Brightening triplet excitons enable high-performance white-light emission in organic small molecules via integrating n–π*/π–π* transitions

In situ high-pressure optical properties of IPAThe single IPA molecules in the twisted layers were arranged in zigzag forms via hydrogen-bond interaction and the twisted layers were offset face-to-face staked along the normal direction (Fig. 1a). The intrinsic intermolecular hydrogen-bond interactions and π–π staking motifs endow the solid IPA materials with RTP emission by SOC and HFC effect20,35,36. Under the 355 nm laser excitation, a blue-light emission with PLQY of 19% was observed based on the pristine configuration, and accompanied by a green RTP emission after turning off the laser excitation (Fig. 1b). The initial PL spectrum of IPA was composed of duel emission bands with a primary strong one centered at 402 nm and another weak one in the range of 450–720 nm under the 355 nm laser excitation (Supplementary Fig. 1a). After turning off the laser excitation, the lower-energy band was measured located ~530 nm (Supplementary Fig. 1b). The time-resolved PL decays of these two species indicate their lifetimes of 6.81 ns (402 nm) and 888.6 ms (530 nm), respectively (Supplementary Fig. 2). These properties reflected that the short-lived emission should originate from singlet excited state, while the long-lived one belonged to triplet excitons emission. The corresponding fluorescence quantum yield (ΦF) and phosphorescence quantum yield (ΦP) of the pristine sample separately reached 12% and 7% (Supplementary Fig. 3a).Fig. 1: The molecular structure and photoluminescence (PL) properties of IPA.a The crystal structure of IPA. b Photographs of the pristine and pressure-treated IPA taken at different time intervals before and after turning off the laser excitation (355 nm) at ambient conditions. The photographs (UV off) were taken immediately after turning the UV light off. A mass of quenched products (decompression from 20.0 GPa) was harvested by high-pressure experiments in a Walker-type large-volume press. c PL spectra of IPA upon compression to 9.0 GPa using symmetric diamond anvil cell (DAC) devices under the 355 nm laser excitation. d PL spectra of IPA (left) at ambient conditions and after pressure was released from 18.0 GPa. The corresponding PL photographs (right) were taken in DAC devices. e Changes of chromaticity coordinates at ambient conditions (0.22, 0.22) and after releasing pressure from 18.0 GPa (0.28, 0.36). f PL (blue solid line), fitting fluorescence (green dashed line), and fitting phosphorescence (orange dashed line) spectra of the targeted IPA treated by the Walker-type large-volume press.Considering the fluorescence-dominated weak blue-light emission and limited phosphorescence efficiency of the initial sample, high-pressure processing was expected to alter the singlet/triplet excitons distribution by directly regulating the intermolecular interactions and staking configuration. In situ high-pressure PL measurements were thus carried out up to 18.0 GPa with a pressure transmitting medium (PTM) of silicone oil. The emission of the sample experienced a stark persistent increase with the maximum value exceeding 9.7 times from 1 atm to 9.0 GPa and then followed by a decrease in intensity (Fig. 1c and Supplementary Fig. 1c, e). A weak white-light emission with chromaticity coordinates of (0.30, 0.38) was observed at 18.0 GPa (Supplementary Fig. 4). With the decrease of pressure, the PL intensity gradually enhanced (Supplementary Fig. 1d, e). Upon releasing pressure to the ambient conditions, the targeted sample displayed a high-performance white-light emission with PLQY of 75% and chromaticity coordinates of (0.28, 0.36) (Fig. 1d, e). We also explored the high-pressure optical evolutions with different PTMs, including liquid argon (Supplementary Fig. 5) and nitrogen (Supplementary Fig. 6). Their PL evolutions under high pressure still experienced a similar trend to that of silicon oil, where the white-light emission was also harvested after releasing pressure completely. These similar results manifested that the presently used PTMs (silicone oil, liquid argon, and nitrogen) could not influence the current results: the PL enhancement under high pressure and the harvest of white-light emission after pressure treatment. Furthermore, the treated sample still displayed a high-performance white-light emission after ~400 days (Supplementary Fig. 7). In order to further explore the influence of pressure treatment on RTP properties, the relevant characterization of RTP emission was performed. We measured the yellowish-green afterglow emission at ~573 nm (Supplementary Fig. 8) after turning off the UV lamp, which could be captured by the naked eye even lasting for 8 s (Fig. 1b, Supplementary Fig. 9, and Supplementary Movie). The time-resolved PL decay measurements of the targeted IPA revealed a longer phosphorescence lifetime of 900.2 ms (573 nm) (Supplementary Fig. 2). The fluorescence peak was fitted at 481 nm (Fig. 1f). ΦP and ΦF unexpectedly increased to 40% and 35% from 7% (1 atm) and 12% (1 atm) (Supplementary Fig. 3b and Supplementary Table 1).We further analyzed the in situ high-pressure ultraviolet-visible (UV-vis) absorption profiles to elucidate the electronic behaviors (Supplementary Fig. 10). At ambient conditions, the absorption peak was located at 295 nm, where the absorption edge was at ~307 nm. With the increase of pressure to ~21.9 GPa, the absorption edge exhibited a continuous redshift. After pressure was completely released, the absorption peak at ~317 nm in the targeted sample redshifted by about 22 nm compared with the pristine sample. We then estimated the bandgap changes before and after pressure treatment. We found that the bandgap narrowed from 4.20 eV to 3.91 eV, illustrating the reduction of relevant excited-states energies in targeted IPA.Structural evolution of IPA under pressureFollowing the optical changes above, we devoted efforts to the corresponding structural evolution to explore the mechanism. We have conducted in situ high-pressure infrared (IR) spectra in Supplementary Fig. 11a. The vibration mode at 1696 cm−1 belonged to the stretching vibration of C=O bonds37, which redshifted by 14 cm−1 below 20.0 GPa. This suggested that the O–H···O=C hydrogen bonds were continuously strengthened with the increase of pressure38,39. More importantly, the C=O stretching vibrational mode in the pressure-treated IPA was redshifted by 6 cm−1 compared to the original state (Fig. 2a). This demonstrates that the enhanced hydrogen bonds have survived down to the ambient conditions after pressure treatment. It is noted that the hydrogen-bond interaction in IPA refers primarily to the electrostatic interaction between oxygen and hydrogen atoms. Such enhancement of the hydrogen bonds indicates that the proton donor of hydrogen atoms has a stronger attraction to the lone pair (n) electrons on oxygen atoms. Thus, the n-orbitals can be stabilized to the lower energies by the enhanced electrostatic environments36.Fig. 2: Crystal structure evolution upon compression and decompression.a IR spectra of IPA in the region of C=O stretching vibrational mode ν (C=O) at ambient conditions and after pressure was released from 20.0 GPa. b Molecular packing along the b-axis (left). The schematic diagram of the parallel misalignment angle (σ) of IPA (right). c The compression rate of three lattice constants (a, b, c) at different pressures. d Time-of-flight (TOF) neutron diffraction patterns of IPA-d6. The TOF neutron diffraction patterns of the pristine (blue line) and recovered (magenta line) samples were collected at BL11 PLANET in the MLF at J-PARC. The asterisks indicate the peaks of the Pb marker. The neutron diffraction patterns of the pristine (green line) and recovered (orange line) samples were collected at the High-pressure neutron diffractometer (Fenghuang) at the CMRR neutron science platform. e Pressure-dependent hydrogen-bond distances D17···O9 (d1) and D21···O5 (d2) evolution of IPA. f Pressure-dependent σ evolution of IPA. The d1, d2, and σ of IPA-d6 were determined by Rietveld refinement of neutron diffraction patterns. The error bars in (e) and (f) represent the standard error in the Rietveld refinement.In order to further explore the structural variation upon compression and decompression, we tracked in situ high-pressure angle-dispersive synchrotron X-ray diffraction (ADXRD) experiments from 1 atm to 19.5 GPa (Supplementary Fig. 11b). Upon compression, all Bragg diffraction peaks shifted continuously along the large 2-theta direction and no new peaks appeared. It indicated that the structure did not undergo phase transformation during the whole compression process. We performed Rietveld refinements of ADXRD patterns to obtain detailed lattice parameters upon compression and decompression (Supplementary Fig. 12). The initial IPA possesses monoclinic symmetry with space group P21/c at ambient conditions. As the pressure increased, lattice parameters a, b, and c decreased. The β also underwent continuous reduction upon compression, which appeared to be in favor of a lessened spatial overlap of the π-stacked motif (Fig. 2b).Given that organic IPA consists solely of light-weight elements such as C, H, and O, which displayed a weaker scatter ability in ADXRD experiments. We further performed in situ time-of-flight (TOF) neutron diffraction to delve more deeply into the structural evolution upon compression and decompression (Supplementary Figs. 13–16, Supplementary Table 2, and Fig. 2c). As pressure was increased from ambient conditions to 19.8 GPa, all diffraction patterns moved toward smaller d-spacing, indicating lattice compression (Supplementary Fig. 13 and Fig. 2c). In addition, several peaks of (033), (110), and (022) planes in the recovered neutron spectrum located at smaller d-values (the blue and magenta lines in Fig. 2d). We also conducted experiments on the recovered sample after removing the gasket (the green and orange lines in Fig. 2d). The consistent shift in the (033), (110), and (022) planes indicated that these changes were indeed derived from the pressure treatment engineering (Supplementary Fig. 17). Notably, the orientations of (033) and (022) lattice planes are almost perpendicular to the direction of hydrogen bonds, the evolution of their d-spacings is closely related to the change of hydrogen-bond distance (Supplementary Fig. 18). Therefore, the reduced d-spacings of these almost vertical planes in the targeted IPA could further confirm the result of strengthened hydrogen bonds after pressure treatment38. In addition, attributed to such reduced hydrogen-bond distances, coupled with the intensified parallel-displaced π staking arrangement, the d-spacing of the (110) plane also exhibited a decrease in the recovered sample.In this regard, we move forward to investigate the changes of the correlative intermolecular distances D17···O9 (d1), D21···O5 (d2), and the parallel misalignment angle (σ) to assess the hydrogen bonds and π-stacked configuration of IPA upon compression and decompression. We found that both d1 and d2 displayed a decreasing trend with the increase of pressure, suggesting the strengthened hydrogen bonds at high pressure40 (Fig. 2e). Moreover, the d1 of the recovered sample decreased from 1.73(4) Å to 1.49(5) Å and d2 decreased from 1.77(4) Å to 1.69(5) Å. Therefore, the hydrogen bonds in the recovered IPA were stronger than those ones before pressure treatment. The σ also tended to decrease upon compression (Fig. 2f), and the lesser σ value of 65.7(9)° rendered a lessened spatial overlap of the π-stacked motif in the targeted IPA (Supplementary Table 3). Meanwhile, the intermolecular perpendicular distance between benzene rings was compressed from 3.47 Å to 3.43 Å. The intermolecular distance d(C···C) between benzene rings did not change significantly, which varied from 3.77(5) Å to 3.75(5) Å (Supplementary Fig. 19). In this regard, the electronic coupling between the π-stacked molecules should following these changes and then contribute to the targeted phosphorescence enhancement.SOC, HFC, and ΔE
ST effect of IPANotably, the decreased hydrogen-bond distances would drive strengthened electrostatic interactions between the proton and the lone pair electrons, leading to the consequent changes in the ISC process36. The coupling of the radical ion pairs (RIP) excitons derived from π–π staked IPA molecules is expected to be influenced by the alteration of π–π stacking arrangement20. Besides, the parallel-displaced configuration also played a significant role in minimizing energy difference ΔE(π, π*) of the bonding and antibonding π-type molecular orbitals41, thereby resulting in the alteration of ΔEST. Drawing upon the aforementioned considerations, we systematically analyze the influence of these structural changes on the ISC process from the following three aspects: SOC, HFC, and ΔEST. To begin with, the narrowed hydrogen-bond distance was considered a vital factor for efficient ISC. To decipher the mechanism, we performed the SOC coefficients (ξ) and the natural transition orbitals (NTOs) on the singlet and triplet states based on the pristine and pressure-treated IPA (Fig. 3a and Supplementary Table 4). Before pressure treatment, the S1 and high-lying triplet (T5 and T6) states mainly exhibited the transition character of 1(n, π*) and 3(n, π*) (Supplementary Fig. 20). The low-lying Tn (1 ≤ n ≤ 4) states were primarily assigned to the transition character of 3(π, π*). After pressure treatment, the T5 and T6 states underwent a noticeable change from 3(n, π*) in the original IPA to the mixture of 3(n, π*) and 3(π, π*) in pressure-treated IPA (Fig. 3b). According to the El-Sayed’s rule42, a superior SOC between S1 to Tn (5 ≤ n ≤ 6) states was expected to be achieved in the targeted sample for efficient ISC. In fact, the calculated ξ(S1-T5) and ξ(S1-T6) for targeted IPA were 29.18 cm−1 and 7.71 cm−1, respectively. These values were both larger than those of pristine IPA (with ξ(S1 − T5) of 28.53 cm−1 and ξ(S1 − T6) of 0.08 cm−1). Such enhanced SOC could efficiently promote the ISC process for enhanced RTP emission. In addition, we found that the pressure treatment also endowed ξ enhancements of other triplet states (Supplementary Table 4). This further boosts the population of triplet excitons, allowing for substantial utilization of triplet-state energies in treated IPA. Therefore, the effective mixture of n–π*/π–π* transition configurations tremendously promotes the increase of SOC for the accelerated spin-flipping process (Fig. 3c). A high fraction of triplet excitons was then generated for drastic phosphorescence enhancement.Fig. 3: The SOC and NTOs of the pristine and pressure-treated IPA.a The calculated ξ (ξ(S1 − T6), ξ(S1 − T5), ξ(S1 − T4), ξ(S1 − T3), ξ(S1 − T2), and ξ(S1 − T1)) for the pristine and pressure-treated IPA. b The NTOs of T5 and T6 in the pristine and treated IPA. At 1 atm, the T5 and T6 states feature 3(n, π*) transition. After pressure treatment (Released), the T5 and T6 states feature a mixture of 3(π, π*) and 3(n, π*) transitions. c Schematic representation of hydrogen bonds acting on the ISC process. The strengthened hydrogen bonds promote the mixture of 3(π, π*) and 3(n, π*) transitions, thus accelerating the ISC process.Additionally, the misaligned π–π stacking arrangement was prone to alter the coupling of the RIP excitons, which further influenced the HFC effect on the ISC process. To assess the impact of pressure treatment engineering on HFC, we carried out the magnetic-field PL (MPL) spectra of the pristine and pressure-treated IPA (Fig. 4a). The MPL intensity of the pristine sample changed with the magnetic flux density modulation of up to ~2 T. This magnetic-field effects proved the presence of RIP excitons and the HFC mechanism in the initial IPA20,43. However, after pressure treatment, we noticed minimal variation in pressure-treated IPA. The little influence of the external magnetic field on PL indicated that the RIP excitons were almost absent in the pressure-treated IPA43. In order to explore the relationship between the parallel-displaced arrangement and the radical pairs, we conducted the molecular orbitals of π–π staking dimers before and after pressure treatment. We found that the pristine IPA mainly exhibited charge-transfer-like π–π* transition between two IPA molecules, thus benefiting the presence of weakly coupled radical pairs. However, the pressure-treated IPA mainly shows an intramolecular n–π* transition, which is not conducive to the existence of radical pairs and the supplemental HFC channel (Supplementary Fig. 21). Therefore, the effect of HFC was negligible after pressure treatment owing to the misaligned π–π stacking arrangement.Fig. 4: The effects of HFC and singlet-triplet energy gap (ΔEST) on the ISC process of IPA before and after pressure treatment.a Magnetic-field effects on the PL intensity of IPA. The data at 1 atm and released represent the mean value from five and three experiments, respectively. The error bars are ±s.e.m. b The energies of singlet and triplet states of IPA before and after pressure treatment. c The calculated ΔEST (ΔE(S1 − T6), ΔE(S1 − T5), ΔE(S1 − T4), ΔE(S1 − T3), ΔE(S1 − T2), and ΔE(S1 − T1)) for the pristine and targeted IPA. d Proposed energy transfer processes for fluorescence and phosphorescence in IPA before and after pressure treatment. The narrowed ΔEST effectively accelerated the ISC process after pressure treatment.Besides the HFC effect, this parallel-displaced configuration also contributes to minimizing energy difference ΔE(π, π*). Notably, the enhanced hydrogen bonds could stabilize the n-orbitals to the lower energies by the enhanced electrostatic environments. These modifications would consequently induce variations in the relative energy levels of singlet and triplet states, along with changes in ΔEST, which is closely associated with the ISC process. Therefore, the energies for both singlet and triplet excited states, as well as ΔEST were calculated in detail. We converged on investigations of the lowest singlet state (S1) and six triplet states (T1–T6) below S144. Compared with the initial states, the treated sample featured various degrees of lowered energy levels, coinciding with the experimental fluorescence and phosphorescence emission redshift (Fig. 4b and Supplementary Table 5). Meanwhile, the ΔE(S1 − T6), ΔE(S1 − T5), ΔE(S1 − T4), ΔE(S1 − T3), and ΔE(S1 − T2) decreased after pressure treatment (Fig. 4c, d and Supplementary Table 6). Based on the energy band theory45,$${K}_{{{\rm{ISC}}}}\propto {\xi }^{2}*{\exp [-({\varDelta E}_{{{\rm{ST}}}})^{2}]}$$
(1)
a reduced ΔEST would gift a higher degree of bright triplet excitons through an accelerated ISC process (where the KISC is the ISC rate constant). Undoubtedly, the ISC rate was expected to be further expedited followed by these reduced energy gaps.In this regard, the pressure treatment engineering synergistically improved SOC and lowered the ΔEST via strengthened hydrogen bonds, as well as parallel-displaced π–π staking motif, leading to an efficient spin-flipping process. Moreover, the strengthened hydrogen bonds would also suppress the non-radiative dissipation, benefiting the emission enhancement. In order to quantitatively analyze the effects of the pressure treatment engineering on ISC and radiative transition, we calculated the ISC, radiative, and non-radiative decay rates in the light of the standard methods using the measured quantum yields and lifetimes (Supplementary Table 1). According to the following equation16,$${\varPhi }_{{{\rm{P}}}}={\varPhi }_{{{\rm{ISC}}}}\frac{{K}_{{{\rm{P}}}}}{{K}_{{{\rm{P}}}}+{K}_{{{\rm{nr}}}}}$$
(2)
$${\tau }_{{{\rm{P}}}}=\frac{1}{{K}_{{{\rm{P}}}}+{K}_{{{\rm{nr}}}}}$$
(3)
After pressure treatment, the estimated rate constant of the phosphorescence non-radiative decay (Knr) remarkably decreased from 1.047 s−1 to 0.667 s−1. The significant increase of KP from 0.079 s−1 to 0.444 s−1 made a further contribution to the higher ΦP (40%) in the targeted sample. Meanwhile, the fluorescent non-radiative decay Knr′ remained a considerable decrease from 0.119 ns−1 to 0.037 ns−1 after an entire compression cycle of 1 atm to 20.0 GPa, resulting in the enhancement of ΦF from 12% to 35%. Moreover, we received a rapid ISC rate (0.059 ns−1) in the pressure-treated sample from the original 0.010 ns−1, which sufficiently triggered the balanced contribution of singlet and triplet excitons to the white-light emission.