The Ln(III) ions possess unique photophysical properties, which have led to a lot of interest in recent years.2-3,132,287 Lanthanides are the elements of atomic numbers 57–71, in the first period of the f-block of the periodic table, see Figure 3.1.56 The most stable lanthanide oxidation state is +3; these ions are hard Lewis acids, possessing relatively high charge densities and have a strong electrostatic nature to their bonding. Therefore, ligand design usually comprises of hard Lewis bases, such as amines, carboxylic acids, hydroxyl groups and nitrogen-containing heterocycles.3 The Ln(III) ions have a large and varied range of coordination numbers of between 6 and 12, with the most studied ions, Eu(III) and Tb(III), commonly displaying a coordination number of 9.7 The Ln(III) ions comprising these complexes, upon saturation of their coordination sphere, are protected from vibrational coupling increasing the light absorption profile by ‘antenna effects’.289 The characteristic photophysical features of the ions allow for the monitoring of the self-assembly process in solution via relatively simple spectroscopic techniques such as UV-vis absorption, luminescence spectroscopy, especially as NMR techniques can be rendered complicated due to the paramagnetic Ln(III) ions.290 Upon complexation, the unique emission spectra of the Ln(III) ion may be sufficiently modulated such that complex formation as a function of metal addition to ligand (or vice versa) may be evaluated using non-linear regression analysis software such as SPECFIT and ReactLab Equilibria.21 These software
Figure 3.1 Partial energy diagrams for Ln(III) ions (main luminescent levels in red and fundamental levels in blue). Figure reproduced from Ref.56 Copyright 2005: Royal Society of Chemistry
programs can be used to analyse the experimental binding isotherms observed in both the ground and the excited states, with the view of determining both the speciation and the stability constants for the formation of various self-assembly equilibria in solution.
The unique photophysical properties of Ln(III) ions are beneficial in the sense that the excited states are long-lived, long wavelength emission, sharp characteristic line-like emission bands and large pseudo-Stokes shifts.288 These unique properties of the Ln(III) ions arise from the shielding of the 4f orbitals by the 5s2 and 5p6 orbitals resulting in little vibrational coupling with the environment and as a result they interact only very weakly with ligand orbitals. Emission bands are therefore narrower and ion-specific, leading to relatively pure emission colours.56,132 Pure f→f transitions are forbidden and as a result, excited state lifetimes are several orders of magnitude longer than those of the background fluorescence of the biological media, for example.291 Long lived excited states are useful in sensing as time-gating can overcome drawbacks arising from light scattering and auto-fluorescence140 while emission beyond tissue absorption wavelengths improves signal quality. Ln(III) ions are being increasingly used as useful reporter groups for monitoring certain chemical reactions and as non-cytotoxic fluorescent probes.287,292-295
As direct excitation of Ln(III) emission is not efficient due to the small dipole strength of f→f transitions, the Ln(III) excited state can be populated alternatively using a sensitising chromophore, ideally with a high absorption coefficient and Förster overlap,140 through a
process called the ‘antenna effect’,140,288 see Figure 3.2(A). The modified Jablonski diagram in Figure 3.2(B)56 shows the sensitising antenna absorbing electromagnetic radiation, followed by a change in spin multiplicity of the excited state. The energy is then transferred to the Ln(III) centre289,296 and from there it can be dispersed either by emission of light (luminescence) or via a variety of non-radiative deactivation mechanisms. The energy difference between the triplet excited state of the ligand and the accepting level of Ln(III) must be sufficient (>1700 cm−1) such that thermally activated energy back-transfer is minimised.291
The efficiency of a given ligand to sensitise the luminescence of a Ln(III) ion is given by:
𝑡𝑜𝑡 = 𝜂𝑠𝑒𝑛𝑠𝐿𝑛𝐿𝑛 = 𝜂𝑠𝑒𝑛𝑠𝜏𝜏𝑜𝑏𝑠
𝑟𝑎𝑑 Equation 1
56
where tot is the overall quantum yield of the metal centered luminescence measured upon ligand excitation relative to a standard such as Cs3[Eu(dpa)3] in Tris buffer (with 𝑡𝑜𝑡 = 24±2.5% for Cs3[Eu(dpa)3] under excitation at 279 nm).58𝐿𝑛𝐿𝑛 is the intrinsic quantum yield determined upon direct metal excitation, ηsens is the antenna-to-ion energy transfer efficiency and 𝜏𝑜𝑏𝑠 and 𝜏𝑟𝑎𝑑 are the observed and radiative lifetimes, respectively. For Eu(III) complexes tot can be measured by relative method to which the absorbance and emission
intensity of the sample are compared according to:
𝑡𝑜𝑡= 𝑥𝑟= 𝐸𝑟𝐸𝑥×𝐴𝑥(𝜆𝑥)𝐴𝑟(𝜆𝑟)×𝐼𝑥(𝜆𝑥)𝐼𝑟(𝜆𝑟)× (𝑛𝑥
𝑛𝑟)
2 Equation 2
where subscript r – reference and x – sample; E – integrated luminescence intensity; A –
absorbance at the excitation wavelength; I – intensity of the excitation light at the same
wavelength, n – refractive index of the solution. The estimated error for quantum yield
determination using the relative method is ±10%.
Figure 3.2 (A) The ‘Antenna’ effect. (B) Simplified Jablonski diagram for Ln(III) absorption– emission process.
Having measured tot,,sens and 𝐿𝑛𝐿𝑛can be calculated using Equation 1 once radlifetime
is obtained. This is achieved using Equation 3: 1
rad= 𝐴𝑀𝐷,0×
𝐼𝑡𝑜𝑡
𝐼𝑀𝐷× 𝑛
3 Equation 3
where nis the refractive index of the solvent, AMD,0 is the spontaneous emission probability
for the 5D0→7F1 transition
in vacuo, AMD,0 = 14.65 s–1 and 𝐼𝑡𝑜𝑡
𝐼𝑀𝐷is the ratio of the total area of
the corrected Eu(III) emission spectrum to the area of the 5D0→7F1 band.61,297 The value of
AMD,0 was found from the theoretically calculated dipole strength and with the aid of Judd- Ofelt theory. The 5D0→7F1 transition in the Eu(III) emission spectrum is the only magnetic dipole transition and has no electric dipole contribution. Therefore, this transition is practically independent of the changes of the Eu(III) coordination environment. The reason
R, 𝐿𝑛𝐿𝑛 and sens are not accessible in the case of Tb(III) is because this metal does not possess a purely magnetic dipole transition like the 5D0→7F1 in Eu(III).
The quantum yield of the luminescence step (𝐿𝑛𝐿𝑛) expresses how well the radiative process compete with non-radiative processes.𝐿𝑛𝐿𝑛essentially depends on the energy gap between the lowest lying excited (emissive) state of the metal ion and the highest sublevel of its ground multiplet. The smaller this gap, the easier is its deactivation by non-radiative processes, for instance, through vibrations of bound ligands or solvent molecules, particularly those possessing high energy vibrations such as O–H. In this way, the luminescence of Ln(III) ions can be quenched via the transfer of energy from the Ln* excited state to solvent molecules, such as H2O and CH3OH, through molecular vibrations when these species are inside the coordination sphere. The minimisation of this process can be achieved through the relevant saturation of coordination sites with chelating ligands. The number of metal-bound protic solvent molecules is proportional to the rate of quenching. Horrocks et al. developed an equation to estimate the hydration state of a Eu(III) or Tb(III) complex based on the observation that O–D isotopic oscillators reduce the excited state lifetimes of Eu* and Tb* to a far lesser extent than O–H oscillators.298-299 This equation was further modified by Parker et al.300 to include the contribution of N-H and C-H oscillators and the following describes the hydration state of Eu(III) in H2O and deuterium oxide i.e.
𝑞𝐸𝑢(𝐼𝐼𝐼)= 1.2 {( 1 𝜏𝐻2𝑂−
1
𝜏𝐷2𝑂) − 0.25 − 0.075𝑥} Equation 4
where τ is the lifetime of the complex in solution and accounts for the oscillations of water and deuterium oxide bonds. 𝑥 accounts for the number of C-H and N-H oscillators directly
bound to the Eu(III) metal. A q values of 0 suggests that there are no solvent molecules in the inner coordination sphere which is ideal for systems whose importance rely on their luminescence. By complexing with appropriate multidentate ligands assembled around the metal centre it is expected that the Ln(III) sits in a fully saturated environment, shielded from solvent molecules, with a coordination number of 9, for example through tris-terdentate ligand binding in the case of btp. Hence, the expected self-assembly formation is of a 1:3 M-L complex. The formation and characterisation of several btp-derived Ln(III) complexes will be presented in the next section.