D 6 h Symmetric Radical Donor–Acceptor Nanographene Modulated Interfacial Carrier Transfer for High-Performance Perovskite Solar Cells

Abstract

Open AccessCCS ChemistryRESEARCH ARTICLES13 Dec 2022D6h Symmetric Radical Donor–Acceptor Nanographene Modulated Interfacial Carrier Transfer for High-Performance Perovskite Solar Cells Can Wang†, Yifeng Gao†, Zhen-Lin Qiu†, Ping-Ping Sun, Naoyuki Shibayama, Zilong Zhang, Qiu Xiong, Fangbin Ren, Shui-Yang Lien, Lusheng Liang, Jiaoxia Zhang, Yuan-Zhi Tan and Peng Gao Can Wang† CAS Key Laboratory of Design and Assembly of Functional Nanostructures, Fujian Provincial Key Laboratory of Nanomaterials, Fujian Institute of Research on the Structure of Matter, Chinese Academy of Sciences, Fuzhou, Fujian 350002 Laboratory for Advanced Functional Materials, Xiamen Institute of Rare Earth Materials, Haixi Institute, Chinese Academy of Sciences, Xiamen 361021 University of Chinese Academy of Sciences, Beijing 100049 †C. Wang, Y. Gao, and Z.-L. Qiu contributed equally to this work.Google Scholar More articles by this author , Yifeng Gao† CAS Key Laboratory of Design and Assembly of Functional Nanostructures, Fujian Provincial Key Laboratory of Nanomaterials, Fujian Institute of Research on the Structure of Matter, Chinese Academy of Sciences, Fuzhou, Fujian 350002 Laboratory for Advanced Functional Materials, Xiamen Institute of Rare Earth Materials, Haixi Institute, Chinese Academy of Sciences, Xiamen 361021 School of Materials Science and Engineering, Jiangsu University of Science and Technology, Zhenjiang 212003 †C. Wang, Y. Gao, and Z.-L. Qiu contributed equally to this work.Google Scholar More articles by this author , Zhen-Lin Qiu† Collaborative Innovation Center of Chemistry for Energy Materials, State Key Laboratory for Physical Chemistry of Solid Surfaces, Engineering Research Center for Nano-Preparation Technology of Fujian Province, Department of Chemistry, College of Chemistry and Chemical Engineering, Xiamen University, Xiamen 361005 †C. Wang, Y. Gao, and Z.-L. Qiu contributed equally to this work.Google Scholar More articles by this author , Ping-Ping Sun Department of Chemistry, School of Science, Hainan University, Haikou 570228 Google Scholar More articles by this author , Naoyuki Shibayama Faculty of Biomedical Engineering, Graduate School of Engineering, Toin University of Yokohama, Yokohama, Kanagawa 225-8503 Google Scholar More articles by this author , Zilong Zhang CAS Key Laboratory of Design and Assembly of Functional Nanostructures, Fujian Provincial Key Laboratory of Nanomaterials, Fujian Institute of Research on the Structure of Matter, Chinese Academy of Sciences, Fuzhou, Fujian 350002 Laboratory for Advanced Functional Materials, Xiamen Institute of Rare Earth Materials, Haixi Institute, Chinese Academy of Sciences, Xiamen 361021 Google Scholar More articles by this author , Qiu Xiong CAS Key Laboratory of Design and Assembly of Functional Nanostructures, Fujian Provincial Key Laboratory of Nanomaterials, Fujian Institute of Research on the Structure of Matter, Chinese Academy of Sciences, Fuzhou, Fujian 350002 Laboratory for Advanced Functional Materials, Xiamen Institute of Rare Earth Materials, Haixi Institute, Chinese Academy of Sciences, Xiamen 361021 University of Chinese Academy of Sciences, Beijing 100049 Google Scholar More articles by this author , Fangbin Ren Xiamen University of Technology, Xiamen 361024 Google Scholar More articles by this author , Shui-Yang Lien Xiamen University of Technology, Xiamen 361024 Google Scholar More articles by this author , Lusheng Liang CAS Key Laboratory of Design and Assembly of Functional Nanostructures, Fujian Provincial Key Laboratory of Nanomaterials, Fujian Institute of Research on the Structure of Matter, Chinese Academy of Sciences, Fuzhou, Fujian 350002 Laboratory for Advanced Functional Materials, Xiamen Institute of Rare Earth Materials, Haixi Institute, Chinese Academy of Sciences, Xiamen 361021 Google Scholar More articles by this author , Jiaoxia Zhang *Corresponding authors: E-mail Address: [email protected] E-mail Address: [email protected] E-mail Address: [email protected]. School of Materials Science and Engineering, Jiangsu University of Science and Technology, Zhenjiang 212003 Google Scholar More articles by this author , Yuan-Zhi Tan *Corresponding authors: E-mail Address: [email protected] E-mail Address: [email protected] E-mail Address: [email protected]. Collaborative Innovation Center of Chemistry for Energy Materials, State Key Laboratory for Physical Chemistry of Solid Surfaces, Engineering Research Center for Nano-Preparation Technology of Fujian Province, Department of Chemistry, College of Chemistry and Chemical Engineering, Xiamen University, Xiamen 361005 Google Scholar More articles by this author and Peng Gao *Corresponding authors: E-mail Address: [email protected] E-mail Address: [email protected] E-mail Address: [email protected]. CAS Key Laboratory of Design and Assembly of Functional Nanostructures, Fujian Provincial Key Laboratory of Nanomaterials, Fujian Institute of Research on the Structure of Matter, Chinese Academy of Sciences, Fuzhou, Fujian 350002 Laboratory for Advanced Functional Materials, Xiamen Institute of Rare Earth Materials, Haixi Institute, Chinese Academy of Sciences, Xiamen 361021 University of Chinese Academy of Sciences, Beijing 100049 Google Scholar More articles by this author https://doi.org/10.31635/ccschem.022.202202433 SectionsSupplemental MaterialAboutAbstractPDF ToolsAdd to favoritesDownload CitationsTrack Citations ShareFacebookTwitterLinked InEmail Imbalanced charge-carrier extraction remains an issue aggravating interfacial charge accumulation and recombination. More hopping transport channels could accelerate the extraction of charge. Here, we demonstrated an effective “bridging interface” strategy between the perovskite/2,2′,7,7′-tetrakis(N,N-di-p-methoxyphenylamine)-9,9′-spirobifluorene (spiro-OMeTAD) that modulates interfacial charge transfer and improves hole mobility using radical-containing donor–acceptor nanographenes (D–A NGs) possessing electron-deficient perchlorinated NGs and electron-rich aniline derivatives. The fully delocalized backbone of nanographene formed a conjugated bridge for intermolecular charge transfer and generated stable radical cations, verified by electron spin resonance. Lamellar and π–π stacking orientation of D–A NGs also provided advantageous hopping transport channels. Besides favorable charge transfer within D–A NGs, systematic explorations indicated a strong interface coupling and noticeable charge transfer across the D–A NGs and perovskite interface, where electrons would flow from D–A NGs to perovskite, and holes would flow from perovskite to D–A NGs. Moreover, the hole mobility of spiro-OMeTAD was also enhanced because the D–A NGs would diffuse into the spiro-OMeTAD layer. As a result, planar n–i–p perovskite solar cells modified by D–A NG-OMe/D–A NG-tBu delivered champion power conversion efficiencies (PCEs) of 23.25% and 23.51%, respectively. Download figure Download PowerPoint Introduction The power conversion efficiency (PCE) of perovskite solar cells (PSCs) progressed rapidly from 9.7% to 25.7% within a short time, showing enormous potential for next-generation photovoltaic technology.1–3 However, the frequently occurring imbalanced charge-carrier extraction in PSCs caused interfacial charge accumulation and recombination, thereby severely limiting the approach of PSCs to their theoretical efficiency. By employing electron beam-induced current and Kelvin-probe force microscopy, researchers have observed more efficient extraction of electrons over holes in p–i–n-type devices.4,5 The consequently accumulated holes at the interface of the perovskite/hole transport layer (HTL) create an electrical potential barrier that cripples the short-circuit current.5 Thus, a faster and balanced hole extraction is indispensable for highly efficient PSCs. Interfacial carrier extraction is drastically influenced by the interfacial energy-level alignment (ELA) and built-in electric field.6–11 A benign interface must selectively extract the majority carriers while blocking the minority carriers by means of a large Schottky barrier at the heterojunction.6 For a regular n–i–p device, the perovskite/HTL heterojunction is the transfer rate-limiting interface because a small majority carrier band offset at this interface can trigger a considerable upwards band bending of hole quasi-Fermi level, resulting in an exponential increase in the recombination rates.12 Modulating the ELA via specifically designed interlayers has been considered one of the most effective approaches to improve the hole transferability, including dipole molecule,13–15 graded HTL,16 π-conjugated organic ligands,17,18 and other pyridine/triphenylamine/thiophene derivatives.19–22 Apart from ELA, HTL with higher conductivity and hole mobility would also help the interfacial hole extraction.23,24 Since 2,2′,7,7′-tetrakis(N,N-di-p-methoxyphenylamine)-9,9′-spirobifluorene (spiro-OMeTAD), as the most popular HTL in regular n–i–p PSCs, has poor intrinsic hole mobility (∼10−5 cm2 V−1 s−1), lithium bis(trifluoromethanesulfonyl)-imide (Li-TFSI) is commonly needed to aid the formation of [spiro-OMeTAD]•+ radical cations in the presence of oxygen.25–27 These long-lived charged species,26 promoted charge hopping,28,29 and increased the conductivity/hole mobility.30 Various dopants were also utilized to generate the radical cations through intermolecular charge transfer.31–33 Xia et al.31 demonstrated a dipole-derived doping process that uses phenylamine cation (PHC) as the single dopant of spiro-OMeTAD, generating p-doping via intermolecular dipole and creating radicals on the spiro-OMeTAD. Spiro-OMeTAD employing PHC exhibited more rational ELA, faster interface carriers transfer, and further relieved trap-induced recombination. Lu et al.34 directly used free radical-containing compounds as p-dopants for spiro-OMeTAD to enhance the conductivity and tune the work function. Recently, Gao et al.35 developed an ion-modulated radical doping strategy (using organic radicals as the dopant and ionic salts as the doping modulator) for effectively doping spiro-OMeTAD without host-oxidation, which decoupled conductivity and work function tunability. Although the radical-doping strategy could also be found in other hole transport materials (HTMs), its application in spiro-OMeTAD is still less explored and needs further studies to achieve faster and balanced hole extraction.36–39 Inspired by the pioneer works, we have demonstrated in this work that an efficient “conjugated bridge” linked interface between perovskite and spiro-OMeTAD to realize the modulation of ELA and improve hole mobility. Two radical-containing donor–acceptor nanographenes (D–A NGs) possessing electron-deficient perchlorinated NGs and electron-rich aniline derivatives were utilized as the interfacial bridge. The delocalization effect of the nanographene core could induce an intermolecular charge transfer and stabilize the radical cations. At the same time, the peripheral aniline substituents with tert-butyl (D–A NG-tBu) or methoxy (D–A NG-OMe) at the para-position could tune the donating ability and keep the solubility of the D–A NGs. Furthermore, when the spiro-OMeTAD layer was deposited atop the D–A modified perovskite, the NGs could diffuse into the HTL due to good solubility in chlorobenzene (CB) to enhance the conductivity and hole mobility due to the adventitious radical nature. Moreover, the favorable frontier orbital levels of D–A NG-tBu relative to perovskites led to suitable ELA, thereby facilitating the hole extraction and blocking the backflow of electrons. Notably, through differential charge density calculation, we proved a strong interface coupling and noticeable charge transfer across the D–A NGs and the perovskite interface. Finally, by modulating the interfacial carrier transport, planar PSCs modified by D–A NG-OMe/D–A NG-tBu presented champion PCEs of 23.25% and 23.51%, respectively. Experimental Methods Materials SnO2 colloid precursor [tin(iv) oxide, 15% in H2O colloidal dispersion] was bought from Alfa Aesar Chemical Co., Ltd. (Shanghai, China), and lead iodide (99.99%) was purchased from TCI Development Co., Ltd. (Shanghai, China). Formamidinium iodide (FAI) and methylammonium bromide (MABr) were new types of perovskite units synthesized according to a previous report.40 Similarly, methylammonium chloride (MACl) was synthesized by reacting methylamine and hydrochloric acid. The spiro-OMeTAD was purchased from Derthon Technology Co., Ltd. (Shenzhen, China). Lithium bis(trifluoromethanesulfonyl)imide (Li-TFSI) salt (99.95%), 4-tert-butylpyridine (tBP) (96%), lead bromide (99.99%), CB (99.8%), N,N-dimethylformamide (99.8%), and dimethyl sulfoxide (99.9%) were purchased from Sigma-Aldrich (Shanghai, China). Fluorine-doped tin-oxide (FTO) substrates were obtained from Advanced Electronic Technology Co., Ltd. (Liaoning, China). Synthesis of D–A NG-tBu and D–A NG-OMe D–A NG-OMe was prepared according to the previous literature.41 D–A NG-tBu was prepared, as follows: Perchlorinated hexa-peri-hexabenzocoronene (50 mg, 0.044 mmol), 4-tert-Butylaniline (79 mg, 0.53 mmol), Cs2CO3 (172 mg, 0.53 mmol), Pd2(dba)3 (12 mg, 0.013 mmol), and racemic-2,2′-Bis(diphenylphosphino)-1,1′-binaphthyl (16 mg, 0.026 mmol) were introduced into a 25 mL reaction tube under argon. Then toluene (5.0 mL) was added. After stirring at 105 °C for 24 h, 50 mL dichloromethane (DCM) was added. The organic phase was washed with water (50 mL) and dried over anhydrous MgSO4. After removing the solvent, the crude products were separated by a silica column using DCM/petroleum ether (60–90 °C) (1.5∶1) as eluent. Then the isolated D–A NG-tBu was further purified by high-performance liquid chromatography (HPLC) using the JAIGEL-2.5H column (Japan Analytical Industry Co., Ltd., Tokyo, Japan), and chloroform as eluent. Finally, 36 mg of D–A NG-tBu (45%) was obtained as a reddish-brown solid. 1H NMR (500 MHz, CDCl3, δ): 7.35 (d, 12H), 7.02 (d, 12H), 6.63 (s, 6H), 1.33 (s, 54H) ppm. Matrix-assisted laser desorption ionization time-of-flight mass spectrometry (MALDI-TOF-MS) m/z calcd for C102H84Cl12N6, 1818.29; found, 1818.29. Device fabrication The FTO substrate was first cleaned in deionized water, acetone, and ethanol for 15 min. Then, the drying FTO substrates were treated with oxygen plasma for 15 min. 100 μL SnO2 solution was spin-coated on the substrates at 3000 rpm for 30 s and then annealed at 150 °C for 30 min. (FAPbI3)0.92(MAPbBr3)0.08 perovskite precursor solution was prepared by dissolving FAI, MABr, MACl, PbI2, and PbBr2 in a mixed solvent dimethylformamide∶dimethyl sulfoxide (DMF∶DMSO) = 8∶1) and stirred overnight. As for deposition, perovskite precursor solution was spin-coated on FTO/SnO2 sample at 1000 rpm for 10 s and then at 5000 rpm for 25 s. 160 μL of CB was dropped at the last 5 s, and then the sample was annealed at 150 °C for 10 min. 50 μL of D–A NGs solution (0.25–3 mg/mL in CB) was spin-coated onto the perovskite layer at 4000 rpm for 30 s and heated at 100 °C for 5 min. Subsequently, 50 μL spiro-OMeTAD solution (72.3 mg spiro-OMeTAD in 1 mL CB with additives, including 17.5 μL solution of Li-TFSI (520 mg mL−1 in acetonitrile) and 28.5 μL of tBP) was spin-coated on the cooled substrates at 4000 rpm for 30 s. Finally, 120 nm of the silver electrode was thermally deposited under 5 × 10−5 Pa. Results and Discussion Energy level, structure, and radical generation of D–A NGs D–A NG-tBu was synthesized following a similar method to our previously reported D–A NG-OMe using palladium-catalyzed Buchwald–Hartwig C–N cross-coupling reaction ( Supporting Information Figure S1a).41 Mass spectra and 1H NMR spectrum of D–A NG-tBu were shown in Supporting Information Figure S1b,c. We performed density functional theory (DFT) at the B3LYP/6-31G(d) level to describe the molecular orbitals of D–A NGs ( Supporting Information Figure S2). The lowest unoccupied molecular orbitals (LUMOs) were generally distributed over the whole NG. In contrast, the highest occupied molecular orbitals (HOMOs) were mainly located at the peripheral anilino units and the core of NG, indicating the fully π-conjugated structure. The HOMO levels are calculated to be −5.00 and −5.16 eV for D–A NG-OMe and D–A NG-tBu, respectively. The shallower HOMO level of D–A NG-OMe is ascribed to the stronger electron-donating ability of methoxyl in peripheral anilino groups, which agreed well with our differential pulse voltammetry (DPV) results ( Supporting Information Figure S3). Meanwhile, the UV–vis absorption peaks of D–A NG-OMe also presented a gradual bathochromic shift compared with that of D–A NG-tBu for the above reason ( Supporting Information Figure S4). The absorption peak around 450 nm is triggered by intramolecular charge transfer from anilino groups to the inner NG, and the absorption peaks occurring around 300 nm reveal the π–π* transition.42 Combining the DPV and UV–vis results, we depicted the energy level diagram of perovskite, D–A NGs, and spiro-OMeTAD, as shown in Supporting Information Figure S5. It was apparent that the HOMO level of D–A NG-OMe was shallower than that of spiro-OMeTAD, forming a slight hole transport barrier from perovskite to spiro-OMeTAD. Nevertheless, the insertion of D–A NG-tBu could optimize the ELA between perovskite and spiro-OMeTAD, thereby facilitating the hole extraction. The single-crystal structure of D–A NG-OMe and D–A NG-tBu is presented in Figure 1a,b, and crystallographic parameters are summarized in Supporting Information Table S1. Their corresponding packing information is presented in Supporting Information Figures S6–S8. Although the peripheral aniline groups hampered π–π stacking in the solid state, the hydrogen atoms on peripheral aniline of D–A NG-tBu could connect with other peripheral anilines of adjacent molecules through C–H···π interactions (2.89 Å), providing advantageous hopping transport channels ( Supporting Information Figure S8).42 The orientation of the D–A NGs over the layer was studied by grazing incidence wide-angle X-ray scattering (GIWAXS) experiments, as shown in Figure 1c,d. Interestingly, D–A NG-OMe showed edge-on orientations on Si substrates, while D–A NG-tBu exhibited a random morphology.43,44 D–A NGs showed the lamellar structure around 0.35 Å−1 and π–π stacking structure around 1.4 Å−1 (the corresponding distances are around 18 and 4.5 Å, respectively). D–A NGs were found to form π–π stacks, advantageous for electrical conductivity. Furthermore, by combining with single-crystal data, we gained insight into the GIWAXS data. Supporting Information Figure S9a shows the powder X-ray diffraction (PXRD) pattern of D–A NG-tBu derived from single-crystal data, in which the signals of 2θ = 4.74°, 5.57°, and 10.05° correspond to q (q = 2 π/d) values of 0.34, 0.39, and 0.62 Å−1 respectively. Similarly, Supporting Information Figure S9b showed the PXRD pattern of D–A NG-OMe derived from single-crystal data. Note that scattering peaks in GIWAXS were not found in the single crystals ( Supporting Information Figure S9), indicating that the D–A NGs interacted with and oriented to the substrate surface. Figure 1 | Single-crystal structure of D–A NG-OMe (a) and D–A NG-tBu (b). GIWAXS of D–A NG-OMe (c) and D–A NG-tBu (d) film deposited on Si substrate. ESR results of D–A NG-OMe (e) and D–A NG-tBu (f). ESP graph of D–A NG-OMe (g) and D–A NG-tBu (h). Download figure Download PowerPoint Electron spin resonance (ESR) was performed to confirm the existence of stable radicals in these D–A NGs,20 from which we observed a Lorentz broadening of D–A NG-tBu samples (Figure 1f). The ESR line shape, expressed by the Lorentz function indicated a mobile spin, and the ESR line shape, expressed by the Guess function indicated a static spin.26 Therefore, Lorentz’s broadening of D–A NG-tBu was attributed to the dynamic intermolecular charge transfer.35,45 Somewhat nonsymmetric line shape was observed in the D–A NG-OMe sample (Figure 1e), which was well fitted using two Lorentz functions, as shown in Supporting Information Figure S10. The two ESR components originated from domains with different molecular orientations, shortening the spin-lattice relaxation time.46,47 The generation of stable radicals in these molecules indicated active intermolecular charge transfer. Furthermore, the electron accumulation on nitrogen (N) atoms revealed by electrostatic potential (ESP) supported our speculation of the possible formation route of radicals, as illustrated in Figure 1g,h. The fully delocalized backbone of NG forms a conjugated bridge for the π electrons. The electron-withdrawing chlorine at the ortho-position of anilino units induced the charge transfer of unpaired electrons from the N atom to the central NG. The para-substituted electron-donating group (-OMe or -tBu) stabilized the unpaired single electron on the N atom, forming a stable radical ( Supporting Information Figure S11). The effect of D–A NGs on perovskite: modulating carrier interface transport and passivating the trap density X-ray diffraction (XRD) measurements of the perovskite films with/without D–A NGs were performed ( Supporting Information Figure S12a). All the samples revealed similar diffraction peaks, where the prominent peaks at 14.1° and 28.3° were assigned to cubic perovskite structures of (001) and (002) planes.48 The diffraction peak at 12.7° belonged to the PbI2 phase.49 As expected, all the perovskite films also showed alike UV–vis absorption spectra ( Supporting Information Figure S12b). GIWAXS patterns of perovskite modified with D–A NGs are shown in Supporting Information Figure S13. Apart from the signals of D–A NGs, brighter scattering rings were assigned to the polycrystalline 3D perovskite. We confirmed that the D–A NGs layers showed the same orientation on the Si substrate and perovskite layers.50 To achieve an in-depth understanding of the interaction between D–A NGs (-tBu and -OMe) and perovskite layers, we used first-principles calculations (calculation details are shown in the Supporting Information) to identify regions where electrons are localized in atomic/molecular systems and explore the charge transfer at the molecular level. The optimized geometries of the two D–A NGs based on their single-crystal structures were used to form the adsorbed models on the perovskite. Based on the stable structures, where D–A NGs are located at the top of the (001) surface of FAPbI3 perovskite layers in the cubic phase, the electron localization function (ELF) of the perovskites with and without D–A NGs are depicted in Figure 2a–c. The I atoms located at the top layer of perovskite in good symmetric growth become asymmetric after contacting D–A NGs. The distribution of electron localization density indicated that an interaction exists between I atoms and -tBu and/or -OMe groups, which could induce interface electron transfer between D–A NGs and perovskite layers. We found that the D–A NGs layers formed a lamella structure and a π–π stack on the perovskite layer from GIWAXS measurements. These structures were not observed in the single crystals of D–A NGs, providing evidence that interactions between D–A NGs and perovskite crystals occurred and that the calculated and experimental results were consistent. Figure 2 | (a) ELF of the bare perovskites. (b) ELF of the interaction between perovskite and D–A NG-OMe. (c) ELF of the interaction between perovskite and D–A NG-tBu. Differential charge density between perovskite and D–A NG-OMe (d)/D–A NG-tBu (e). (f) The average charge changes of –R1/–R2 group of D–A NGs (pink background: D–A NG-OMe; green background: D–A NG-tBu) and I atom of perovskite. Download figure Download PowerPoint The differential charge density results provided further evidence that a strong interface coupling and noticeable charge transfer arose across the D–A NGs and perovskite interface. With Bader Charge Analysis, we showed the differential charge density between perovskite and D–A NGs in Figure 2d,e, where the yellow and blue colors represented the electron accumulation and depletion, respectively. The peripheral aniline groups with -tBu and -OMe (R1 and R2, Supporting Information Figure S14) show a prominent trend to lose electrons by transferring to the top layered I and Pb atoms. To quantitatively evaluate the extent of charge transfer across the interface, Bader charges analysis on each atomic charge before and after D–A NGs depositing on the perovskite layer in the charge transferred region (the red dotted area in Supporting Information Figure S14) was considered.51,52 The change of atomic charge is visualized in Figure 2f. After introducing D–A NGs, the average atomic charge of I atoms has changed from −0.584 to −0.351 (D–A NG-OMe case) and −0.574 to 0.836 (for D–A NG-tBu case). Undoubtedly, I atoms have obtained lots of charges after introducing D–A NGs, which means that I vacancy might be passivated.53 Moreover, the average charge change of H atoms in both D–A NGs showed a decreasing trend; the O atom was almost unchanged in D–A NG-OMe and the average C atoms in D–A NG-OMe obtain negative charges while donating negative charges in D–A NG-tBu. Comparing the single -OMe and -tBu groups in both R1 and R2 branches, the -tBu group showed a greater electron-donating ability (−0.017 to −2.031, −0.075 to −1.607) than the -OMe group (0.495 to 0.437, 0.486 to 0.424). After combining the phenyl group with single -OMe and -tBu groups, the charges of both R1 and R2 branches have significantly dropped, showing that the phenyl group possesses a great negative charge donation ability. These changes proved that the existence of D–A NGs induced a strong driving force at the interface where electrons would be transferred from D–A NGs to perovskite and holes would be transferred from perovskite to D–A NGs. Furthermore, given that organic–inorganic hybrid perovskite was an ionic semiconductor, its charged ions/carriers could migrate under bias/illumination. In this context, the electrons from D–A NGs could balance the positive charge (I vacancy) in perovskite that would otherwise trap electrons, especially when operating under illumination conditions with high charge carrier densities ( Supporting Information Figure S15). Therefore, the introduction of D–A NGs could be an effective method for enhancing the charge injection at the interfaces of PSCs and enhancing the short current.54,55 Moreover, we speculated that the D–A NGs “layer” might be discontinuous due to its low concentration and good solubility in CB solvent, thus, could not make the band bend. The interaction between D–A NGs and perovskite was further verified by X-ray photoelectron spectroscopy (XPS) measurements and the C1s spectrum was shown in Supporting Information Figure S16. As shown in Figure 3a, the Pb 4f7/2 (138.43 eV) and Pb 4f5/2 (143.32 eV) peaks of pristine perovskite shift to lower binding energies at 138.27 and 143.12 eV after incorporating D–A NGs. Meanwhile, the I 3d5/2 (619.27 eV) and I 3d3/2 (630.75 eV) in the pristine perovskite film shifted to lower positions at 619.09 and 630.56 eV (Figure 3b). Regrettably, the binding energy difference between D–A NG-OMe and D–A NG-tBu was trivial. Therefore, these shifts could be ascribed to the electron transfer from D–A NGs to perovskite, increasing the electron density of perovskite,56 consistent with the above DFT calculations. Figure 3 | (a) XPS results of Pb 4f core levels spectra of the pristine perovskite surface and perovskite/D–A NGs surface. (b) XPS results of I 3d core levels spectra of the pristine perovskite surface and perovskite/D–A NGs surface. PL (c) and TRPL (d) spectra of perovskite with or without D–A NGs on the glass substrate. (e) The hole mobility from space-charge limited current (SCLC) measurement. (f) TRPL of perovskite films with spiro-OMeTAD. Download figure Download PowerPoint Furthermore, steady-state photoluminescence (PL) and time-resolved photoluminescence (TRPL) measurements were also used to analyze the carrier recombination dynamics. Compared with the bare perovskite, the PL intensities of D–A NG-OMe and D–A NG-tBu-modified films have been remarkably increased, demonstrating efficient passivation of nonradiative trapping centers in the perovskite surface (Figure 3c).57 Meanwhile, the prolonged average carrier lifetimes (control: 90 ns; D–A NG-OMe: 279 ns; D–A NG-tBu: 679 ns) further validated the reduced trap