Fluorescence, Phosphorescence, or Delayed Fluorescence?âA

Subscriber access provided by University of South Dakota

C: Plasmonics; Optical, Magnetic, and Hybrid Materials

Fluorescence, Phosphorescence, or Delayed Fluorescence? – a Theoretical Exploration on the Reason Why a Series of Similar Organic Molecules Exhibit Different Luminescence Types Yingchen Duan, Li-Li Wen, Ying Gao, Yong Wu, Liang Zhao, Yun Geng, Guo-Gang Shan, Min Zhang, and Zhong-Min Su J. Phys. Chem. C, Just Accepted Manuscript • DOI: 10.1021/acs.jpcc.8b06533 • Publication Date (Web): 18 Sep 2018 Downloaded from http://pubs.acs.org on September 19, 2018

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 24 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

Fluorescence, Phosphorescence, or Delayed Fluorescence? – a Theoretical Exploration on the Reason Why a Series of Similar Organic Molecules Exhibit Different Luminescence Types Ying-Chen Duan†, Li-Li Wen†, Ying Gao†, Yong Wu‡, Liang Zhao†, Yun Geng*†, Guo-Gang Shan*†, Min Zhang† and Zhong-Min Su*†

†

Institute of Functional Material Chemistry, Faculty of Chemistry, Northeast Normal University, Changchun, P. R. China ‡

School of Pharmaceutical Sciences, Changchun University of Chinese Medicine, 1035 Boshuo Road, Changchun, P. R. China E-mail: [email protected]; [email protected]; [email protected] Abstract

In contrast to the traditional view that the small organic molecules emit fluorescence, more and more experiments manifest their special luminescence types, such as the thermally activated delayed fluorescence (TADF) and room-temperature phosphorescence. Why the similar organic molecules exhibit different luminescence types is focused on in this work based on density functional theory/time dependent density functional theory (DFT/TDDFT) calculations on a series of small organic molecules with phenoxazine or carbazole as donor and diphenyl-triazine as acceptor. The deep analysis of the geometrical and electronic structures shows how the structure, especially for the donor-acceptor dihedral angle, determines the singlet-triplet energy gap and the property of excited state. The explorations on the electron-hole pairs of natural transition orbitals and the contribution of the key heteroatom (N) to different molecular orbitals reveal the distinct electron transition processes of excitation to singlet and triplet states, explain the reason for different energy level distribution of excited states, and identify which pairs have more favorable intersystem crossing for these molecules. While the calculations of spin-orbit coupling and reorganization energy display the efficiency of the different luminescence type. Meanwhile, considering the potential application of TADF materials in organic light emitting diodes, we also respectively modified the phosphorescent molecule and the prompt fluorescent molecule through introducing methyls to increase the steric hindrance and realize the perpendicular orientation of donor and acceptor unit, and finally to screen the excellent TADF molecules 1 / 24

ACS Paragon Plus Environment

The Journal of Physical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

theoretically.

Introduction In recent years, the luminescence type of small organic molecules has been developed from traditional fluorescence to delayed fluorescence and phosphorescence. Thermally activated delayed fluorescence (TADF)1, 2 materials which perform up-conversion through reverse intersystem crossing (RISC) from triplet excitons achieve as high internal quantum efficiency of 100% as phosphorescent transition metal complexes. The non-noble metal Cu complexes or small organic molecules as common emitting materials in this mechanism have the advantages of both low cost and high quantum efficiency3-7. Phosphorescence materials with long lifetime could eliminate the interference of background fluorescence8, and the discovery of phosphorescent pure organic molecules enriches this sort of materials9-11. Thanks to the advantage in mass production, these two types of materials have favorable prospect in practical application. Compared with the traditional fluorescence molecules, TADF and pure organic phosphorescence molecules have higher internal quantum efficiency because the utilization of triplet excitons. They generally have stronger spin-orbit coupling (SOC) than traditional molecules, and have heteroatom such as N or O assisting the intersystem crossing (ISC) and RISC between excited singlet and triplet states5-7, 9-15. However, their luminescent mechanisms are different. The former radiates after the excited electron undergoes ISC from singlet to triplet state and RISC back to singlet state, while the latter radiates directly after the excited electron performs ISC to the triplet state at room temperature. What’s the difference between them on geometrical and electronic structures? What’s the reason for the different properties of their triplet states? And which factors influence the final luminescence type? These questions arouse our interests to explore their common and different properties from theoretical perspective although there have been works focusing on their respective luminescent mechanism either in experiment or on theory16-21. Based on the questions above, we chose a series of small molecules with similar geometry structures but different luminescence types (see Figure 1) to conduct a theoretical research. The phenothiazine-2,4,6-triphenyl-1,3,5-triazine (PXZ-TRZ) designed by Chihaya Adachi’s group is an efficient TADF molecule22. The 4,6-diphenyl-2-carbazole-1,3,5-triazine (DPhCzT) synthesized by Wei Huang’s group was identified to emit ultralong phosphorescence, whose H-type aggregation was speculated to extend the lifetime of the excited states. The 9-(4-(4,6-diphenyl-1,3,5-triazin-2-yl)phenyl)-9H-carbazole (Cz-TRZ) reported by Mounggon Kim et al. is a traditional fluorescent molecule23. Additionally, we also add 4,6-diphenyl-2-phenothiazine-1,3,5-triazine (DPhPXZT) to have a theoretical comparison. Because both TADF and room-temperature phosphorescence involve the 2 / 24


Page 2 of 24

Page 3 of 24 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


crossing process between singlet and triplet states, we detailedly investigated ISC/RISC and their determining factors, expecting to ascertain why these molecules have different luminescence types. Furthermore, organic light emitting diodes (OLEDs) is a significant application of organic semiconductor molecules, whose widespread use is still restricted by the efficiency and stability of the luminescent materials. Because of the high internal quantum efficiency of TADF materials, we predict they have tremendous potential application prospect in OLEDs. According to the theoretical analysis in this article, we attempt to modify the low efficient normal fluorescent molecule and the phosphorescent molecule and make them turn to TADF molecules. Therefore, we designed a series of molecules through introducing methyls to enlarge steric hindrance and change the structure, and finally adjust the energy difference between the lowest singlet and triplet excited state (S1 and T1). At the same time, we assessed their TADF performances from theoretical perspective and synthesized a representative compound to prove the strategy.

Figure 1. Molecule structures of PXZ-TRZ, DPhCzT, Cz-TRZ and DPhPXZT.

Methodology and computational details Geometrical and electronic structures of all S0 state molecules were calculated in Gaussian 09 program package24. Considering the calculations are for pure organic molecules, we adopted the 6-31G* basis set which is enough to describe the common organic small molecules25, 26. As we know it, the long-range corrected hybrid functional CAM-B3LYP is suitable for describing the charge transfer excited state27, so we used it to carry out the geometry calculation of excited states taking their distinct charge transfer characterization into account. We should mention that in order 3 / 24



Page 4 of 24

to reduce the errors during the structure comparison between S0 and excited states and the calculation of reorganization energy, we employed the CAM-B3LYP functional in S0 state calculation as well. In addition, the S0 geometries optimized by functional CAM-B3LYP basically match the X-ray diffraction crystal structure obtained from the experimental reports22, 28 and the results are close to those calculated by functional B3LYP which is commonly used for organic small molecules (see Table S1), certifying the reliability of the chosen functional. Because the energy differences of the excited states are crucial to estimate the RISC ability, we should use appropriate functional to obtain relatively accurate excitation energy. However, the conventional density functional often seriously underestimates the charge transfer transition energy29, 30, which goes against our calculation of the molecules with obvious donor and acceptor fragments. Range-separated exchange (RS) density functionals can lead to a balance between the electron correlation that describes the delocalization features and the exact-exchange effect that depicts the localization features by searching for the optimal range-separation parameter ω, thus the electronic structures could be described accurately31, 32. According to the research of Jean-Luc Brédas’s group33, long-range corrected functional LC-ωPBE can make accurate prediction of ionization potentials (IPs) and electron affinities (EAs). Furthermore, the Tamm-Dancoff approximation (TDA) method34 can mingle the part of configuration interaction with single excitations (CIS) which overestimates the excitation energy to offset the underestimation of charge transfer transition energy in triplet excitation energy calculation. In this way, more accurate singlet-triplet energy gap ∆EST could be obtained. Hence we adopted TDA-LC-ωPBE/6-31+G* to conduct all electronic structure calculation of the excited states. The exchange term is divided into a short-range domain characterized by density functional theory (DFT) [the former on the right of formula (1)] and a long-range domain characterized by Hartree-Fock (HF) [the latter on the right of formula (1) ] :

=

∙ ( )

+

∙ ( )

(1)

Here r12 represents the interelectronic distance, α and β are constants (0≤α≤1，0≤β≤1), erf(ωr12) denotes the error function, and ω is the reciprocal of the demarcation point where the exchange term separates into long-range and short-range domains. In order to get the optimal ω value, ω must make the exact functional obey the exact Kohn− Sham (KS) or generalized KS (GKS) theory35, which means for an N-electron system, the opposite number of the HOMO energy value must equal the vertical IP, and J = |εH(N) + IP(N)| should be as small as possible36. For a donor-acceptor system, in addition to the IP that close relates to the donor fragment, the vertical EA in connection with the acceptor should be considered as well. And the latter of an N-electron system could be taken as the IP of an (N+1)-electron system37. Therefore, 4 / 24


Page 5 of 24 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


we should adjust ω to make the value of the following equation as small as possible:

= ∑ ( + ) + IP( + i)

(2)

Figure S1 shows the relation between ω and the sum of HOMO energy and ionization energy deviation J2 at the electronic state of N and (N+1) electrons for PXZ-TRZ, DPhCzT, Cz-TRZ and DPhPXZT. It is seen that the deviations are minimalized when the ω values are near 0.18. On the basis of the diagram we can have further calculation and obtain more precise ω values. The J2 values corresponding to the ω values are presented in Table S2 and the optimal ω values are screened out. We also use this way to find the optimal ω values of the modified molecules. The SOC matrix elements were calculated in ADF2013.01 program package38 and the zeroth-order regular approximation (ZORA) method39 was used to consider the relativistic effects for all atoms under all electron TZP basis set. All overlap integrals of the norms of HOMO and LUMO, overlap integrals of the norms of natural transition orbitals (NTOs) and the centroid distance between NTOs were obtained in Multiwfn program40. The reorganization energy from S1 to S0 and from T1 to S0 were computed in DUSHIN program package41, 42. Furthermore, we should note in passing that the purpose of this article is to explore the basic mechanism for different luminescence types at the level of single molecule, the intermolecular interaction or the packing effect is not involved herein.

Results and discussion 1. Comparative analysis of the molecular luminescence properties 1.1Geometric structure and frontier molecular orbital Table 1 The bond lengths (Å) and dihedral angles (°) between donor and acceptor at optimized S0, S1 and T1 states, respectively, together with their difference (∆) at the CAM-B3LYP/6-31G* level. bond length / S1 ∆(S1-S0) T1 ∆(T1-S0) S0 dihedral angle 1.426 1.434 0.008 1.404 -0.022 PXZ-TRZ N2-C3 1.382 1.466 0.084 1.389 0.007 N2-C3 DPhCzT 1.413 1.412 0.000 1.413 0.000 N2-C3 Cz-TRZ 1.379 1.477 0.098 1.486 0.106 DPhPXZT N2-C3 4.03 124.13 38.45 PXZ-TRZ C1-N2-C3-C4 85.68 89.71 C1-N2-C3-N4 -17.65 -37.90 -20.25 -20.20 -2.55 DPhCzT 0.13 52.67 0.47 C1-N2-C3-C4 52.21 52.34 Cz-TRZ -80.04 -64.32 -55.14 DPhPXZT C1-N2-C3-N4 -9.18 -89.22 5 / 24



As our previous work certifies, geometric and electronic structures have great influence on the energy gap between singlet and triplet and the relevant ISC and RISC processes43, thus they are fully analyzed firstly in this work. For the four investigated molecules, the bond lengths and dihedral angles between donor and acceptor units at the S0 and excited states as well as the variation are listed in Table 1. It shows both the N2-C3 bond lengths and C1-N2-C3-N4 dihedral angles in PXZ-TRZ and Cz-TRZ are larger than those in DPhCzT and DPhPXZT. The larger dihedral angles in PXZ-TRZ and Cz-TRZ are ascribed to the steric hindrance between the H atoms in central phenyl and those in phenothiazine or carbazole. For DPhCzT and DPhPXZT, the formation of π bond between C3 and N2 atoms leads to their shorter bond length, which is related to the electron in p orbital of C3 atom vertical to the plane of triazine as well as the lone pair electrons of N2 atom perpendicular to the plane consisting of N2 atom and the three adjacent C atoms. This is clearly illustrated by the HOMO and LUMO maps in Figure 2, wherein the distinct bonding π orbital between C3 and N2 in DPhCzT and DPhPXZT and n orbital in PXZ-TRZ and Cz-TRZ are observed. Comparing with the S0 state structures, the geometries at excited states (S1 and T1) perform diverse changes, which may result in their different luminescence styles to some extent. Firstly, for the TADF molecule PXZ-TRZ and traditional fluorescence molecule Cz-TRZ, the structures of S1 state change little compared with S0 state referring to the small ∆(S1-S0) in both bond length and dihedral angle (see Table 1), which may favor fluorescence emission from S1 to S0. However, for PXZ-TRZ the structure of T1 state changes a lot, especially the C1-N2-C3-N4 dihedral angle, which may hinder its phosphorescence emission from T1 to S0 since the geometry relaxation between S0 and excited states would increase non-radiation energy dissipation. Secondly, for the organic phosphorescence molecule DPhCzT, S1 structure changes more than T1 structure according to the much smaller variations in N2-C3 bond length and C1-N2-C3-N4 dihedral angle at T1 state than those at S1 state compared with S0 state, which indicate its more favorable phosphorescence emission from T1 to S0. Finally, does the little change in geometry for both S1 and T1 compared with S0 state impart the favorable fluorescence and phosphorescence for Cz-TRZ? It will be further explained in the next section.

Figure 2. Calculated HOMO and LUMO diagram of the investigated molecules. 6 / 24


Page 6 of 24

Page 7 of 24 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


It is well known that the electronic structure has a dominant influence on the photoluminescence property, thereby it is important to have a clear insight on the frontier molecular orbitals of these compounds. It is illustrated in Figure 2 that the HOMOs mainly locate at phenothiazine or carbazole, and the LUMOs mainly delocalize at triazine and benzene rings. For the TADF molecule PXZ-TRZ, the small HOMO-LUMO overlap is resulted from the near orthogonal orientation of phenothiazine relative to triazine and benzene ring. For the HOMOs of DPhCzT and DPhPXZT, there are still some distributions at triazine in addition to carbazole or phenothiazine induced by the smaller C1-N2-C3-C4 dihedral angle, which increase their respective HOMO-LUMO overlaps. For Cz-TRZ, the LUMO distribution resembles that of PXZ-TRZ, but the relative smaller C1-N2-C3-C4 dihedral angle than PXZ-TRZ gives rise to its overlapped distribution with HOMO at benzene ring. One of the frontier molecular orbital distinctions of these four compounds is their respective HOMO-LUMO overlaps, which may cause different photoluminescence properties. Decreasing as much as possible the HOMO-LUMO overlap is favorable to getting smaller ∆EST, which is one important characteristic of TADF materials44. Therefore, we calculated the orbital overlap integral 〈ΨH|ΨL〉 for all compounds and list them in Table S3 to quantitatively scale the HOMO-LUMO overlap. We see PXZ-TRZ has the smallest overlap integral value among these four compounds, which drops its ∆EST and promotes the RISC. This brings us an inspiration that we could modify the molecules to increase the steric hindrance in other ways to turn the donor-acceptor relative position from planar to nearly perpendicular and obtain TADF features. While Cz-TRZ showing the highest overlap integral is suspected to related with its traditional fluorescence. The DPhCzT and DPhPXZT where the benzene ring is absent also show much larger overlap integral than PXZ-TRZ. 1.2 The excited-state property The excited-state properties such as excitation energy, transition configuration and hole-electron distribution at excited state could directly reflect the luminescence of one compound. Firstly, the ∆EST value influences the ISC and RISC ability, which imparts whether it is easy to reach triplet excited state and go back. Thus the excitation energies of S1 and T1 as well as the corresponding vertical ∆EST values for each molecule were calculated and the results are collected in Table 2. Table 2 Calculated excitation energy of S1 (ES/eV) and T1 (ET/eV) as well as the vertical singlet-triplet energy gap ∆EST(eV) for each molecule in S0 geometry with their optimal ω values at the LC-ωPBE/6-31+G* level. PXZ-TRZ DPhCzT Cz-TRZ DPhPXZT ω 0.1845 0.1785 0.1798 0.1760 ES 2.904 3.839 3.678 3.680 ET 2.883 3.466 3.146 3.381 ∆EST 0.015 0.373 0.532 0.300 7 / 24



Comparing the results, we find that the energy gap of PXZ-TRZ is an order of magnitude smaller than the other three molecules, and the ∆EST values of the latters do not meet the qualification of typical TADF molecules for which the ∆EST is less than 0.1 eV45, although the values are also pretty small (< 0.6 eV). It is noted in passing that our calculated ∆EST value of PXZ-TRZ is close to its calculated value in the report of Jean-Luc Brédas’s group as well as the experimental value, and their order of magnitudes are the same46, 47. According to the equation of the ∆EST expressed by the exchange energy J44, the minimization of the ∆EST can be realized by decreasing the HOMO-LUMO overlap. Therefore, the separation of HOMO and LUMO significantly reduces the ∆EST of PXZ-TRZ, while Cz-TRZ has the largest ∆EST among these molecules, which results from the more HOMO-LUMO overlap in Cz-TRZ than in DPhPXZT or DPhCzT. For small organic molecules, when ∆EST are less than 0.4 eV and the SOC degrees between singlet excited states and triplet excited states are similar, a reduction of one order of magnitude in ∆EST will lead to a rise of several orders of magnitude in RISC rate46. Thus the SOC effect of PXZ-TRZ would have weak influence on its RISC process because of its so tiny ∆EST (

Fluorescence, Phosphorescence, or Delayed Fluorescence?âA

Recommend Documents

Fluorescence, Phosphorescence, or Delayed Fluorescence?âA