Regulation of Thermally Activated Delayed Fluorescence to Room-Temperature Phosphorescent Emission Channels by Controlling the Excited-States Dynamics via J- and H-Aggregation

Article abstract of DOI:10.1002/anie.202103192

Control of excited-state dynamics is key in tuning room-temperature phosphorescence (RTP) and thermally activated delayed fluorescence (TADF) emissions but is challenging for organic luminescent materials (OLMs). We show the regulation of TADF and RTP emissions of a boron difluoride β-acetylnaphthalene chelate (βCBF₂) by controlling the excited-state dynamics via its J- and H-aggregation states. Two crystalline polymorphs emitting green and red light have been controllably obtained. Although both monoclinic, the green and red crystals are dominated by J- and H-aggregation, respectively, owing to different molecular packing arrangements. J-aggregation significantly reduces the energy gap between the lowest singlet and triplet excited states for ultra-fast reverse intersystem crossing (RISC) and enhances the radiative singlet decay, together leading to TADF. The H-aggregation accelerates the ISC and suppresses the radiative singlet decay, helping to stabilize the triplet exciton for RTP.

Full text of DOI:10.1002/anie.202103192

A Journal of the Gesellschaft Deutscher Chemiker
Angewandte
Chemie
International Edition
www.angewandte.org
Accepted Article
Title: Regulation of Thermally Activated Delayed Fluorescence
to Room Temperature Phosphorescent Emission Channels
by Controlling the Excited-States Dynamics via J- and H-
Aggregation
Authors: Hongbing Fu, Shuai Li, Liyuan Fu, Xiaoxiao Xiao, Hua Geng,
Qing Liao, and Yi Liao
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: Angew. Chem. Int. Ed. 10.1002/anie.202103192
Link to VoR: https://doi.org/10.1002/anie.202103192

10.1002/anie.202103192
Angewandte Chemie International Edition
RESEARCH ARTICLE
Regulation of Thermally Activated Delayed Fluorescence to Room
Temperature Phosphorescent Emission Channels by Controlling
the Excited-States Dynamics via J- and H-Aggregation
Shuai Li†,[a,b] Liyuan Fu†,[b] Xiaoxiao Xiao,[b] Hua Geng,[b] Qing Liao,[b] Yi Liao,[b] and Hongbing Fu*[a,b]
[a]
Dr. S. Li, Prof. H. B. Fu
Institute of Molecule Plus, Scholl of Chemical Engineering and Technology, Tianjin University, and Collaborative Innovation Center of Chemical Science
and Engineering (Tianjin), Tianjin 300072, China
[b]
Dr. S. Li, L. Fu, X. Xiao, Prof. H. Geng, Prof. Q. Liao, Prof. Y. Liao, Prof. H. B. Fu
Beijing Key Laboratory for Optical Materials and Photonic Devices, Department of Chemistry, Beijing Advanced Innovation Center for Imaging Technology,
Capital Normal University, Beijing 100048, China
E-mail: hbfu@cnu.edu.cn
[†]
These authors contributed equally to this work.
Supporting information for this article is given via a link at the end of the document.
Abstract: Control of excited-state dynamics plays a central role in
tuning room-temperature phosphorescence (RTP) and thermally
activated delayed fluorescence (TADF) emissions, but remaining
challenging in the exploration of organic luminescent materials
(OLMs). Herein, we demonstrated the regulation of TADF and RTP
emissions of a boron difluoride β-acetylnaphthalene chelate (βCBF2)
by controlling the excited-state dynamics via its J- and H-
aggregation states. Two crystalline polymorphs emitting green and
red light have been controllably obtained. Although both belong to
the monoclinic system, the green and red crystals are dominated by
J- and H-aggregation, respectively, due to different molecular
packing arrangements. Experimental and theoretical studies show
that J-aggregation not only significantly reduces the energy gap
between the lowest singlet and triplet excited states for the
realization of ultra-fast reverse intersystem crossing (RISC) process,
but also enhances the radiative decay of singlet, together leading to
TADF. The H-aggregation accelerates the ISC and meanwhile
suppresses the radiative decay of singlet, helping to stabilize the
triplet exciton for RTP. These results deepen the understanding of
the effect of aggregation on the luminescence mechanism of organic
solids, and provide new strategy to tuning TADF and RTP OLMs.
where ξST is the spin–orbit coupling (SOC) constant and ∆EST
the energy gap between the involved singlet and triplet states, kB
is the Boltzmann constant, and T is temperature. In the past few
years, several methods have been developed to achieve organic
RTP: 1) introducing aromatic carbonyl groups, heteroatoms or
heavy halogen atoms to improve the SOC constant ξST thus to
accelerate ISC;[4] 2) using polymerization, metal-organic
framework coordination, and formation of rigid crystals to
effectively suppress the non-radiative decay of low-lying triplet
excitons.[5] In the meantime, the design of TADF emitters has
been proposed mainly based on electron donor-acceptor (D-A)
compounds, in which spatial separation of the highest occupied
and the lowest unoccupied molecular orbital (HOMO and LUMO)
effectively reduce the singlet-triplet splitting energy (ΔEST),
therefore facilitating up-conversion of the dark triplet exciton to
the luminescent singlet exciton through RISC.[6] Compared to
RTP, TADF requires fast RISC. That is, controlling of excited-
state dynamics plays a central role in tuning RTP and TADF
emissions. This, however, remains challenging in the exploration
of OLMS, owing to the limited understanding of luminescence
mechanisms.
Note that researchers worked on RTP pursued persistent
luminescence
for
imaging
and
anti-counterfeiting
Introduction
applications,[2c,7] while those worked on TADF preferred to
shorten the exciton lifetime in order to decrease the singlet-
triplet and triplet-triplet annihilation that leads to OLED efficiency
roll-off and device degradation.[8] Mostly, persistent RTP under
ambient condition has been achieved in solid (crystalline) states
by stabilization of the triplet state through H-aggregation and by
prevention of the penetration of oxygen in the solid-state,[2c,2e]
whereas TADF has been demonstrated exclusively in the guest-
host doping films. As a matter of fact, the relationship between
the properties of TADF and aggregate states remain largely
unexplored. Recently, Yang and co-workers demonstrated co-
existence of TADF and RTP dual emission pathways in the
crystal form, which are ascribed to separated monomer and
dimer states, respectively.[9] Chi and co-workers reported
dynamical triplet-exciton behaviors arising from different
conformers in multiple crystalline polymorphs, including RTP,
Organic luminescent materials (OLMs) have attracted
extensive attention of researchers due to their unique electronic
structures and optical properties.[1] In particular, room-
temperature phosphorescence (RTP) and thermally activated
delayed fluorescence (TADF) materials are of great importance
for applications in organic light emitting diodes (OLEDs),
sensors,
bio-imaging
and
anti-counterfeiting
agents.[2]
Phosphorescence is generated through two consecutive
intersystem crossing (ISC) steps, i.e., S1→T1 non-radiative ISC
and T1→S0 radiative transition, which are both spin-forbidden
processes. Differently, up-conversion from T1 to S1 through
reverse intersystem crossing (RISC) gives rise to TADF.
According to the perturbation theory, the rate of ISC (or RISC),
kISC(RISC), is expressed by the equation (1),[3]
TADF,
mechanochromic
luminescence
(MCL),
and
mechanoresponsive luminescence (MRL).[10] It is widely
accepted that molecular aggregates of different packing
arrangements determine the photophysics of OLMs. Therefore,
푘퐼푆퐶(푅퐼푆퐶) ∝ 2 exp (−△퐸
(1)
푆푇
)
T
푆푇
푘
퐵
1
This article is protected by copyright. All rights reserved.

10.1002/anie.202103192
Angewandte Chemie International Edition
RESEARCH ARTICLE
exploration of the role of molecular aggregates in controlling
RTP and TADF emissions is highly important.[11] In this paper,
we realized the regulation of TADF and RTP properties through
the polymorphism of organic molecules, and revealed the
relationship between crystal stacking and triplet excitons. We
have designed and synthesized a new molecule (E)-1-(((4-
nitrophenyl)imino)methyl)naphthalen-2-olate•BF2, named βCBF2,
(Figure 1a), with the difluoroboron naphthalene as an electron
donor (D) moiety and the nitrobenzene as an electron acceptor
(A) moiety. Two crystalline polymorphs emitting green and red
light have been controllably obtained. Although both belong to
the monoclinic system, the green and red crystals are dominated
by J- and H-aggregation, respectively, due to different molecular
packing arrangements. Experimental and theoretical studies
show that J-aggregation not only significantly reduces the
energy gap between the lowest singlet and triplet excited states
for the realization of RISC process, but also enhances the
radiative decay of singlet for the realization of TADF. The H-
aggregation accelerates the ISC and meanwhile suppresses the
radiative decay of singlet, helping to stabilize the triplet exciton
for RTP. Our results provide insights into the modulation of
TADF or RTP behavior through J- and H-aggregates.
polymethylmethacrylate (PMMA) matrix further confirmed the
TADF properties (Figure S5). Note that the PL spectrum at 77 K
exhibited multimodal emissions at 465, 488 nm and 590, 640 nm
(orange line in Figure 1b and Figure S3b). The TRPL spectrum
at 77 K on a nanosecond time scale shows an emission
maximum approximately at 488 nm (green line in Figure S4c),
which decays with a lifetime of 1.90 ± 0.015 ns (blue line in
Figure 1d) and thus can be attributed to the fluorescence
emission from S1 to S0. Furthermore, the TRPL spectrum
recorded with a delay time of 10 ms presents the main emission
peak at 590 nm and a shoulder at 640 nm (red line in Figure
S4c), which decay with a lifetime as long as 976 µs (inset in
Figure 1d) and is ascribed to the phosphorescent emission from
T1 to S0. Therefore, the S1 and T1 energies can be estimated
from the fluorescence and phosphorescence spectra at 77 K
and are determined to be 2.65 and 2.1 eV, respectively (Table 1),
giving rise to a ΔEST as large as 0.55 eV. Note that such a large
energy gap between S1 and T1 makes the RISC directly from T1
to S1 almost impossible. As shown below by theoretical
calculation results, TADF of βCBF2 in dilute DCM might be
enabled by an ultra-fast RISC between high-lying Tn and S1
states.[12]
Results and Discussion
The UV/Vis absorption and photoluminescence (PL)
spectrum of βCBF2 in dilute dichloromethane (DCM) solution
(10-5 M) at room temperature are shown in Figure 1b. Table 1
summarizes the related photophysical data. The absorption
spectrum (black line) display a strong absorption peak at 412 nm
with molar extinction coefficient (ε = 13030 L M-1 cm-1, Figure
S1), and a shoulder at 354 nm. The PL spectrum (green line)
shows a structureless peak around 488 nm with a quantum yield
Φ = 401%. In order to further study the properties of excited
states, we measured the time-resolved PL (TRPL) using a
streak camera (see Figure S2). The PL at 488 nm exhibits a bi-
exponential decay in air, with a short lifetime of 1.16 ± 0.012 ns
and a long lifetime of 0.73 ± 0.03 µs, respectively (see the inset
in Figure 1c also Figure S3a). Moreover, the prompt spectrum is
similar to the delayed one (Figure S2), and both are identical to
the steady-state PL spectrum (green line in Figure 1b).
Therefore, the PL of βCBF2 DCM solution exhibits TADF.
Moreover, in the case of the de-oxygen conditions, the lifetime of
the delayed component was extended to 7.83 ± 0.008 µs (green
line in Figure 1c and Figure S3b), revealing the sensitivity of the
triplet exciton to oxygen.
To further confirm the occurrence of TADF, we measured
the temperature-dependent steady-state PL and TRPL. First of
all, the PL intensity decreased significantly with the decrease of
temperature (Figure S4a). And the PL intensity at 77 K is only
about 1/5 of that at 298 K. Meanwhile, the long-lived component
in the PL decay decreases with decreasing the temperature
(Figure S4d and orange line in Figure 1c for 77 K), confirming
the TADF observed at 298 K (green line in Figure 1c). The
photophysical behavior of a thin-film with molecules doped in the
Figure 1. a) Chemical structures of βCBF2. Inset show photographs of βCBF2
solution in DCM under 365 nm ultraviolet lamp at room temperature and 77 K.
b) Steady-state absorption and PL spectra of βCBF2 in dilute DCM solutions.
The green and orange lines represent PL spectra recorded at RT and 77 K. c)
The PL decay of βCBF2 in DCM without O2 at RT (green line) and 77 K
(orange line). Inset: PL decay curves of βCBF2:DCM solution in air at room
temperature. d) PL decay and fitted curves of fluorescence at 488 nm (blue
line) and phosphorescence at 590 nm (red line) at 77k. e) Energy level
diagrams and spin–orbit coupling (SOC) constant for βCBF2. f) Hole-electron
analysis (holes and electrons are shown in lilac and green, respectively) of the
S1, T1, and T2 for βCBF2 in DCM based on the optimized geometries.
2
This article is protected by copyright. All rights reserved.

10.1002/anie.202103192
Angewandte Chemie International Edition
RESEARCH ARTICLE
Figure 2. a) PL spectra of green (green line) and red crystals (red line), respectively. Inset: photographs of the green and red crystals under UV excitation. b, c)
Transient PL decay of green (b) and red crystals (c) at RT (blue line) and 77 K (red line). Inset: PL decay curve of fluorescence emission measured at 510 nm.
Table 1. Photophysical parameters of βCBF2 monomers in the dilute solution and solid-state aggregation.
Sample
solution
λF [nm]
λPhos [nm]
τF [ns]
τDF [μs]
τPhos [μs]
kr [s-1][b]
kISC [s-1][b]
ES/ET [eV][c]
ΔEST [eV][d]
Φpl [%]
401
590[a]
1.16
0.73
976[a]
2.54/2.10
0.44
2.04×108
3.52×108
488
3.04×108
6.31×105
6.18×108
2.44×109
green crystals
red crystals
514
505
558
598
0.56
0.41
0.93
－
2.41/2.22
2.45/2.07
0.19
0.38
261
－
9 μs
2.81
[a] Phosphorescence spectra and lifetime measured at 77k. [b] kr, and kSC represent the rate constant of radiative (S1→S0), intersystem-crossing (ISC) rate
constant (S1→T1), respectively. [c] Singlet (ES) and triplet (ET) energies estimated from the emission spectra at 77 K respectively. [d] ΔEST = ES- ET.
Density functional theory (DFT) and time-dependent DFT
(TD-DFT) at B3LYP/6-31G(d) level combined with polarization
continuum model (PCM) were used for theoretical calculation to
explore the mechanism of TADF characteristics of βCBF2
molecule. The ground-state structure exhibit a dihedral angle ()
between the difluoroboron naphthalene plane and the
nitrobenzene plane about  = 36.9 (Table S1). And the
calculated vertical energy is 3.09 eV, consistent with the
absorption maximum at 412 nm. Upon geometry optimization,
the adiabatic energies of S1 and T1 have been determined to be
2.60 and 1.94 eV (Figure 1e), corresponding well to F = 488 nm
(2.54 eV) and Phors = 590 nm (2.10 eV) (see Table 1). These
results validate the calculation methods adopted in our study. It
can be seen from Figure 1e that there are two triplet excited
states (T1 and T2) below S1. The adiabatic energy difference
between S1 and T2 is only 0.11 eV, allowing the RISC process to
occur between S1 and T2. According to the analysis of the
density distribution of "hole" and "particle" (Figure 1f), in the S1
state, the hole distribution is mainly controlled by the
naphthalene ring donor part, and the electron distribution is
mainly located in the acceptor unit of nitrobenzene, exhibiting
obvious CT excitation characteristics. This CT characteristic is
further confirmed by the positive solvatochromism, i.e., the
gradually red-shifted PL peaks with increasing the polarity of the
solvents (Figure S6). In contrast, the T1 state is actually locally
excited (LE) in nature. Notably, the T2 state demonstrates a
coexistence of LE and CT features (HLECT), which is 0.11 eV
lower than S1 (see Figure 1f and Figure S7). Moreover, we found
that S1 and T2 have similar geometric structures with  = 26.4
and 27.7, respectively (Figure 1f and Table S1). The SOC value
of 1.40 cm−1 between S1 and T2 is much larger than that of 0.45
cm−1 between S1 and T1 (Figure 1e). Therefore, we speculate
that the effective RISC between T2↔S1 is responsible for
efficient TADF observed for βCBF2 molecules in DCM.
What is interesting is that slow evaporation of the solution of
βCBF2 in DCM at 5 and 30 C generates two polymorph single-
crystals (see the supporting information for the detailed process
and Figure S8). Upon ultraviolet light (330-380 nm) excitation,
these two polymorphs exhibit green and red emissions, as
indicated by their -PL images of the insets of Figure 2a. Single-
crystal X-ray diffraction (XRD) measurements indicate that they
belong to two different polymorph crystals (Figure S9). Figure 2a
depicts the PL spectra of pure-phase green and red crystals.
Compared with the PL spectrum of dilute solution, the PL of
green crystals show a slightly red-shifted maximum at 514 nm
with Φ = 261% (Table 1).[13] In sharp contrast, the PL of red
crystals exhibit a maximum at 598 nm, which is very similar to
the phosphorescence spectrum of the solution measured at 77 K
(Figure S4b), together with a weak peak around 505 nm;
meanwhile, the PL quantum yields is significantly reduced to Φ =
2.81%.
To verify the PL origin of the two crystals, we measured
temperature-dependent PL and TRPL spectra. For comparison,
their related photophysical parameters were also summarized in
Table 1. Similar to the DCM solution, the PL of green crystals
decays bi-exponentially with a prompt component of 0.560.008
ns (inset in Figure 2b) and a delayed component of 0.93  0.05
μs (blue line in Figure 2b). However, at 77 K, the delayed
3
This article is protected by copyright. All rights reserved.

10.1002/anie.202103192
Angewandte Chemie International Edition
RESEARCH ARTICLE
component significantly reduced (Figure S10b and red line in
Figure 2b); meanwhile, the PL intensity at 77 K decreases to
about 1/2 of that at 298 K (Figure S10a). Therefore, the green
crystals of βCBF2 also exhibit TADF just like its dilute solution.[14]
Notably, the phosphorescence of green crystals measured at 77
K by setting a delay gate of 10 s was found at 558 nm with a
lifetime of 0.05  0.008 ms (Figure S10c and d), giving rise to
ΔES-T = 0.19 eV, much smaller than that of monomers in the
dilute solution (Table 1).
crystals that possess a density of (1.209 g /cm3), while planar
structure give rise to tight molecular packing in red crystals as
evidenced by a larger density of 1.536 g /cm3 (Table S2).
It is widely accepted that the optical and electronic
properties of solid-state OLMs is significantly influenced by
molecular aggregate states, which depend strongly on the
molecular packing arrangement. The Kasha’s exciton model
represents a milestone, leading to often found descriptions of
“head-to-tail” arrangement for J-aggregates and “face-to-face
arrangement” for H-aggregates in literatures.[15] As shown in
Figure 3c and 3d, βCBF2 molecules are packed in an anti-
parallel arrangement in both green and red crystals, forming
dimer-like structures. In green crystals, owing to large twisted
molecular conformation, the nitrobenzene units of the two
molecules are anti-parallelly packed by face to face with a
separation of 3.3 Å, while the difluoroboron naphthalene groups
point to opposite directions without any overlapping (labeled as
g-dimer, Figure S16), which seriously affects the π-π interaction
between adjacent molecules. The longitudinal displacement
between the two molecules is 4.99 Å (Figure S16), and the
intermolecular center-to-center distances (dc–c) is 8.84 Å (Figure
S17). We then used the energy splitting method to calculate the
exciton-coupling between neighbored molecules, which works
well for identical molecular dimer.[16] The sign of the coupling is
determined by the relative orientation of the transition dipole
moments within the selected molecular pair, that is, negative for
J-type and positive for H-type couplings. According to our
calculation results, the head-to-tail arrangement of g-dimer
exhibit a negative exciton coupling of Δε = -16 meV, which is
dominant among total seven selected molecular pairs (Figure
S18), suggesting moderate J-type coupling in green crystals. In
sharp contrast, it can be seen from Figure 3d that in red crystals,
planar βCBF2 molecules lead to anti-parallel packed dimers
(labeled as r-dimer) with a - distance of 3.39 Å (Figure S19).
And further crystal analysis revealed that the pitch angle was
90°, and there was no longitudinal displacement between
adjacent molecules, showing typical H-aggregate characteristics
(Figure S20). Indeed, the exciton coupling value of r-dimer is
calculated to be positive ca. +91 meV, much larger than other
selected dimers, indicating strong H-aggregation effect in red
crystals (see Figure S21).
In sharp contrast, the red crystals exhibited distinct RTP
phenomenon. As shown in Figure S12, the PL spectra of red
crystals remain the same with decreasing the temperature,
presenting a maximum around 598 nm; meanwhile, the PL
intensity at 77 K is significantly enhanced about 15 times (peak
top intensity) higher than that at 298 K (Figure S11a). It should
be noted that these temperature-dependent behaviors observed
for red crystals are completely different from characteristics of
TADF. Furthermore, the PL lifetime at 77 K is determined to be
0.44  0.02 ms (red line in Figure 2c), longer than 9  0.8 s at
298 K (blue line in Figure 2c). Therefore, the PL of red crystals
around 598 nm originates from phosphorescence. Notably, the
PL of red crystals also exhibit a very weak band around 505 nm,
which does not exhibit long-lived component (Figure S12) and
can be ascribed to fluorescence with a lifetime about 0.410.01
ns (inset of Figure 2c and Table 1). Therefore, the ΔES-T = 0.38
eV of red crystal is larger than that of green crystal but smaller
than that of the dilute solution (Table 1).
Figure 3. a, b) Single crystal structures of (a) green crystal and (b) red crystal.
To model the effect of solid-state environment, the quantum
mechanics/molecular mechanics (QM/MM) method was used to
evaluate a single molecule (QM level) embedded in the crystal
lattice, with the surrounding molecules defined as the rigid MM
part (Figure S22 a and b). We found that the geometric changes
of molecules in different electronic states (S0, S1 and T1)
characterized by the root mean square displacement (RMSD)
are quite small[17], indicating small geometry relaxation owing to
strong intermolecular interactions in crystals (Figure S22c).
Note that QM/MM method is not suitable for calculating
high-lying triplet Tn states for large systems such as dimers.
Therefore, DFT and TD-DFT-D3 calculations at B3LYP-D3/6-
31g(d) level were performed to calculated the vertical energy
levels of monomers and dimers extracted from green and red
single-crystal structures. It can be seen from left panels of
Figure 4a and 4b that both the calculated S1 and T1 energies of
c, d) The molecular packing mode of (c) green crystal and (d) red crystal.
In order to understand the conversion mechanism between
TADF and phosphorescence of βCBF2 in different polymorph
crystals, the single crystal structures were determined (CCDC
Nos. 2045851 and 2045853). In the green crystal, βCBF2 shows
a highly twisted molecular conformation with  = 40.4° (Figure
3a), close to  = 36.9 optimized for S0 in dilute solution. While in
red crystal, βCBF2 adopts a more planar conformation with  =
19.8° (Figure 3b), approximately similar to  = 21.6° optimized
for T1 in dilute solution. In both cases, there are abundant
intermolecular interactions that solidify molecular conformations
in green and red crystal, including π···π and X-H···π (X = B, F,
O) interactions (Figure S14). It should be noted that the distorted
molecular structure leads to loosely molecular packing in green
4
This article is protected by copyright. All rights reserved.

10.1002/anie.202103192
Angewandte Chemie International Edition
RESEARCH ARTICLE
twisted g-monomer ( = 40.4°) is blue-shifted than that of planar
r-monomer ( = 19.8°). This is consistent with the observation of
Phors of green crystal at 558 nm and Phors of red crystal at 598
nm (Table 1). However, the calculated ΔEST between S1 and T1
are similar between g-monomer (0.92 eV) and r-monomer (1.02
eV).
Remarkbly, the calculated emission oscillator strength (f) of
S1 states for g-dimer is 0.64, much larger than 0.006 for r-dimer,
suggesting enhanced radiative decay in g-crystal and
suppressed radiative decay in r-crystal according to the Einstein
spontaneous emission expression of 푘푟,퐹 = 푓 퐸푔2/1.499, where f
is the dimensionless oscillator strength and Eg is the energy gap
in units of cm-1.[18] It can be seen from Table 1 that the kr,F of g-
crystal (3.04×108 s-1) is indeed much larger than that of r-crystal
(6.31×105 s-1, Table 1), consistent with Kasha's theory. That is,
H-aggregates can suppress fluorescence, which facilitates
crossover between systems and enhances phosphorescence.
On the contrary, J-aggregates are beneficial to fluorescence.[19]
Based on the above discussion, we proposed energy
diagrams of Figure 4e and 4f. The twisted molecular
conformation induces the molecular arrangement of J-
aggregates in the g-crystal. The formation of J aggregates
significantly reduces ΔEST and accelerates the rate of radiative
decay kr,F. The accelerated kr,F deactivates those singlet excitons
generated through RISC process enabled by small ΔEST, leading
to TADF of g-crystal (Figure 4e). In sharp contrast, the nearly
planar molecular structure results in H-aggregation in r-crystal
(Figure 4f). Similar to the case of g-crystal, both ISC and RISC
might be fast in r-crystal. [The large number of transition
channels in r-crystal and related SOC values (0.65 cm-1 for S1→
T4, 0.26 cm-1 for S1→T6, 0.33 cm-1 for S1→T8, 0.93 cm-1 for S1→
T10,) even lead to enhanced ISC (Figure S23).] However, even
though the RISC process can take place, the suppressed
radiative decay of H-aggregates recycles those delayed
generated singlets into triplet manifold again via enhanced ISC,
and subsequent Internal conversion within triplet manifold helps
to stabilize T1 states. Finally, large SOC value of 4.16 cm-1 for T1
→S0 leads to an increase in the radiation rate T1 to S0 and thus
promotes phosphorescence emission in r-crystal (Figure 4f).
Figure 4. a, b) The vertical energy levels for monomer and dimer of green (a)
and red (b) based on TD-B3LYP-D3/6-31G(d) calculations in gas phase. c, d)
Nature transition orbital analysis of S1 state for the dimers of green (c) and red
(g). e, f) Energy level diagrams for g-crystal (e) and r-crystal (f) emissions.
Figure 4c show the frontier molecular orbitals (FMOs) of the
lowest singlet of green dimer. The hole and electron distributions
of S1 are located on the difluoroboron naphthalene and the
nitrobenzene parts, respectively, showing obvious CT feature. In
sharp contrast to the delocalization of S1 of the green dimer,
hole and electron distributions of S1 are more localized and
extended over the whole red dimer (Figure 4d), showing LE
characteristics, also is T1 (see Figure S25). As compared with
those of monomers (left panels of Figure 4a and 4b), the energy
level splitting in both g-dimer (right panel of Figure 4a) and r-
dimer (right panel of Figure 4b) caused by molecular
aggregation significantly reduces the energy level of S1, but has
little effect on T1. Moreover, the calculated ΔEST = 0.48 eV of g-
dimer between S1 and T1 is found to be smaller than that of ΔEST
= 0.68 eV of r-dimer, consistent with the experimental results
(Table 1). It should be noted that the number of high-lying triplet
states between ΔEST  ±0.3ev increases in both the cases of g-
and r-dimers (Figure 4a and b), suggesting that more ISC
channels are available in crystalline states. As a matter of fact,
the kISC values of both g-crystal and r-crystal are faster that of
the solution (Table 1). Therefore, consideration of only ISC and
RISC processes cannot explain the distinct difference between
TADF of g-crystal and RTP of r-crystal.
Conclusion
In summary, we demonstrated that the emission
characteristics of TADF and RTP are affected by the
aggregation state. We designed and synthesized a novel
organic molecule βCBF2, and obtained green and red crystalline
polymorphs through simple temperature control. Importantly, the
two polymorphs exhibit different photophysical processes, i.e.,
green crystals exhibit TADF properties and red crystals shows
the behavior of RTP. Single crystal data indicate that there are
significant differences in the molecular stacking arrangements
between the two types of polycrystalline, with green and red
crystals dominated by J- and H- aggregates, respectively.
Experimental and theoretical investigations further elucidate that
J-aggregates can significantly reduce ΔEST to achieve ultra-fast
RISC process, followed by enhanced radiative decay leading to
TADF. The H-aggregates accelerate the ISC and meanwhile
suppress the radiative decay of singlet, helping to stabilize the
triplet exciton for RTP. These results deepen the understanding
of the effect of aggregation on the luminescence mechanism,
and provide new strategy to tuning TADF and RTP OLMs.
5
This article is protected by copyright. All rights reserved.

10.1002/anie.202103192
Angewandte Chemie International Edition
RESEARCH ARTICLE
Zhou, S. Li, L. Fu, Y. Wu, C. Gu, Q. Liao, H. Fu, Acs Photonics 2019, 6,
3208-3214.
Acknowledgements
[13] W. Z. Yuan, P. Lu, S. Chen, J. W. Y. Lam, Z. Wang, Y. Liu, H. S. Kwok,
Y. Ma, B. Z. Tang, Adv. Mater. 2010, 22, 2159-2163.
This work was supported by the National Natural Science
Foundation of China (Grant Nos. 22090022, 21873065,
21833005 and 21790364), the Beijing Natural Science
Foundation of China (Grant No. 2162011 and 2192011), the
High-level Teachers in Beijing Municipal Universities in the
Period of 13th Five-year Plan (Grant No. IDHT20180517 and
CIT&TCD20180331), the Open Fund of the State Key
Laboratory of Integrated Optoelectronics (IOSKL2019KF01),
Capacity Building for Sci-Tech Innovation-Fundamental
[14] F. B. Dias, K. N. Bourdakos, V. Jankus, K. C. Moss, K. T. Kamtekar, V.
Bhalla, J. Santos, M. R. Bryce, A. P. Monkman, Adv. Mater. 2013, 25,
3707-3714.
[15] a) S. Ghosh, X.-Q. Li, V. Stepanenko, F. Wuerthner, Chem.- Eur. J.
2008, 14, 11343-11357; b) F. Wuerthner, T. E. Kaiser, C. R. Saha-
Moeller, Angew. Chem., Int. Ed. 2011, 50, 3376-3410.
[16] C.-P. Hsu, Acc. Chem. Res. 2009, 42, 509-518.
[17] T. Lu, F. W. Chen, J Comput Chem 2012, 33, 580-592.
[18] a) N. J. Hestand, F. C. Spano, J Chem Phys 2015, 143, 244707; b) N. J.
Hestand, F. C. Spano, Chem. Rev. 2018, 118, 7069-7163.
[19] Z. Shuai, D. Wang, Q. Peng, H. Geng, Acc. Chem. Res. 2014, 47,
3301-3309.
Scientific
Research
Funds
(025185305000/210
and
009/19530050162), Youth Innovative Research Team of Capital
Normal University (009/19530050148), Beijing Advanced
Innovation Center for Imaging Technology (009/19530011009),
the Ministry of Science and Technology of China (Grant No.
2013CB933500 and 2017YFA0204503).
Keywords: polymorphs • thermally activated delayed
fluorescence • room temperature phosphorescence • crystal
engineering • J- and H- aggregates
[1]
[2]
a) Z. Chi, X. Zhang, B. Xu, X. Zhou, C. Ma, Y. Zhang, S. Liu, J. Xu,
Chem. Soc. Rev. 2012, 41, 3878-3896; b) D. Frath, J. Massue, G.
Ulrich, R. Ziessel, Angew. Chem.-Int. Edit. 2014, 53, 2290-2310.
a) Y. Liu, C. Li, Z. Ren, S. Yan, M. R. Bryce, Nat. Rev. Mater. 2018, 3,
18020-18039; b) Z. Yang, Z. Mao, Z. Xie, Y. Zhang, S. Liu, J. Zhao, J.
Xu, Z. Chi, M. P. Aldred, Chem. Soc. Rev. 2017, 46, 915-1016; c) Z. An,
C. Zheng, Y. Tao, R. Chen, H. Shi, T. Chen, Z. Wang, H. Li, R. Deng, X.
Liu, W. Huang, Nat. Mater. 2015, 14, 685-690; d) J. Jin, H. Jiang, Q.
Yang, L. Tang, Y. Tao, Y. Li, R. Chen, C. Zheng, Q. Fan, K. Y. Zhang,
Q. Zhao, W. Huang, Nat. Commun. 2020, 11, 842; e) J. Yuan, S. Wang,
Y. Ji, R. Chen, Q. Zhu, Y. Wang, C. Zheng, Y. Tao, Q. Fan, W. Huang,
Materials Horizons 2019, 6, 1259-1264.
[3]
[4]
a) J. L. Bredas, D. Beljonne, V. Coropceanu, J. Cornil, Chem. Rev.
2004, 104, 4971-5003; b) K. Schmidt, S. Brovelli, V. Coropceanu, D.
Beljonne, J. Cornil, C. Bazzini, T. Caronna, R. Tubino, F. Meinardi, Z.
Shuai, J.-L. Bredas, J. Phys. Chem. A. 2007, 111, 10490-10499.
a) S. Cai, H. Shi, J. Li, L. Gu, Y. Ni, Z. Cheng, S. Wang, W.-w. Xiong, L.
Li, Z. An, W. Huang, Adv. Mater. 2017, 29, 1701244; b) C. Feng, S. Li,
X. Xiao, Y. Lei, H. Geng, Y. Liao, Q. Liao, J. Yao, Y. Wu, H. Fu, Adv.
Opt. Mater. 2019, 7, 1900767 ; c) S. M. A. Fateminia, Z. Mao, S. Xu, Z.
Yang, Z. Chi, B. Liu, Angew. Chem.-Int. Edit. 2017, 56, 12160-12164.
a) T. Ogoshi, H. Tsuchida, T. Kakuta, T.-a. Yamagishi, A. Taema, T.
Ono, M. Sugimoto, M. Mizuno, Adv. Funct. Mater. 2018, 28, 1707369;
b) X. Yang, D. Yan, Chem. Sci. 2016, 7, 4519-4526; c) D. Li, F. Lu, J.
Wang, W. Hu, X.-M. Cao, X. Ma, H. Tian, J. Am. Chem. Soc. 2018, 140,
1916-1923; d) K. Narushima, Y. Kiyota, T. Mori, S. Hirata, M. Vacha,
Adv. Mater. 2019, 31. 1807268.
[5]
[6]
[7]
M. Y. Wong, E. Zysman-Colman, Adv. Mater. 2017, 29. 1605444.
Z. He, H. Gao, S. Zhang, S. Zheng, Y. Wang, Z. Zhao, D. Ding, B.
Yang, Y. Zhang, W. Z. Yuan, Adv. Mater. 2019, 31, 1807222.
[8]
[9]
a) L. Gan, K. Gao, X. Cai, D. Chen, S. Su, J. Phys. Chem. Lett. 2018, 9,
4725-4731; b) Y. Yuan, Y. Hu, Y. Zhang, J. Lin, Y. Wang, Z. Jiang, L.
Liao, S. Lee, Adv. Funct. Mater. 2017, 27, 1700986.
L. Zhan, Z. Chen, S. Gong, Y. Xiang, F. Ni, X. Zeng, G. Xie, C. Yang,
Angew. Chem.-Int. Edit. 2019, 58, 17651-17655.
[10] W. Li, Q. Huang, Z. Mao, J. Zhao, H. Wu, J. Chen, Z. Yang, Y. Li, Z.
Yang, Y. Zhang, M. P. Aldred, Z. Chi, Angew. Chem.-Int. Edit. 2020, 59,
3739-3745.
[11] N. J. Hestand, F. C. Spano, Chem Rev 2018, 118, 7069-7163.
[12] a) H. Uoyama, K. Goushi, K. Shizu, H. Nomura, C. Adachi, Nature 2012,
492, 234-238; b) T. Chen, L. Zheng, J. Yuan, Z. An, R. Chen, Y. Tao, H.
Li, X. Xie, W. Huang, Sci. Rep. 2015, 5, 10923; c) H. Huang, Z. Yu, D.
6
This article is protected by copyright. All rights reserved.

10.1002/anie.202103192
Angewandte Chemie International Edition
RESEARCH ARTICLE
Entry for the Table of Contents
J-aggregates significantly reduce the energy gap between the lowest singlet (S1) and triplet (T1) excited states and contribute to the
reverse intersystem crossing (RISC) process; while H-aggregates will accelerate ISC and suppresses radiation transition of S1 to
realize phosphorescence. This provides crucial information for deeply understand of the internal mechanism of the excited state
dynamics controlled by different aggregates.
7
This article is protected by copyright. All rights reserved.