Role of Linker Functionality in Polymers Exhibiting Main‐Chain Thermally Activated Delayed Fluorescence

Abstract Excellent performance has been reported for organic light‐emitting diodes (OLEDs) based on small molecule emitters that exhibit thermally activated delayed fluorescence. However, the necessary vacuum processing makes the fabrication of large‐area devices based on these emitters cumbersome and expensive. Here, the authors present high performance OLEDs, based on novel, TADF polymers that can be readily processed from a solution. These polymers are based on the acridine‐benzophenone donor–acceptor motif as main‐chain TADF chromophores, linked by various conjugated and non‐conjugated spacer moieties. The authors’ extensive spectroscopic and electronic analysis shows that in particular in case of alkyl spacers, the properties and performance of the monomeric TADF chromophores are virtually left unaffected by the polymerization. They present efficient solution‐processed OLEDs based on these TADF polymers, diluted in oligostyrene as a host. The devices based on the alkyl spacer‐based TADF polymers exhibit external quantum efficiencies (EQEs) ≈12%, without any outcoupling‐enhancing measures. What's more, the EQE of these devices does not drop substantially upon diluting the polymer down to only ten weight percent of active material. In contrast, the EQE of devices based on the monomeric chromophore show significant losses upon dilution due to loss of charge percolation.

The crude product was further purified by flash chromatography on flash silica gel (Hex:EtOH 50:1) and a subsequent preparative SEC in CHCl 3 . The product was obtained as a yellow solid (88 mg, 0.09 mmol, 52%).
The combined organic fractions were washed with water and HCl solution (5 mL, 2 M). The organic phase was dried over MgSO 4 and the solvent was evaporated under reduced pressure.
The combined organic fractions were washed with water and HCl solution (5 mL, 2 M). The organic phase was dried over MgSO 4 and the solvent was evaporated under reduced pressure.
The combined organic fractions were washed with water and HCl solution (5 mL, 2 M). The organic phase was dried over MgSO 4 and the solvent was evaporated under reduced pressure.
The crude product was further purified by soxhlet extraction (1) methanol, 2) isopropanol, 3) hexane 4) CHCl 3 ) and a subsequent preparative SEC in CHCl 3 . The desired product was obtained as a yellow solid (156 mg, 62 %). 1  The proton NMR spectra of the TADF-polymers are displayed in Figure S1. The GPC traces and results are respectively depicted and listed in Figure S2 and Table S1.   The molecular structure of p-pTFF-C 2 F 5 SIS is depictured in Figure S3. phenylene) (p-pTFF-C 2 F 5 SIS).

S2. Cyclic voltammetry and ultraviolet photoelectron spectroscopy
Cyclic voltammetry (CV) measurements were performed in dichloromethane solutions using Bu 4 NPF 6 as a supporting electrolyte and ferrocene as an internal reference. The voltammograms are shown in Figure S4.  internal reference was added after the scan and the CV was remeasured.
The materials were further investigated by ultraviolet photoelectron spectroscopy (UPS). For this, thin films of the materials were spin-cast onto silicon wafers, covered with 2 nm of chromium and 50 nm of gold. The resulting UPS spectrum shows the intensity of the detected electrons plotted against the binding energy. For each of the samples, three UPS spectra were recorded for different spots of the film to increase accuracy. The spectra are given in Figure  S5. For the evaluation of the UPS ionization potential , the onsets on both ends of the spectra were obtained from the intersection of two linear fits (see Figure S5 b-c). At high binding energies close to the energy of the excitation light, the secondary electron cut-off is obtained. The energies on the other end of the spectra give the HOMO edge ( ( ) ). With and ( ) , the ionization potential was calculated using: ( ( ) ), with Planck's constant and is the frequency of the excitation radiation.
The LUMO was estimated through addition of the optical gap: . Figure S5. UPS spectra of the TADF materials prepared in this work.

S3. Spectroscopy
The normalized steady state optical absorption (UV-vis) and luminescence spectra of the TADF materials are combined in Figure S6a

S4 Devices
The current-voltage ( ) behavior of hole only (HO) devices of the TADF materials is given as a function of temperature in Figure S8. In Figure S9 the HO current-voltage data has been fitted by the drift-diffusion model mentioned in the main document, in order to quantify charge carrier mobility and trap density. The current-voltage ( ) behavior of electron only (EO) devices of the TADF materials is given for = 295 K in Figure S10. The results obtained from curve fitting of the EO data is summarized in Table S2.   The performance of the OLED (dual carrier) devices has been summarized in Table S3.   In order to investigate if the insulating polystyrene as a host has a negative impact on charge transport and hence OLED performance, we fabricated two series of devices using a conventional semiconducting host. For the first set, TCTA was used as a host (devices E), while the second set comprises a mixed host of TCTA and TAPC (devices F), which has been used in combination with TADF polymers in solution processing. [1,3] For both sets of devices the doping concentration of the TADF materials is 10 %, resulting for the latter series in a ratio of TADF:TCTA:TAPC = 10:65:25. The EMLs were spincoated from chlorobenzene solution and, due to the low solubility of TCTA and TAPC, the solutions had to be heated to 80 °C to fully dissolve the materials. The J-V-L characteristics and the external quantum yield (EQE) of these devices is depicted in Figure S11. Overall, the EQE of these devices is somewhat lower in comparison to the devices based on polystyrene as a host (see main document). In Figure S12 we plot the normalized electroluminescence (EL) spectra of the OLED devices based on our TADF emitters, diluted in polystyrene at various ratios. The shape and position of the band is similar for all materials and consistent with the emission of the monomeric DMAC-BP chromophore. Interestingly, there is a small but clear blue-shift upon increasing the dilution. We suspect that this is a solvatochromic effect, known to occur for TADF emitters, on account of the charge transfer nature of their excited state. [4] The blue shift suggests that the environment of the chromophores becomes less polar upon increasing the dilution, which leads to a destabilization of the charge transfer state and hence a widening of the band gap.

S5. Blend film morphology
To check for possible phase separation of the TADF-polymer:polystyrene blends during solution casting, we recorded AFM images of dried spin-coated films. No signs of phase separation were detected, probably owing to the fact that we maintain a reasonably high mixing entropy by using a low molecular weight polystyrene (see Experimental Section), as explained in the main text. Figure S13 shows the AFM images for the films of Tol-MAc-BP:polystyrene and P(C6-MAc-BP):polystyrene as representative results. No structuring was discerned and the films are smooth. As an illustrative comparison, Figure S13d reveals a blend of P(C6-MAc-BP) and the cyclo olefin copolymer Zeonex® 480 to exhibit phase separation during spincasting, clearly evidenced by the round dark domains, which represent