Dual emission and its λ-ratiometric detection in analytical fluorimetry. Pt. I. Basic mechanisms of generating the reporter signal

The wavelength-ratiometric techniques gain increasing popularity in fluorescence probing and sensing for providing inner reference to output signal and removing instrumental artefacts, in this way increasing the sensitivity and reliability of assays. Recent developments demonstrate that such approach can allow achieving much more, with the application of broad range of novel molecular and nanoscale fluorophores (luminophores), exploring the whole power of photophysical and photochemical effects and using extended range of assay formats. Simplicity of detection and potentially rich content of output data allows realizing these techniques in different simplified, miniaturized and multiplexing devices. The latter issues are discussed in Pt. II of these series.


Introduction. Why we need wavelengthratiometric sensing and imaging?
The simplest way to explore analytical techniques based on fluorescence is just to record the analyteinduced changes of fluorescence intensity at a single wavelength. The light intensity is detected when it increases (OFF-ON) or decreases (ON-OFF) as a function of analyte concentration (figure 1). Such analytical signal based on quenching-dequenching of fluorescent probe can be easily achieved in different ways, which makes the assays very popular [1]. However, fluorescence intensity commonly presented in relative units proportional to the number of emitted quanta does not have absolute meaning. Therefore, for obtaining any quantitative result, the output signal must always be calibrated as a function of analyte concentration, and if the calibration is made, it can be applied only to particular measurement in particular conditions [2]. The sources of errors and artifacts can be of different origin: -Originating from the probe properties, such as photobleaching, chemical degradation or uncontrolled static quenching by some impurity. Resulting fluctuation or even degradation of fluorescence signal is hard to control.
-Originating from errors in applied probe concentration (on which the response depends directly) and its distribution in the applied medium. For instance, in cellular research, some dyes can be absorbed by specific organelles and some detracted from them, making impossible to trace the distribution of analyte.
-Originating from the measurement system. The results depend on the instrument, its optical configuration and performance. There can be fluctuations in the intensity of excitation light source and sensitivity of detector making unreliable the measurements of output fluorescence intensity [2].
The wavelength ratiometry (see figure 1) allows removing the important problems existing in singlechannel recording of fluorescence [3][4][5] and offers obtaining much more reliable quantitative data [6]. The analytical signal here is formed of two or more intensities at different wavelengths, in which at least one changes in proportion to analyte concentration. The most efficient is the design when the dual emission comes from a single fluorophore. The internally calibrated dual emission can also be obtained from two or more simultaneously excited independent or interacting emitters. All that allows recording the information on probe-analyte interaction as the λratiometric signal.
Due to the presence of internal calibration, these methods respond to analyte concentration in a wellcontrollable manner, demonstrating insensitivity to many artefacts. The undesired probe modification or degradation changing the fluorescence can be accounted or even eliminated simply because in different assays the fluorescence at two wavelengths decreases/ increases in proportion, and therefore the output information may not be distorted. Importantly, within the determined range for providing analysis, the dependence of fluorescence response on concentration of probe itself may be eliminated. The ratiometric methods allow internal compensation for the variability and instability of instrumental factors. These factors change proportionally the signal in two (or more) channels. Consequently, the λ-ratiometric signal should be insensitive to variations in intensity of the light source, to sensitivity of detector or to sample geometry. The intensity ratio should be reproducible on an instrument with a different optical arrangement, light source intensity, slit widths, etc.
Thus, the two-band ratiometric signal can be easily calibrated in analyte concentration, providing improved performance [7]. The construction of systems generating two sets of signals and obtaining their ratio can be plentiful and diverse. As we will see below, the λ-ratiometry fits ideally to two different principles in operation of fluorescence probes-to molecular reactivity and molecular recognition. Both molecular [8] and nanoscale [9,10] fluorophores can be applied in their realization.
Rather than increasing or reducing the intensity of a single color, the λ-ratiometric probes produce a range of colors that are easily distinguishable and recorded by a simple instrumentation. Moreover, a clear change of color is well-suited for visual detection, sometimes with the 'naked eyes' [11,12]. This possibility is hard to achieve when only the brightness at single wavelength is detected.
The calibrated λ-ratiometric signal allows achieving informative imaging of distribution of analyte when the probe concentration and its distribution in sample cannot be easily controlled. Providing imaging in any heterogeneous system, one can manipulate with probe concentration (observing its distribution as the change of its absolute intensity) together with the intensity ratio (applied for determination of analyte concentration). Therefore, the λ-ratiometric imaging formed of distributions of concentrations of different analytes in complex biological environments become very popular [13][14][15][16].
The present review discusses different possibilities for recording and analysing the intensity ratios, in order to help the researchers to find the avenues for their most efficient use in the development of advanced sensor technologies. The reader may choose between different approaches in λ-ratiometric sensing that are based on different mechanisms producing spectroscopic change and using single, double and multiple fluorescence emitters. In Pt. II of these series [17] we outline different practical aspects of this versatile methodology, analyze different sources of error and provide many successful examples of its application. The prospects for further development of λratiometry due to its multiplexing and multimodal capability and easy application of miniaturized devices are also discussed.

Basic principles and definitions
According to the most general principle 'no interaction-no information' [18], every analytical application of fluorescence should involve reversible or irreversible, covalent or noncovalent interaction of analyte with molecular or nanoscale device that provides the reporting signal.
2.1. Molecular reactivity and molecular recognition -two different principles in operation of fluorescence probes When interacting with analyte, the fluorescence probe may change its chemical structure resulting in detectable change of its emission. It can serve as λratiometric probe if as a result of chemical reaction, initially fluorescent probe generates a new band in its bright emission. One can observe its initial emission band intensity decreasing and the new band of reaction product increasing. Their ratio provides the measure of analyte concentration. Fluorescence probes operating in this way are called the λ-ratiometric reactivity-based probes or chemodosimeters [19,20]. Selectivity that allows recognizing analyte in complex mixtures is determined by its unique reactivity towards the probe [21]. Since the analyte-induced reactions are basically the ground-state phenomena, the two emitting reactant and analyte-modifies species should be display different both excitation and emission spectra (see the case a,c in figure 3 below).
Such chemodosimetry is of special importance for the detection and imaging of reactive oxygen, nitrogen and sulfur species in living cells [22][23][24]. Small size of these analyte molecules makes difficult their affinitybased sensing, but their high reactivity allows realizing a number of chemical transformations with different dyes. In some of these cases, the primary response is of ON-OFF manner, involving electron transfer reaction [25], and the λ-ratiometry is achieved by coupling a molecular or nanoscale reference [26].
Disadvantages of this kind of probes are in the irreversible manner of their reactions, which proceed in time and often with relatively slow reaction rate. The probes are required in relatively high concentrations and they are unsuitable for the rapid kinetic detection of analytes. They can be considered as a new generation of fluorescent detection reagents in analytical chemistry. Figure 2 depicts the specific features of application of these probes.
Quite different are the properties of molecular or nanoscale probes that interact with target analyte reversibly (see figure 2). They do not change their structure on this interaction and can be re-used for many times. Therefore, they can be called sensors in strict sense of this term [18]. Upon exposure to tested sample, only a part (often, a small number) of the target analyte species bind to a sensor, and this binding can be sufficient to tell us about the true target concentration. In this case, the testing is not limited to size of the sample. Imaginary, it can be even the whole sea, to which the sensor can be immersed for continuous monitoring its pollution.
For such sensor operation, the molecular recognition should operate in the conditions of equilibrium between bound and unbound analyte. The term molecular recognition is commonly applied to operation of these sensors. It indicates the formation of multiple noncovalent specific interactions between molecules of chemical or biological origin [27,28]. Such interactions could be hydrogen bonding, metal coordination forces, van der Waals interactions, π-π stacking, etc [29]. Their formation-disruption should be translated to fluorophore that provides reporting signal [18].
The equilibrium established on reversible binding allows a simple description based on mass action law. Kinetics of ligand binding is expressed by a rate constant k 1 and that of dissociation of the complex by rate constant k 2 , and the equilibrium can be characterized by a ratio of these constants. This can be either the binding constant K b = k 1 /k 2 or its reverse function, the dissociation constant K d = k 2 / k 1 . K b is often called also the stability or affinity constant, it is expressed in reverse molarity units (M −1 ), whereas K d = 1/K b is expressed in molar (M) units. Thus, K d expressed in concentration units is a convenient characteristics of a sensor characterizing the dynamic range of sensed concentrations. This range is not large; it is roughly one order of magnitude below and one order above K d . Therefore, K d as the sensor property should fit the range of analyte concentrations to be measured. If someone wants to provide sensing outside this concentration range, the sensor with different affinity to analyte must be used.
Thus, the application of principle of molecular recognition allows obtaining true sensors that by exploring the regularities existing between the bound and unbound species allow determining their whole number, keeping in mind that only the bound species produce the sensor response signal.
2.2. Selection of ground-state forms and/or excitedstate reactions in probe/sensor response Each light-emitting molecule stays in the electronically ground state and switches to the excited state upon absorption of light quantum. Intermolecular interactions that can be valuable for probing and sensing can be formed both in the ground and excited states, but their spectroscopic manifestations are different ( figure 3).
The positions of fluorescence excitation and emission spectral bands are determined by the energies of correspondent electronic transitions, and their variation-by shifting in equilibrium between correspondent ground-state or excited-state forms. So, the transitions between different ground-state forms are reflected in dependence of fluorescence emission on excitation wavelength, F(λ ex ). The interactions with analyte may generate the differences in the ground-state energies between free and bound forms producing the difference in their absorption (and excitation) spectra. Thus, the two forms of the dye belonging to two states of the system and differing by their intermolecular interactions can be considered as different species with their own characteristic excitation and emission spectra. Therefore, the interplay of these spectra originating from unbound and bound forms generates the λ-ratiometric signal that can be recorded in excitation spectra ( figure 3(A)).
The excited-state reactions (reversible or irreversible) can generate new bands in fluorescence emission, F(λ em ). The ground-state interactions in this case are the same. Therefore, the excitation spectra do not change, and the probe can be excited by light of a single wavelength. The spectral change in fluorescence appears as a result of establishing the equilibrium between analyte-free and bound forms in the excited state ( figure 3(B)), demonstrating the changes in fluorescence emission spectra, but not in excitation spectra.
Thus, for generating the λ-ratiometric response we possess two possibilities, the ground-state selection or the excited-state reaction. This can be done by recording either the excitation or the emission spectra [30]. The latter possibility is commonly preferable, since it allows the convenient use of a single-wavelength light source.

Introduction of reference signal in intensitybased sensing and imaging
The frequently used approach in λ-ratiometry is the introduction of light-emitting reference that serves for  The energy diagrams and schematic illustration of the changes in excitation and emission spectra in the cases of ground-state and excited-state equilibrium. In the case of ground-state equilibrium (A), the two species, 1 and 2, are independent absorbers and emitters, so both their excitation (hν 1 a and hν 2 a ) and emission (hν 1 F and hν 2 F ) energies differ. This allows observing dual excitation spectra changing in proportion to the bound analyte. Since they emit independently, two bands in emission spectra can also be observed. In the case of excited-state reaction (B), the excitation energy hν 1 a is the same but the emission energies (hν 1 F and hν 2 F ) differ, which allows observing two emission bands.
providing additional independent channel of information from that of reporter dye. The latter may operate based on ON-OFF or OFF-ON switching ( figure 1(E)). The role of reference is solely to indicate its own presence in particular medium or at particular site.
The reference dye is usually excited at the same wavelength as the reporting dye, but it should possess strongly different emission spectrum (with minimal spectral overlap) and of comparable intensity to that of reporter band. One must account that the two dyes behave quite independently regarding chemical degradation and photobleaching. Different rates of these processes change their recorded ratio of intensities with time. Likewise, quenching in conditions of tested medium and other unaccountable factors may produce errors in λ-ratiometric recording.
When the reporting and reference dyes are located at a short distance (e.g. coupled within the same conjugate or nanocomposite), there is a problem for their emission to be independent in view of the possibility of excited-state reaction, such as electron or energy transfer, between them. An experience exists, how to choose the dyes for avoiding their 'cross-talking' [31][32][33]. Different types of coupled organic dyes, the hybrids of nanoscale emitters may be quite efficient with this approach.

Realizing dual-triple emission from a single dye
Multiple ground-state forms of a single fluorophore can generate dual or triple emission, and transitions between them can be induced by target analyte (see figure 3). The analyte can interact with fluorophore directly, influence its electronic system or provide conformational change [6].
Multiple emission states in single-component compounds can be observed when they demonstrate the excited states transformations [34]. These transformations are reversible in a sense that in a cycle [groundstate excitation → excited-state reaction → relaxation to reaction product ground state → transition to initial ground state] the same initial ground-state species are restored. This ensures their use many times in the sensing process. Switching between these excited-state forms has to be coupled with sensing event, offering many possibilities (figure 4).
3.1. Excited-state intramolecular charge transfer (ICT) Intramolecular charge transfer occurs at a small distance within the dye molecule by redistribution of electronic charge and creating (or destroying) the strong dipoles in the excited states [35]. The energy of electronic transition changes and because of that we observe the shifted fluorescence bands (figure 5). Typical ICT dyes contain an electron-donating group (often a dialkylamino group) and an electron-withdrawing group (often, carbonyl) located at opposite sides of aromatic molecules. The addition of an analyte can cause its binding to the donor or acceptor sites and result in alteration of the dipole strength of the donor-acceptor species, leading to spectral shifts [36][37][38]. If the analyte is a charged compound or if its binding is coupled with the variation of charge in sensor vicinity, then such switching is easy, generating the two-band response.
Internal rotation between two fragments of fluorophore molecule may stabilize the ICT state by generating the twisted intramolecular charge transfer (TICT) [35]. The switching from initially excited to TICT state can be used in sensing [39]. The problem with the application of such dyes is in occurring often significant quenching of their strongly Stokes-shifted TICT emission [40,41].
The other effect influencing fluorescence spectra of ICT dyes is electrofluorochromism that is known as internal Stark effect observed in fluorescence [42,43]. This phenomenon is caused by direct interaction of the ground-and excited-state dipoles with the applied electric field that can occur on molecular scale due to the presence of nearby charges. It allows designing the so-called molecular voltage sensors [44]. Specially designed dyes can be applied for recording the electrostatic potential in biomembranes [45]. Thus, inducing or removing the charged group in vicinity of dye may induce sensor response.

Making-breaking of hydrogen bonds in the ground and excited states
The presence of H-bond donor and acceptor groups coupled to aromatic structures offers new possibilities for λ-ratiometric sensing based on formation/breaking of intermolecular H-bonds. These bonds may influence the intramolecular charge-transfer (ICT) character of the excited states with correspondent spectral shifts. Such effects are usually smaller than that on protonation (section 3.3) and occur in the same direction [53]. In excitation spectra, on interaction with the H-bond donor at the acceptor site one can observe the shifts of spectra to longer wavelengths and on interaction with proton acceptor at the donor site -to shorter wavelengths.
Carbonyl groups attached to π-electronic system can be found in many organic dyes. They are the strongest H-bonding sensors that are sensitive to protic solvent environment [54]. The electron lone pairs on oxygen atoms allow arrangement of two such bonds with proton donors, the stronger and weaker ones [55]. Their formation is clearly seen in fluorescence spectra (figure 6) and can generate strong λ-ratiometric signal.
Generation of dual or even triple fluorescent signal based on modulation of intermolecular hydrogen bonding could be a very important possibility for fluorescence ratiometric reporting. It should be accounted, however, that the formation/disruption of these bonds can frequently lead to fluorescence quenching [56].

Proton transfer to solvent or to external acceptor
Excited states of dye molecules differ from correspondent ground states by the ability to donate proton to the medium or to accept proton. Protonation-deprotonation of groups attached to fluorophore (usually aromatic hydroxyls) may change dramatically the spectroscopic properties leading to the appearance of new bands in the absorption and emission spectra [57]. These reactions occur with the participation of solvent water molecules or other proton donors/ acceptors interacting with fluorophore. In aqueous solutions, two or more forms based on protonation equilibrium can be observed with the appearance of new bands in absorption and emission spectra (figure 7).
In many organic molecules, both proton-bound and proton-dissociated forms are strongly fluorescent and demonstrate different positions of their spectra [58] being suited for λ-ratiometric sensing. Many fluorophores conjugated with proton-dissociating chemical groups (phenolic hydroxyl, carboxylic and sulfonic acids) are used as fluorescent pH probes [59]. Presently the most extensively used dyes are the benzo [c]xanthene derivatives, such as C-SNAFL-1 [60,61]. For providing the λ-ratiometric recording in an extended pH range, a combination of two fluorophores [62] or attaching several dissociating groups to a single fluorophore [63] must be applied.
The λ-ratiometric recording of protonation behavior of fluorescent dyes has found application not only in pH sensing but also on a broader scale, when it is coupled with other reactions induced by analyte binding in water. For instance, it can be coupled with the recognition of charged species [64]. The fact that the pH-dependent change in ionization of dye functional groups is local and can respond to the change in association-dissociation equilibrium of macromolecules without change of medium pH can be efficiently used in sensing technologies.

Excited-state intramolecular proton transfer (ESIPT)
The dyes exhibiting ESIPT demonstrate very strong (often by 100 nm and more) shifts of emission of proton-transfer forms to longer wavelengths and, therefore, are very attractive for the design of λratiometric reporters [65,66]. Meantime, realization  The intensity ratio between N * and T * emissions can respond to analyte binding in a limited number of cases: (a) When the ESIPT reaction is gated by transformations between two or more ground-state forms as a result of chemical reaction [65], conformational change or action of any factor resulting in disruption or perturbation of the H-bond as an ESIPT pathway (e.g. deprotonation/protonation reaction in proton donor or acceptor [70]).
(b) When the initially excited normal (N * ) state is stabilized by intermolecular interactions to possess  . Illustration of proton dissociation equilibrium leading to pH sensing in the absorption and fluorescence spectra. At higher pH values, one may observe the interplay of two absorption bands due to simultaneous presence of proton-bound and proton-dissociated forms. In the excited state, the equilibrium at this pH is commonly shifted to proton-dissociated form (photoacids), leading to one band in fluorescence emission. At lower pH, the ground-state equilibrium is shifted to proton-bound form, and the absorption spectrum is represented by a single band. In the excited state at this pH, one may reach the conditions of equilibrium between two forms (proton-bound and dissociated), that are represented by correspondent fluorescence bands. the same or close energy as the product tautomer (T * ) state [71]. Shifting the dynamic equilibrium to one of N * or T * forms can be modulated by analyte binding influencing the energetics of ESIPT reaction, [72,73].
(c) When the ESIPT reaction exhibits slow kinetics on the time scale of emission that allows the N * form to be present in spectra. In general, perturbation of πelectronic sub-system activates/suppresses the ESIPT kinetics or modulates the ESIPT equilibrium.
A very attractive feature can be observed if the dynamic N * ↔ T * equilibrium is established very fast on the time scale of emission. Then the λ-ratiometric response is robust against the thermal and collisional quenching [74,75]. Simply, two emission channels are quenched in proportion without change of λ-ratiometric signal. Such unique feature is extremely valuable for applications in different chemical sensing and imaging technologies [6,76,77].

Dual fluorescence and phosphorescence emissions
Phosphorescence is the emission from triplet state that is characterized by strong Stokes shifts, long emission lifetimes (up to milliseconds and seconds) and strong temperature-dependent quenching. The dyes that display strong room-temperature phosphorescence with the steady-state intensity comparable to that of fluorescence can be used in λ-ratiometric sensing [78,79].
Several types of materials are efficient in λ-ratiometric sensing. There are metal-free organic dyes [80,81], often containing bromine and iodine heavy atoms [82], platinum and palladium ions incorporated into porphyrins [83], chelating complexes of ruthenium [84,85] and different nanoscale composites [86]. The reporting signal is usually connected with the quenching of phosphorescent component. For oxygen sensing in solutions [87] and in vivo ratiometric hypoxia detection [88], such fluorescent-phosphorescent dyes can be applied, since oxygen is a collisional quencher of long-lived emission.
Summarizing, the dual-emissive molecular probes are very attractive for sensing and imaging. However, they should demonstrate the ground-state or excitedstate transformations that can provide band-shifting and generation of new bands; these bands should be brightly emissive.

Generation of ratiometric fluorescence response in the systems of two or more interacting emitters
Traditional ratiometric fluorescent probes based on composing different dyes and fluorescent nanoparticles are very popular despite their essential drawbacks, such as time-consuming and complex preparation and complicated modification or conjugation. However, in contrast to single dual-emitting dyes, they offer greater variability of structures providing broader choice for sensor engineering.

Excitation energy transfer (EET) to fluorescent acceptor
When two molecules or nanoparticles are located at a close distance and one (donor) becomes excited, it can transfer its excitation energy to unexcited one (acceptor) and the latter becomes emissive. We can observe the case depicted in figure 1(D)-the analyte-induced perturbations result in variations of both emissions in opposite directions, leading to wavelength-resolved ratiometry. This process is called the excitation energy transfer (EET).
There are two basic energy transfer mechanisms. The Dexter energy transfer (DET) is an electron exchange process that requires overlap of the electron wavefunctions of energy donor and acceptor [89,90]. Sharp exponential decays with the distance on the scale of one nm allows calling DET the short-range energy transfer. It can be realized in molecular dyads and commonly results in quenching. Recently the systems displaying through-bond energy transfer were demonstrated [91][92][93], where fluorescent chemosensors are composed of three main parts: energy donor, energy acceptor, and rigid linker. This approach is promising for designing efficient λ-ratiometric sensors and imagers [94] and deserves further development.
In contrast, the Förster resonance energy transfer (FRET) that is presently very popular in sensor design [95][96][97], allows easy achieving the λ-ratiometric recording on a broader scale [98]. In this case, two or more dye molecules can exchange their excited-state energies due to long-range dipole-dipole resonance interaction. The energy acceptor should not be excited directly but has to be strongly fluorescent. The FRET efficiency depends on two major factors: the distance and mutual orientation of donor and acceptor (to allow the non-radiative resonant dipole-dipole coupling) and the overlap integral of emission spectrum of the donor with the absorption spectrum of the acceptor, as the condition of resonance ( figure 9).
The fact that the fluorescence response depends strongly on the properties of both donor and acceptor and on their relative distance and orientation allows many possibilities for its modulation and adaptation to the desired conditions. The dependence of transfer efficiency on the distance R is very steep, as R −6 . Thus, for the observation of FRET, the direct contact between donor and acceptor is not needed and they can interact through space being located at separation distances within 1-10 nm from each other.
The weak points of FRET technology should be indicated. Its realization usually needs labeling with two strongly fluorescent dyes serving as donor and acceptor, which complicates the probe synthesis [4]. The overlap of absorption spectra of donor and acceptor may cause significant problems due to undesirable direct excitation of the acceptor. In addition, variation of noncovalent intermolecular interactions may shift the absorption and fluorescence spectra, influencing the overlap integral in uncontrollable manner. One has also to keep in mind that the coupled emitters may have unequal chemical stability, photostability or uncontrolled perturbation by external quenchers, which may cause systematic errors.
The most attractive feature of FRET is its steep distance dependence on the scale of 5-10 nm. There are many possibilities to modulate this distance by target analyte and in this way to obtain λ-ratiometric sensor response ( figure 9). The conformational change in double labeled sensor may occur under the influence of analyte [99,100]. Also enzymatic splitting of covalent bond between two labeled units [101,102] and substitution of labeled competitor in a complex with the labeled molecular sensor [103] may be used as the signal for sensing.
In addition, one can change the spectral overlap that determines the FRET efficiency by variation of the properties of both donor and acceptor, thus increasing the sensor sensitivity [104] and modulating the λratiometric response [105]. In this respect, several possibilities can be outlined. First is the target-induced change of the absorption spectrum of the acceptor. A variety of cassettes of two fluorophores can be constructed, in which the target binding changes the overlap integral [106].
The other possibility is realizing 'cascade transfer' [107]. In a sequence of dyes, FRET can occur in cascade manner, in which an energy acceptor can serve as the donor to another dye with lower excitation and emission energy. In 'FRET-gating' [108,109] the donor and the acceptor are in proximity but their spectral overlap is insufficiently small for the transfer. An introduction of the third partner that can serve an acceptor to primary donor and a donor to terminal acceptor results in the appearance of efficient transfer. Including/removing of this third component can generate a sensing signal producing a strong separation of donor and acceptor emissions.
A variety of donor-acceptor pairs for FRET applications have been developed, satisfying the requirement of at least partial overlap of the donor emission spectrum the absorption spectrum of the acceptor. Selection of these partners can be made among different molecular and nanoscale emitters. The most popular in this respect are the organic dyes [110,111]. Increase in application of nanocompositions is observed in recent years [112]. They include semiconductor quantum dots [113,114] and other nanoparticles in combination with organic dyes [115,116] or silver nanoclusters [117].

Excited-state interaction between two emitters: the excimers
When a molecule absorbs light, its electronic properties change dramatically. The fluorophore may participate in reactions that are not observable in the ground state. Particularly it can make a reversible complex with molecule like itself but residing in the ground state. These excited dimeric complexes are called the excimers. Pyrene demonstrates the bright excimer emission spectrum that is very different from that of monomer [118,119]. It is broad, shifted to longer wavelengths (from ∼400 nm to ∼485 nm) and, in contrast to monomer, it is not structured. Therefore, switching from monomer to excimer can provide the λ-ratiometric signal that is very convenient for observation. Bringing together or forcing apart the monomers generates sensor response [120].
There are different examples of formation/disruption of excimers coupled with conformational changes in sensor molecules. Sensors based on excimers are popular in detecting ions and also of neutral molecules, such as glucose [121]. The flexible oligo-or polynucleotides are actively explored in DNA hybridization techniques and recognition of specific mRNA sequences [122,123]. In a quantitative detection of ATP, the excimer-forming constructions are used.
Thus, the use of pyrene-based cyclophanes is an illustrative example (figure 10).

Collective effects between the dyes: exciton coupling and diffusion
With the rapid development of nanotechnologies, new possibilities became open for fluorescence sensing. The fluorescent nanosensors are becoming ultrabright and ultra-sensitive due to exploration of effects of plasmonic enhancement, upconversion, superquenching, etc Detailed description of these phenomena can be found elsewhere [18]. Here we briefly outline excitonic effects in fluorophore ensembles that are useful for λ-ratiometric sensing.
By definition, molecular excitons are mobile neutral quasi-particles in the form of electron-hole pairs that appear in solid bodies and molecular associates upon electronic excitation [125,126]. Excitons can be considered as the propagating collective excitations in a system of interacting molecules, in which the participating fluorophores do not behave as individual emitters and the whole system responds (emits light or is quenched) as one unit [127].
Specific excitonic effects appear in regular associates of similar dye molecules, even their dimers, leading to quite characteristic spectra of J and H aggregates that differ dramatically from the spectra of constituting monomers [128,129]. These aggregates differ between themselves by specific arrangements of monomers. In H-aggregates, the spectra are broad and shifted to shorter wavelengths, and in J-aggregates the shift is to longer wavelengths, with essential band sharpening. Self-assembly and disassembly of these aggregates is in the background of operation of JC-1 dye, the most popular λ-ratiometric sensor for membrane potential in mitochondria [130].
In textbooks, the Förster resonance energy transfer, FRET, is commonly presented as a binary event occurring between single donor and single acceptor. Meantime, when many such donors and acceptors assemble within the critical distance in a small volume, they lose their individuality and new collective effect of exciton propagation appears. High level of ordering is not required. The energy transfer occurs as incoherent hopping, in which the excitation energy jumps randomly as a diffusive energy transport between molecular-type forms [18].
In all these cases the phenomenon of superquenching can be realized [18], which may provide strong enhancement of sensor response. The rapid development of techniques for obtaining the dyedoped nanoparticles made of silica [131], organic polymers [132] and dendrimers [133] offer new possibilities for improvement of sensor properties.

Molecular fluorophores used in λratiometry
Small organic molecules can satisfy nearly all possible requirements for fluorescent reporters in λ-ratiometric sensors. Many photophysical and photochemical processes can be realized with them [18], see section 3. They can be used as single dyes demonstrating wavelength shifting and new band forming; they are basic components in EET pairs and ensembles. They can be constituents of conjugated polymer nanoparticles and silica-based and polymer-based nanocomposites. Their absorption and fluorescence spectra may occupy the whole range of electronic transitions-from UV to near-IR. Low chemical stability and photostability [134] may be revealed as their weak point. The most popular are the dyes demonstrating intramolecular excited-state reactions of charge and proton transfer.  [124]. The sensing mechanism involves the encapsulation of a nucleic acid base between two pyrene rings, which disrupts the monomer-excimer equilibrium. The nature of the spacer and its connection pattern to pyrene rings were designed to achieve high selectivity for ATP.

Dyes demonstrating intramolecular charge transfer
Intramolecular charge transfer (ICT) is realized between two distant but electronically coupled functional groups of atoms, one serving as electron donor and the other as electron acceptor (see figure 6). These groups are integrated into π-electronic system and upon excitation shift their electronic density, creating strong dipoles. Their excitation energy is reduced, resulting in fluorescence band shifted to longer wavelengths. Additional shift is observed if these excited-state dipoles are stabilized by dipoles of the polar environment suggesting their application as solvent polarity sensors [135]. Prodan [51] is their popular example. Its properties were improved by different substitutions of the donor and acceptor group in the naphthalene core [136] and by using anthracene [137] and pyrene [138] instead of naphthalene core. New probes on the basis of fluorene [139] also demonstrated their improved features.
Phenoxazone dye Nile Red demonstrates solventdependent shifts both in absorption (from 480 to 590 nm) and fluorescence (from 530 to 660 nm) spectra on transition from hexane to water. Its derivatives are used for covalent labeling of proteins and peptides [49], also inside the living cells [140]. Due to very strong spectroscopic changes, this dye and its derivatives [141] are in active use not only as indicators of polarity but also for constructing chemosensors [142]. In phospholipid membranes [51] and in the environment of transmembrane receptors [143] they are efficient indicators of lipid order.

The excited-state intramolecular proton transfer (ESIPT) dyes. Deprotetion of reactive groups
The development of ESIPT-based fluorescence probes (see section 3.4) is particularly attractive for the design of λ-ratiometric sensors [65,66]. This reaction requires unique properties of the dyes, such as close location and intramolecular H-bonding interaction between a hydrogen bond donor (-OH and NH 2 ) and a hydrogen bond acceptor (=Nand C=O). The formation/disruption of these bonds can be coupled with deprotection reactions and with conformational changes. Those are the benzazole dyes ( figure 11(a)). Their two-band switching, dependent on pH and binding of ions disrupting the ESIPT process, find use in sensing [65,145].
An idea of using the ESIPT-activating effect of analyte-induced bond splitting is illustrated in figure 11(b). A λ-ratiometric fluorescent probe for Mercury ions was developed based on the analyte-promoted hydrolysis of a vinyl ether derivative of 2-(benzothiazol-2-yl)phenol in a buffer solution, leading to deprotection of ESIPT donor group ( figure 11(c)). The probe with a marked fluorescence change from blue to cyan responds selectively to Mercury species over various other metal ions [144].

The dyes coupling ICT and ESIPT reactions
The strongest perturbation of dye electronic structures can be achieved when the excited-state process is the reaction between the ICT state, initially excited as the N * state, and the product of ESIPT reaction T * state [67]. Transforming even small ICT-provided shifts into ESIPT effects of strong variation of relative intensities of two well-resolved emission bands, these dyes can be considered as extremely efficient amplifiers of ICT response [67,146,147].
The designed organic dyes belonging to the family of 3-hydroxychromones (3HCs) satisfy these requirements in the optimal way [66,70,72]. Structural modifications of parent fluorophore allow variation of spectroscopic properties of these dyes in very broad ranges. In figure 12 the dyes of two series based on 3-hydroxyflavone (left) and 2-benzofurylchromones (right) are shown together with correspondent fluorescence spectra. The dyes arranged in line with the increase of their excited-state dipole moments demonstrate shifting the range of λ-ratiometric sensitivity to lower polarities. Their other series, in which naphthofuranyl [150] or thiophenyl [151] groups are attached to chromone heterocycle extend the range of their unique properties.
The application of a great number of 3HC derivatives in their free forms or on covalent labeling of proteins, peptides and nucleotides allow outlining the general trends in exploration of their ultrasensitive response: -Response on the scale of polarity. Within a very narrow scale of variation of polarity in their environment, they exhibit an ultrasensitive twoband ratiometric response, up to complete switching between N * to T * bands in emission. This range can be adjusted by chemical substitutions [149] covering the range from very low [73] to very high [152] polarity values.
-Response to hydrogen bond donation potential (proticity) of the environment, which in biomolecular structures can also be a measure of hydration. The proton acceptor carbonyl group can serve as H-bond sensor. The probes can sense the true concentration of H-bond partners existing in the ground state [54], which can be achieved by spectral deconvolution [30].
-Response to the magnitude and direction of the local electric field on molecular scale (electrochromic modulation of ESIPT. This allows providing the sensing response to the presence of closely located charges. The methods for determining the changes of membrane potentials [153][154][155] and detection of the early steps of apoptosis [14] are based on such response.
Thus, the molecules of the 3HC family can respond to several, the most important, types of intermolecular interactions. Based on these findings, the concept of multiparametric sensing was developed [72]. It is successfully applied in different sensor technologies.
On a general scale, organic dyes play major role in designing the response units of chemical sensors and biosensors based on λ-ratiometry. They offer tremendous possibilities of choice due to their huge number, diversity of their spectroscopic properties and possibilities of variation of these properties by chemical modifications.

Wavelength-ratiometry with nanoscale structures
Nanoclusters, nanoparticles and nanocomposites have rapidly become very popular as reporters in fluorescence sensing and imaging [18]. Their advantage is the possibility of manifold increase of brightness and  typically much higher chemical stability and photostability [156]. The change of their emission color can be observed even with the naked eye [12,157] and on recording with a cellphone [158] and on test paper strips [159], see Part II. Meantime, when scanning recent literature [160,161], one may observe that the realization of their advantages in λ-ratiometry on nanoscale is rather poor. They are frequently used only as the reference emitters or quenchers. The reason for that is clear. Direct coupling of sensing event with photophysical transformations of their spectra, as in organic dyes, is complicated and not always possible. Their benefit is different: in many cases the collective photophysical effects [9,10], such as superquenching, make them the key players in different nanocomposites.

Ratiometric fluorescent composites based on quantum dots
The role of semiconductor quantum dots (QDs) in sensing technologies is significant due to their rather narrow and size-dependent spectra. Depending on the material, they may be finely tuned covering the broad range of wavelengths in visible and near infrared [114,116]. Their excitation range is very broad, extending to the UV. This means that in combination with other QDs of different colors they can make the reporter-reference pair (case E in figure 1) excited at the same wavelength. In combination with organic dyes and other fluorophores possessing localized excitation bands, they can form ideal FRET pairs, in which, by selecting the excitation wavelength, direct excitation of the acceptor is avoided (case D in figure 1).
The hybrid probes are typical for CDs. Thus, comprising two sizes of cadmium telluride QDs emitting red and green fluorescence, respectively, the sensor was devised for on-site visual determination of copper ions [157]. The red-emitting QDs were embedded into silica nanoparticles and served as the reference, and the green-emitting ones were covalently linked onto the surface and were functionalized to be selectively quenched by the analyte. The continuous color change from green to red is clearly observed by the naked eye ( figure 13).
Combinations of organic dyes as FRET donors and semiconductor QDs as acceptors have been reported for λ-ratiometric detection of a wide variety of analytes [114,116]. The compositions, in which QDs serve as excited-state energy donors and lanthanides europium and terbium as the acceptors, deserve attention. Targeting antibiotic tetracycline, Eu 3+ and Tb 3+ ions can be chelated by this analyte, and QDs provide the transfer of energy with strong fluorescence enhancement [162,163].

Carbon dots and their compositions
Carbon dots (CDs) are fluorescent nanoparticles obtained by carbonization of different organic materials. Their spectroscopic behavior is very different from that of semiconductor QDs [164] and is similar to organic dyes with large Stokes shifts [165]. Different polar groups can be found on their surface, the composition of which depends on material of their origin [166]. Depending on this composition, CDs can emit light in different colors, from blue-green to near-IR. Their surface groups allow easy modification and functionalization, which is convenient for making the efficient sensors [167,168]. The major mechanism of CDs fluorescence response is the electron-transfer quenching [169] in the form of superquenching [164]. They can be used in combination with different emitters for making λ-ratiometric sensors [170,171] serving as both energy-transfer donors and acceptors [172].
Co-assembly of CDs with organic dyes was performed by many authors and resulted in versatile λratiometric sensing and imaging technologies exploring the FRET mechanism [172,173].
Combination of CDs with quantum dots allows strong separation of their emission wavelengths. Observing two emissions in blue-green and red ranges is convenient for both donor-acceptor (case D in figure 1) and reporter-reference (case E in figure 1) types of response [170,174].
Atomically precise metal nanoclusters of gold, silver and copper stabilized in proper environments are strong luminescence emitters [175]. Usually in their composites with CDs, the reporting function belongs to nanoclusters, and the CDs are used as the reference [176,177]. Interesting is the use of lanthanide-carbon dot composites. The ions of europium and terbium possess structured luminescence bands in the visible range, the intensities of which can be increased by proper chelating units. These structures forming composites with CDs make very efficient λ-ratiometric sensors [178][179][180]. Their change of bright emission colors may be sufficient for visual analyte detection.

Ratiometric sensing with conjugated polymers
The emissive conjugated polymers demonstrate considerable potential as the basis for λ-ratiometric sensors [181,182]. Their extremely high brightness [183] is determined by polarizable π-electronic system extending along the conjugated backbone [184] that behaves as one unit, demonstrating the collective effect of 'superquenching' [185].
A strongly amplified response to the analyte binding can be achieved by coupling this effect with the excited-state energy transfer or electron transfer. Polymer dots are compact nanostructures made of these polymers [186,187]. The sensing nanocomposites can be made by noncovalent incorporation of organic dyes into the dots or by their covalent binding to the chains [188].
Conjugated polymer nanoparticles doped with a Mercury-responsive rhodamine spirolactam derivative enable aqueous detection of Hg 2+ with two-color λ-ratiometric response [189], figure 14. The dyes are nonfluorescent until they are irreversibly modified by Mercury ions converting them to fluorescent rhodamines. The polymer fluorescence is switched from green-yellow to orange-red via FRET to the rhodamine dyes. The light-harvesting capabilities of the donor allows the conjugated polymer chains to induce strongly enhanced intensity to FRET-excited dyes.

Metal-organic frameworks
Metal-organic frameworks (MOFs) are ordered nanoporous hybrid materials that can be self-assembled from their corresponding inorganic metal ions/clusters with organic linkers [190], which allows formation of an extended, crystalline framework with nanoscale porosity ( figure 15). MOFs can combine the inherent physical and chemical properties of both inorganic and organic photonic units that allow their potential applications in luminescence-based sensing and imaging [192,193].
There are diverse emitting mechanisms for generating tunable dual-color emission that can be realized in MOFs [193]. They include the metal-centered (MC) emission (e.g. ligand-to-metal charge transfer (LMCT)) and ligands-centered (LC) emission (e.g. metal-to-ligand charge transfer (MLCT)). Usually, these processes in the excited states are complete, and we have to observe only one band in emission. However, a specific analyst can disrupt the charge transfer, allowing the λ-ratiometry.
The MOF framework can also incorporate the ligand demonstrating its own emission or participating in energy transfer [194]. Thus, luminescent behavior can arise from the ligand unit, metal ions, or the cavity guest molecules. Based on different strategies of synthesis, the MOFs-based dual-emissive λ-ratiometric sensors are categorized into three classes: MOFs' intrinsic dual-emission, single-emissive MOFs with incorporated fluorophore and non-emissive MOFs with two fluorophores [193].
The luminescence properties of MOFs are very sensitive to their structural characteristics, coordination environment, nature of the pore surfaces, and their interactions with guest species through coordination bonds, π-π interactions, hydrogen bonding, etc, thus providing solid rationale to develop luminescent MOF sensors [195]. The ability of both inorganic and organic components of structure to participate in emission of different types of λ-ratiometric sensors can be realized [196]. Thus, the MOFs building blocks can incorporate the dyes exhibiting the excited-state intramolecular proton transfer in the sensors for water vapor [197].
Lanthanide MOFs, as the most common luminescent MOFs [193,198]. The units incorporated into them can participate in charge, electron or excitedstate energy transfer [199]. The diversity of presently available Ln-MOFs, leading to broad diversity of emission colors [200], make them powerful candidates in developing ratiometric detection systems [201].
Fluorescent nanoparticles, such as carbon dots, can be incorporated as the guests in the network structure [202,203]. In formed nanocomposites, they can work together with lanthanides [204] and serve as the donors on transferring the excitation energy to organic dye [205]. Spectacular change of light emission color from red to blue was observed for MOFbased nanocomposites responding to the presence of Figure 14. The conjugated polymer nanoparticles enabling ultrasensitive detection of Mercury ions in water [189]. The Mercury-responsive nanoparticles are composed of the conjugated polymer poly [9,9- water molecules in organic solvents [191], figure 15. The probing for water in organic solvents demonstrated dramatic color change that can be recorded as the ratio of light intensity at 420 nm to that at 623 nm, increasing linearly with the increase of water content.
Concluding this section, it can be stated that the nanostructures formed of organic and/or inorganic luminophores open new possibilities in sensing technologies. In addition to increase of surface area, number of reactive groups and porosity of some structures, new diverse means appear for multicolor outputs in fluorescence reporting. In order to produce λ-ratiometric signal, any photophysical or photochemical mechanism is realizable here for generating or suppressing new emission bands, modulating their brightness and position on the wavelength scale. In addition to discussed above, novel types of nanostructures, such as silicon quantum dots [206], sulfur nanostructures [207] and boron nitride-based dots [208] may become efficient in these technologies. Additional possibilities that can be realized on nanoscale generating amplified response include the following: (a) Plasmonic enhancement of λ-ratiometric signal.
Metallic nanostructures exhibit the ability to increase the fluorescence of closely located molecular and nanoscale light emitters due to generation of localized plasmons [209,210]. This effect is size-dependent and spectrally selective, and its range of optimal action differs for gold, silver and copper nanoparticles that can be included in nanocompositions.
(b) Superquenching in a system of conjugated fluorophores. When the dyes are linked in excitonic structures or coupled performing efficient EET exchange, they can be quenched as a single unit. This allows increasing dynamic range from strongly enhanced emission of nanoparticles to complete darkness. This effect is very efficient for J-aggregates [211,212], conjugated polymers [213] and dye-doped fluorescent nanoparticles [214].
(c) Light-harvesting (antenna effect). In λ-ratiometric sensing, one can operate with nanoparticles demonstrating FRET in such a way that many FRET donor molecules are assembled in a nanoparticle together with few number of acceptors [215]. Then the light absorption by donors is great and this increases the fluorescence intensity of acceptors, resulting in a total increase of brightness.

Upconversion nanoparticles and nanocomposites
The upconversion nanoparticles (UCNPs) possess unique features converting the long-wavelength excitation to shorter-wavelength emission by taking advantage of long-lived intermediate energy states of lanthanide ions [216,217]. The physical term 'upconversion' means the excitation from already populated excited state to a higher energy state, from which the emission can proceed at shorter wavelength. Presently, the most popular for various applications are nanocrystalline NaYF 4 host lattices doped with Yb 3+ , Er 3+ and Tm 3+ [218,219]. They are excited in the near-IR (e.g. by 980 nm diode laser) and emit light in the form of several narrow bands in the UV, blue, green, red, or near-IR regions.
The unique position of UCNPs in sensing and imaging technologies is due to their characteristic features. Primarily, their luminescence is free from background, since no other presently known natural or men-made materials can demonstrate such mode of anti-Stokes excitation and emission. Figure 15. The dual-emission nanohybrid sensor designed for a sensitive and visual detection of water contents in organic solvents [191]. (a) The sensor Eu-MOFs/N,S-CD is formed by incorporation of blue light-emitting nitrogen and sulfur co-doped CDs (N,S-CDs) into the framework of red-emitting and water-stable Eu-MOFs with nanoscale porosity. In organic solvents, the sensor demonstrates red emission. Upon interaction with water molecules, the emission color is changed to blue. (b) The visual change of emission color from red to blue on increasing the concentration of water in dimethylformamide.
The presence in UCNPs of multiple emission lines together with the possibility to manipulate with their relative intensities open new possibilities in λ-ratiometric sensing. They can be realized in full when the UCNP nanoparticles are integrated with various type of optically active materials such as gold nanostructures, quantum dots and organic dyes [216]. Energy transfer and electron transfer processes generated in them can be coupled with both analyte-recognition and analyte-induced reactive events. For λratiometric sensing, different types of nanocomposites can be formed, in which the UCNPs serve as the references with their very sharp characteristic emission bands.
The UCNPs are ideal excited-state energy donors to any molecular or nanoscale acceptors [220]. Due to extremely narrow emission bands of luminescent lanthanide ions, no donor emission can be detected at the wavelengths of sensitized emission of the acceptor. In addition, no acceptor can be excited directly because the excitation is in the near-IR. This means that the donor and acceptor emission channels do not overlap, which is almost impossible to achieve with organic dyes. Meantime, the use of these nanoparticles as the energy transfer acceptors allows overcoming their weak point -a weak light absorbance of lanthanide ions [221]. Their conjugation with multiple strongly near-IR absorbing organic dyes, acting as light harvesters and providing the excitation energy to emissive sites, allows making them strong light absorbers [222,223]. Plasmonic enhancement of UCNP emission is also possible. When induced by gold nanoparticles, the modulation of relative band intensities can be observed, selectively enhancing the red/green intensity ratio [224].
The fact that UCNPs form a number of emission bands that are sharp and can be well-separated on the wavelength scale suggests the construction of λ-ratiometric fluorescence nanosensors based on selective modulation of emission of these bands. In this case, the UCNPs play a passive role and the analyte-responsive contributor provides the switching of emission of different colors. Commonly, this is realized in two ways, by the excitation energy transfer quenching and by exploration of the inner filter effect [225].
The idea of spectrally selective energy transfer quenching is illustrated in figure 16(a). The UCNP incorporates a naphthalimide-based colorimetric probe A that changes dramatically its absorption spectrum upon interaction with Hg 2+ ion and being the acceptor of its short-wavelength emission, switches to become the acceptor of long-wavelength emission [226]. This composite nanoprobe uses the upconversion luminescence ratio at 451 nm and 361 nm to demonstrate high selectivity and sensitivity toward Mercury ions. The EET quenching in composite UCNPs can be also provided in near-IR range, which demonstrates important advantages in biological imaging [228].
The realization of inner filter effect (IFE) does not need the presence of analyte-responsive factor in the composite UCNP-based nanoscale probe, which may substantially simplify the assays. The IFE requires only the overlap between the absorption band of the absorber and the emission bands of the fluorophores in the detection system. The analyte-responsive factor has to be present in solution in sufficiently high concentration. This factor can operate by the shift of its absorption spectrum covering only part of spectrally distributed spectrum of UCNP and by such spectrallyselective screening of UCNP emission generate the ratiometric response of the whole system. Since it is not the quenching but screening of emission, it changes the intensity but does not alter the lifetime of emission.
Thus, characteristic feature of this technique is the use the strong light absorbers as a target-binding and responsive element, instead of fluorophores or light quenchers. They can be organic dye-metal ligand complexes [229] and gold nanoparticles [230]. The later change dramatically their extinction on aggregation. Detecting different analytes, such as uric acid, the IFE agents can be generated in enzyme [231] and enzyme mimicking [232] reactions.
Example of dye-UCNP system for selective detection of fluoride anion is presented in figure 16(b). For the detection of fluoride ion, the λ-ratiometric fluorescence sensor, incorporating Yb 3+ , Er 3+ , and Tm 3+ co-doped NaYF 4 upconversion nanoparticles emitting at 546, 657, 758 and 812 nm were used [227]. The curcumin dye served as specific recognition element in nanocomposites with UCNPs. When Fions are added, the absorption peak of curcumin demonstrates a bathochromic shift, causing an upconversion fluorescence quenching at 546 and 657 nm through inner filter effects (IFE), whereas the upconversion emission at 758 and 812 nm remained unchanged. Thus, the fluorescence ratio I 546 /I 758 is inversely proportional to Fconcentration.
The spectrally selective energy transfer quenching and the IFE-based spectrally selective screening can be realized not only in UCNPs but also with other emitters, such as two-color QD or CD composites [233][234][235]. The unique spectroscopic properties of UCNPs demonstrate strong advantages in application of these techniques.

Conclusions and prospects
This Tutorial Review has to demonstrate the advantages of λ-ratiometry that combines simplicity and high information content when intensity ratio, rather than intensity itself, is used to quantitatively determine the presence of analytes. Such approach is used not only in fluorescence but also in Raman scattering, photoacoustics [9] and, probably, in other areas of analytical science. High sensitivity of fluorescence method, together with other attractive features, allows strong move in development towards miniaturization and multiplexing of assays. Focusing on materials for optimal realization of λ-ratiometry, we can emphasize two trends of further development: -Substantial increase of brightness that could allow sensing the change of emission color by naked eye. For molecular dyes, it is limited by their absorption cross-section and quantum yield of fluorescence but become possible with nanoparticles. Thus, with dye-doped nanocompositions, one can obtain a fluorescence signal, which in some cases is 10 4 to 10 5 times stronger [131].
-Extension of the wavelength range of response that could decrease or even eliminate the overlap of fluorescence bands, the intensities of which form the ratio. Such extension of λ-ratiometric fluorimetry to near-IR range [236,237] for ESIPT performing dyes [238] and members of green fluorescent protein (GFP) family [239] was achieved recently, and these properties can be realized in a variety of nanoscale systems.
Rapid development of λ-ratiometry has transformed it into powerful branch of fluorescence sensing technology with new design and adaptation of existing fluorescence emitters, modification of chemical assays and bioassays for monitoring the environment and clinical diagnosis, new instrumentation and data analysis. These issues together with selected examples of applications will be discussed in Part II of these series [17].

Data availability statement
The data that support the findings of this study are available upon reasonable request from the authors. The data that support the findings of this study are available upon reasonable request from the authors.

Funding
This research has received no external funding.

Conflicts of interest
The author declares no conflict of interest. Figure 16. Examples of using the UCNPs in ratiometric fluorescence sensing. (a) The structure of colorimetric probe A and its attachment to the surface of UCNP [226]. Its absorption spectrum changes from UV-violet to blue on binding of Hg 2+ ion. Also shown is the spectroscopic effect of this binding. The UV-vis absorption spectrum of probe A (3.9 × 10 -5 mol/L) shifting dramatically to longer wavelengths covers the short-wavelength emission of water-soluble DMSA-UCNPs (0.2 mg mL −1 ) excited at 980 nm. The EET quenching changes the ratio of intensities between UCNP emission bands generating ratiometric response. (b) The UV-vis spectra of the curcumin in the absence (red dashed line) and presence of Fanion (red solid line). This change induces the wavelengthselective screening of UCNP emission due to inner filter effect [227]. The upconversion fluorescence spectrum of UCNPs (black dashed line) under excitation with 980 nm laser changes relative intensities in UCNPs-curcumin mixed system with the addition of Fion generating the ratiometric response. The upconversion fluorescence intensity ratios (I 550 /I 758 ) of UCNPs-curcumin versus different concentrations of Fand the linear fits are also shown.