction of the analyte with the sensor material. The unequivocal evidence of the involvement of the IFE mechanism in the sensory response is the intersection of the spectra.

The second, also relatively simple, response mechanism involves either the coordination of the antenna molecule to the lanthanide complex or the elimination of the antenna molecule and its replacement with a non-antenna analyte molecule. In the first case, a turn on response is observed, and in the second, a turn off. Proof for this

#### **Figure 5.**

*Response mechanism associated with the internal filter effect: a) - the effect of the analyte on the luminescence of the sensor molecule; b) Jabłoński diagram when the analyte absorbs excitation energy; c) Jabłoński diagram for the absorption of the radiation emitted by the analyte.*

mechanism can derive from analytical and spectroscopical data confirming the change in the structure of the complex.

The third possible mechanism, **analyte switching antenna** (**ANA**) is associated with a change in the structure of the ligand, "turning on" or "turning off" the antenna function (**Figure 6a**). In this case, the analyte changes either the absorption efficiency of the ligand (A) (**Figure 6b**), or the energy of the singlet and triplet levels (*S\**, *T\**) (**Figure 6c**), or the sensitization efficiency (*kT ET*) (**Figure 6d**), but the analyte itself should not have absorption in the region of excitation or emission of the sensor and should not have a direct effect (antenna sensitization or quenching) on the Ln3+ ion, otherwise other possible mechanisms (effect of an internal filter, energy transfer, vibrational quenching of luminescence) take place in parallel with this mechanism or dominate it. In a typical case related to this mechanism, the sensor molecule contains an antenna group, a macrocyclic analyte receptor as a spacer, and a lanthanide ion in a suitable environment. The coordination of Mn+ by a macrocyclic receptor leads to a change in the electronic structure of the ligand, although the cation itself (usually an alkali or alkaline earth metal) does not affect the luminescence of other Ln3+ compounds.

Depending on the nature of the substance, both "turn off" and "turn on" responses can be realized. It is not easy to prove the existence of an ANA mechanism; it is necessary to show the formation of a ligand-analyte bond by analytical methods and to detect a change in the efficiency of specific processes using spectroscopic, kinetic, and/or calculation methods.

#### **Figure 6.**

*Response mechanisms associated with an analyte-switched antenna (ANA). An analyte with its o antenna function is inserted into the structure of the complex (a). As a result, we have a change in the absorption efficiency (b), of the energy of the singlet or triplet level (c), and finally of the sensitization efficiency (d).*

#### **Figure 7.**

*Response mechanisms associated with the control of luminescence quenching processes (quencher addition/ elimination, QAE). Suppression of nonradiative relaxation of the singlet, triplet states, and the excited state of lanthanide (a). Controlling vibrational quenching by introducing or removing X-H bonds (b).*

The fourth possible mechanism is the addition or elimination (due to oxidation, substitution, or another process) of the quencher group during the interaction of the sensor with the analyte (quencher addition/elimination, QAE mechanism) (**Figure 7**). Bonds with suitable vibrational energy (OH, NH, CH) can act as quenching groups: in fact, the excited state of Ln3+ or the ligand upon interaction with the analyte can effectively undergo phonon relaxation. This mechanism involves the transformation of the sensor molecule into a new compound, which can be confirmed by analytical

chemistry and spectroscopy. A "turn off" or a "turn on" response is observed depending on whether the groups are attached or split off. When a quenching group (for example, H2O) is replaced by an antenna ligand, two possible processes simultaneously contribute: removal of the quencher and enhancement of sensitization by the new antenna molecule or the reverse process.

The fifth possible mechanism can be observed when the sensor material is a bimetallic system containing two lanthanum ions, usually Tb3+ and Eu3+. Due to the proximity of radii and chemical properties, in these compounds, the two cations, as a rule, occupy the same crystallographic positions and are statistically distributed. If both Ln3+ ions are capable of emission, the excitation energy is transferred between the ions, and the rate constant of this process can be determined from kinetic data [23]. If the analyte affects the energy transfer constant between the ions, then a sensory response appears, which manifests itself as a change in the ratio of the integral luminescence intensities of the lanthanum ions. The proof of this mechanism requires a careful analysis of spectroscopic and kinetic data and cannot be extended to any "ratiometric" sensors, since in many cases, at least in the case of detection of CH, OH, and NH bonds, the response is not associated to a change in the efficiency of sensitization of one lanthanide ion by another, but with more efficient quenching of the luminescence of one of the ions.

Finally, also materials that are potentially usable as thermometric sensors can be based on REE compounds. However, the discussion on this topic is beyond the scope of this review. Lanthanide-containing luminescent thermometers are the subject of special reviews and monographs [24, 25].

The internal nature of the f-shells of lanthanides makes them less susceptible to the effects of the crystal field, however, a change in the geometry of the coordination environment can still be detected in several cases by careful analysis of the luminescence spectra. This principle is applicable in gauge sensors, in which compression of the crystal under pressure leads to a reduction in the Ln-O [26, 27] or Ln-F [28, 29] distances, which shifts the bands in the emission spectrum.

## **4. Turn off sensors**

Recent papers describing luminescent sensor materials have reported clearly that "turn off sensors" dominate other types of response. This can be easily explained by the fact that luminescence quenching due to IFE, nonradiative relaxation of the excited states of the complex, or formation of a not-luminescent complex (static quenching) are caused by a large number of analytes with a different chemical nature [30]. Turn off sensors are poorly usable in real devices since it is difficult to detect signals at a low analyte concentration.

Effective internal filters can be either intensely colored compounds (due to d-d transitions typical of d-cations as Cu2+ or Co2+, or to charge-transfer bands as in hydrolyzed Fe3+ aqua ions, Cr2O7 <sup>2</sup>�, CrO4 <sup>2</sup>�, MnO4 � anions).

The nonradiative relaxation of the excited state of the complex obeys the Stern-Volmer equation:

$$\frac{I\_0}{I} = \frac{PLQY\_0}{PLQY} = \frac{\tau\_0}{\tau} = \mathbf{1} + K\_{SV}^D[Q].$$

where *K<sup>D</sup> SV* – is the dynamic Stern-Volmer constant [31]. The higher the *K<sup>D</sup> SV* value, the higher the sensitivity of the sensor material with respect to the quenching analyte. In the case of dynamic quenching, in addition to the luminescence intensity (and quantum yield), the lifetime of the excited state of Ln3+ is significantly reduced. Dynamic quenching is caused by the excitation energy adsorption from the quencher molecule without the formation of a phosphor-quencher bond.

Static quenching is caused by the formation of a non-luminescent complex formed by the luminophore and quencher molecules. The bond between the two species can be either covalent or hydrogen or stacking interaction. Static quenching is associated with nonradiative deactivation of the singlet or triplet state of the ligand, which weakens the antenna sensitization. As a result, during static quenching, the observed lifetime of the excited state (τobs) of the lanthanide ion can slightly decrease, but to a lesser extent with respect to the dynamic quenching. For static quenching, the Stern-Volmer equation is also applicable:

$$\frac{I\_0}{I} = \mathbf{1} + K\_{SV}^S[Q].$$

In many cases, these two types of quenching complement each other, and by combining the equations for the static and dynamic cases the following generalized equation can be derived:

$$\frac{I\_0}{I} = \mathbf{1} + \left(K\_{SV}^D + K\_{SV}^S\right)\mathbf{Q} + K\_{SV}^D \times K\_{SV}^S \times \mathbf{Q}^2$$

The nonlinear dependence of I0/I on the quencher concentration Q indicates a complex quenching mechanism.

Measurements carried out with a temperature control also make it possible to distinguish between static and dynamic quenching mechanisms, since for a static mechanism, an increase in temperature leads to a decrease in *KSV* and for a dynamic one, to an increase [31].

The luminescence quenching can occur via excitation energy transfer (ET) from the ligand or metal through FRET (Förster Resonance Energy Transfer), DEE (Dexter Electron Exchange), or PET (Photoinduced Electron Transfer) mechanisms [32] (**Figure 8**) which imply the deactivation of the excited state of the molecule by transferring the excitation energy to the quencher molecule, whose LUMO is lower in energy with respect to the excited electron of the donor molecule.

#### **Figure 8.**

*Schematic representation of the mechanisms of deactivation of excited states. Left - Förster mechanism (FRET), center – Dexter mechanism (DEE), right – photoinduced electron transfer (PET) mechanism.*

FRET [33] implies nonradiative energy transfer from an excited donor molecule to an unexcited acceptor molecule, followed by radiative or nonradiative relaxation of the latter. The typical distances between the donor and acceptor molecules usually are 20–60 Å, and the efficiency of the Förster transfer decreases very rapidly with increasing distance (efficiency is proportional to r<sup>6</sup> ), which makes it possible to measure the distance between particles. A transfer is described as the emission and absorption of a virtual photon [34].

DEE [35] includes two electronic transitions: an excited electron from the donor molecule to the LUMO of the acceptor molecule, and an electron from the HOMO of the donor to the HOMO of the acceptor. This creates a hole in the HOMO of the donor molecule [36]. These two processes proceed either simultaneously or sequentially. This mechanism is observed at much shorter distances (less than 10 Å) and is associated with the need for direct overlapping of the wave functions of the fluorophore and quencher. The efficiency of DEE quenching also decreases very rapidly with increasing distance (proportional to e<sup>r</sup> ).

Another common mechanism of nonradiative relaxation of an excited state is PET [37] involving the formation of a charge-transfer complex (exciplex). An exciplex is formed when an electron moves from a LUMO donor to a LUMO acceptor, followed by radiative or nonradiative relaxation of the latter. An increase in the distance between the donor and acceptor has a much smaller effect on the efficiency of PET quenching with respect to FRET and DEE mechanisms.

To distinguish between these mechanisms, special quantum chemical methods are used to determine the HOMO and LUMO energies, as well as the dependence of the quenching efficiency on the distance between the donor and acceptor. Often, the real quenching mechanism has not been studied in detail, the PET or FRET quenching being only postulated.

The most common analytes detected with lanthanide-containing "turn off" sensors are listed in **Table 3**. Data on the mechanisms of sensory response are not always reported, moreover, in some cases, the author's interpretation of the mechanism may not be entirely correct. In several cases a static quenching mechanism associated with analyte-sensor bonds due to hydrogen bonds, stacking, and other weak interactions is suggested. Unfortunately, in most works, a more detailed mechanism of such static quenching has not been reported.




#### **Table 3.**

*Proposed sensory response mechanism in turn off sensors employed to detect different analytes.*

It should be noted that the most common mechanism is IFE. Typical excitation spectra of luminescent lanthanide complexes lie in the near UV region in the range 250–400 nm [4, 6, 145]. In this region, intense absorption of several types of analytes takes place, which makes such materials nonselective, although the selectivity has not been investigated in most of the articles.

The most efficient materials for the detection of various analytes are listed in **Table 4**. Materials with high *Ksv* values (>30,000) are selected as illustrative examples, making these materials potentially promising for practical applications.




#### **Table 4.**

*The most sensitive "turn off" touch materials.*

The detection of Fe3+, d-metal ions and also d-metal-containing anions with strong oxidizing abilities can be performed by simpler methods with respect to the production of the luminescent sensors. At the same time, the determination of impurities of toxic metals (lead, mercury) in waters or foods seems to be a much more urgent task, and there is hope that other results on the determination of such analytes will appear soon. Nitroaromatic compounds are found in a large number of explosives and ammunition, and proposed sensor materials often exhibit very high sensitivity to them. At present, there are some commercial luminescent "turn off" detectors of nitroaromatic compounds, and although trade secrets prevent disclosure of the compositions of specific materials, lanthanide-containing complexes seem to be promising in this direction. As for the detection of other organic molecules, due to the prevalence of absorption in the near UV region, the industrial production of these materials requires increased selectivity.

Based on the analysis of literature data, we can propose the following algorithm for establishing the sensory response mechanism for "turn off" sensors:


transition. If the response completely disappears, then only the IFE mechanism is in place.


## **5. Conclusions**

The luminescence of lanthanide coordination compounds is an unusual and multistage process that imposes numerous restrictions on the structure of both ligands and complex compounds. Interruption of any of the intermediate stages due to the *effect of an internal filter*, *nonradiative deactivation* of the excited states of the ligand and metal, *changes in the energy of the triplet level* of the ligand, or other factors dramatically affect the luminescence brightness. This presents a serious challenge in the development of efficient luminescent materials for light-emitting devices, but it can be an elegant key for the development of sensitive sensor materials for the detection of analytes of various natures.

The main problem with *turn off* materials is *low selectivity*, since dissimilar analytes such as nitroaromatic compounds and Fe3+ cations, for example, cause exactly the same response. Now, a new challenge for researchers is to find the way to increase the selectivity of the response, which requires the use of modern approaches of supramolecular chemistry and crystal chemical engineering.

In the following chapter, we will focus on consideration of two less common, but more technologically advanced and promising areas in lanthanide-based luminescent chemical sensors – *turn on* and *ratiometric* materials.

## **Acknowledgements**

University of Camerino is gratefully acknowledged.
