Improving electron transportation and operational lifetime of full color organic light emitting diodes through a “weak hydrogen bonding cage” structure

Efficient electron-transporting materials (ETMs) are critical to achieving excellent performance of organic light-emitting diodes (OLEDs), yet developing such materials remains a major long-term challenge, particularly ETMs with high electron mobilities (μeles). Herein, we report a short conjugated ETM molecule (PICN) with a dipolar phenanthroimidazole group, which exhibits an electron mobility of up to 1.52 × 10−4 cm2 (V−1 s−1). The origin of this high μele is long-ranged, regulated special cage-like interactions with C–H⋯N radii, which are also favorable for the excellent efficiency stability and operational stability in OLEDs. It is worth noting that the green phosphorescent OLED operation half-lifetimes can reach up to 630 h under unencapsulation, which is 20 times longer than that based on the commonly used commercial ETM TPBi.


Ⅱ. General Methods
Prior to the measurement, all of the two target molecules were obtained as pure products by vacuum sublimation.TPBi used in the device was purchased from Jilin OLED Material Tech Co., Ltd., and HATCN, TAPC, TCTA, Bepp2 and MADN were purchased from Xi'an Polymer Light Technology Corp.The optimized geometry and electron density distribution of the frontier molecular orbitals (FMOs) were obtained based on density functional theory (DFT) method of B3LYP/6-31G (d, p) using the Gaussian 09 package.Thermal gravimetric analysis (TGA) of the materials was performed using a Perkin-Elmer thermal analysis system at a ramp rate of 10 °C/min under nitrogen atmosphere.The differential scanning calorimetry (DSC) data of the materials in the range of -50 to 350 °C were tested under a heating rate of 10 °C/min using NETZSCH (DSC-204) instrument.In a nitrogen-purged dichloromethane (DCM) or N,N-Dimethylformamide (DMF) solution at room temperature, the Cyclic voltammetry (CV) analysis was measured.All the potentials with ferrocene/ferrocene + (Fc/Fc +) as standard, and tetra-n-butyl-ammonium hexafluorophosphate (TBAPF6, 0.1 m in acetonitrile) as the supporting electrolyte.The Ultraviolet Visible (UV-vis) absorption spectra of the solutions and films were scanned with a Hitachi U-4100 spectrophotometer.The photoluminescence (PL) spectra of the solutions and films were recorded with a Hitachi U-4600 spectrometer.Single crystal x-ray diffraction data were acquired by a Rigaku RAXIS-PRID diffractometer.ITO patterned anode with a sheet resistance of 5~20 Ω/square was used as the substrate.Before passing into a deposition chamber, the ITO substrate was scrubbed with Hellmanex TM III, followed by ultrasonic treatment with acetone solvent, Hellmanex TM III, and deionized water for 15 minutes each, dried at 120 °C in an oven.The clean ITO substrate is surface oxygenated and subsequently passed into the vacuum vapor deposition bin.
The cathode layers and organic layers were deposited in different chambers with a base pressure of less than 1.6×10 -4 Pa.The single-electron devices of the ETMs were fabricated with the structure of ITO/LiF (1 nm)/TPBi (10 nm)/ETMs (50 nm)/LiF (1 nm)/Al (100 nm), and the single-hole devices of the ETMs were fabricated with the structure of ITO/HATCN (20 nm)/ETMs (50 nm)/HATCN (20 nm)/Al (100 nm), in which lithium fluoride The EQEs were calculated from the L, J and EL spectra.The calculation formula (Formula S1) is as follows: where L (cd m -2 ) is the total luminance of the device, I (A) is the current, λ (nm) is EL wavelength, I(λ) is the relative EL intensity at each wavelength and is obtained by measuring the EL spectrum, K(λ) is the Commision International de L'Eclairage chromaticity (CIE) standard photopic efficiency function, e is the charge of an electron, h is the Planck's constant, c is the velocity of light.All the above data were measured in the forwardviewing direction without using any light out-coupling technique.
The efficiency roll-offs were calculated from maximum EQE (EQEmax) and the EQE at luminescence of 1000 cd m −2 (EQE1000).According to equation following (Formula S2): Ⅲ. Supplementary Results and Discussion

SUPPORTING INFORMATION
3 We prepared green PhOLEDs with Bepp2 as the electron transport layer (ETL) as an example to analyze the reason for the long lifetime of OLEDs with PICN as the ETL.Firstly, we found that Bepp2-based green PhOLEDs also have a long lifetime of up to 526 h (Figure S21), which is much higher than the lifetime of OLED devices when TPBi is used as the ETL (30 h), but does not reach the lifetime of OLED devices when PICN is used as the ETL (630 h).Also, Bepp2 (10 -4 cm 2 /(V•s)) and PICN (1.52 × 10 -4 cm 2 /(V•s)) have higher μeles than TPBi (1.68 × 10 -6 cm 2 /(V•s)) and provide more balanced carrier mobility when prepared into devices, which perhaps largely avoids polariton-exciton collisional bursts, thus providing a more stable luminescence environment and enabling longer device lifetimes when Bepp2 and PICN are used as ETLs.In addition, although Bepp2 and PICN have similar μele, longer lifetimes exist for devices using PICN, so we believe that carrier balance is one of the factors for this long lifetime.In order to reveal more fully the differences in the lifetimes of TPBi and PICN OLEDs, we vaporised pure films of the two materials separately and explored their morphological stability by atomic force microscopy (AFM).As shown in Figure S13, the AFM result and the root-mean-square (RMS) value of the freshly vapour-deposited PICN film and the AFM result and RMS value of the film after 96 h in the glove box were tested, respectively.
When the films are freshly prepared, they all show very smooth surface, with the root-RMS in the order of PICN (2.35 nm) < TPBi (2.65 nm).In addition, DSC was tested in the temperature range of 25~300 ℃, and it was found that PICN has no obvious glass transition temperature, and there is no cold crystallization peak during the secondary heating process, which also indicates that the material has good morphological stability.
Thus the morphology of PICN only show a little change when placed in the glove box for 96 h, while the deformation of TPBi (RMS=4.27nm) is much more severe than that of PICN, which would be harmful for devices by causing a severe interface separation.In summary, the long lifetime of PICN is caused by a more balanced carrier mobility and a more stable morphology.

Figure S7 .
Figure S7.(a) Normalized UV-vis absorption spectra of the two molecules in Different Polar Solvents.(b) Normalized PL spectra of the two molecules in Different Polar Solvents.(c) and (d) Lippert-Mataga solvatochromic model of the two molecules.

Figure S9 .
Figure S9.Fluorescence and phosphorescence spectra of the two molecules in 10 −5 M toluene at 77K.

Figure S10 .
Figure S10.The cyclic voltammetry (CV) curves of the two molecules in DCM solution (oxidation section) and DMF solution (reduction section).

Figure S11 .
Figure S11.(a) and (b) are the thermogravimetric analysis (TGA) curves of the two molecules.(c) and (d) are the differential scanning calorimetry (DSC) curves of the two molecules.

Figure S12 .
Figure S12.The morphology of PICN, PINH and TPBi films measured right after and 96 h in the glove box after the film preparation.

Figure S13 .
Figure S13.(a) PICN and (b) PINH single crystal structures with torsion angles and minimum distance diagram of the conjugate planes.

Figure S14 .
Figure S14.Estimation of the combination energy of the weak hydrogen bond cage of PICN.The calculation method is b3lyp/6-31g(d,p) considering the basis set superposition error (BSSE) correction and dispersion correction (GD3).

Figure S17 .
Figure S17.Normalized EL spectra of PINH at different voltages.

Figure S22 .
Figure S22.Lifetime curve of the green PhOLED at a fixed current density with an L0 of 1000 cd/m 2 .