Liangwei
Ma‡
,
Muyu
Cong‡
,
Siyu
Sun
and
Xiang
Ma
*
Key Laboratory for Advanced Materials and Joint International Research Laboratory of Precision Chemistry and Molecular Engineering, Frontiers Science Center for Materiobiology and Dynamic Chemistry, School of Chemistry and Molecular Engineering, East China University of Science and Technology, Shanghai 200237, China. E-mail: maxiang@ecust.edu.cn
First published on 4th April 2025
Designing and optimizing efficient organic room-temperature phosphorescent (RTP) materials remains a captivating yet challenging endeavour due to the inherent difficulties in generating and stabilizing triplet excitons. Here, we report a suite of highly efficient phosphors characterized by near-unity intersystem crossing (ISC) yields. Surprisingly, upon doping these dyes into a polyvinyl alcohol matrix, their phosphorescence quantum yields (ΦP) spanned a wide range from 2.7% to 69.6%, governed by the position of the methyl substituent. Theoretical calculations and experimental results indicate that the variation in phosphorescence efficiency is primarily due to the strong electron–phonon coupling caused by the positional variation of the methyl substituents, rather than common factors such as ISC or energy levels. These findings provide a new insight into the design of high-performance organic RTP dyes.
In this work, we systematically demonstrate the significant impact of EPC on the performance of phosphorescent dyes. As the position of the substituents changes, the ISC efficiency of these methyl-substituted thiochroman-4-one derivatives (MTOs, Scheme 1) approaches 100%, with the ISC rate constant (kISC) gradually increasing. In contrast, the phosphorescence quantum yield (ΦP) decreases from 69.6% (MTO-1@PVA) to 2.7% (MTO-4@PVA) with the increase of kISC. Temperature-dependent emission spectra show that materials with lower ΦP are more temperature-sensitive, indicating stronger EPC. Theoretical simulations reveal that MTO-4 has a Huang–Rhys factor of 23.06 at 82 cm−1, while the Huang–Rhys factors of the other three compounds do not exceed 1.0, indicating that MTO-4 has significantly stronger EPC. Combined with the temperature-dependent experiments, it is determined that the significant decrease in phosphorescence efficiency is primarily due to the excessively strong EPC in MTO-4, which causes molecular vibrations to exceed the threshold that the polyvinyl alcohol (PVA) matrix can effectively suppress. This work highlights the critical role of EPC in the design of efficient RTP materials and provides new insights into the mechanisms governing phosphorescence performance.
![]() | ||
Scheme 1 Chemical structure of target compounds and the simplified Jablonski diagram for the transition process from the lowest excited state in aromatic compounds. |
The comprehensive synthetic protocols for MTO compounds are elaborately outlined in the ESI.† Their chemical structures were meticulously authenticated through 1H NMR, 13C NMR, and HRMS analyses. Upon dissolution in dichloromethane (DCM), the four dyes displayed strikingly analogous absorption spectra (Fig. 1a), with their maximum absorption peaks consistently centred around 350 nm. The subtle shifts in the maximum emission wavelengths can be attributed to the subtle perturbations induced by the varying methyl substituent positions on the energy levels of the frontier orbitals. This observation is consistent with the fact that the methyl group lacks π-orbitals to conjugate with the frontier orbitals of thiochroman-4-one. However, a marked disparity was observed in the fluorescence intensities among the dyes as the methyl substituent position varied (Fig. 1a). The fluorescence quantum yield (ΦF) was found to be 1.19% in DCM when the methyl group occupied the 6-substituent position (Table S1†), whereas it was virtually quenched (ΦF = 0.25%) when situated at the 5-substituent position. Furthermore, substantial differences in the fluorescence lifetimes of these compounds were discernible (Fig. 1b and S1†), indicative of substantial variations in the relaxation dynamics of the lowest singlet excited state (S1).
To unravel the unusual impact of the methyl group on the photophysical properties, density functional theory (DFT) and time-dependent DFT (TDDFT) calculations were conducted using the G09 program with the M062x/aug-cc-pVTZ level of theory.39,49 As depicted in Table S2,† the S0 → S1 transition for these compounds predominantly involved HOMO/HOMO-2 → LUMO transitions. Despite their similar molecular frameworks (Fig. S2†), the contribution from the HOMO → LUMO transition, which characterizes a π → π* transition, progressively dwindled from 20.1% to 0% as the methyl substituent position varied. Notably, the HOMO-2 → LUMO transition, constituting the dominant contribution to the S0 → S1 transition, represents a typical n → π* transition. Thus, the S0 → S1 transition for these compounds is classified as a forbidden transition due to the non-conservation of orbital angular momentum. Correspondingly, with the decline in the HOMO → LUMO transition contribution, the f of these compounds plummeted from 0.0165 to 0.0002 (Table S2†). Conversely, the S0 → S2 transition for these compounds was primarily driven by the HOMO → LUMO transition, representing a spin-allowed process. Consequently, the S0 → S1 transition is designated as a dark state.50 Hence, the maximum absorption peak for these compounds is attributed to the S0 → S2 transition. The shoulder peak observed at 370 nm in the absorption spectrum of MTO-1 is assigned to the S0 → S1 transition due to its weak yet non-negligible f value.
The S1 → S0 transition mirrored the trends observed for the S0 → S1 transition (Table S2†). For MTO-1, the S1 → S0 transition was primarily governed by the HOMO → LUMO transition (95.4%), with a minor contribution from the HOMO-2 → LUMO transition (2.2%). However, the contribution from the HOMO → LUMO transition significantly declined to 24.1% for MTO-2, further dwindling to 18.5% and 8.6% for MTO-3 and MTO-4, respectively. Concurrently, the HOMO-2 → LUMO transition contribution escalated to 62.3%, 67.6%, and 76.9% for MTO-2, MTO-3, and MTO-4. As a result, the fluorescence transition for these compounds transformed into a forbidden process as the methyl substituent position varied. Comparable phenomena were also evident in different solvents. The consistent maximum absorption and emission wavelengths of MTO-1 across various aprotic solvents with differing polarities supported the π → π* nature of the transition (Fig. 1c). An anomalous redshift in both absorption and emission spectra was observed for these compounds in protic solvents, accompanied by a marked increase in emission intensity (Fig. 1d). Correspondingly, the ΦF increased to 6.67% in methanol. This abnormal enhancement and redshift were attributed to hydrogen bonding interactions between the solvent and the carbonyl group of MTOs.51
Previous studies have established thiochroman-4-one derivatives as promising phosphorescent dyes with high ISC efficiency.18,52,53 Hence, the photophysical properties of these dyes were further investigated in 2-methyltetrahydrofuran at 77 K. As depicted in Fig. 1e, a broad emission band spanning 420–620 nm, adorned with a fine structure, was discernible in the steady-state spectra. Remarkably, the phosphorescence spectra of these compounds closely resembled their corresponding steady-state emission spectra. Most intriguingly, no appreciable fluorescence emission was detected at 77 K, strongly suggesting an efficient ISC process for these compounds.
Subsequently, these compounds were embedded in a PVA matrix, known to effectively suppress non-radiative transitions and isolate oxygen,15,54 thus enabling the fabrication of RTP materials. Powder X-ray diffraction analysis of these films (Fig. S3†) revealed an absence of discernible diffraction peaks, signifying their amorphous nature. Analogous to the behaviour of the dyes in solution, the maximum absorption peaks for these films were found around 350 nm (Fig. 2a). The steady-state emission spectra of the films displayed structureless emission bands spanning 380–700 nm (Fig. 2b). Specifically, MTO-1@PVA showcased a maximum emission wavelength at 485 nm, accompanied by a shoulder peak at approximately 420 nm. Notably, the shoulder peak vanished in the phosphorescence emission spectrum, implying that it corresponded to the weak fluorescence emission of MTO-1@PVA, while the dominant emission peak at 485 nm was attributed to the phosphorescence emission. This attribution was further substantiated by lifetime measurements performed at different temperatures (Fig. S4†). The phosphorescence lifetime of MTO-1@PVA was measured to be 21.92 ms at 297 K (Fig. 2c). Its photoluminescence quantum yield (ΦPL) was calculated to be 69.6% (Fig. S5† and Table 1). The excellent agreement between the excitation and absorption spectra (Fig. 2a) confirmed the origin of phosphorescence from MTO-1. Similarly, MTO-2@PVA displayed a much weaker shoulder peak around 420 nm. In contrast, MTO-3@PVA displayed a structureless emission peak at 460 nm, which was confidently assigned to the phosphorescence emission. The ΦPL values for MTO-2@PVA and MTO-3@PVA were determined to be 47.7% and 48.8%, respectively. Thus, it can be concluded that MTO-1@PVA, MTO-2@PVA, and MTO-3@PVA are essentially pure phosphorescent materials, a rarity among metal-free substances. However, MTO-4@PVA exhibited a distinct behaviour. As evident in Fig. 2b, the maximum emission wavelength of MTO-4@PVA was positioned at 435 nm in the steady-state emission spectrum, while its phosphorescence emission wavelength was identified as 470 nm, implying that the phosphorescence emission of MTO-4 was significantly weaker than its fluorescence emission within the PVA matrix. Correspondingly, the ΦPL of MTO-4@PVA was measured to be a mere 2.7%. The phosphorescence lifetime for MTO-4@PVA was found to be 12.0 ms. The fluorescence lifetimes of all films under scrutiny ranged from 1.44 to 2.47 ns (Fig. 2d). It is particularly noteworthy that MTO-1@PVA and MTO-4@PVA exhibited comparable decay patterns and lifetimes (2.34 and 2.47 ns, respectively). Meanwhile, the lifetime spectra and parameters of MTO-2@PVA closely resembled those of MTO-3@PVA.
λ Abs/nm | λ F/nm | λ P/nm | τ F/ns | τ P/ms | Φ PL/% | Φ ISC/% | k ISC/s−1 | k P/s−1 | k nr/s−1 | |
---|---|---|---|---|---|---|---|---|---|---|
a
![]() ![]() ![]() ![]() |
||||||||||
MTO-1@PVA | 356 | 423 | 488 | 2.34 | 21.92 | 69.6 | 99.999 | 1.50 × 1010 | 31.75 | 13.87 |
MTO-2@PVA | 349 | n.d. | 470 | 1.46 | 10.28 | 47.7 | 99.995 | 1.75 × 1010 | 46.40 | 50.88 |
MTO-3@PVA | 345 | n.d. | 464 | 1.44 | 11.77 | 48.8 | 99.996 | 2.05 × 1010 | 41.46 | 43.50 |
MTO-4@PVA | 347 | n.d. | 468 | 2.47 | 12.00 | 2.7 | 99.995 | 2.55 × 1010 | 2.25 | 81.08 |
To delve deeper into the influence of the methyl group substituent position on the ISC process, TDDFT and ORCA calculations were also performed.18,23 As previously discussed, the S1 → S0 transition in MTOs progressively transitioned from a π → π* to an n → π* character with the change in the methyl group position. However, no conspicuous alteration in the frontier orbital distribution was observed (Fig. S2 and Table S2†). The energy levels of T1 and T2 exhibited slight fluctuations with the varying substituent positions (Table S2†). Notably, T2 represented the highest triplet state, often being either close to or lower in energy than the S1 state. Hence, S1 → T1 and S1 → T2 transitions were identified as plausible ISC pathways for generating triplet excitons. Fig. 2e reveals that the hole in the S0 → T1 transition of MTOs is primarily localized on the sulfur atom and benzene rings, corresponding to π orbitals, while the electron is primarily distributed over the benzene and carbonyl moieties, associated with π* orbitals. Although the electron distribution in the S0 → T2 transition mirrors that of the S0 → T1 transition, the hole is primarily contributed by the n orbital of the carbonyl moiety. Consequently, the S1 → T1 ISC process evolved from a forbidden to an allowed transition as the substituent position changed, according to El-Sayed's rule. Conversely, the S1 → T2 ISC process transformed from an allowed to a forbidden transition with the variation in the substituent position. ORCA calculations further substantiated the significant impact of the methyl substituent position on the SOC, which is intricately linked to the ISC process.
Femtosecond TA spectra of MTOs in the PVA matrix were recorded and are presented in Fig. 3 and S6.† No ground state bleach (GSB) signal was observable due to their absorptions lying outside our detection wavelength range, nor was any stimulated emission (SE) signal detected. A distinct excited state absorption (ESA) signal was promptly detected for MTO-1@PVA post-excitation (Fig. 3a). The initially formed ESA band featured a peak around 620 nm with a shoulder at approximately 525 nm (Fig. 3b). Within the first 200 ps, it evolved into a long-lived ESA band with peaks at 470 nm and 610 nm. Analogous TA spectra were also evident for MTO-2@PVA and MTO-3@PVA (Fig. S6a and b†). Remarkably, MTO-4@PVA displayed a markedly different TA spectrum compared to the other three films (Fig. S6c†). A distinct ESA signal spanning 400–600 nm, with peaks at 405 nm and 485 nm, was observed. The initial ESA band transformed into a long-lived ESA band spanning 400–600 nm, centred at 450 nm, within the first 100 ps (Fig. S6f†). Decay-associated difference spectra (DADS) for MTO-1@PVA were extracted via global analysis of the femtosecond TA data (Fig. S7†). The first decay-associated difference spectrum was attributed to the absorption of the singlet state (Fig. S7,† black line), displaying a broad signal in the 410–700 nm range with peaks around 520 nm and 620 nm, consistent with species present in the first 300 ps. The subsequent DADS was assigned to the absorption of the triplet state (Fig. S7,† red line), given its slow generation and decay within the 7 ns detection window. This spectrum exhibited a broad signal in the 410–700 nm range, with peaks around 450 nm and 615 nm. Comparable DADS were also discernible for the remaining three films (Fig. S6†). The decay kinetics curves of the singlet excited state and the generation kinetics of the triplet state of MTOs in the PVA matrix are compiled in Fig. 3c. Notably, the decay rate of the singlet excited state showed a pronounced increase from MTO-1 to MTO-4. The near-symmetric kinetics curves of decay and generation implied a highly efficient ISC process. The ISC yields (ΦISC) of MTOs were estimated to be >99.9% based on the kinetic parameters (Fig. 3d).55 The kISC for MTO-1 was calculated to be 1.50 × 1010 s−1, increasing to 1.75 × 1010 s−1 and 2.05 × 1010 s−1 for MTO-2 and MTO-3, respectively. MTO-4 displayed the highest kISC, reaching up to 2.55 × 1010 s−1. However, the ΦP of MTOs in the PVA matrix decreased from 69.6% to 2.7% as kISC increased. The TA data indicate that the ISC in the MTO series of compounds is highly efficient. Therefore, the phosphorescence quenching observed in MTO-4@PVA is unlikely to be due to issues with the generation of triplet excitons. Instead, strong non-radiative transitions are likely the primary cause.
To exclude the possibility of oxygen-induced non-radiative transitions in the PVA matrix, the emission spectra of MTO-4@PVA were initially measured under different atmospheres. As depicted in Fig. S8,† no noticeable change in phosphorescence intensity was observed when the film atmosphere was switched from air to argon. The phosphorescence intensity of MTO-4@PVA significantly intensified in the steady-state spectrum upon decreasing the temperature from 297 K to 77 K (Fig. 3e). Specifically, the area of the emission peak grew by a factor of 12.1 with decreasing temperature. In comparison, MTO-1@PVA, MTO-2@PVA, and MTO-3@PVA exhibited increases of 1.1, 2.1, and 1.6 times, respectively. A pronounced elongation in the phosphorescence lifetime was also observed as the temperature dropped from 297 K to 77 K. The emission spectrum areas of MTOs in the PVA matrix at various temperatures were meticulously analysed (Fig. S9†) and plotted in Fig. 3f. The area of MTO-1@PVA showed a marginal increase when the temperature increased from 77 K to 97 K. A distinct decline, signifying the onset of triplet exciton quenching by molecular thermal motion, ensued once the temperature exceeded 217 K. Similar trends were observed for the remaining three films. The areas of MTO-2@PVA, MTO-3@PVA, and MTO-4@PVA started to diminish when the temperature surpassed 157 K, 177 K, and 97 K, respectively. Notably, MTO-4@PVA displayed a much steeper decrease rate compared to the other three films. Intriguingly, the transition temperature points followed the same trend as ΦP (Fig. 3g). The lower transition temperature points implied that triplet excitons in these films were more susceptible to quenching by molecular thermal vibration, hinting at a large knr.
To further understand the non-radiative pathways and the underlying mechanisms responsible for the phosphorescence quenching, we further used the DFT calculations to quantitatively assess the vibrational freedom of these compounds at the B3LYP/def2SVP level. The reorganization energy versus different normal vibrational modes is computed to evaluate the feasibility of a phosphor in its T1 state to decay via nonradiative decay (Fig. 4a). Compared to MTO-1, MTO-2 shows a slight increase in vibration at a low-frequency mode (377 cm−1), whereas MTO-3 primarily exhibits a slight increase in high-frequency vibration modes. The subtle changes in the vibrational modes of MTO-2 and MTO-3 may explain the decrease in their ΦP. MTO-4 has a remarkably high reorganization energy of 1891.3 cm−1 at a vibrational mode of 82 cm−1. Additionally, the reorganization energies at 565 cm−1 and 1192 cm−1 are both greater than 500 cm−1. The significant increase in vibrational modes may explain the strong non-radiative transitions in MTO-4. Further analysis of the root mean square deviation (RMSD) of this series of compounds revealed that, with the change in the position of the substituents, the difference between the optimal configurations of the T1 state and the S0 state gradually increases (Fig. 4b). Specifically, MTO-1 has the smallest RMSD value of 0.08, whereas MTO-4 has the largest, at 0.33—an increase of about four times. The large RMSD value of MTO-4 indicates stronger molecular vibrational behavior, providing more opportunities for non-radiative relaxation. We further calculated the Huang–Rhys factors of this series of compounds to characterize their EPC. As shown in Fig. 4c, the Huang–Rhys factors of MTO-1, MTO-2, and MTO-3 for various vibrational modes do not exceed 1. However, MTO-4 has four vibrational modes with Huang–Rhys factors exceeding 1.0, including a staggering 23.06 at 82 cm−1. Such a high Huang–Rhys factor indicates strong EPC in MTO-4. Therefore, despite having similar electronic structures and energy levels in the S0 and T1 states compared to the other three compounds, the T1 state of MTO-4 is highly susceptible to non-radiative deactivation.
Footnotes |
† Electronic supplementary information (ESI) available. See DOI: https://doi.org/10.1039/d5sc02149a |
‡ These authors contributed equally to this work. |
This journal is © The Royal Society of Chemistry 2025 |