Design of water-soluble lanthanide luminescent probes for the selective detection of HOCl

Abstract

The synthesis and photophysical properties of tacn- or pyclen-based macrocyclic complexes featuring extended thioanisole-picolinate antennas are reported. These systems efficiently sensitize Eu(III) luminescence, whereas Tb(III) emission is quenched due to back-energy transfer (BET). In all lanthanide-containing complexes (Eu, Tb and Gd), the sulfide donor group was readily oxidized by HOCl to the corresponding sulfoxide or sulfone. Upon oxidation, the Tb(III) complexes became highly luminescent and act as OFF–ON probes when excited at the isosbestic point (302 nm). Finally, a water/buffer soluble Tb(III) complex combining ether and sulfide antenna was prepared, exhibiting remarkable sensitivity and selectivity towards HOCl detection.

---

Introduction

Hypochlorous acid (HOCl), the active species in household bleach, is a powerful disinfectant. It is a weak acid with a pK_a of 7.54 at 25 °C and its conjugate base is hypochlorite (ClO⁻).¹ HOCl is a strong oxidant that reacts with a wide variety of biomolecules including amino acids, nucleic acids, protein co-factors such as NAD⁺ and unsaturated lipids.^2,3 Among amino acids, sulfur-containing residues such as cysteine and methionine are the primary targets, although lysine and other aromatic residues, i.e. tryptophan, tyrosine, phenylalanine and histidine, are also affected. Cysteine, featuring a –SH end-group, can undergo oxidative dimerization to form disulfide or can be oxygenated to form sulfenates (RSO⁻), sulfinates (RSO₂⁻), and sulfonates (RSO₃⁻). In contrast, methionine, featuring an –SMe end-group, can be oxidized into sulfoxide (RS(O)Me) and sulfone (RS(O)₂Me). Lysine reacts through its side-chain amino group to form chloramine, another potent oxidizing agent. Aromatic side chains of amino acids can undergo chlorination and other biological targets, including nucleobases, co-factors and lipids, also form chlorinated products. HOCl plays a crucial role in the innate immune response, being produced in large amounts within the phagosomes of macrophages and neutrophils by myeloperoxidases from H₂O₂ and Cl⁻.⁴ At lower concentrations, HOCl has also been proposed to act as a mediator in biological signaling.⁵

The biological significance of HOCl has motivated the development of fluorescent probes for its detection. Most reported systems rely on organic fluorophores, whose emission is modulated by the presence of HOCl, typically through an irreversible chemical transformation (activity-based probes).^6–8 Several mechanisms have been exploited for the selective detection of HOCl, including the oxidation of phenols to quinones, oximes to aldehydes, alkenes to aldehydes and sulfides to sulfoxides or sulfones.

Luminescent lanthanide(III) (Ln(III)) complexes are attractive alternatives to organic fluorophores for biological applications.⁹ They exhibit sharp emission bands at fixed wavelengths, characteristic of each lanthanide, and are largely independent of the surrounding environment. Moreover, they possess long-lived emission lifetimes in the microsecond-to-millisecond range. These properties make them easy to distinguish from the endogenous organic fluorescence of biological systems, which is characterized by broad emission bands and short-lived signals (in the nanosecond range).¹⁰ Because f–f transitions are Laporte forbidden, efficient sensitization of Ln(III) complexes relies on the antenna effect.¹¹ A light-harvesting chromophore, referred to as an antenna, is positioned in close proximity to the Ln(III) center. Upon photon absorption, the antenna transfers its energy to the Ln(III), thereby populating its emissive excited state. For a given Ln(III) ion, the antenna must be carefully selected: the donor state, most often the first triplet excited state, should lie approximately 2000–5000 cm⁻¹ above the Ln(III) emissive excited state to ensure efficient energy transfer while preventing back-energy transfer (BET).^12,13 Compared with organic-fluorophores, Ln-based HOCl-responsive probes are still rare.^9,14,15 Most reported examples come from the group of Yuan. One strategy for designing HOCl-responsive Ln(III) complexes employed an aminophenyl-functionalized antenna, which quenched the excited chromophore, likely via photoinduced electron transfer, thereby blocking the Ln(III) sensitization.^16–19 Upon reaction with HOCl, the quenching moiety is cleaved, restoring efficient Ln(III) sensitization, and thereby switching on the Eu(III) emission. Another strategy used a Eu(III) complex bearing two β-diketonate ligands that are oxidized by HOCl. This oxidation releases Eu(III), quenching its emission.^20,21 These probes were used for HOCl imaging in cells by time-gated microscopy with UV excitation (λ_ex = 330–380 nm). Sørensen and coworkers also reported Ln(III) complexes (Ln = Tb or Eu) incorporating a proximal hydroquinone, where reaction with HOCl altered the Ln(III) emission, although the underlying mechanism was not determined.²²

Over the past decade, many groups focused on developing luminescent lanthanide bioprobes for microscopy imaging. A major drawback of visible-emitting Ln(III) complexes is their reliance on UV excitation for luminescence sensitization, which can damage biological samples. For this reason, current commercial microscopes are not equipped with such high-energy light sources. This limitation can be overcome by using two-photon (2P) absorbing antennas, which shift the excitation wavelength from the UV to the red/near-infrared region, significantly less damaging for in vivo studies.^23–25

We have recently reported LnL₁ complexes (Ln = Eu, Tb, Gd) that comprise a triazacyclononane (tacn) macrocycle bearing three picolinate groups extended with a thioanisolyl moiety (Chart 1).²⁶ The 2P-absorbing 4-(4-methylthiophenyl)picolinate antenna efficiently sensitizes the luminescence of Eu(III) but not that of Tb(III). As mentioned above, sulfides readily react rapidly with HOCl to yield sulfoxides and other poly-oxygenated derivatives.^2,27,28 Oxygenation of the 4-(4-methylthiophenyl)picolinate antenna is therefore expected to alter its sensitizing properties. In this study, we first examined the luminescence response of the LnL₁ and LnL₂ complexes shown in Chart 1 to assess their potential as HOCl-sensitive probes. Building on the limitations identified for these two complexes, namely poor water solubility and a blue-shift in absorption upon oxidation, we then designed and evaluated an improved structure, TbL₃, as a HOCl-responsive probe potentially suitable for 2P-microscopy imaging.

Chart 1 Structures of the discussed and studied Ln(III) complexes.

Results and discussion

Synthesis

The syntheses and characterization of the EuL₁, TbL₁ and GdL₁ complexes based on tacn macrocyclic ligands bearing three thioanisolyl antennas have been described recently.²⁶ The pyclen derivatives LnL₂ have been prepared according to reported procedures²⁴ but using the thioanisolyl picolinate antenna. The synthesis of L₂ is based on the key compound 1 (Scheme 1), described earlier,^29,30 which consists of a pyclen macrocycle, regiospecifically N-functionalized at a lateral position with an acetate pendant arm. The pyclen derivative 1 was reacted with mesylated thioanisolyl antenna 2 in acetonitrile with potassium carbonate under reflux for 24 h. As the reaction proceeded slowly, an additional amount of activated antenna 2 was added, and the mixture was refluxed for further 24 h to drive the conversion. After cooling, the reaction mixture was filtered, evaporated, and purified on an alumina column, affording the ligand precursor L₂′ in 48% yield. The methyl ester-protected precursor L₂′ was then dissolved in a THF : MeOH 4 : 1 mixture and saponified with potassium hydroxide under stirring at 70 °C for 16 h. After solvent removal, and elimination of salts by centrifugation, the product was subjected to dialysis, yielding compound L₂ as a white powder in quantitative yield. The final lanthanide complexes were prepared by dissolving L₂ in water, adjusting the pH to 6.3–6.4 with HCl or KOH and adding 1.1 equiv. of the corresponding europium and terbium chloride hexahydrate salts. The reaction mixtures were stirred for 48 h under reflux (Ln = Eu) or at r.t. (Ln = Tb), then centrifuged and purified by C₁₈ reverse-phase flash chromatography with a H₂O : MeCN gradient. The complexes were isolated as white solids with yields depending on the lanthanide ion and reaction conditions (85% for EuL₂ and 10% for TbL₂).

Scheme 1 Synthesis of pyclen-based ligand L₂ and LnL₂ complexes (Ln = Eu, Tb).

The water soluble tacn-based TbL₃ complex featuring two distinct antennas has been prepared through a multistep synthesis (Scheme 2). The thioanisolyl picolinate antenna A₂ was first synthesized from the polyethyleneglycol (PEG) functionalized 4-bromothioanisole 3.³¹ Miyaura borylation of 3 with bis(pinacolato)diboron afforded the boronic ester 4 in 91% yield which was subsequently employed in a Suzuki–Miyaura coupling with the methyl ester of 6-(hydroxymethyl)-4-iodopicolinate 5, affording the target SPEG-functionalized antenna A₂ in 73% yield. Alcohol A₂ was then quantitatively converted to the mesylate 6, suitable for tacn alkylation (vide infra). Alternatively, antenna A₂ could be obtained via a similar Suzuki reaction with dimethyl p-iododipicolinate followed by monoreduction of a resulting diester A₂′ to the alcohol A₂ with sodium borohydride (Scheme S1 in the SI). However, this route resulted in a significantly lower yield (34% vs. 73%) due to competitive reduction of both methyl esters in A₂′ yielding undesired bis-alcohol byproducts. The second antenna with the twisted charge-transfer character, was proved to sensitize bright Tb(III) emission in a tacn-trispicolinate macrocyclic environment with a 74% quantum yield in water.³² The mesylate of the twisted antenna 7 was used for the bis-alkylation of tacn in the presence of Hünig's base in a 1 : 1 dry MeOH : CHCl₃ mixture under reflux for 24 h. Using a strict 1 : 2 stoichiometric ratio of tacn to activate antenna 7 and a weaker organic base instead of alkali carbonates allowed the formation of bis-alkylated tacn derivative 8, which was purified via column chromatography on neutral alumina (43%). The remaining secondary amine of the bis-alkylate 8 was reacted with the mesylate of the SPEG-functionalized antenna 6 using sodium carbonate as a base in dry acetonitrile upon heating to 60 °C for 24 h. The formed [2 + 1] ligand precursor L₃′ required identical chromatographic purification as in the previous step, affording the target molecule in diminished yield (24%). The protected ligand L₃′ was saponified with sodium hydroxide in a methanolic-aqueous 1 : 4 mixture at 70 °C for 24 h and then reacted in situ with terbium chloride after methanol removal and pH adjustment to 6.0 with HCl and NaOH. Complexation was performed at r.t. over 72 h. TbL₃ was isolated as a colorless solid in 93% yield after extraction with dichloromethane.

Scheme 2 Synthesis of SPEG-functionalized antenna A₂, ligand precursor L₃′ and the TbL₃ complex.

Oxidation of the thioanisolyl antenna

The oxidation of the A₁ sulfide functional group to sulfoxide and sulfone moieties by HOCl was studied using UV-Vis absorption spectroscopy and high-resolution mass spectrometry (HRMS). The thioanisolyl antenna A₁ experienced a rapid reaction of the sulfide function upon treatment with 3 equiv. of HOCl in MeOH (C = 0.1 mM) as evidenced by the pronounced blue shift of the UV-Vis absorption transition of the charge-transfer (CT) band from λ_abs = 312 nm of A₁ to 270 nm of A₁^ox (Fig. 1). HRMS analysis of the cuvette titration products confirmed the formation of the mono-oxidized sulfoxide A₁^ox (Fig. S1, m/z [M + H]⁺ = 306.0795 for C₁₅H₁₆NO₄S along with the corresponding sodium adduct). The reaction kinetics of the mono-oxidized product is very fast, and in our hands, the reaction was completed within the mixing time.^27,33 The over-oxidized sulfone A₁^diox was also detected (m/z [M + Na]⁺ = 344.0562 for C₁₅H₁₅NNaO₅S). To characterize the fully oxidized product of A₁, an in situ oxidation was carried out by stirring 0.028 mmol of A₁ (C = 10 mM) a 1 : 1 MeOH : HOCl (5% aqueous solution mixture) at r.t. for a longer time, i.e. 72 h, to ensure complete formation of the end-product. The isolated product of the oxidation was A₁^diox, showing a further blue-shifted absorption maximum at λ_abs = 259 nm (Fig. 1). HRMS data confirmed the quantitative conversion to the sulfone product (Fig. S2, m/z [M + H]⁺ = 322.0742 for C₁₅H₁₆NO₅S and the corresponding sodium adduct). In the ¹H NMR spectrum, the di-oxidized species A₁^diox were characterized by a deshielded methyl signal (3.11 ppm), shifted by almost 0.6 ppm compared to the sulfide precursor.⁴ The pyridine and other aliphatic protons remained largely unaffected, while the aromatic signals of the phenyl ring conjugated to the sulfone group appeared ∼0.5 ppm downfield, consistent with reduced electron density at the sulfur in the +6 oxidation state (Fig. S3).

Fig. 1 (Top) Oxidation of antenna A₁ to A₁^ox and A₁^diox by HOCl. (Bottom) UV-Vis absorption spectra of antenna A₁ in MeOH before and after HOCl oxidations (C = 0.1 mM).

To rationalize the experimentally observed hypsochromic shift of the lowest-energy charge-transfer band upon stepwise oxidation of antenna A₁, we performed a theoretical study based on Time-Dependent Density Functional Theory (TD-DFT, see the SI for computational details). The calculated transitions were consistently red-shifted by about 12–24 nm relative to the experiment (Table S1). Nevertheless, the difference between A₁ and A₁^ox in the theoretical absorption maxima of the most red-shifted bands matched the experimental value, i.e. 42 nm (λ_abs = 336 nm and 294 nm for A₁ and A₁^ox, respectively). According to the calculations, the sulfone derivative A₁^diox exhibited an absorption transition at λ_abs = 271 nm, which represents a 23 nm blue-shift compared to A₁^ox. This shift was somewhat larger than the experimental value (Δλ_abs = 11 nm). The nature of the transitions was found to be predominantly HOMO → LUMO in A₁ (98.3%) and to a slightly lesser extent in the sulfoxide form (93.3%). By contrast, the charge-transfer character was markedly reduced in the sulfone form (78.2%, Tables S1 and S2). This trend was also reflected in the molecular orbital representations: in the di-oxidized antenna A₁^diox, the majority of the electron density is no longer localized around the sulfur atom, thereby hindering effective charge transfer from the donor (methylphenylsulfone) to the acceptor (picolinate) moieties of the molecule (Fig. S4–S6).

Responsive lanthanide complexes

Comparable HOCl oxidation studies were carried out on representative examples of both lanthanide complex families LnL₁ and LnL₂ (Ln = Eu, Gd, Tb) featuring three and two thioanisolyl moieties, respectively. In HRMS, the major peaks corresponded to the addition of three oxygen atoms in the case of the tacn-based complexes and two oxygen atoms in the case of the pyclen-based complexes (Fig. S7–S11). Together with the short time between HOCl addition and analysis, this suggests that sulfoxide is formed first on each antenna (sulfone would require a longer reaction time, vide supra). In both the tacn- and pyclen-based systems, TbL₁^ox and EuL₂^ox, minor amounts of over-oxidized species were also detected, namely sulfone/sulfoxide or bis-sulfone derivatives.

The photophysical properties of EuL₁ and TbL₁ in diluted MeOH solutions have been reported previously (Fig. 2 and Table 1).²⁶ Their absorption spectra show an Intra-Ligand Charge Transfer (ILCT) transition from the electron-donating thioanisolyl moieties to the electron-withdrawing pyridine group with a maximum at 330 nm. The EuL₁ complex showed a long excited-state lifetime (1 ms), high quantum yield (44%) and strong brightness (25 000 M⁻¹ cm⁻¹), making it particularly attractive for both 1P- and 2P-luminescence applications.²⁶ In contrast, TbL₁ exhibited poor emissive properties, dominated by ligand-centered fluorescence, with only a weak Tb(III) emission signal detectable in the red tail of the fluorescence emission. This weak Tb(III) emission was attributed to inefficient sensitization of Tb(III) by the thioanisolyl antenna, caused by an efficient back-energy transfer (BET) from the Tb(III) ⁵D₄ excited state (20 500 cm⁻¹) to the triplet associated with the thioanisolyl antenna (20 600 cm⁻¹). The small energy gap between these states (∼100 cm⁻¹) was estimated using the corresponding GdL₁ complex to determine the position of the ligand triplet excited state (Fig. 3a).

Fig. 2 Absorption (black), excitation (blue, λ_em = 614 nm for Eu and 541 nm for Tb) and steady-state emission spectra of LnL₁ (Ln = Tb, green; Ln = Eu, red) in methanol solutions at 293 K before (top, λ_ex = 330 nm) and after (bottom, λ_ex = 283 nm) addition of excess HOCl.

Fig. 3 (a) Time-gated emission spectra (delay = 50 µs) in a methanol : ethanol 1 : 4 mixture at 77 K of GdL₁ before (blue, λ_ex = 330 nm) and after (pink, λ_ex = 290 nm) addition of an excess of HOCl; the coloured lines indicate the energy position of the triplet state. (b) Perrin-Jablonski diagram of the Tb(III) luminescence sensitisation in LnL_1–3 complexes before and after oxidation by HOCl. CT = charge transfer state, ET = energy transfer, BET = back energy transfer, T₁ = first triplet excited state.

Table 1 Photophysical properties of LnL₁ and LnL₂ complexes (Ln = Tb and Eu) in methanol and TbL₃ in PBS buffer (pH 7.4) in the absence or in the presence of an HOCl excess

Complex (solvent)	Initial complexes^a		After addition of HOCl^b		Titration
	λ ^absmax (nm) (ε, M^−1 cm^−1)	Φ (%) (τ, ms)	λ ^absmax (nm)	τ (ms)	I F/I0@λex (nm)
a λ ex = 330 nm, precision for ε = ±10%. b λ ex = 283 nm. c 2.9% for Tb emission only.
TbL1 (MeOH)	330 (53 000)	1.6 (<0.02)	281	1.29	×36@301
EuL1 (MeOH)	330 (53 000)	44 (0.98)	281	1.03	÷420@370
TbL2 (MeOH)	328 (45 000)	0.9 (<0.02)	279	1.54	×406@302
EuL2 (MeOH)	329 (45 000)	46 (1.33)	279	1.34	÷730@370
TbL3 (PBS)	305 (31 000)	4.6^c (<0.02)	282	1.13	×15@298

---

The photophysical properties of the thioanisolyl antenna complexes changed drastically upon addition of excess HOCl. First, the ILCT transition was blue-shifted by 42 nm. This shift reflected oxidation of the thioether fragment, which suppressed its donor character and thereby reduced the ILCT contribution to the absorption band (vide supra). Second, the luminescence properties of the EuL₁^ox complex were essentially unchanged with identical lifetime and emission profile, indicating that the oxidized antenna still efficiently sensitized Eu(III) and that the coordination sphere of the central ion was preserved. In contrast, the photophysical behavior of TbL₁^ox was profoundly different. The ligand-centered transition disappeared completely, while an intense Tb(III) emission pattern emerged, accompanied by an increased luminescence lifetime of 1.29 ms. This indicated that the oxidized antenna was now able to efficiently sensitize Tb(III).

To rationalize the strong enhancement of 4f–4f luminescence, the position of the oxidized ligand triplet state was determined using the GdL₁^ox complex via low temperature time-gated emission spectroscopy. The long-lived triplet emission of GdL₁^ox was significantly blue-shifted to 442 nm compared to 485 nm for GdL₁, corresponding to a higher-energy triplet excited state at 22 600 cm⁻¹ (Fig. 3a). The resulting larger energy gap between this triplet and the Tb(III) accepting level (2100 cm⁻¹) effectively suppressed BET and eliminated triplet-mediated oxygen quenching (Fig. 3b). Consequently, TbL₁^ox became strongly luminescent owing to the removal of this major non-radiative deexcitation pathway.

Similar results were obtained with the pyclen-based EuL₂ and TbL₂ complexes as the same antenna was involved in the Ln(III) luminescence sensitization process (Table 1 and Fig. S12). While EuL₂ and EuL₂^ox exhibited similar luminescence properties, only the oxidized form of the antenna efficiently sensitizes the Tb(III) emission in TbL₂^ox. The photophysical properties of the complexes were therefore governed primarily by the nature of the antenna, the macrocycle exerting a marginal influence. These findings highlight both series of complexes as promising candidates for sensing applications, which motivated the subsequent titration experiments.

Titration experiments

The titration experiments of EuL₁, TbL₁, EuL₂ and TbL₂ were carried out in diluted methanol solutions [Optical Density (OD) was 0.2–0.4] by successive addition of HOCl equivalents directly into a 1 cm spectroscopic cuvette. The results are reported in Fig. 4, 5, S13 and S14, respectively. The reaction with HOCl was instantaneous; hence, absorption and emission spectra were recorded after each addition. For each titration at a given excitation wavelength (λ_ex), the variation of the luminescence intensity at the local emission maxima (λ_em) as a function of the number of HOCl equivalents was determined and the total variation of I_F/I₀ (I₀ and I_F representing the initial and final luminescence intensities, Table 1) was measured. In all cases, the absorption spectra clearly showed a single isosbestic point at 302 ± 1 nm indicating an equilibrium between two species: the initial (sulfide) and the oxidized (sulfoxide) antenna. The over-oxidized form (sulfone) detected by mass spectrometry played only a minor role in the titration experiments. The number of HOCl equivalents required for complete titration was estimated to be 15 for LnL₁ and 8 for LnL₂, bearing three and two antennae, respectively. So, ca. 4 to 5 equivalents of HOCl per antenna were necessary to fully oxidize all sulfoxide groups. For europium derivatives EuL₁ and EuL₂, both the initial and oxidized forms were emissive. Therefore, luminescence titration was performed at the cut-off wavelength (370 nm) where only the initial form absorbs. Consequently, the decrease in OD at the excitation wavelength (λ_ex) accounted for the observed decrease in Eu(III) luminescence. The europium complexes thus behave as emission on–off probes with luminescence intensity divided by factors 420 and 730 between the initial and oxidized forms for EuL₁ and EuL₂, respectively.

Fig. 4 Titration of EuL₁ by HOCl in MeOH at 293 K: evolution of the absorption spectra (blue to black), evolution of the Eu(III) emission spectra upon 370 nm excitation (red to black), and variation of the 614 nm emission intensity with the number of HOCl equivalents (inset).

Fig. 5 Titration of TbL₁ by HOCl in MeOH at 293 K: evolution of the absorption spectra (blue to black), evolution of the Tb(III) emission spectra upon 302 nm excitation (black to green), and variation of the 542 nm emission intensity with the number of HOCl equivalents (inset).

The terbium complexes, TbL₁ and TbL₂, were almost non-emissive in their initial state. A luminescence titration was therefore performed by irradiation at the isosbestic point where an equal amount of light is absorbed throughout the oxidative photoluminescence titrations (iso-absorption point). Upon oxidation, the TbL₁^ox and TbL₂^ox complexes became strongly emissive, with Tb(III) emission intensities increasing progressively during the titration. The luminescence signal was enhanced by factors of 36 and 406 for TbL₁ and TbL₂, respectively.³⁴ These complexes thus act as highly sensitive off-on luminescent probes. In conclusion, these titration experiments established a proof-of-concept for this class of complexes as responsive luminescent probes for HOCl detection. The europium derivatives behaved as on–off emission probes upon excitation at 370 nm, while the terbium compounds showed off-on response when irradiated at the isosbestic point (302 ± 1 nm). However, two limitations remain. First, the sulfide antenna does not efficiently sensitize Tb³⁺ luminescence, while the sulfoxide antenna does not absorb above 320 nm. As a result, these probes are unsuitable for 2P microscopy, which requires longer excitation wavelengths (i.e. 2P excitation ≥ 690 nm, corresponding to 1P absorption ≥ 345 nm). Second, the LnL₁ and LnL₂ probes are soluble in methanol but not in water, preventing their direct application in biological media.

Water-soluble HOCl probe

To address these drawbacks, complex TbL₃ (Chart 1) was designed featuring two kinds of sensitizing antennae, (4-alkoxy-2-methylphenyl)picolinate and 4-(4-alkylthiophenyl)picolinate, bearing ether and sulfide electron donating groups, respectively. The latter is intended to act as a HOCl-responsive antenna as described above, while the former is expected to remain insensitive to HOCl and provide absorption up to 355 nm. Lanthanide complexes with such an ether-based antenna were previously shown to be suitable for 2P microscopy.³² The addition of short PEG oligomers ensured the solubility in aqueous solutions of the final Tb(III) complex. TbL₃ was studied in PBS buffer at pH 7.4. Its absorption spectrum exhibits a maximum at 307 nm with a shoulder at ca. 285 nm and extends up to 370 nm as expected because of the sulfide-containing antenna. In its initial state, TbL₃ exhibits a modest Tb(III) emission (Φ = 2.9%) and a short Tb(III) emission lifetime (<0.02 ms) along with a residual antenna fluorescence (Φ = 1.7%), similar to what was observed for TbL₁ and TbL₂ (Table 1). However, compared with those complexes, the Tb³⁺ emission is stronger and the Tb(III)/antenna emission ratio is larger (63%), indicating a less efficient quenching by the BET mechanism in the case of TbL₃ due to the presence of a single sulfide antenna. The Tb(III) emission quantum yield of TbL₃ is ca. 25 times lower than that of the parent complexes featuring three ether-based antennas (τ = 1.36 ms, Φ = 74% in water).³ Therefore, a single sulfide antenna is sufficient to effectively quench Tb(III) emission through a BET mechanism in oxygenated water. The excitation spectrum of TbL₃ is blue-shifted compared with the absorption spectrum, indicating weaker Tb(III) sensitization efficiency of the sulfide donor antenna compared with its oxygen-donor counterpart.

TbL₃ was reacted with a 4-fold excess of HOCl in PBS. LC/MS analysis confirmed the complete consumption of TbL₃ and the formation of a new complex, TbL₃^ox, showing a +16 mass increase, consistent with the oxidation of the sulfide to a sulfoxide (Fig. S15). The absorption spectrum of TbL₃^ox showed a ca. 20 nm blue-shift relative to TbL₃ (Fig. 6) with a maximum at 285 nm and extending up to 355 nm, in line with the reduced CT character of the sulfide donor. Upon oxidation, an intense Tb(III) emission was recovered (Φ = 29%) with a long luminescence lifetime (1.13 ms) and no residual antenna emission (Fig. 6). This behavior is consistent with the suppression of the BET mechanism upon oxidation of the sulfide antenna. Furthermore, the excitation spectrum of TbL₃^ox closely matched its absorption profile, in contrast to TbL₃, indicating that both the ether- and sulfoxide-based antennas efficiently sensitize Tb(III) luminescence.

Fig. 6 Absorption (black), excitation (blue, λ_em = 541 nm) and steady-state emission spectra of TbL₃ in PBS pH 7.4 at 293 K before (top, λ_ex = 298 nm) and after (bottom, λ_ex = 298 nm) addition of a 4-fold excess of HOCl.

Then, the titration of TbL₃ with HOCl was performed in PBS at pH 7.4 and monitored by absorption and emission spectroscopy (Fig. S16 and Fig. 7A). In the absorption spectra, the low-energy ILCT band progressively decreased, while a high-energy band emerged, with the evolution directly proportional to the amount of HOCl added. An isosbestic point was observed at 312 nm. Complete oxidation of TbL₃ occurred at ca. 1.8 eq. of HOCl. Emission monitoring (λ_ex = 298 nm) showed a similar evolution: Tb(III) emission increased proportionally with HOCl concentration up to ca. 1.8 eq. beyond which a plateau was reached (Fig. 7A, left). Upon full oxidation, the Tb(III) emission intensity increased 14.7 fold, demonstrating that TbL₃ acts as a turn-off/on probe for HOCl in PBS buffer when excited at ca. 300 nm, as observed for the parent probes TbL₁ and TbL₂. To evaluate the role of the (4-alkoxy-2-methylphenyl)picolinate antenna, which absorbs around 350 nm regardless of the redox state, the titration was also monitored using the 350 nm excitation wavelength. Interestingly, Tb(III) emission still increased upon HOCl addition when excited at this wavelength (Fig. 7A, right). In this case, an 8.5-fold luminescence enhancement was observed, compared to 14.7-fold under higher-energy excitation (298 nm). This behavior highlights the possibility of using TbL₃ for 2P-microscopy using 700 nm excitation.

Fig. 7 (A) Evolution of the Tb(III) emission upon titration of TbL₃ by HOCl in PBS pH 7.4 at 293 K under 298 nm (left) and 350 nm excitation (right). Emission (λ_ex = 350 nm) of solutions of TbL₃ (5 µM) after addition of 2 eq. of HOCl, O₂˙⁻, H₂O₂, tBuOOH, HO˙, tBuOO˙ and ONOO⁻ (black bars) and after subsequent addition of 10 eq. HOCl (grey bars).

Finally, the response of TbL₃ to other oxidizing species (O₂˙⁻, H₂O₂, tBuOOH, HO˙, tBuOO˙ and ONOO⁻) was investigated in PBS at pH 7.4. None of these oxidants produced a significant enhancement of TbL₃ emission, in contrast to HOCl. Moreover, after exposure to these oxidants, TbL₃ remained responsive to HOCl, confirming the absence of interference from competing oxidants in solution (Fig. 7B). Therefore, TbL₃ operates as a highly sensitive and selective off-on responsive probe for HOCl in water, supporting further developments for microscopic applications in HOCl detection.

Conclusion

This work reports the synthesis and photophysical properties of tacn- or pyclen-based macrocyclic complexes featuring an extended thioanisolyl-picolinate antenna. The chromophore efficiently sensitizes Eu(III), whereas Tb(III) emission is quenched by BET to the low-lying triplet state. Oxidation of the sulfide donor group by HOCl yields the corresponding sulfoxide, inducing a blue shift in absorption and a significant increase in the triplet-state energy of the oxidized antenna. Consequently, the oxidized Tb(III) complexes become strongly luminescent. Titration experiments revealed complementary sensing behaviors: Eu(III) complexes behave as ON–OFF probes when excited on the red-tail of the absorption band (370 nm), due to the blue shift upon HOCl oxidation, while Tb(III) complexes act as OFF–ON probes when excited at the isosbestic point (302 nm), reflecting enhanced sensitization by the oxidized antenna. Furthermore, a water or buffer-soluble Tb(III) complex was prepared by combining ether and sulfide antenna, extending the excitation window to 350 nm and making the complex suitable for two-photon microscopy. Reactivity studies toward other oxidants (O₂˙⁻, H₂O₂, tBuOOH, HO˙, tBuOO˙ and ONOO⁻) confirmed its excellent selectivity for HOCl. These results open new avenues for the design of bioconjugated Tb-based bioprobes for in cellulo HOCl detection using 2P-microscopy.

Author contributions

Conceptualization: O. S. and O. M.; investigation: L. B., A. N., L. C., D. A., S. R. K., K. P. M., and B. L. G.; validation: O. S., O. M., B. L. G., M. B., R. T., and J. M. K.; writing (original draft): O. M., S. K. R., O. S., and R. T.; writing (review & editing): all authors.

Conflicts of interest

There are no conflicts to declare.

Data availability

The data supporting this article have been included as part of the supplementary information (SI). Supplementary information is available, including detailled synthetic procedures and complete chemical characterizations, photophysical measurement details and computational chemistry results. See DOI: https://doi.org/10.1039/d5dt02832a.

Acknowledgements

The authors acknowledge the Agence Nationale de la Recherche (LANTEN: ANR-21-CE29-0018) for financial support. AN, JKM and OS thank the Labex ARCANE and CBH-EUR-GS (ANR-17-EURE-0003).

References

1. J. C. Morris, J. Phys. Chem., 1966, 70, 3798–3805.
2. C. C. Winterbourn, in Encyclopedia of Radicals in Chemistry, Biology and Materials, Wiley, 2012.
3. M. J. Gray, W.-Y. Wholey and U. Jakob, Ann. Rev. Microbiol., 2013, 67, 141–160.
4. S. J. Klebanoff, J. Leukocyte Biol., 2005, 77, 598–625.
5. J. M. Pullar, M. C. M. Vissers and C. C. Winterbourn, IUBMB Life, 2000, 50, 259–266.
6. T. Yudhistira, S. V. Mulay, Y. Kim, M. B. Halle and D. G. Churchill, Chem. – Asian J., 2019, 14, 3048–3084.
7. D. Wu, L. Chen, Q. Xu, X. Chen and J. Yoon, Acc. Chem. Res., 2019, 52, 2158–2168.
8. S. Lin, C. Ye, Z. Lin, L. Huang and D. Li, Talanta, 2024, 268, 125264.
9. C. Alexander, Z. Guo, P. B. Glover, S. Faulkner and Z. Pikramenou, Chem. Rev., 2025, 125, 2269–2370.
10. J.-C. G. Bünzli and S. V. Eliseeva, in Lanthanide Luminescence: Photophysical, Analytical and Biological Aspects, ed. P. Hänninen and H. Härmä, Springer Berlin Heidelberg, Berlin, Heidelberg, 2011, pp. 1–45.  DOI:10.1007/4243_2010_3.
11. S. I. Weissman, J. Chem. Phys., 1942, 10, 214–217.
12. A. P. S. Samuel, J. Xu and K. N. Raymond, Inorg. Chem., 2009, 48, 687–698.
13. A. D'Aléo, F. Pointillart, L. Ouahab, C. Andraud and O. Maury, Coord. Chem. Rev., 2012, 256, 1604–1620.
14. D. Mouchel Dit Leguerrier, R. Barré, J. K. Molloy and F. Thomas, Coord. Chem. Rev., 2021, 446, 214133.
15. N. Kwon, D. Kim, K. M. K. Swamy and J. Yoon, Coord. Chem. Rev., 2021, 427, 213581.
16. Y. Xiao, R. Zhang, Z. Ye, Z. Dai, H. An and J. Yuan, Anal. Chem., 2012, 84, 10785–10792.
17. Y. Xiao, Z. Ye, G. Wang and J. Yuan, Inorg. Chem., 2012, 51, 2940–2946.
18. X. Liu, Z. Tang, B. Song, H. Ma and J. Yuan, J. Mater. Chem. B, 2017, 5, 2849–2855.
19. Z. Wang, Y. Huang, B. Song, Y. Shi and J. Yuan, Inorg. Chem., 2025, 64, 8685–8693.
20. H. Ma, B. Song, Y. Wang, D. Cong, Y. Jiang and J. Yuan, Chem. Sci., 2017, 8, 150–159.
21. H. Ma, B. Song, Y. Wang, C. Liu, X. Wang and J. Yuan, Dyes Pigm., 2017, 140, 407–416.
22. E. Del Giorgio and T. J. Sørensen, Molecules, 2020, 25, 1959.
23. A. Picot, A. D'Aléo, P. L. Baldeck, A. Grichine, A. Duperray, C. Andraud and O. Maury, J. Am. Chem. Soc., 2008, 130, 1532–1533.
24. N. Hamon, A. Roux, M. Beyler, J. C. Mulatier, C. Andraud, C. Nguyen, M. Maynadier, N. Bettache, A. Duperray, A. Grichine, S. Brasselet, M. Gary-Bobo, O. Maury and R. Tripier, J. Am. Chem. Soc., 2020, 142, 10184–10197.
25. K. P. Malikidogo, T. Charnay, D. Ndiaye, J. H. Choi, L. Bridou, B. Chartier, S. Erbek, G. Micouin, A. Banyasz, O. Maury, V. Martel-Frachet, A. Grichine and O. Sénèque, Chem. Sci., 2024, 15, 9694–9702.
26. D. Akl, L. Bridou, M. Hojorat, G. Micouin, S. R. Kiraev, F. Riobé, S. Denis-Quanquin, A. Banyasz and O. Maury, Eur. J. Inorg. Chem., 2024, 27.
27. X. L. Armesto, M. Canle L, M. I. Fernández, M. V. García and J. A. Santaballa, Tetrahedron, 2000, 56, 1103–1109.
28. V. Lebrun, J.-L. Ravanat, J.-M. Latour and O. Sénèque, Chem. Sci., 2016, 7, 5508–5516.
29. M. Le Fur, M. Beyler, E. Molnar, O. Fougere, D. Esteban-Gomez, G. Tircso, C. Platas-Iglesias, N. Lepareur, O. Rousseaux and R. Tripier, Inorg. Chem., 2018, 57, 2051–2063.
30. M. Le Fur, E. Molnar, M. Beyler, O. Fougere, D. Esteban-Gomez, O. Rousseaux, R. Tripier, G. Tircso and C. Platas-Iglesias, Inorg. Chem., 2018, 57, 6932–6945.
31. B. Chartier, N. Hamon, D. Akl, A. Sickinger, L. Corne, G. Micouin, A. Banyasz, O. Maury, S. Erbek, V. Martel-Frachet, A. Grichine, M. Beyler, O. Sénèque and R. Tripier, Chem. – Eur. J., 2025, 31, e02044.
32. A. T. Bui, A. Roux, A. Grichine, A. Duperray, C. Andraud and O. Maury, Chem. – Eur. J., 2018, 24, 3408–3412.
33. V. Lebrun, J.-L. Ravanat, J.-M. Latour and O. Sénèque, Chem. Sci., 2016, 7, 5508–5516.
34. The difference in the enhancement factor is explained by the initial emission spectra. TbL₂ is almost non emissive and therefore the initial Tb emission intensity is very low, upon oxidation the Tb luminescence is restorered resulting in a more important enhancement factor.

---

Footnote

† Present address: Faculté des Sciences et Technologies, Université de Goma, B.P. 204 Goma, R. D. Congo.

---