Highly luminescent carbazole-functionalized tris(tribromophenyl)methyl radicals with stable circularly polarized photoluminescence Authors: Larissa Schöneburg, Markus Gross, Philipp Thielert, Julia Zolg, Mona E. Arnold, Philipp A. Schuster, Bernhard Putz, Sabine Richert, Alexander J. C. Kuehne Published: 2026 ABSTRACT Brominated trityl radicals form stable enantiomers and present promising candidates for molecular qubits in spintronic and quantum technologies. However, their inherently weak photoluminescence (ϕ < 3 %) limits optical spin readout. Here, we report the synthesis of three carbazole-functionalized tris(2,4,6-tribromophenyl)methyl (TTBrM) radicals via Buchwald-Hartwig cross-coupling. These chiral open-shell emitters exhibit high photoluminescence quantum yields of up to ϕ = 72%, comparable to their chlorinated analogues, with red-shifted emission (λPL = 646-688 nm) due to enhanced charge-transfer character. Enantiomeric resolution via chiral chromatography yields stable atropisomers (Trac > 90 °C) with strong circular dichroism and circularly polarized luminescence showing dissymmetry factors (gabs, glum) of order 10⁻⁴. EPR confirms spin localization on the trityl unit (g ≈ 2.006) with phase memory times Tm of 1.5 – 1.6 µs, while donor methylation improves photostability (t1/2 up to 49 min). These bright, chiral TTBrM derivatives offer a significant advance toward stable molecular qubits. Introduction Triphenylmethyl or trityl radicals represent one of the few classes of stable molecules with unpaired (odd) electrons that exhibit photoluminescence1–3. Typically, the trityl radical is halogenated so that the hemispheres above and below the p-orbital are screened at the methyl-carbon, where the unpaired electron resides4,5. The halogenation forces the phenyl rings into a propeller conformation, such that the trityl radicals become chiral (atropisomerism). However, in the case of chlorination, the resolved enantiomeric tris(2,4,6-trichlorophenyl)methyl (TTM) radicals can convert from the left- to the right-handed propeller and vice-versa (see Fig. 1).Fig. 1Previously established stable organic radicals.TTM, TTBrM, and TTM-Cz with ϕ = 88%, and the enantiomerically stable TTBrM-Cz with ϕ = 72%. Previously established stable organic radicals. TTM, TTBrM, and TTM-Cz with ϕ = 88%, and the enantiomerically stable TTBrM-Cz with ϕ = 72%. The thermal barrier for this racemization is so low, that it occurs at room temperature within minutes – just long enough to determine a dissymmetry factor for the circularly polarized photoluminescence (CPL) of glum ≈ 5 × 10−4 6. Bromination yields the tris(2,4,6-tribromophenyl)methyl (TTBrM) radical, which is stable in its axial chirality up to about 60 °C, with similar glum of 7 × 10–4 (see Fig. 1)7. Unfortunately, TTM and TTBrM exhibit only small photoluminescence quantum yields of ϕ = 3.1 – 0.8 %, because the molecules are C3 symmetric and radiative transitions are forbidden8. Nonetheless, TTBrM has been reported to exhibit long coherence times, making it an interesting molecular qubit9. To increase ϕ, TTM has been coupled to electron donating substituents, disrupting the C3 symmetry by producing a charge transfer (CT) excited state10–13, strongly increasing its ϕ to 88 % (Fig. 1)14,15. Rendering this donor unit chiral produces light-emitting radicals with CPL, although with glum only of order 10–4 16. Unfortunately, donor-functionalized TTBrM has not been synthesized to date, due to the reduced reactivity of bromine in radical-mediated nucleophilic aromatic substitution reactions, typically applied to attach the donor unit to a trityl unit. Moreover, the subsequent conversion to the radical is more difficult for TTBrM than for TTM because the required deprotonation of the central methine unit is not as effective with the more electropositive bromides. Accordingly, we have recently established that palladium-catalyzed cross-coupling can be employed to couple the radical, after selective iodine or bromine introduction to the para-site of TTM17–19. Realizing such donor functionalized TTBrM radicals would open up pathways to a new class of stable and highly luminescent trityl propellers. In contrast to previous reports with CPL in donor functionalized TTM radicals16,20,21, the interesting CPL characteristics would not originate from the donor or a surrounding matrix but from the trityl propeller itself. Unfortunately, such donor-functionalized TTBrM radicals have not been synthesized, despite their interesting properties that have to date only been investigated theoretically22–24. Here, we demonstrate that carbazole (Cz)-functionalized TTBrM radicals can be obtained via Buchwald-Hartwig cross-coupling followed by the deprotonation-oxidation mechanisms. The resulting TTBrM-Cz radicals exhibit a strong ϕ of up to 72 % similar to TTM-Cz. Enantiomeric resolution yields stable radicals with CPL and glum of up to 6.5 × 10–4, opening up this new class of light-emitting radicals for further exploration. Results And Discussion Synthesis of donor functionalized TTBrM radicals To generate the carbazole functionalized TTBrM compounds, we start from synthesizing the closed-shell HTTBrM radical precursor by Friedel-Crafts alkylation (the synthesis is described elsewhere)7. Notably, for the preparation of HTTBrM, an alternative protocol employing hexafluorobenzene as the reaction medium, affords markedly improved reproducibility and, owing to the reduced reaction temperature (80 °C versus 120 °C in the melt), minimizes competing side reactions (see SI). In contrast to the related TTM-Cz compound, Buchwald-Hartwig coupling between the TTBrM radical and the carbazole derivatives (Cz, MeCz, Me2Cz) is not successful. Therefore, we perform Buchwald-Hartwig coupling between the Cz-derivatives and the HTTBrM-precursor and subsequently convert to the radical, via the established two-step deprotonation and oxidation sequence (Fig. 2)16,25,26.Fig. 2Donor functionalized TTBrM radicals.Pathway to synthesize carbazole functionalized TTBrM radicals via Buchwald-Hartwig coupling of Cz to the HTTBrM radical precursor, followed by deprotonation and oxidation. The 3,6-positions are successively methylated to adjust the donor strength of the carbazole moiety. Donor functionalized TTBrM radicals. Pathway to synthesize carbazole functionalized TTBrM radicals via Buchwald-Hartwig coupling of Cz to the HTTBrM radical precursor, followed by deprotonation and oxidation. The 3,6-positions are successively methylated to adjust the donor strength of the carbazole moiety. In optimizing this conversion, we find that sterically demanding bases (such as potassium tert-butoxide) are not suitable for deprotonation. By contrast, finely ground potassium hydroxide in combination with 18-crown-6 proves effective. The crown ether complexes the potassium ion and therefore activates the hydroxide ion, enabling deprotonation and formation of the TTBrM-anion under relatively mild conditions27. The oxidation to the radical is straight forward using p-chloranil (see Fig. 2 and the synthetic procedures in the SI). This approach represents a practical route to TTBrM-based radicals bearing electron-rich aryl substituents in high yields > 90 %, expanding the accessible scope of this class of TTBrM compounds. Optical, electronic, and spin properties of the TTBrM radicals To gain deeper insight into the fundamental electronic and photophysical properties of the radicals, a combination of electron paramagnetic resonance (EPR) spectroscopy, optical, electrochemical, and computational studies are employed. Continuous wave (cw)-EPR spectroscopy reveals broad signals without hyperfine splitting, with g values of 2.0063 to 2.0059, see Fig. 3a. The unsubstituted TTBrM radical (without donor moiety) shows a nearly identical value of g = 2.0065. By contrast, TTM and TTM-Cz exhibit hyperfine splitting and show a similarly small shift in g for the Cz donor attachment; however, at slightly lower values of 2.0035 and 2.0032, respectively (see the whole set of cw-EPR spectra in Fig. S1 in the SI)8. The small shift in the g-values upon Cz substitution indicates that the unpaired electron remains localized on the trityl (TTM/TTBrM) core, with negligible delocalization into the donor unit. By contrast, the significant increase in g-value upon halogen substitution (Cl → Br) can be attributed to enhanced spin-orbit coupling (SOC) introduced by the heavier bromine atoms (cf. Fig. S2 in the SI). These findings confirm that the g-value is primarily governed by local SOC effects at the radical center, while distant donor groups such as Cz have only minimal influence unless directly involved in spin delocalization28.Fig. 3Magnetic, computational, optical, and electrochemical data.a cw-EPR signal recorded at the X-band (9.75 GHz) at 295 K of TTBrM-Cz (black), TTBrM-MeCz (blue), TTBrM-Me2Cz (red), and TTM-Cz (grey) as a reference. b Electron spin echo (ESE) detected field-swept EPR spectra recorded at the Q-band (34 GHz) in frozen toluene solution at 80 K. c NTOs of the D1 transition for the three donor-functionalized TTBrM radicals on the PBE1PBE/6-311 + + G** level of theory. d Absorption spectra in cyclohexane solutions (c ~ 10–5 mol L–1) (dashed lines) and normalized photoluminescence spectra in cyclohexane solutions (c ~ 10–5 mol L–1) (solid lines). e Normalized cyclic voltammograms of the radicals in methylene chloride at a scan rate of 10 mV s–1 with [Bu4N]+[PF6]– (0.1 M) as electrolyte. Magnetic, computational, optical, and electrochemical data. a cw-EPR signal recorded at the X-band (9.75 GHz) at 295 K of TTBrM-Cz (black), TTBrM-MeCz (blue), TTBrM-Me2Cz (red), and TTM-Cz (grey) as a reference. b Electron spin echo (ESE) detected field-swept EPR spectra recorded at the Q-band (34 GHz) in frozen toluene solution at 80 K. c NTOs of the D1 transition for the three donor-functionalized TTBrM radicals on the PBE1PBE/6-311 + + G** level of theory. d Absorption spectra in cyclohexane solutions (c ~ 10–5 mol L–1) (dashed lines) and normalized photoluminescence spectra in cyclohexane solutions (c ~ 10–5 mol L–1) (solid lines). e Normalized cyclic voltammograms of the radicals in methylene chloride at a scan rate of 10 mV s–1 with [Bu4N]+[PF6]– (0.1 M) as electrolyte. Previous studies have shown that the TTBrM radical exhibits significantly extended phase memory times Tm compared to TTM, rendering TTBrM interesting as a molecular qubit9. To determine Tm and the spin-lattice relaxation time T1, we perform pulse EPR measurements in frozen toluene solution at 80 K (see Fig. 3b and Figs. S2–S4 in the SI). We are able to reproduce the reported T1 for TTBrM, obtaining a value of 0.05 ms (literature: T1 = 0.07 ms) (Fig. S3). However, the measured phase memory time Tm = 1.5 ± 0.1 µs is significantly shorter than the reported literature value (Tm = 17.0 µs) (Fig S4)9. For comparison, the corresponding relaxation times that we determine for TTM (T1 = 0.80 ms; Tm = 4.0 ± 0.1 µs) and TTM-Cz (T1 = 1.05 ms; Tm = 4.1 ± 0.1 µs) are markedly longer than those of TTBrM. The overall reduction in spin coherence is not unexpected for brominated radicals relative to their chlorinated analogues. This is attributed to bromine's higher atomic mass, which enhances SOC and thus increases spin relaxation rates, leading to shorter phase memory times. The discrepancy between our measured and the reported Tm values for TTBrM may arise from solvent effects, variations in temperature and measurement conditions, or from differences in sample purity. When we functionalize the TTBrM with our three different Cz-based electron donors, we observe slightly increased T1 values compared to the unsubstituted TTBrM with T1 = 0.07 ms, 0.09 ms, and 0.10 ms, respectively, for TTBrM-Cz, TTBrM-MeCz, and TTBrM-Me₂Cz (Table 1 and Figure S3). No significant variation in Tm is observed with Cz functionalization and with successive methylation of the Cz-donor, with Tm values ranging between 1.5 ± 0.1 µs and 1.6 ± 0.1 µs (see Fig. S4). These coherence times can be slightly increased when measuring in deuterated toluene, with Tm ~ 2.1 µs, while Tm of the TTM-Cz reference is increased to 4.7 ± 0.1 µs (Fig. S5 in the SI). The results demonstrate that methylation of the carbazole does not negatively impact the phase memory times.Table 1Electronic and spin resonance characterizationλabs/ nmλPL/ nmε/ L mol cm-1\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\phi$$\end{document}ϕ / %gτ / ns(in Ar)kr /106 s-1knr / 106 s-1t1/2 / minTTBrM-Cz390;6236463009722.006338.89 ± 0.0218.57.22TTBrM-MeCz391;6396793839702.005937.36 ± 0.0218.88.015TTBrM-Me2Cz391;6556884679622.006032.57 ± 0.0219.011.749All optical measurements are performed in cyclohexane solutions (c ~ 10–5 mol L–1). The extinction coefficients for the D0 → D1 transition are determined from the mean of three individual measurements at defined concentrations. The g values are obtained by cw-EPR. The PL lifetimes and decay rates kr and knr are determined by TCSPC and the half-life t1/2 by photostability measurements, measured in toluene, under UV-irradiation. Table 1Electronic and spin resonance characterizationλabs/ nmλPL/ nmε/ L mol cm-1\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\phi$$\end{document}ϕ / %gτ / ns(in Ar)kr /106 s-1knr / 106 s-1t1/2 / minTTBrM-Cz390;6236463009722.006338.89 ± 0.0218.57.22TTBrM-MeCz391;6396793839702.005937.36 ± 0.0218.88.015TTBrM-Me2Cz391;6556884679622.006032.57 ± 0.0219.011.749All optical measurements are performed in cyclohexane solutions (c ~ 10–5 mol L–1). The extinction coefficients for the D0 → D1 transition are determined from the mean of three individual measurements at defined concentrations. The g values are obtained by cw-EPR. The PL lifetimes and decay rates kr and knr are determined by TCSPC and the half-life t1/2 by photostability measurements, measured in toluene, under UV-irradiation. Electronic and spin resonance characterization All optical measurements are performed in cyclohexane solutions (c ~ 10–5 mol L–1). The extinction coefficients for the D0 → D1 transition are determined from the mean of three individual measurements at defined concentrations. The g values are obtained by cw-EPR. The PL lifetimes and decay rates kr and knr are determined by TCSPC and the half-life t1/2 by photostability measurements, measured in toluene, under UV-irradiation. To further investigate the effect of methylation of the Cz-unit on the optical properties, we turn to UV-vis absorption spectroscopy of our compounds in cyclohexane solution (Fig. 3d). The absorption spectra are typical for donor functionalized trityl radicals, with a strong absorption band in the near UV spectrum, here at λabs ~ 390 nm (λabs ~ 375 nm for TTM-Cz)14,29. Moreover, we also observe the typical weak absorption feature in the red spectral region, at λabs = 623 nm for TTBrM-Cz, 640 nm for TTBrM-MeCz, and 655 nm for TTBrM-Me2Cz (Fig. 3d and Table 1). With increasing methylation, the donor strength of the Cz units will increase because of the inductive +I-effect of the methyl groups. It is therefore understandable that with increasing methylation, we are increasing the Cz-donor strength and λabs is systematically red shifted. For successive methylation of the Cz-donor, the extinction coefficient ε of this red band increases, with respect to the strong band in the near UV. The spectral shift and stronger absorption suggest an increasing charge-transfer (CT) character of the absorption and a decreasing energy gap between the highest doubly occupied molecular orbital (HDMO) and the singly occupied molecular orbital (SOMO) for the increasing donor strength (HDMO-SOMO gap). To better elucidate the photophysical properties, UV-vis-absorption and photoluminescence spectra are recorded in solvents of increasing polarity (cyclohexane → ethyl acetate). While the absorption spectra remain less affected by the increased polarity of the solvents, the photoluminescence spectra exhibit a clear and systematic bathochromic shift, consistent with a stabilized CT excited state. In strongly polar solvents, the emission intensity is significantly reduced, indicative of enhanced non-radiative decay (Figs. S12–S14 in the SI). Concomitantly, in the UV-vis spectra, the low-energy absorption band becomes broader and less structured with increasing polarity. These observations provide direct experimental evidence for the CT character of the D1 transition (cf. Fig. 3c). To rationalize these observations, density functional theory (DFT) and time-dependent (TD-)DFT calculations are performed. The computed excitation energies (Eex ≈ 1.9–2.0 eV) and oscillator strengths (f ≈ 0.11–0.13) slightly increase with rising donor strength (Table 2). The dihedral angles between the donor and acceptor units (γDA) of the DFT optimized ground state of the radicals are very similar across the series (TTBrM-Cz: γDA = 68.4°, TTBrM-MeCz: γDA = 67.0°, TTBrM-Me2Cz: γDA = 66.6°), indicating that electronic rather than geometric effects dominate the observed trends (see Fig. 3c and Table 2). Thus, the photophysical behavior appears to be primarily driven by electronic tuning through the donor substitution, rather than changes in conjugation or steric orientation.Table 2Computational, electrochemical, and chiroptical parametersDonor IE / eVEex / eVEox / V (vs. Ag/Ag+)fγDA/ °gabs / 10–4glum / 10–3BCPLTTBrM-Cz7.52.00.890.1168.42.70.530.57TTBrM-MeCz7.31.90.850.1267.02.90.580.69TTBrM-Me2Cz7.11.90.830.1366.63.10.650.94Ionization energies (IE) for the donor moieties are obtained by DFT as the difference between the cationic and the neutral form. The excitation energies (Eex) and oscillator strengths (f) are computed using TD-DFT. Oxidation potentials (Eox) are determined by cyclic voltammetry vs. Ag/Ag+. The dihedral angles between the donor and acceptor units (γDA) are determined in Avogadro using DFT-optimized geometries. The dissymmetry factors (gabs, glum) and brightness of circularly polarized luminescence (BCPL) are obtained from CD and CPL measurements, maximum values are shown here. Table 2Computational, electrochemical, and chiroptical parametersDonor IE / eVEex / eVEox / V (vs. Ag/Ag+)fγDA/ °gabs / 10–4glum / 10–3BCPLTTBrM-Cz7.52.00.890.1168.42.70.530.57TTBrM-MeCz7.31.90.850.1267.02.90.580.69TTBrM-Me2Cz7.11.90.830.1366.63.10.650.94Ionization energies (IE) for the donor moieties are obtained by DFT as the difference between the cationic and the neutral form. The excitation energies (Eex) and oscillator strengths (f) are computed using TD-DFT. Oxidation potentials (Eox) are determined by cyclic voltammetry vs. Ag/Ag+. The dihedral angles between the donor and acceptor units (γDA) are determined in Avogadro using DFT-optimized geometries. The dissymmetry factors (gabs, glum) and brightness of circularly polarized luminescence (BCPL) are obtained from CD and CPL measurements, maximum values are shown here. Computational, electrochemical, and chiroptical parameters Ionization energies (IE) for the donor moieties are obtained by DFT as the difference between the cationic and the neutral form. The excitation energies (Eex) and oscillator strengths (f) are computed using TD-DFT. Oxidation potentials (Eox) are determined by cyclic voltammetry vs. Ag/Ag+. The dihedral angles between the donor and acceptor units (γDA) are determined in Avogadro using DFT-optimized geometries. The dissymmetry factors (gabs, glum) and brightness of circularly polarized luminescence (BCPL) are obtained from CD and CPL measurements, maximum values are shown here. Accordingly, the computed natural transition orbitals (NTOs, Fig. 3c) reveal a growing spatial separation between hole and electron densities along the series, reflecting the increased CT character from TTBrM-Cz to TTBrM-Me2Cz. The excellent agreement between simulated and experimental UV-vis spectra further supports the accuracy of the computational model and confirms the CT nature of the transitions (Fig. 3d and Figs. S9–S11 in the SI). We hypothesize that the CT state is stabilized by the stronger electron-donating ability of the methyl-substituted donor units, where the inductive +I-effect raises the energy of the HDMO (in the radical ground state), while localizing it predominantly on the donor. This is in agreement with our calculated ground-state frontier molecular orbitals of all three radicals, whereas the lowest unoccupied molecular orbital (LUMO) remains localized on the TTBrM unit (Figs. S6–S8 in the SI). This observation is also consistent with the decreasing ionization energies (IE) determined specifically for the isolated Cz-moieties. For these donor fragments we compute the IE via TD-DFT as the energy difference between their cationic and neutral forms. A decrease in IE is consistent with increasing electron donating strength of the Cz-groups (see Table 2). To further support our hypothesis, we perform cyclovoltammetry (CV) using a three-electrode-setup with Ag/Ag+ as a reference and [Bu4N]+[PF6]– (0.1 M) as the electrolyte in anhydrous methylene chloride (Fig. 3e and Fig. S15 in the SI). All TTBrM-Cz radicals exhibit two well-defined and electrochemically distinct redox processes. A reversible reduction at approximately –0.7 V that can be attributed to the TTBrM radical center. Naturally, this feature is absent in the cyclic voltammograms of the corresponding individual carbazoles. By contrast, the oxidation wave is assigned to the carbazole donor unit and shows a systematic cathodic shift upon increasing methyl substitution, progressing from Eₒₓ = +0.89 V for TTBrM-Cz to +0.85 V for TTBrM-MeCz and +0.83 V for TTBrM-Me2Cz (vs. Ag/Ag+). This trend closely follows the donor strengths and is consistent with the oxidation potentials of the respective carbazoles (Fig. S15 in the SI), as well as with the computed ionization energies (Table 2). Chiroptical properties of the TTBrM radicals Like TTBrM, also our Cz-donor functionalized TTBrM radicals can be resolved into their respective enantiomers using chiral high-performance liquid chromatography (HPLC) (see details in the SI, Table S3 and Figs. S17–S19). We obtain excellent separation between the HPLC peaks on a chiral Daicel Chiralpak® IB N-3 column, and we receive the resolved enantiomers (see HPLC trace as insets in Fig. 4a–c). In contrast to TTM derivatives, our Cz-functionalized TTBrM radicals are stable against racemization at room temperature and the enantiomers exhibit circular dichroism (CD) with nicely mirror-imaged spectra and a maximum dissymmetry factor of gabs = 3.1 × 10–4 (Table 2 and Fig. 4a-c). TD-DFT systematically underestimates gabs, within the typical error margins of the method (see Table S2 in the SI). We can clearly observe the Cotton effect in the CD spectra at λabs = 318 nm and 416 nm, originating from the chirality of the TTBrM radical propellers. By comparing the theoretical CD spectrum of the (P)-enantiomers (obtained by TD-DFT) with the experimental data, the blue curves in Fig. 4 can be assigned to the (M)-enantiomer, and the red curves to the (P)-enantiomer (Fig. 4a–c). We test the thermal stability of our enantiomers by heating the sample in 10 °C steps and recording the CD spectrum. All three radicals, TTBrM-Cz, TTBrM-MeCz, and TTBrM-Me₂Cz, are stable up to a temperature of Trac > 80 °C. After four hours at this temperature, only 5% of the pure enantiomer has racemized (Fig. S20 in the SI).Fig. 4Chiroptical data and photostability measurements.a–c Circular dichroism spectra of a TTBrM-Cz, b TTBrM-MeCz, and c TTBrM-Me2Cz in cyclohexane. For assigning of the obtained spectra to the respective (M)- and (P)-enantiomers, we calculated the CD spectrum of the (P)-enantiomer using TD-DFT. The calculated spectra are scaled along the y-axis to match the experimental values. d–f Circularly polarized luminescence as ΔI of d TTBrM-Cz in toluene, e TTBrM-MeCz, and f TTBrM-Me2Cz both in cyclohexane with additional Gauss Fit as a guide to the eye. g Photostability measurements in degassed toluene under UV irradiation (400 nm at 70 mW/cm2) with TTBrM-Cz (black), TTBrM-MeCz (blue), TTBrM-Me2Cz (red), and TTM-Cz (grey) as a reference. Chiroptical data and photostability measurements. a–c Circular dichroism spectra of a TTBrM-Cz, b TTBrM-MeCz, and c TTBrM-Me2Cz in cyclohexane. For assigning of the obtained spectra to the respective (M)- and (P)-enantiomers, we calculated the CD spectrum of the (P)-enantiomer using TD-DFT. The calculated spectra are scaled along the y-axis to match the experimental values. d–f Circularly polarized luminescence as ΔI of d TTBrM-Cz in toluene, e TTBrM-MeCz, and f TTBrM-Me2Cz both in cyclohexane with additional Gauss Fit as a guide to the eye. g Photostability measurements in degassed toluene under UV irradiation (400 nm at 70 mW/cm2) with TTBrM-Cz (black), TTBrM-MeCz (blue), TTBrM-Me2Cz (red), and TTM-Cz (grey) as a reference. The effect of the donor strength variation by methylation is also observed in the photoluminescence (PL) spectra of the prepared cyclohexane solutions of our donor functionalized TTBrM radicals. TTBrM-Cz exhibits a photoluminescence maximum at λPL = 646 nm, TTBrM-MeCz at λPL = 679 nm, and TTBrM-Me2Cz at λPL = 688 nm (Fig. 3d). Compared to TTM-Cz with λPL = 628 nm the brominated derivatives exhibit substantially red-shifted emission, consistent with the increased CT character and the reduced optical gaps. Despite these shifts, all compounds display high ϕ, ranging from 62 % for TTBrM-Me2Cz to 70 % for TTBrM-MeCz to 72 % for TTBrM-Cz, indicative of efficient radiative decay. With such high ϕ, we also investigate the circularly polarized photoluminescence (CPL) of our chiral radical propellers. Surprisingly, we are not able to determine CPL in cyclohexane solutions, despite having recorded clear CD spectra in this non-polar solvent. We turn to toluene solutions, in which we can record mirror-imaged CPL spectra for the resolved enantiomers with glum dissymmetry factors of order 10–4 (Table 2 and Fig. 4d–f). These values are in excellent agreement with glum values determined via TD-DFT after geometrical optimization of our donor functionalized TTBrM radicals in the excited state (Table S2). To explain the lower ϕ for the methylated radicals, we hypothesize that methylation could open up non-radiative relaxation channels. To elucidate these relaxation processes, photoluminescence lifetime measurements (τ) are performed to extract the radiative (kr) and non-radiative (knr) rate constants (Figs. S21–S23 and Table 1). Whereas the kr is almost constant for the three radicals, knr clearly increases for successively methylated radicals (TTBrM-MeCz, TTBrM-Me2Cz), suggesting that non-radiative decay pathways are most suppressed in TTBrM-Cz, leading to the enhanced \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\phi$$\end{document}ϕ. Interestingly, photostability testing, where we irradiate degassed cyclohexane solutions of our radicals with a 5 W UV-lamp (λPL = 400 nm, E = 70 mW / cm2), reveals a dramatic increase in the stability of the TTBrM radical with methylation of the Cz-donor. In cyclohexane solution, the donor-functionalized TTBrM radicals exhibit photostability half-lives (t1/2) increasing from about 1 s for TTBrM-Cz to t1/2 = 3 s for TTBrM-MeCz to t1/2 = 24 s for TTBrM-Me2Cz (Fig. 4c and Figs. S24–S26 in the SI). The higher photostability of the methylated donors may arise from their larger knr, and shorter excited-state lifetimes. This larger knr could reduce the formation of reactive intermediates under photoexcitation. These short photostability half-lives may also be responsible for the insufficient signal during the relatively long CPL measurements. By contrast, TTM-Cz shows an intermediate stability half-life of t₁/₂ = 4 s, highlighting the stabilizing effect of the shielding methyl groups on the Cz-donor functionalized TTBrM radical. Changing the solvent to degassed toluene dramatically increases the photostability, yielding t1/2 = 123 s for TTBrM-Cz, 902 s for TTBrM-MeCz, 2593 s for TTBrM-Me2Cz and 558 s for TTM-Cz, indicating that the higher polarity of toluene stabilizes the excited state of the radicals. In summary, we have introduced highly luminescent TTBrM-radicals with stable chirality and good photostability. During the submission of this manuscript, we became aware of another study reporting the synthesis of a TTBrM-Cz radical30. By overcoming the intrinsic low emissivity of TTBrM, while preserving spin coherence and axial chirality, our work removes a key obstacle toward optical spin readout in TTBrM-based molecular qubits. As such, these donor functionalized TTBrM radicals represent interesting building blocks for spintronic applications and for molecular qubits. Methods Detailed description of the methods used, synthetic procedures, and details on the computational study are provided within Supplementary Information. The authors have cited additional references within the Supplementary Information31–33. Chemicals and solvents All chemicals were used as received without further purification. Sodium tert-butoxide (NaOtBu) was obtained from abcr. 9H-Carbazole (Cz), Pd₂(dba)₃, anhydrous chloroform (CHCl₃), hexafluorobenzene, and potassium hydroxide (KOH) were purchased from Aldrich. 1,3,5-Tribromobenzene, 18-crown-6, and p-chloranil were supplied by Apollo Scientific. (tBu)₃PH·BF₄, 3-methyl-9H-carbazole (MeCz), and 3,6-dimethyl-9H-carbazole (Me2Cz) were obtained from BLD. Aluminum bromide was purchased from TCI. All reactions are performed under an inert atmosphere. Flash column chromatography Purification via column chromatography is carried out with Silica 60 (particle size: 0.04-0.063mm) provided by Macherey-Nagel; the solvents were distilled beforehand. Aside from that, a Puriflash Interchim 450 with a silica column F0025, particle size 30 µm is used. High resolution mass spectrometry (HRMS) Atmospheric pressure chemical ionization (APCI), laser desorption/ionization (LDI), and matrix-assisted laser desorption/ionization (MALDI) high-resolution mass spectrometry (HRMS) is performed on a Bruker solariX Fourier transform ion cyclotron resonance (FTICR) mass spectrometer. For MALDI, trans-2-[3-(4-tert-butyl-phenyl)-2-methyl-2-propenylidene]malononitrile (DCTB) is used as a matrix. Nuclear magnetic resonance (NMR) spectroscopy NMR spectra are obtained by using the Bruker AVANCE NEO 400 or AVANCE NEO 600, and CD2Cl2 is chosen as the solvent. UV/vis absorption spectroscopy UV-Vis spectra are measured with a Perkin Elmer Lambda 365. Fluorescence spectroscopy Photoluminescence spectra are obtained by Perkin Elmer FL6500. Photoluminescence quantum yields The quantum yield is determined on a HAMAMATSU Quantaurus-QY. All optical data are measured in cyclohexane (spectrophotometic grade) with a concentration of ~10–5 M. Cyclic voltammetry For the cyclovoltammetry (CV) measurements, a potentiostat from Ivium Technologies (IviumStat.h) was used. A platinum disc was used as the working electrode, a platinum wire as the counter electrode, and a non-aqueous Ag/Ag+ electrode as the reference. Chiral stationary phase HPLC The AGILENT analytical series with a chiral stationary phase of Daicel CHIRALPAK® IB N-3 (250 ×4.6 mm, particle size 3 µm) and Phenomenex Lux® i-Amylose-3 (250 ×4.6 mm, particle size 5 µm) column at a flow rate of 0.5–1 ml/min and a concentration of 1 mg/ml is used to separate the enantiomers. Electronic circular dichroism and circularly polarized luminescence spectroscopy The Olis DSM 245 spectrophotometry system is used for measuring chiroptical properties. For ECD measurements a fixed slit width of 5 mm and an integration time of 0.5 s were used. CPL spectra were recorded with slits of 3.16 mm, an excitation wavelength of 390 nm and an integration time of 2 s. All solutions were deoxygenated prior to measurement. DFT The theoretical calculations were performed with Gaussian 16 (Revision C.01)33. The geometry optimization was performed on the PBE1PBE/6-311 + + G** level of theory. To account for non-covalent interactions, the empirical dispersion correction GD3BJ is applied. Furthermore, the solvent environment is simulated using the SMD continuum model with cyclohexane as the solvent (SCRF = SMD, Solvent = Cyclohexane). TD-DFT is performed on the same level of theory and under the same conditions, for N = 50 states. Excited state calculations were subsequently carried out at the same level of theory and geometry optimization of the excited state was performed using the TD-DFT results as a starting point with an increased SCF convergence limit (scf = (maxcycle = 3000)). The number of computed states was set to five excited states with the first excited state as the reference state (TD = (NStates = 5, Root = 1)). Supplementary information Supplementary Information Transparent Peer Review file Supplementary Information Transparent Peer Review file