Azulene has emerged as a promising candidate for constructing advanced organic optoelectronic materials due to its unique optoelectronic properties [
1-
6]. In recent years, azulene-based materials have been extensively investigated in organic field-effect transistors (OFETs), solar cells, and other related applications [
7-
14]. As illustrated in
Figs. 1a and
b, the asymmetric frontier molecular orbitals (FMO) of azulene reduce electron repulsion between unpaired electrons, resulting in a narrow energy gap between its ground state (S
0) and first excited state (S
1). This reduced energy gap leads to a visible-range absorption band (S
0 → S
1 transition), endowing azulene with its characteristic blue color [
15-
17]. Furthermore, the significant energy difference between S
1 and S
2 (∆
E(S
2-S
1) > 10,000 cm
−1) coupled with orbital asymmetry slows the non-radiative S
2 → S
1 transition while enhancing the radiative S
2 → S
0 pathway. This gives rise to the anti-Kasha fluorescence behavior observed in azulene [
18,
19]. However, this unique photophysical property also confines the fluorescence of azulene and its derivatives to the short-wavelength region with inherently weak emission intensities [
20-
25]. For example, the reported azulene-based BN-heteroaromatics exhibit photoluminescence quantum yield (PLQY) of less than 0.5% (Fig. S1a in Supporting information) [
24].
As illustrated in
Fig. 1a, replacement of the CH moiety at the 1-position of azulene with a N atom yields its heteroanalog, 1-azaazulene. Density functional theory (DFT) calculations demonstrate that N substitution profoundly alters the electronic structure and FMO energetics (
Fig. 1b). Compared to pristine azulene (HOMO/LUMO = −5.20/−1.82 eV,
Egap = 3.38 eV), 1-azaazulene exhibits stabilized HOMO (−5.93 eV) and LUMO (−2.16 eV) levels, with the bandgap widening to 3.77 eV due to a more pronounced stabilization of the HOMO level. Time-dependent DFT (TD-DFT) calculations reveal that while 1-azaazulene retains asymmetric frontier orbitals, the S
0-S
1 energy gap becomes substantially wider than that of the azulene. Notably, the reduced S
2-S
1 energy separation accelerates the non-radiative S
2 → S
1 conversion compared to azulene, effectively suppressing the anti-Kasha S
2 → S
0 radiative transition. Such changes in electronic structure and FMO energetics render 1-azaazulene a promising scaffold for designing Kasha-compliant fluorescent materials with red-shifted emission profiles compared to its quinoline isomer (Fig. S2 in Supporting information) [
26-
29].
Boron difluoride (BF
2) complexes represent a prominent class of luminescent materials [
30-
35]. The lone pair electrons on the N atom of 1-azaazulene enable effective coordination, making it an ideal ligand for constructing BF
2 complexes [
36-
40]. In this study, four asymmetric 1-azaazulene-based BF
2 complexes (Ph-NN, Ph-ON, Ca-NN, and Py-NN,
Fig. 1c) were designed, synthesized, and systematically investigated for their photophysical properties. The luminescence wavelength and efficiency of these complexes were modulated by structural variations in the ancillary coordination groups. In contrast to the anti-Kasha fluorescence observed in azulene derivatives, all four 1-azaazulene constructs display S
1 → S
0 emission fluorescence. Ph-NN exhibits an emission peak at 705 nm, while Py-NN achieves a PLQY of 19.3%. Furthermore, the organic light-emitting diode (OLED) fabricated with Py-NN as the emitter
via a sensitization strategy delivers a maximum external quantum efficiency (EQE
max) of 3%.
The synthetic routes for Ph-NN, Ph-ON, Ca-NN, and Py-NN are illustrated in
Scheme 1. First, 2-chloro-1-azaazulene (
S1) was synthesized according to the literature procedure [
41,
42]. Subsequently, Suzuki-Miyaura coupling reactions between
S1 and various arylboronic acids
S2a-
S2d were performed, affording intermediates
S3a-
S3d in 67%-85% yields. Notably, for reaction of
S1 and
S2d, treatment with sodium methoxide during workup enabled
in situ deprotection of the Boc group, directly affording intermediate
S3d. To prevent potential interference of the amino group during BF
2 complex formation,
S3a was substituted with an alkyl chain to obtain intermediate
S4a. Finally, exposure of intermediates
S4a and
S3b-
d to excess BF
3·Et
2O in toluene, in the presence of TEA/DIPEA, afforded 1-azaazulene-based BF
2 complexes Ph-NN, Ph-ON, Ca-NN, and Py-NN with 50%-62% yields. These compounds were characterized by nuclear magnetic resonance (NMR) spectroscopy and high-resolution mass spectrometry (HRMS). The full characterization data can be found in Supporting information.
The structures of Ph-NN, Ph-ON, Ca-NN, and Py-NN were further confirmed through X-ray single crystal analysis (
Fig. 2, Tables S1-S4 in Supporting information). The single-crystal structures reveal that all four compounds possess highly planar backbones, with the largest dihedral angles in the fused rings being less than 4.95° (Fig. S3 in Supporting information). The crystal packing diagrams in
Fig. 2 and Figs. S4-S7 (Supporting information) demonstrate that in Ph-NN, Ph-ON, Ca-NN, and Py-NN, adjacent molecules are arranged through antiparallel stacking of their conjugated backbones. The stacking distances, ranging from 3.32 Å to 3.47 Å, suggest that all four molecules have strong
π-
π interactions. These pronounced
π-
π interactions lead to a decrease in the PLQY of the thin films due to aggregation-caused quenching (ACQ) effects (Table S5 in Supporting information). As shown in Fig. S3, in the four BF
2 complexes, the average bond lengths (1.38–1.42 Å) of the 1-azaazulene and its connected aromatic rings are close to the bond length of aromatic C=C bonds, indicating that the ligand backbones maintain good aromaticity. The NICS(1)
zz values (Fig. S8 in Supporting information) and the anisotropy of the induced current density (AICD) plots (Fig. S9 in Supporting information) further confirm the aromatic characteristics of the 1-azaazulene-based ligand backbones. Notably, the N→B coordination bonds in these four molecules (1.55–1.58 Å) are significantly shorter than those reported in asymmetric BF
2 complexes [
43,
44]. The short N→B bond lengths reflect strong dative interactions between nitrogen and boron, enhancing the complex’s stability and ensuring the integrity of the fluorescent materials under various conditions.
To investigate the photophysical properties of Ph-NN, Ph-ON, Ca-NN, and Py-NN, we performed absorption spectroscopy, steady-state, and time-resolved emission spectroscopy under oxygen-free conditions at a concentration of 1 × 10
−5 mol/L (
Table 1). As shown in
Fig. 3a, all four molecules exhibit intense absorption bands in the 300–400 nm range, primarily attributed to S
0 → S
2 transitions of 1-azaazulene units. However, they exhibit distinct absorption differences in the long-wavelength region. Ph-ON and Py-NN display maximum absorption peaks at 487 and 501 nm, respectively, with pronounced shoulder peaks, indicating dominant localized excitations. In contrast, Ph-NN and Ca-NN show significantly red-shifted absorption at 597 and 555 nm, accompanied by broader absorption bands lacking distinct shoulders, which suggests excitations with pronounced charge-transfer (CT) characteristics. Notably, these absorption bands display significant redshifts compared to analogous quinoline derivatives [
43,
45,
46], which may originate from the asymmetric electron distribution and narrower FMO energy gaps of the 1-azaazulene core. Remarkably, the molar absorption coefficients for the S
0 → S
1 transitions of these compounds (
ε > 10
4 L mol
−1 cm
−1) are substantially higher than those of typical azulene-based molecules (
ε < 10
4 L mol
−1 cm
−1).
As depicted in
Fig. 3b, Ph-NN exhibits an S
1 → S
0 emission peak at 705 nm, a rare feature among small-molecule BF
2 complexes, with a Stokes shift (
S) of 108 nm. However, its PLQY is only 0.3%, likely due to ultrafast nonradiative decay. Time-resolved emission spectroscopy (Fig. S10a in Supporting information) reveals an exceptionally short luminescence lifetime (
τ) of 0.18 ns for Ph-NN, yielding a radiative transition rate (
kr) of 1.67 × 10
7 s
−1 and a nonradiative rate (
knr) of 5.54 × 10
9 s
−1 (
Table 1). In contrast, Ph-ON and Ca-NN display blue-shifted emission peaks at 552 nm (
S = 74 nm) and 653 nm (
S = 98 nm), with improved PLQYs of 2.9% and 2.1%, respectively. Their luminescence lifetimes increase to 0.84 ns (Ph-ON) and 0.59 ns (Ca-NN), corresponding to
kr/
knr values of 3.45 × 10
7 s
−1/1.16 × 10
9 s
−1 and 3.56 × 10
7 s
−1/1.66 × 10
9 s
−1 (Figs. S10b and c in Supporting information). Their enhanced PLQYs arise from both increased
kr and reduced
knr. Significantly, Py-NN demonstrates the highest luminescence efficiency among the four 1-azaazulene-based BF
2 complexes, exhibiting an emission peak at 560 nm (
S = 59 nm) with a PLQY of 19.3%. The relatively small Stokes shift and high emission efficiency suggest dominant localized-state emission in Py-NN. Time-resolved photoluminescence spectrum (Fig. S10d in Supporting information) confirms a prolonged luminescence lifetime of 2.03 ns, corresponding to
kr = 9.51 × 10
7 s
−1 and
knr = 0.40 × 10
9 s
−1. These results indicate that by altering the structure of the groups attached to the 1-azaazulene, it is possible to tune the photophysical properties including absorption spectrum, emission spectrum, and emission efficiency of 1-azaazulene-based BF
2 complexes.
DFT calculations at the B3LYP/6–31G+(d,p) level were employed to investigate the electronic properties of the four 1-azaazulene-based BF
2 complexes, with the calculated data summarized in Table S6 (Supporting information). As illustrated in
Fig. 4, the FMO of Ph-NN, Ph-ON, Ca-NN, and Py-NN exhibit significantly uneven electron density distributions, a notable contrast to most BF
2 complexes [
47-
49]. Specifically, the LUMO electron density is primarily distributed over the 1-azaazulene unit, while the HOMO electron density is mainly localized on the aniline, phenol, carbazole, and pyrrole moieties. For Ph-ON and Py-NN, especially the latter, the HOMO electron density also has a small distribution on the 1-azaazulene. This pronounced asymmetry in molecular orbital distribution suggests a propensity for charge-transfer (CT) transition absorption in these compounds, while Ph-ON and Py-NN may also exhibit localized transition absorption. As shown in
Fig. 4, the calculations indicate that by altering the structure of the group connected to the 1-azaazulene, the FMO energetics and band gap of the 1-azaazulene-based BF
2 complexes can be effectively modulated, thereby tuning the absorption and emission wavelengths of the molecules.
Since the B3LYP functional may underestimate the energy of charge-transfer electronic states [
50], the long-range corrected M062X functional was employed to study the excited states of 1-azaazulene-based BF
2 complexes using the 6–31G(d) basis set. TD-DFT calculations show that the oscillator strengths (
f) for the S
0 → S
1 transition of Ph-NN, Ph-ON, Ca-NN, and Py-NN are 0.3299, 0.3009, 0.3458, and 0.4973, respectively. The increased
f could originate from the superior hole-electron overlap of S
1 state of Py-NN compared with the other three molecules. This indicates that Py-NN likely exhibits accelerated radiative decay rate (
kr). In addition, Py-NN demonstrated a relatively large energy gap (Δ
E = 2.75 eV) and smallest reorganization energy (λ
M = 3942 cm
−1), which is conducive to decelerating nonradiative decay rate (see Supporting information for details). Accordingly, Py-NN with large
kr and small
knr likely exhibits high PLQY, which agrees well with the experimental results. In the electron-hole overlap maps of the S
1 excited state and atomic transition density matrix (TDM) heat maps for the S
0 → S
1 transition (Fig. S10e in Supporting information), Ph-ON and Py-NN show substantial electron-hole overlap, primarily localized on the 1-azaazulene unit. This localization is favorable for the formation of localized excited states (LE). Additionally, the maps demonstrate that the electronic excitation process is chiefly focused on electron transfer among the atoms within the 1-azaazulene unit.
Cyclic voltammetry (CV) and differential pulse voltammetry (DPV) tests were performed on Ph-NN, Ph-ON, Ca-NN, and Py-NN, as shown in Figs. S11a-d and Table S8 (Supporting information). Electrochemical calculations determined the HOMO/LUMO levels of Ph-NN, Ph-ON, Ca-NN, and Py-NN to be −5.39/−3.61 eV, −6.39/−3.57 eV, −5.42/−3.70 eV, and −5.73/−3.63 eV, respectively. Overall, the FMO energy levels obtained from electrochemical measurements, particularly the HOMO levels, were consistent with the theoretical computational trends. Additionally, thermogravimetric analysis (TGA) was conducted and the results indicated that the thermal decomposition temperatures (Td) for Ph-NN, Ph-ON, Ca-NN, and Py-NN were 273, 246, 371, and 428 ℃ (Fig. S12 in Supporting information), respectively, demonstrating good thermal stability across all compounds.
Owing to the moderate PLQY and superior thermal stability of Py-NN, OLED devices were fabricated to evaluate the electroluminescent performance of 1-azaazulene-based BF
2 complexes with multilayer configurations of ITO/HATCN (5 nm, hole injection layer)/NPB (40 nm, hole-transport layer)/TCTA (10 nm, exciton-blocking layer)/TPBi: 20 wt% thermally activated delayed fluorescence sensitizer 4CZIPN and 1 wt% Py-NN (30 nm, emissive layer)/TPBi (25 nm, electron-transport layer)/LiF (1 nm, electron-injection layer)/Al (150 nm) (
Fig. 5a and Fig. S12). As shown in
Fig. 5b, the device exhibits strong orange emission with the corresponding electroluminescence peak value (λ
em) located at 565 nm, which is close to the photoluminescnece peak of Py-NN. And the OLED exhibited an EQE
max of 3.0% and a turn-on voltage of 5.2 V.
In summary, this study successfully designed and synthesized four asymmetric 1-azaazulene-based BF2 complexes (Ph-NN, Ph-ON, Ca-NN, and Py-NN) to explore their optoelectronic properties. By modifying ancillary coordination groups, the FMO energetics and band gaps of these complexes were effectively tuned, allowing adjustments in their absorption and emission properties. All four 1-azaazulene constructs exhibit S1 → S0 emission fluorescence, in contrast to azulene derivatives, which typically display anti-Kasha fluorescence and low PLQY. Notably, Py-NN demonstrated superior performance, achieving a PLQY of 19.3% with an emission peak at 560 nm. This enhancement is attributed to improved HOMO-LUMO overlap and suppressed nonradiative decay pathways. OLEDs incorporating Py-NN as an emissive dopant exhibited a maximum EQE of 3%, underscoring its electroluminescent potential. These findings demonstrate that 1-azaazulene serves as a promising scaffold for developing fluorescent materials, overcoming the inherent limitations of traditional azulene derivatives and advancing its potential applications in optoelectronic devices. Moreover, 1-azaazulene constructs possess smaller band gaps compared to traditional fluorescent materials with similar-sized π-conjugated systems. This characteristic facilitates the development of long-wavelength emission fluorescent materials, which are desirable for applications requiring deep-red or near-infrared emission.
PDF
Full
82
44
关闭全屏
No. 86 Xueyuan South Road, Haidian District, Beijing
010-62662699(Operation and Technical Support)
qkjq@cast.org.cn
