Bright Circularly Polarized Electrochemiluminescence from Heterobinuclear Ir^III–Au^I Enantiomers

Abstract

The development of efficient circularly polarized electrochemiluminescence (CP‐ECL) probes is still at its infancy and examples are still very limited. Yet, their achievement would enable gathering a readout that carries privileged information on the probe's chiral environment by monitoring luminescence polarization bias with high signal‐to‐noise ratio. Notwithstanding, this is a highly challenging task and requires judicious chemical engineering of chiral ECL‐active emitters. Herein, we aim at expanding the palette of CP‐ECL luminophores by presenting a novel class of enantiopure heterobinuclear Ir(III)–Au(I) complexes, which are investigated thoroughly by means of chemical, structural, and (chiro‐)optical techniques. The ground and excited state properties are also elucidated by using density functional theory (DFT) approaches including spin‐orbital coupling (SOC) perturbation. The chiral‐at‐metal complexes display luminescence with a polarization bias of the emitted light that is function of the helical arrangement of the coordination sphere around the Ir(III) center. Overall, the photo‐ and electro‐active complexes unraveled in this work combine unparallelly high photoluminescence quantum yield in the orange region, excellent circularly polarized luminescence (CPL) brightness up to 4.5 M⁻¹ cm⁻¹ with a notable ECL activity. Finally, these features provide emitters with CP‐ECL efficiency that encompass remarkably by a factor 3.5 that of the well‐known benchmark tris‐(2,2′‐bipyridyl)ruthenium(II).

Efficient and bright circularly polarized electrochemiluminescence (CP‐ECL) is achieved in a class of enantiopure hetobinuclear iridium(III)–gold(I) complexes owing to the combination of high ECL efficiency with stability, good emission dissymmetry factor and high brightness, thus paving the way to the next‐generation of bright CP‐ECL probes.

Introduction

Electrochemiluminescence (ECL) is a radiative process triggered by an exergonic electron transfer reaction between species that are generated electrochemically.^{[ 1} , ^{2 ]} Compared to more classical photo‐luminescence (PL) techniques, ECL possesses the advantage of providing an optical readout with low background signal, owing to the absence of possibly interfering background emission.^{[ 3} , ^{4 ]} Nowadays, ECL represents an outstanding (bio)analytical tool with very precise spatiotemporal resolution and high sensitivity that has found real‐market application in immunoassays for clinical diagnostics and is a fascinating tool in ECL‐based microscopy imaging.^{[ 5} , ⁶ , ⁷ , ⁸ , ⁹ , ^{10 ]} In this framework, the archetypal [Ru(bpy)₃]²⁺ and its related analogues are by far the most investigated and commercially available ECL‐active luminophores.^{[ 1} , ¹¹ , ^{12 ]} Though they possess limited tunability in terms of redox potentials and emission wavelength and suffer from relatively low values of photoluminescence quantum yield (PLQY) that typically do not exceed 5%. Overall, these limitations represent severe drawbacks in view of further developing ultrasensitive analytical techniques that go beyond the current state‐of‐the‐art. More recently, cyclometalated Ir(III) derivatives have emerged providing an interesting alternative as ECL luminophores,^{[ 13} , ¹⁴ , ¹⁵ , ¹⁶ , ¹⁷ , ^{18 ]} due to the combination of enhanced modulation of redox potentials over a wider electrochemical range, tunability of the emission spectra, ease of functionalization and excellent PLQY.^{[ 19} , ²⁰ , ²¹ , ^{22 ]}

Besides the typical ECL characteristics encompassing emission wavelength, light intensity, and Faradaic current, additional features may emerge, and circular polarization of the emitted light arises as an added value readout when chiral luminophores are employed. Indeed, the information gathered by monitoring polarization bias will provide an additional degree of freedom over techniques based on a more classical emission intensity. In this context, CPL, which is the differential emission of circularly polarized light with opposite handedness upon generation of an emitting excited state, has been already widely used to track the interactions between specific probes and different analytes, such as proteins,^{[ 23} , ^{24 ]} amino acids,^{[ 25} , ²⁶ , ^{27 ]} DNA,^{[ 28 ]} chiral ions,^{[ 29} , ³⁰ , ^{31 ]} and small molecules.^{[ 32 ]} More recently, it has been used for detecting species in chiral environment and complex matrices, often presenting different CPL profiles and intensities, by means of CPL‐based microscopy.^{[ 33} , ^{34 ]} The extent of the circular polarization of the emitted light is quantified by the PL dissymmetry factor, g _PL, defined by the following Equation (1):

$$g_{PL}=\frac{2\left(I_L-I_R\right)}{I_L+I_R}$$ (1)

where I _L and I _R are the intensity of left‐ and right‐ circularly polarized light, respectively. Values closer to the theoretical limits ±2 indicate a high degree of circular polarization. For molecular species, g _PL is often in the order of |10⁻⁴|–|10⁻³|,^{[ 35 ]} with a few notable exceptions, such as in the case of some lanthanide or (assembled) transition metal complexes.^{[ 36 ]}

In the vast majority of the examples, CPL is the result of a photoexcitation process (sometimes named CP‐PL), but it can also be triggered by an ECL process, giving rise to CP‐ECL. On one hand, this novel technique may be exploited to determine the stereochemistry, including absolute configuration and enantiopurity, of the ECL luminophore. Even more, it would allow one to gather privileged information about the chiral species that closely interact with the ECL probe with high selectivity. The latter aspect is of particular interest when developing ECL‐based bioassays, as biologically relevant species, such as bio‐(macro)molecules and bioactive substances, are often chiral and typically present in homochiral form.

CP‐ECL is still a rather unexplored technique and it was first demonstrated by Blok et al.^{[ 37 ]} To date, it has been studied for chiral pyrene‐appended organic molecules,^{[ 38} , ^{39 ]} for an analogue of [Ru(bpy)₃]²⁺ complex,^{[ 37} , ^{40 ]} and for Au nanoclusters,^{[ 41 ]} showing g _ECL dissymmetry factor (defined analogously to the g _PL) in the order of 10⁻³–10⁻².

For the first time, we herein aim at merging the great advantages provided by ECL‐active Ir(III) emitters with their good CPL activity arising from chiral‐at‐metal enantiopure derivatives. This approach provides state‐of‐the‐art CP‐ECL luminophores that display the combination of high ECL efficiency with stability, good emission dissymmetry factor, and high brightness, paving the way to the next‐generation of bright CP‐ECL probes.

Results and Discussion

Tris‐cyclometalated Ir(III) complexes possess either Λ or Δ configuration depending on the helical arrangement of the ligand scaffolds around the metal center, and the vast majority of the examples deal with a racemic mixture. The schematic synthetic pathway employed for the synthesis of the enantiopure heterobimetallic complexes Δ‐IrAu‐3 and Λ‐IrAu‐3 is depicted in Scheme 1. Full details about the experimental synthetic procedures along with the chemical characterization carried out by means of ¹H, ¹³C{¹H} NMR spectroscopy and high‐resolution electrospray ionization mass spectrometry (HR‐ESI‐MS) are available in the Supporting Information (see Figure S1–S8).

Scheme 1.

Synthetic pathway employed for the preparation of enantiopure complexes Δ‐IrAu‐3 and Λ‐IrAu‐3.

The synthesis of the starting chloro‐bridged dimer as a racemic mixture, rac‐[Ir(pbtz)₂(μ‐Cl)]₂, where (pbtz) is the cyclometalating 2‐phenylbenzothiazole ligand, was carried out following procedures reported elsewhere.^{[ 42 ]} The enantiomer resolution in either Λ,Λ or Δ,Δ‐[Ir(pbtz)₂(μ‐Cl)]₂ isomer was based on the strategy proposed by Meggers and co‐workers to achieve enantiopure bis‐cyclometallated iridium(III) complexes.^{[ 43 ]} The chiral auxiliary ligand (S)‐2‐(2′‐hydroxyphenyl)‐4‐isopropyl‐2‐thiazoline, namely (S)‐Aux, was synthesized as previously reported in the literature.^{[ 44 ]} Upon coordination reaction of rac‐[Ir(pbtz)₂(μ‐Cl)]₂ with (S)‐Aux in refluxing CH₂Cl₂ and with NEt₃ as the base, the so‐formed mixture of diastereomers (Δ_Ir,S _C)‐[Ir(pbtz)₂(Aux)] and (Λ_Ir,S _C)‐[Ir(pbtz)₂(Aux)] was resolved with classical column chromatography. The chemical analyses of the obtained compounds are in agreement with literature data.^{[ 43 ]} Reaction with a 2 M HCl aqueous solution with either (Δ_Ir,S _C)‐[Ir(pbtz)₂(Aux)] or (Λ_Ir,S _C)‐[Ir(pbtz)₂(Aux)] yielded the μ‐chloro bridged dimer [Ir(pbtz)₂(μ‐Cl)]₂ with (Λ,Λ) or (Δ,Δ) stereochemistry, respectively, which were then used as enantiopure starting materials for the next synthetic steps.

The successful synthesis of the enantiopure dimers was also confirmed by solving the single crystal X‐ray structure of the derivative (Δ,Δ)‐[Ir(pbtz)₂(μ‐Cl)]₂ obtained by slow diffusion of n‐hexane into a CH₂Cl₂ solution (see ORTEP plot in Figure S9 and data in Table S1 of the Supplementary Information). Subsequently, chlorine abstraction was carried out by reaction with AgPF₆ in MeOH at room temperature followed by the addition of the zwitterionic metalloligand [(IPr)Au^I(IMesAcac)], where IPr is 1,3‐(2,6‐diisopropylphenyl)‐2H‐imidazol‐2‐ylidene and IMesAcac is 1,3‐dimesityl‐5‐acetylimidazol‐2‐ylidene‐4‐olate,^{[ 45} , ^{46 ]} in agreement with our previously reported procedure (Scheme 1).^{[ 47} , ^{48 ]} The target enantiopure complexes Δ‐IrAu‐3 and Λ‐IrAu‐3 were obtained in good reaction yield as PF₆ ⁻ salt upon precipitation with a saturated aqueous solution of KPF₆. For comparison purposes, the corresponding racemic mixture rac‐IrAu‐3 was prepared as well. Single crystals of Λ‐IrAu‐3 suitable for X‐ray diffraction analysis were grown by slow diffusion of n‐hexane into an acetone solution of the complex and the ORTEP plot is displayed in Figure 1 (see Table S2 for crystallographic refinement parameters). The iridium(III) center adopts a distorted octahedral geometry with the two nitrogen atoms of the pbtz ligand arranged in a N,N‐trans fashion. The coordination sphere is completed by the κ² O,O‐[(IPr)Au^I(IMesAcac)] metalloligand. The near‐zero Flack parameter value of 0.015(1) confirmed the absolute configuration Λ around the Ir atom.

Figure 1.

ORTEP plot of the enantiopure complex Λ‐IrAu‐3 with thermal ellipsoids shown at a 50% probability level and obtained by single‐crystal X‐ray diffractometric analysis (CCDC 2415225). Acetone solvent molecules and PF₆ ⁻ anions are omitted for clarity. Selected bond lengths (Å): Ir(1)–N(1) = 2.058(5), Ir(1)–N(2) = 2.043(5), Ir(1)–C(9) = 1.995(6), Ir(1)–C(22) = 2.012(6), Ir(1)–O(1) = 2.184(4), Ir(1)–O(2) = 2.152(4), Au(1)–C(28) = 2.014(6), Au(1)–C(50) = 2.023(6). Atom color code: carbon (gray), nitrogen (light blue), iridium (dark blue), sulfur (yellow).

The optical properties of the complex rac‐IrAu‐3 were first investigated in dilute CH₂Cl₂ solution in both air‐equilibrated and degassed conditions at a concentration of 3 × 10⁻⁵ M. The spectra are displayed in Figure 2 and the corresponding photophysical data are listed in Table 1. The recorded decay traces can be found in Figure S10. In the electronic absorption spectrum, the complex displays intense (ε = 3.1–3.6 × 10⁴ M⁻¹ cm⁻¹) and partially structured bands in the region λ _abs = 250–350 nm ascribed to the singlet ligand centered (¹LC) transitions involving the cyclometalated (pbtz) ligands and the aromatic moieties of the NHC scaffolds. At lower energy in the region λ _abs > 350 nm, the absorption bands with much lower intensity (ε = 1.5–0.4 × 10⁴ M⁻¹ cm⁻¹) can be assigned with confidence to the electronic processes with admixed ligand‐to‐ligand and metal‐to‐ligand charge transfer (¹LLCT/¹MLCT) character. Finally, the weak band with maximum at λ _abs = 526 nm and with ε = ca. 500 M⁻¹ cm⁻¹ is ascribed to the formally spin‐forbidden triplet‐manifold MLCT (³MLCT) that becomes partially allowed due to the presence of the strong SOC effect exerted by the heavy metal atom (Table S5). Upon photoexcitation, a sample of the complex in CH₂Cl₂ displays intense PL in the orange region with a structured profile peaking at λ _em = 545, 589 nm and with PLQY of 65.2%. This emission arises from a ³MLCT excited state mainly localized onto the Ir(pbtz)₂ scaffold as suggested by the relatively long‐lived excited state lifetime of τ = 1.68 µs, the large value of the radiative rate constant, k _r, estimated to be as large as 3.9 × 10⁵ s⁻¹ and the efficient quenching exerted on both PLQY and lifetime observed by the triplet dioxygen molecules in air‐equilibrated samples (Table 1).

Figure 2.

Left: Electronic absorption (solid trace) and PL emission (dashed trace) spectra recorded for complex rac‐IrAu‐3 in CH₂Cl₂ at a concentration of 3 × 10⁻⁵ M at room temperature. The emission spectrum was recorded upon λ _exc = 450 nm. Right: UV–vis absorption spectra computed for complex IrAu‐3 in CH₂Cl₂ as the solvent with (empty symbols) and without (filled symbols) spin‐orbit (SO) perturbation. Curves obtained for Δ and Λ enantiomer are shown with blue and red colors, respectively.

Table 1.

Photophysical data recorded for complex rac‐IrAu‐3 in diluted CH₂Cl₂ solution at a concentration of 3 × 10⁻⁵ M.

λ max (ε) [nm, (103 M−1 cm−1)]	λ em [nm]	PLQY a) (%)	τ a) [ns]	PLQY b) (%)	τ b) [µs]	k r [105 s−1]	k nr [105 s−1]
265 (36.2), 300 (31.1), 314 (31.9), 327 (33.1), 352sh (15.2), 395sh (7.2), 432 (5.4), 463 (4.4), 526 (0.5)	545, 589	13.3	331	65.2	1.68	3.9	2.1

sh denotes a shoulder.

^a) air‐equilibrated samples;

^b) degassed samples. k_r = PLQY/τ and k_nr = (1 – PLQY)/τ.

The structural and optical properties of IrAu‐3 were investigated employing density functional theory (DFT) and time‐dependent DFT (TD‐DFT) formalism, respectively. The molecular structures of both enantiomers were fully optimized employing the B3LYP functional. Further computational details can be found in the experimental section of the Supporting Information. Overall, the computed structures and the absorption spectra are in good agreement with the experimental data (see Figures 2 and S11–S14, Table S3 and Supporting Discussion #1 of the Supporting Information).

To gain deeper insights into the excited states involved in the emission, the lowest triplet excited states of the complex were fully optimized. On the first triplet potential energy surface (T₁ PES), a first minimum, namely T_1a, results from the geometry optimization of the Franck–Condon T₁ state and displays mixed L₁L₁CT/MLCT (with L₁ = pbtz) character (Figure 3). A second minimum on the T₁ PES, namely T_1b, was found which has a mixed ML₂CT/L₁L₂CT character, L₂ being the acac ligand (Figure 3). The computed emission wavelength is λ _em,theo = 588 and 613 nm for T_1a and T_1b, respectively (Table 2). T_1a is more stable than T_1b by 0.085 eV according to the comparison of the stabilization energies, and the former corresponds most likely to the emitting state observed experimentally in good agreement with the experimental findings (Table 1). Inclusion of SOC affects the properties of T_1a and T_1b to a smaller extent and yields a bathochromic shift of a few nanometers of the emission wavelength and a small splitting of the three components of the triplet states (Table 2). The computed rate constants are consistent with the experimental ones.

Figure 3.

Electron density difference maps (EDDMs) between ground and excited states at the triplet states optimized geometries for T_1a (left) and T_1b (right) computed for complex Λ‐IrAu‐3. Electronically depleted and enriched areas are colored in red and green, respectively.

Table 2.

Distortion, emission and stabilization energies in eV and emission wavelength in nm calculated for triplet states at T_1a and T_1b geometry. Emission energies, emission wavelengths, and radiative rate constants obtained for the three triplet substates in the frame of the perturbative SOC formalism are provided in [eV], [nm] and [s⁻¹], respectively. See Supporting Information for energy definition.

	T1a	T1b
Edist	0.269	0.439
Eem	2.108	2.023
Estab	2.377	2.462
λem	588	613
sub‐state	Eem a) [eV]
Tn,I	2.082	2.001
Tn,II	2.083	2.002
Tn,III	2.093	2.004
λem a) [nm]
Tn,I	596	620
Tn,II	595	619
Tn,III	592	619
k r a) [s−1]
Tn,I	4.87 104	4.28 103
Tn,II	4.75 103	1.25 105
Tn,III	1.21 105	1.53 105
Boltzmann‐average k r	5.35 104	1.74 105

^a) computed including SOC perturbation

Electronic circular dichroism (ECD) spectra of diluted CH₂Cl₂ samples of Δ‐, Λ‐, and rac‐IrAu‐3 complexes are reported in Figure 4a and the corresponding chiroptical data are listed in Table 3. The two enantiomers of IrAu‐3 show a clear polarization bias with mirror image spectral profiles. As expected, the racemic sample does not show any significant ECD signal. For the enantiopure samples, the ECD spectra present intense bands both in the higher energy range at λ _abs = 250–350 nm, which can be ascribed to ¹LC transitions mainly, and in the lower energy part of the spectrum, of main ¹LLCT/¹MLCT character (see above). On the lower energy side, in the region of λ _abs ≈ 460 nm, the corresponding ECD band is positive and negative for the Δ and Λ enantiomer, respectively, in agreement with previous literature data.^{[ 49} , ⁵⁰ , ⁵¹ , ⁵² , ⁵³ , ⁵⁴ , ⁵⁵ , ⁵⁶ , ⁵⁷ , ⁵⁸ , ⁵⁹ , ⁶⁰ , ^{61 ]} Overall, absorbance dissymmetry factors calculated on the main bands, g _abs, defined as by Equation (2) are determined to be ca. 2 × 10⁻³ (Table 3).

$$g_{abs}=\frac{2\mathrm{\Delta }\epsilon \left(\lambda \right)}{\epsilon \left(\lambda \right)}=\frac{2\left({\epsilon }_{\mathrm{L}}\left(\lambda \right)-{\epsilon }_R\left(\lambda \right)\right)}{{\epsilon }_L\left(\lambda \right)+{\epsilon }_R\left(\lambda \right)}$$ (2)

Figure 4.

a) ECD spectra recorded for complexes Δ‐IrAu‐3 (—), Λ‐IrAu‐3 (—), and rac‐IrAu‐3 (—) in CH₂Cl₂ at a concentration of 1 × 10⁻⁵ M; b) computed ECD spectra in CH₂Cl₂ without SO perturbation and computed reduced rotatory strength (bars); c) CP‐PL (filled symbols) and CP‐ECL (empty symbols) spectra normalized for the intensity at the peak maximum recorded for complexes Δ‐IrAu‐3 (—), Λ‐IrAu‐3 (—), and rac‐IrAu‐3 (—). CP‐PL spectra were recorded in CH₂Cl₂ at a concentration of 1 × 10⁻⁵ M upon excitation at λ _exc = 365 nm; CP‐ECL spectra were recorded in degassed CH₃CN at a concentration of 0.25 mM with 10 mM BPO and 0.1 M TBAP under imposing a cathodic potential ‐2.27 V versus Ag wire. A GC disk (diameter = 3 mm), a Pt wire and an Ag wire were used as working electrode, counter‐electrode and pseudo‐reference electrode, respectively; d) ECL spectra recorded for samples of Δ‐IrAu‐3 (—), Λ‐IrAu‐3 (—) and rac‐IrAu‐3 (—) complexes under the same experimental conditions as CP‐ECL spectra in Figure 4c.

Table 3.

Photophysical and chiroptical data along with CP‐ECL efficiency values measured for compounds Δ‐IrAu‐3 and Λ‐IrAu‐3.

compound	λabs [nm] (g abs × 10−3)	g PL (×10−3)	g ECL (×10−3)	B CPL a) [M−1 cm−1]	Φ CP − ECL, r (×10−3)
Λ‐IrAu‐3	462 (–2.6) 333 (+2)	−3.1 ± 0.6	−2.2 ± 0.4	4.5	1.1
Δ‐IrAu‐3	462 (+2.3) 333 (–2)	+2.3 ± 0.6	+2.7 ± 0.4	3.3	1.3

^a) determined at the maximum of the lowest‐lying energy band corresponding to the ¹MLCT absorption.

The computed ECD spectra of the complexes Δ−IrAu‐3 and Λ−IrAu‐3 are represented in Figure 4b. One can clearly distinguish three regions of intense ECD activity, namely at λ = 436, 348, and 330 nm, assigned to S₀ → S₁, S₀ → S₆ and S₀ → S₉‐S₁₀ transitions, respectively (see Table S4 and S6). The shape of the computed ECD spectra is nearly identical to the experimental one with a small hypsochromic shift of the lowest band that is computed at λ = 436 nm versus 463 nm in the experimental spectrum (cf. Figure 4a,b). This is explained by the fact that the employed method for the computation of the ECD spectra neglected the SOC perturbation (see above).

The polarization bias of the lowest‐energy ECD bands is reflected in the signs of the CP‐PL spectra of the CH₂Cl₂ samples of the investigated complexes as shown in Figure 4c. Indeed, upon excitation at λ _exc = 365 nm the Δ and the Λ enantiomers display a positive and a negative broad CP‐PL band, respectively, which retraces the unpolarized emission spectrum (Figures 2a and 4c). On the other hand, the racemic sample displays no neat CP‐PL emission, as expected. The g _PL values were determined to be + 2.3 ± 0.6 × 10⁻³ and –3.1 ± 0.6 × 10⁻³ for Δ and Λ enantiomer, respectively, which are similar to the measured g _abs values (Table 3) and in line with the g _PL reported for other Ir(III) complexes (Figure 5 and Table 3).^{[ 49} , ⁶⁰ , ⁶² , ⁶³ , ⁶⁴ , ^{65 ]}

Figure 5.

Circularly polarized luminescence dissymmetry factor as a function of emission wavelength obtained by means of photo‐ (g _PL, empty symbols) and ECL (g _ECL, filled symbols) determined for samples of complex Δ‐IrAu‐3(—), Λ‐IrAu‐3 (—), and rac‐IrAu‐3 (—) in CH₂Cl₂ at a concentration of 1 × 10⁻⁵ M (CP‐PL) and degassed CH₃CN solution at a concentration of complex of 0.25 mM with 10 mM BPO and 0.1 M TBAP under imposing a cathodic potential ‐2.27 V versus Ag wire (CP‐ECL). A GC disk (diameter = 3 mm), a Pt wire and an Ag wire were used as working electrode, counter electrode, and pseudo‐reference electrode, respectively.

For practical application, to further compare different classes of CPL chromophores, it is useful to determine the CPL brightness, namely B _CPL, following the expression that was recently introduced by some of us,^{[ 66 ]} and defined by the following Equation (3):

$$B_{CPL}=\frac{1}{2}\epsilon \left({\lambda }_{max}\right)\times \mathrm{PLQY}\times |g_{PL}|$$ (3)

where ε(λ _max ) is the molar extinction coefficient determined at the wavelength maximum of a given absorption band. Remarkably, Λ‐IrAu‐3 and Δ‐IrAu‐3 displayed B _CPL values that are 3.3 and 4.5 M⁻¹ cm⁻¹, respectively, thanks to the combination of high PLQY and a rather high g _PL factor. It is noteworthy that these values are very close to the highest ones to date for Ir(III) complexes, as recently reported by Thompson and co‐workers,^{[ 58 ]} and superior by more than five times to the median value typically observed for this class of derivatives, i.e., B _CPL median of ca. 0.7 M⁻¹ cm⁻¹.^{[ 59} , ^{66 ]} These B_CPL values are lower than those shown by other CP‐ECL active materials, such as pyrene excimers (43.2 M⁻¹ cm^{−1 [ 39 ]} and 90 M⁻¹ cm^{−1 [ 38 ]}). However, B_CPL values are not straightforwardly correlated to the CP‐ECL efficiency of the compounds (see below), given the different nature of the excitation process (PL versus ECL).

To investigate their electrochemical properties and test the stability of the electrogenerated species, the racemic mixture of the rac‐IrAu‐3 complex was first characterized by cyclic voltammetry (CV) using a solution of 1 mM rac‐IrAu‐3 in CH₃CN containing 0.1 M of tetrabutylammonium perchlorate (TBAP) as the supporting electrolyte. Details concerning the electrochemical setup and experiments can be found in the Supporting Information. Figure S15 displays the CV of rac‐IrAu‐3 recorded in the full potential window at a scan rate of 0.1 V s⁻¹ in a N₂ purged solution to remove the majority of dissolved O₂. In the positive‐going scan, a main mono‐electronic oxidation, O₁ , is observed at $E_{O_1}^0$ = +1.13 V versus Ag P‐RE, which is related to the oxidation of the iridium metal center, formally from Ir^III to Ir^IV. This process is reversible and displays a peak‐to‐peak separation of ∼ 80 mV, which is slightly larger than the theoretical value expected for a fast one‐electron transfer reaction. The incorporation of the pbtz ligand on the Ir metal slightly affects its oxidation shifting its potential toward a more positive value (∼ 160 mV) than that of a structurally‐related complex bearing a phenylisoquinoline ligand, whose electrochemical characterization was recently reported.^{[ 47} , ^{67 ]} This main oxidation peak is preceded by a smaller pre‐peak at ca. +0.8 V versus Ag P‐RE that was also observed in a series of comparable complexes and that tends to increase when first scanning the potential toward the negative bias reaching increasingly more negative values before reverting the potential scan to investigate the anodic counterpart.^{[ 47} , ^{67 ]} In the negative‐going scan, four successive reduction processes, R_1‐4 , are recorded in the investigated potential window before the discharge of the solvent/electrolyte occurs. The first reduction process, which is preceded by a weak shoulder at –1.77 V versus Ag P‐RE, appeared as partially reversible with a peak potential, $E_{p,R_1}$, at –1.9 V versus Ag P‐RE. However, when the potential is reverted immediately after the first reduction peak, reversibility is attained with a $E_{R_1}^0$= –1.85 V versus Ag P‐RE. R₁ is a monoelectronic process that is immediately followed by three cathodic waves, R_2‐4 , peaking at –2.14, –2.25, and –2.41 V versus Ag P‐RE. These processes appeared to be partially reversible as evidenced during the backward scan. It was reported that these cathodic waves might be reversible or irreversible depending on the ligands.^{[ 47} , ^{67 ]} To highlight the reversibility of the main electrochemical processes, O₁ and R₁ , a CV was also reported in Figure S15 (red trace). These two processes are directly involved in the activation of ECL phenomenon as they represent the electrochemical equivalent of the optical HOMO‐LUMO band gap. The electrochemical characterization shows that both reduced and oxidized forms of the IrAu‐3 complex were stable in solution during at least the time scale of the voltammetric experiments. This allows investigating its ECL and CP‐ECL properties.

The cathodic ECL feasibility was evaluated following the reductive–oxidation pathway in degassed CH₃CN solution containing 1 mM of rac‐IrAu‐3, 0.1 M of TBAP and 10 mM of benzoyl peroxide (BPO), which acted as the sacrificial coreactant. BPO is readily reduced at a mild potential of ca. –0.5 V versus Ag P‐RE with a very broad cathodic current. This one‐electron transfer reaction process is concerted with a bond cleavage to afford a stoichiometric amount of Ph‐CO₂ ^– anion and Ph‐CO₂ ^• radical.^{[ 68} , ^{69 ]} The latter benzoyloxy radical is a strong oxidant with a reported potential of +1.36 V versus Fc/Fc⁺.^{[ 68} , ^{70 ]} In Figure S16, no ECL was observed during or after the electrochemical reduction of BPO until the applied potential enabled the reduction of rac‐IrAu‐3. Once the potential reaches the above‐mentioned shoulder at –1.77 V versus Ag P‐RE, the flux of photons generated by ECL was immediately detected by the PMT detector. The corresponding ECL intensity rapidly increases with the maximum at about –2.04 V before fading at more negative applied potentials. Here, the ECL mechanism in action is governed by the highly exergonic redox reaction taking place between the reduced bimetallic complex and Ph‐CO₂ ^• radical that co‐exist simultaneously inside the reaction‐diffusion layer in the vicinity of the electrode surface. By comparison with the ECL reference system [Ru(bpy)₃]²⁺/BPO in the same experimental conditions, the relative ECL efficiency Φ _ECL,r was evaluated.^{[ 71} , ^{72 ]} Our findings showed that a value of 98% for Φ _ECL,r of IrAu‐3 was achieved, indicating an ECL efficiency nearly identical to the model [Ru(bpy)₃]²⁺ luminophore that is widely used in analytical chemistry and imaging as mentioned above.^{[ 1} , ¹¹ , ^{12 ]} The corresponding ECL spectrum was also recorded through an optical fiber‐equipped spectrophotometer and compared to the corresponding PL profile (Figure S17). These two spectra are almost matching and feature the same peak pattern with two major contributions at λ _em = 545 and 589 nm confirming that the same excited state of the complex is reached upon either PL excitation or ECL trigger. The ECL spectra intensity was investigated as a function of the applied potential in the range from –1.5 to –2.7 V versus Ag P‐RE with an increment of 100 mV between each spectrum (Figure S18). As a result, the ECL spectra can be recorded as long as the applied potential is more cathodic than –1.7 V versus Ag P‐RE, reaching its maximum at –2.2 V versus Ag P‐RE. These results obtained by imposing a constant potential are self‐consistent with the ECL data collected in voltammetric mode (Figure S16).

Prompted by these results, the generation of a CPL bias by means of an ECL trigger was subsequently investigated. The CP‐ECL data were collected following a previously described protocol.^{[ 39 ]} Briefly, a home‐made three‐electrode electrochemical cell integrated in a quartz cuvette^{[ 67} , ⁷³ , ^{74 ]} was placed inside a homemade CPL instrument and the corresponding CP‐ECL signal was recorded in situ upon electrochemical activation so that circular polarization is solely related to the ECL process (i.e., without photoexcitation). To this aim, a 0.25 mM solution of either Δ‐IrAu‐3, Λ‐IrAu‐3 or rac‐Au‐3 was prepared in CH₃CN with 10 mM BPO and 0.1 M TBAP. For this series of experiments, an applied potential of –2.27 V versus Ag P‐RE was selected. In these experimental conditions, the circularly polarized ECL signal was stable and could be recorded for both enantiomers. The typical acquisition time was 190 s per spectrum and at least eight successive spectra were collected and averaged to achieve a good signal‐to‐noise ratio. Under these conditions, the CP‐ECL spectra shown in Figure 4c exhibit a comparable noise level to that obtained via photo‐induced CPL, which appears very satisfactory. The CP‐ECL spectra of Δ‐IrAu‐3 and Λ‐IrAu‐3 (Figure 4c, blue and red traces, respectively) are indeed mirror images with the Δ enantiomer exhibiting a positive CP‐ECL with a maximum at λ _em = 550 nm, whereas the Λ enantiomer shows a minimum with negative CP‐ECL. As reported in previous studies,^{[ 38} , ³⁹ , ^{40 ]} the relative positive and negative CP‐ECL signals displayed by samples of each enantiomer are comparable to the corresponding CP‐PL traces. As a control experiment, the CP‐ECL of rac‐IrAu‐3 (Figure 4c, green trace) was also recorded under the same conditions and it reveals no significant signal with a random (I_L – I_R) signal in the range between λ _em = 450–750 nm. On the other hand, the ECL spectra of the two enantiomers and rac‐IrAu‐3 were virtually identical (Figure 4d). As for ECL emission, CP‐ECL signal can be characterized by a dissymmetry factor, g _ECL, which reflects the predominance of the left‐ and right‐ circularly‐polarized components of the ECL emission for each enantiomer. The values were calculated in the range of the ECL/PL emission band, i.e., between 540 and 620 nm (Figure 5). The average g _ECL value was 2.7 ± 0.4 × 10⁻³ for the Δ‐IrAu‐3 whereas, for Λ‐IrAu‐3, it was –2.2 ± 0.4 × 10⁻³. For the racemic mixture of rac‐IrAu‐3, the average g‐value was negligibly low as expected, being < 1 × 10⁻⁴. It is also worth highlighting that both g _PL and g _ECL values are comparable for a given investigated enantiomer (see Table 3 and Figure 5). Furthermore, to compare the efficiency of different CP‐ECL emitters, it is worth introducing a comprehensive figure of merit, in analogy with similar quantities.^{[ 66} , ⁷⁵ , ⁷⁶ , ^{77 ]} To this aim, one can introduce a CP‐ECL relative efficiency, namely Φ _{CP − ECL, r} , defined by the following equation (Equation 4):

$${\mathrm{\Phi }}_{CP-ECL,r}=\frac{1}{2}{\mathrm{\Phi }}_{ECL,r}\times \left|g_{ECL}\right|$$ (4)

Where Φ _ECL,r is the relative ECL efficiency of the emitter compared to the benchmark [Ru(bpy)₃]²⁺ complex and |g_ECL | is the dissymmetry factor obtained by CP‐ECL measurements. The factor Φ _{CP − ECL, r} can be seen as the fraction of circularly polarized photons per electron injected, using [Ru(bpy)₃]²⁺ complex as a reference to calculate the Φ _ECL,r value of the novel complexes. Considering the values for Φ _ECL,r and |g_ECL | calculated above, the values for the CP‐ECL efficiency Φ _{CP − ECL, r} were determined to be 1.3 × 10⁻³ and 1.1 × 10⁻³ for the Δ‐IrAu‐3 and Λ‐IrAu‐3 complexes, respectively. Previously, we reported the two CP‐ECL properties of a congener of the model [Ru(bpy)₃]²⁺ emitter and the |g_ECL | values were 7 × 10⁻⁴,^{[ 40 ]} yielding a Φ _{CP − ECL, r} value of 3.5 × 10⁻⁴. Remarkably, the CP‐ECL efficiency of the newly reported IrAu‐3 complexes is 3.5 times higher compared to the archetypical [Ru(bpy)₃]²⁺ ECL and CP‐ECL active complex.

Conclusion

In summary, two enantiomers of a chiral‐at‐metal heterobimetallic Ir(III)‐Au(I) complex were designed and synthesized straightforwardly as enantiopure samples along with its racemic counterpart. The complexes displayed intense orange PL and good chiroptical properties and, for the enantiopure samples, both ECD and CPL spectra featured polarization bias that was dependent on the helical chirality of the iridium moiety. Notably, the complexes provided the combination of high CPL brightness, with values that are among the highest reported for iridium emitters, with stable and highly efficient ECL activity. These findings prompted their successful use as bright and efficient CP‐ECL active compounds that outperform the benchmark [Ru(bpy)₃]²⁺ complex. Also, a new figure‐of‐merit, namely CP‐ECL relative efficiency, was proposed to better compare different classes of compounds useful for this still little explored, yet fascinating, technique. Overall, this work unlocks a novel and powerful molecular design strategy towards efficient ECL techniques with a CPL readout that would enable exploring the probe's chiral environment.

Supporting Information

CCDC 2415223 ((Δ,Δ)‐[Ir(pbtz)₂(μ‐Cl)]₂) and 2415225 (Λ‐IrAu‐3) contain the supplementary crystallographic data for this paper. These data can be obtained free of charge from The Cambridge Crystallographic Data Centre via www.ccdc.cam.ac.uk/data_request/cif. The authors have cited additional references within the Supporting Information. ^{[ 76} , ⁷⁷ , ⁷⁸ , ⁷⁹ , ⁸⁰ , ⁸¹ , ⁸² , ⁸³ , ⁸⁴ , ⁸⁵ , ⁸⁶ , ⁸⁷ , ⁸⁸ , ^{89 ]}

Conflict of Interests

The authors declare no conflict of interest.

Supporting information

Supporting information

Supporting information

Ballerini L., Liu M., Arrico L., Voci S., Gourlaouen C., Daniel C., César V., Bellemin‐Laponnaz S., Zinna F., Bouffier L., Polo F., Di Bari L., Sojic N., Mauro M., Angew. Chem. Int. Ed. 2026, 65, e10787. 10.1002/anie.202510787

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.