Functionalization of Donor–π–Acceptor Hole Transport Materials Enhances Crystallization and Defect Passivation in Inverted Perovskite Solar Cells: Achieving Power Conversion Efficiency >21% (Area: 1.96 cm2) and Impressive Stability

Inverted perovskite solar cells (PSCs) mainly adopt polytriarylamine (PTAA) for the hole transport material (HTM), which usually brings about inferior interfacial contact owing to their hydrophobicity, high‐lying highest occupied molecular orbital energy level, and deficiency of passivation groups. Herein, a series of donor–π–acceptor (D–π–A) type small molecules is demonstrated based on 2,2′:6′,2″‐terpyridine (TPy) as the acceptor moiety, benzene ring as the π‐linker, and incorporating various donors to act as HTMs. These TPy‐based molecules coated atop PTAA manipulate the energy level and surface wettability, but the incorporation of the phenoxazine (POZ) donor can be prominent for enhancing charge transport and defect passivation, thereby simultaneously addressing the above‐mentioned issues. The highest power conversion efficiency of 22.36% can be achieved with an open‐circuit voltage (VOC) of 1.15 V, a short‐circuit current density (JSC) of 23.96 mA cm−2, and a fill factor (FF) of 81.16% for the optimized POZ‐TPy‐modified device. Moreover, the power PCE of a large POZ‐TPy‐modified device (1.96 cm2) can still reach more than 21%. These results are among one of the highest efficiencies for inverted PSCs, indicating the enormous potential of POZ‐TPy HTM in future perovskite applications.


Introduction
21][22] Nevertheless, the solar efficiency of inverted PSCs still falls behind those of the conventional PSCs, substantially owing Inverted perovskite solar cells (PSCs) mainly adopt polytriarylamine (PTAA) for the hole transport material (HTM), which usually brings about inferior interfacial contact owing to their hydrophobicity, high-lying highest occupied molecular orbital energy level, and deficiency of passivation groups.Herein, a series of donor-π-acceptor (D-π-A) type small molecules is demonstrated based on 2,2 0 :6 0 ,2 00 -terpyridine (TPy) as the acceptor moiety, benzene ring as the π-linker, and incorporating various donors to act as HTMs.These TPy-based molecules coated atop PTAA manipulate the energy level and surface wettability, but the incorporation of the phenoxazine (POZ) donor can be prominent for enhancing charge transport and defect passivation, thereby simultaneously addressing the above-mentioned issues.The highest power conversion efficiency of 22.36% can be achieved with an open-circuit voltage (V OC ) of 1.15 V, a shortcircuit current density ( J SC ) of 23.96 mA cm À2 , and a fill factor (FF) of 81.16% for the optimized POZ-TPy-modified device.Moreover, the power PCE of a large POZ-TPy-modified device (1.96 cm 2 ) can still reach more than 21%.These results are among one of the highest efficiencies for inverted PSCs, indicating the enormous potential of POZ-TPy HTM in future perovskite applications.
to the comparatively larger open-circuit voltage (V OC ) loss, [23][24][25][26] originating from the trap-states-induced nonradiative recombination of photogenerated charge carriers both in perovskite bulk and/or at the interface.This is due to disordered nucleation and excessively rapid crystallization with random crystal orientation and small grains.[29][30] For example, we have recently developed a natural vitamin B, namely carnitine, which contains dual effective passivating groups (i.e., carboxyl and quaternary ammonium functional groups), hence could simultaneously diminish multiple defects at grain boundaries and surfaces, resulting in a small V OC loss in the inverted PSCs. [31]Moreover, Huang et al. reported that the trap state densities at all interfaces are mainly larger than those within perovskites. [32]The defects at the interface of PSCs are mainly deep-level defects, which would turn into nonradiative recombination centers, thereby degrading the device performance.35][36][37] Given that hole transport materials (HTMs) are usually spincoated prior to the deposition of perovskite film in inverted PSCs, HTM not only plays a critical role in charge transport but also affects interfacial characteristics and acts as a template for perovskite growth.Therefore, developing more efficacious HTMs with passivation function to repair the trap densities at interface of buried layer is expected to augment the interfacial contact and further ameliorate the crystallization of perovskites, thereby reducing the interfacial trap-assisted nonradiative recombination losses, which would eventually enhance the V OC and fill factor (FF) of PSCs.
Pyridine is an extensively used Lewis base-type surface passivating molecule because it can effectively passivate uncoordinated lead ions on the perovskite surface. [38,39]In addition, the introduction of pyridine as the electron-acceptor moiety to assemble donor-acceptor (D-A)-type HTM could raise the dipole moment of the molecule, which is beneficial to the intramolecular charge transfer, while concurrently manipulating the highest occupied molecular orbital (HOMO) energy level.[45][46][47] Accordingly, pyridine derivatives may serve as an ideal building block for constructing high-performing multifunctional HTMs.
Inspired by the above-mentioned viewpoints, we report herein a synthetic methodology to prepare a series of 2,2 0 :6 0 ,2 00terpyridine (TPy)-based D-π-A small molecules with benzene π-bridge unit, namely, DPA-TPy, POZ-TPy, and CBZ-TPy (see Scheme 1), which turns out to be a very promising HTM candidate for efficient PSCs.Notably, the introduction of these predesigned TPy-based small molecules-coated atop PTAA HTM as a buried layer with appropriate energy level cascade, higher hole mobility, and sufficient passivating ability efficaciously diminishes the interfacial trap-assisted nonradiative recombination losses of perovskites.We comprehensively investigate the interplays of TPy-based molecules, PTAA, and perovskites by combining theoretical calculations and experimental studies.The results indicate that the lone-pair electrons in the phenoxazine unit of POZ-TPy forms coordination bonds with Pb 2þ in the perovskites, while the terpyridine moiety of POZ-TPy molecule is able to chemically couple with the phenyl group of PTAA, resulting in the improvement of interfacial properties of PTAA/ perovskites.The devices modified by DPA-TPy, CBZ-TPy, and POZ-TPy deliver a PCE of 20.69%, 21.56%, and 22.36%, respectively, which are consistently higher than that of the reference PTAA device (20.45%).Furthermore, the long-term ambient, thermal, and UV irradiation stabilities of devices modified by POZ-TPy are greatly improved in comparison to the reference PTAA device.

Results and Discussion
The synthetic routes for DPA-TPy, POZ-TPy, and CBZ-TPy are outlined in Scheme 1.Details of the synthetic procedures and compound characterizations are described in the Supporting Information (SI). 4 0 -(4-Bromophenyl)-terpyridine was prepared in good yield by a modified Kröhnke pyridine reaction of 4-bromobenzaldehyde and 2-acetylpyridine, [48] which was subsequently coupled with diphenylamine, phenoxazine, or carbazole via the Buchwald-Hartwig amination to readily afford the corresponding TPy-based molecules.It was expected that the TPy moiety would function as an electron acceptor as well as a passivating unit.
HOMO and the lowest unoccupied molecular orbital (LUMO) energy levels of TPy-based HTMs were determined by cyclic voltammetry (CV) measurements as shown in Figure S1b, Supporting Information.The HOMO energy levels were estimated using the empirical equation onset and E (ferrocene) onset express the onset oxidation potentials of the targeted compounds and ferrocene, respectively.The LUMO energy levels were calculated according to the formula E LUMO = E HOMO þ E g opt .As a result, the HOMO energy levels of DPA-TPy, POZ-TPy, and CBZ-TPy were estimated to be À5.44,À5.33, and À5.70 eV, respectively.In this regard, the subtle modification of donor units with different push-pull electron abilities could efficaciously tune the HOMO energy level of HTMs, so that the resulting HTMs can have a suitable cascade energy level alignment with PTAA and perovskite to propel the driving force for exciton dissociation.Also, the LUMO energy levels of DPA-TPy, POZ-TPy, and CBZ-TPy were calculated to be À2.53,À1.92, and À2.55 eV, respectively.The corresponding energy level diagram of materials is shown in Figure S1d, Supporting Information.
To gain a deeper understanding of the interaction between D-π-A molecules and the PTAA layer, we performed molecular and periodic density functional theory (DFT) and timedependent DFT calculations.The structural optimization and optical excitation energies of adsorbates were carried out at the M062X/6-311 þ G(d,p) level by Gaussian 16 package. [49]he solvent effects were considered using a polarizable continuum model (PCM) with N,N-dimethylformamide (ε = 36.7).The energy minima were confirmed with no imaginary frequencies by vibrational frequency analyses.The calculated optical excitations, compositions, and natural transition orbitals (NTOs) for the optimized adsorbates are shown in Table S1 and Figure S2, Supporting Information.NTO analyses of DPA-TPy and CBZ-TPy show a combination of local excitation (LE) and a certain degree of charge-transfer (CT) transitions from the hole located on DPA and CBZ, respectively, to the electron located on TPy.In contrast, POZ-TPy displays exclusively CT transitions.The CT character is believed to contribute to improved adsorption performance on PTAA, which serves as the donor in the complexes.
Periodic adsorbate and substrate complexes for the DPA-TPy, POZ-TPy, and CBZ-TPy/PTAA model were performed by the Vienna ab initio simulation package (VASP 6.3.1) and employed the projector-augmented waves (PAW) plane-wave basis set with the Perdew-Burke-Ernzerhof (PBE) functional. [50]For all the geometric optimization, total energy and charge density calculations, a 3 Â 3 Â 1 Monkhorst-Pack k-point mesh for sampling of the Brillouin zone was used and cutoff energy of 500 eV was applied.All structures were fully optimized until the energy and residual force converge to 10 À3 eV and 0.02 eV Å À1 , respectively.A vacuum thickness of 20 Å was adopted along the z-axis to avoid the interaction between adjacent unit cells.The adsorption energy (E ads ) between PTAA layer and DPA-TPy, POZ-TPy, and CBZ-TPy was calculated by equation [51] E ads where E sub , E ab , and E sys are the energies of the substrates, adsorbates, and systems (i.e., substrates with adsorbate), respectively.According to this equation, the systems that adsorbate approaching to substrate via the TPy terminal, CBZ-TPy/PTAA, DPA-TPy/ PTAA, and POZ-TPy/PTAA, possess higher E ads value (panels a À c of Figure 1 and Table S2, Supporting Information), which results in more stable adsorption structures.The computational results of adsorption structures are corresponding to electron D-A characters of substrate and adsorbates.As the aspect of charge analysis, charge density difference plots in Figure 1, which evidence the interaction between substrate and adsorbates, was obtained by following equation [52] ρ ads where ρ sub , ρ ab , and ρ sys are the charge densities of the substrates, adsorbates, and substrates with adsorbate, respectively.In addition, we analyzed the sum of Bader charges on the atoms (N1, N2, C1) according to the previous reports, [53,54] which express the shortest distance between these molecules and PTAA.The results of Bader charge analysis, listed in Table S3, Supporting Information, indicate the sum of Bader charges to be 2.24, 2.16, and 2.22 e for POZ-TPy, DPA-TPy, and CBZ-TPy, respectively.A better Bader charge is conducive to molecular adsorption and can effectively passivate the interfacial defects.
We then performed the steady-state photoluminescence (PL) measurement to verify the intimate contact between TPy and PTAA.We observe that the PL peaks of DPA-TPy, POZ-TPy, and CBZ-TPy are located at 447, 553, and 452 nm, respectively (Figure 1d), while the PL peaks of the TPy/PTAA series are all situated at 490 nm (Figure 1e), suggesting that TPy and PTAA undergo the energy transfer, resulting in the same wavelength of PL emission.This result is in good agreement with DFT theoretical calculations.
Figure 2a shows the contact properties between perovskite precursor (DMF:DMSO, 85:15, v/v) and pristine PTAA and PTAA/TPy-modified HTMs.After stabilization for 30 s, the contact angle of PTAA was 59.38°, while the contact angles of PTAA/ DPA-TPy, PTAA/CBZ-TPy, and PTAA/POZ-TPy substrates were 44.26°, 41.26°, and 39.48°, respectively, thus indicating that the surface wettability is strikingly enhanced.The reduction of the surface energy discrepancy between the perovskite precursor and substrate confirms that TPy-based HTMs are favorable for the spontaneous formation of uniform thin films (Figure 2b).This phenomenon can be attributed to the fact that the TPyacceptor unit can chelate with the hydrophobicity of the PTAA surface, acting as a molecular bridge, enhancing the processing performance of perovskite on PTAA while reducing the defects at the PTAA/perovskite interface.This is considered to facilitate the rapid transport of charge carriers and alleviate the degree of device degradation to some extent.
In general, perovskite films with high crystalline orientation often have less defect densities, higher charge carrier mobilities, and lower trap-assisted nonradiative recombination losses, which are the key factors for fabricating high-quality PSCs.The surface morphology of perovskite films with and without TPy-based HTM modification was investigated by scanning electron microscopy (SEM).Clearly, one can observe that the surface of PTAA film presents the pinhole-rich and uneven morphology (Figure 2c).By contrast, the PTAA/POZ-TPy-modified perovskite film (Figure 2f ) reveals more homogeneous surface with larger grain size in the whole region, indicating improved film quality and well-repaired defects at grain boundaries.Larger perovskite grain sizes and fewer grain boundaries represent effective reduction of defect densities, which can suppress trap-assisted nonradiative recombination in PSCs.Meanwhile, the cross-sectional SEM images indicate that dense and uniform perovskites grown atop PTAA/POZ-TPy HTM were free of cracks (Figure S3, Supporting Information).
To gain in-depth insight into the structureÀfunction relationship of the interactions between the TPy-based HTMs and perovskites, X-Ray photoelectron spectroscopy (XPS) measurements were further conducted (Figure 3a,b).Apparently, compared to the perovskite films with PTAA, I 3d peaks of the perovskite films with these three TPy-functionalized HTMs exhibit a slight shift toward the lower binding energies, suggesting the robust interactions between the TPy-HTMs and perovskites, [55] which can effectively enhance the passivation ability to uncoordinated defects and diminish the interfacial nonradiative recombination.Additionally, we analyzed the PbI 2 -containing TPy series using proton nuclear magnetic resonance ( 1 H NMR) spectroscopy to obtain further information, as shown in Figure S17-S19, Supporting Information.The results indicate a significant shift in the pyridine signal for the TPy-series, suggesting that iodide ions interact with pyridine.However, only POZ-TPy was found to exhibit a significant shift in the spectrum of Pb 4f (Figure 3a), indicating that the oxygen atom in POZ interact with lead.
Since HTM in inverted PSCs plays a linchpin role in the crystallization of perovskites, we thus study the X-Ray diffraction (XRD) patterns to assess the quality of the perovskites grown on TPy-modified and PTAA HTMs (Figure 3c).The small peak at 12.6°for PTAA and PTAA/DPA-TPy perovskite films is attributed to the residue of PbI 2 crystals.Nevertheless, for the PTAA/ POZ-TPy and PTAA/CBZ-TPy perovskite films, there is no obvious diffraction peak at 12.6°, suggesting these two perovskite films are free of PbI 2 .In addition, all perovskites display similar diffraction peaks, but subtle differences in peak positions can be scrutinized when the XRD patterns are locally enlarged.The zoomed-in (100) and (200) reflections are shown in Figure 3d, where the perovskites atop TPy-modified HTMs exhibit a distinct blue-shift in comparison to that of PTAA, revealing that the perovskite tends to form a single-phase, [56] which is favorable for reducing defects in perovskites.The improved quality of perovskite film could be attributed to its suitable surface wettability and efficient passivation capability, which may mitigate the defects inside the perovskite and thus diminish the interfacial recombination of charge carriers.
To further realize how POZ-TPy improves the crystal quality, the azimuth angle of the crystal can be obtained by 2D grazing incidence wide-angle X-Ray scattering (GI-WAXS) analysis and integration of the (100) reflection (Figure S4, Supporting Information).At the azimuth angle of 90°, the POZ-TPy-treated perovskite reveals a very sharp peak than that of the control PTAA film (Figure S4c, Supporting Information), manifesting that POZ-TPy produced a more preferred out-of-plane orientation of the perovskite grains.[59][60] These findings demonstrate that the PTAA/POZ-TPy-modified HTM can induce better crystallinity and preferential orientation of perovskite compared to that of the reference PTAA film, which is beneficial for charge transport and device stability (vide infra).
We then further analyzed how PTAA/POZ-TPy affects perovskites by depth-profiling XPS shown in panels (a-d) of Figure 4. Interestingly, the Pb 4f and I 3 d depth signals of PTAA are discontinuous, while those of PTAA/POZ-TPy are tidy and continuous, suggesting that a larger contact angle leads to the discontinuity of the perovskite phases, which in turn causes perovskites exhibiting multiple interfaces and chaotic phases.The results manifest that POZ-TPy not only reduces the contact angle to enhance surface wettability, but also efficaciously anchors the precursor of perovskites, diminishing the interfacial defects of perovskites, hence ameliorating the crystallinity of perovskites.
Figure 5a,b depicts the currentÀvoltage ( J-V ) characteristics and incident photon-to-current efficiency (IPCE) spectra of champion devices with various HTMs, and the relevant photovoltaic parameters are summarized in Table 1.The device fabrication conditions are detailed in Supporting Information.It can be observed that with the introduction of TPy-functionalized HTMs, the performance of PSCs is gradually enhanced, where the FF of the devices exhibits a similar trend as that of the J SC .The control PTAA-based device conveys a champion PCE of 20.45% together with a V OC of 1.12 V, a J SC of 23.46 mA cm À2 , and an FF of 77.83%.Comparatively, the POZ-TPy-modified device achieves an excitingly higher PCE of 22.36% with a V OC of 1.15 V, a J SC of 23.96 mA cm À2 , and an FF of 81.16%, which not only outperforms the reference device, but also is among one of the highest values for dopant-free inverted PSCs.Apparently, the enhancement of all photovoltaic parameters after modification can be attributable to the upgrading film quality of perovskite, lower interfacial energy barrier, and reduction of the trapassisted recombination losses.As anticipated, the J SC values calculated from integration of the IPCE spectra (Figure 5b) are 23.2 and 22.7 mA cm À2 for the POZ-TPy and control PTAA devices, respectively, which match well with those obtained from the J-V measurements, supporting the reliability of the photovoltaic measurement.
In accordance with the above-mentioned results, the defect passivation effects of POZ-TPy should be the main reason for reducing the nonradiative recombination loss.However, except surface defects, the hole extraction/transfer rate between the HTM and perovskite is another essential factor affecting the interfacial nonradiative recombination.That is to say, efficacious hole extraction/transfer can avoid holes from lingering at the interface, resulting in unnecessary charge recombination.We further performed the steady-state PL and time-resolved photoluminescence (TRPL) measurements to explore the effects of TPy-HTM on photogenerated carriers in the perovskite film.Comparatively, the POZ-TPy-modified perovskite film reveals the most effective PL quenching efficiency, suggesting a significant suppression of nonradiative recombination (Figure 4e).Likewise, the corresponding TRPL spectra (Figure 4f ) consist a fast decay component (τ 1 ) and a slow decay component (τ 2 ) acquired by fitting the PL decay with a biexponential function: , where the fast τ 1 lifetime is related to the charge-transfer quenching at interface and the long τ 2 lifetime is attributed to the free charge recombination quenching in the perovskite film. [61,62]It is evident that τ 1 of the perovskite on PTAA/POZ-TPy is shortened to 16.9 ns in comparison with pristine PTAA (17.9 ns).[65] Again, the reduced average lifetime suggests that the POZ-TPy strategy simultaneously deactivates trap-assisted recombination and facilitates charge transport.The reductions in PL intensity and carrier lifetimes lead us to conclude that all TPy-HTM modifiers could boost interfacial charge transfer and inhibit interfacial charge recombination, which is derived from the increase in grain size, enhanced crystallinity, thereby improving the morphology of perovskite films and effectively passivating interfacial defects.
Figure 5c displays the V OC versus light intensity for the control PTAA and PTAA/POZ-TPy-treated devices with a slope of kT/q, where k, T, and q are the Boltzmann constant, temperature in Kelvin, and elementary charge, respectively. [66,67]A larger slope connotes a greater possibility of trap-assisted recombination.Therefore, the slope for the PTAA/POZ-TPy modified device (1.20 kT/q) is lower than that of the reference PTAA counterpart (1.89 kT/q), indicating that the POZ-TPy modifier can alleviate trap-assisted nonradiative recombination, which is instrumental to the V OC improvement.This consequence is consistent with the PL and TRPL results (vide supra).Subsequently, we executed the electrochemical impedance spectroscopy (EIS) measurement to probe the impact of defect densities on the charge transport and recombination dynamics in these devices.Figure 5d exhibits the Nyquist plots together with an equivalent circuit of the control and POZ-TPy-modified devices measured at a bias of 0.8 V in the dark.The Z 0 -intercept part in the high-frequency region and the diameter of semicircle in the low-frequency region are attributable to the series resistance (R s ) and charge recombination resistance (R rec ), respectively. [68]The R s value of pristine PTAA device was 12.7 Ω, while the R s value of PTAA/POZ-TPymodified device dropped to 11.6 Ω, indicating that POZ-TPy modifier is beneficial for hole extraction.Also, the R rec values of pristine PTAA and PTAA/POZ-TPy-treated devices were fitted to be 2.31 Â 10 5 and 3.87 Â 10 5 Ω, respectively.71][72] Capacitance-voltage (C-V ) measurement (Figure 5e) was then carried out to study the influence of the designed POZ-TPy molecule on the built-in potential (V bi ).The large increase in V bi of the POZ-TPy-modified device connotes an elevated driving force for photogenerated carrier separation and extraction, thanks to the large reduction in the trap densities at grain boundaries and surface of perovskites as well as the cascade of energy level alignments at the HTM/perovskite interface, which is advantageous for improving the V OC of PSCs. [73]To gain deep insight into the augmented V OC in the POZ-TPy-treated devices, the PSCs were operated as LED devices at varied bias voltages.Figure 5f presents the EQE EL of PTAA and PTAA/POZ-TPymodified PSCs as a function of injection current density.When the injected current density is equal to the J SC of the device, the EQE EL of PSC modified by PTAA/POZ-TPy is close to 1.82%, which is approximately twofold higher than that of PSC based on PTAA (0.98%).On the whole, a higher EQE EL usually expresses the reduced nonradiative recombination in PSCs, giving rise to a higher V OC of devices.Consequently, the increase in V OC in the PTAA/POZ-TPy-modified PSCs is primarily owing to the significant inhibition of nonradiative recombination within PSCs after passivation of perovskite defects.
To gain further insight into the impact of the designed POZ-TPy modifier on PSC stability, we then executed the UV irradiation, thermal stability, and long-term ambient stability measurements of the reference and champion devices.To investigate storage stability in the dark, we fulfilled long-term aging tests on the unencapsulated PSCs in an ambient environment with a relative humidity of 25 AE 5% (Figure 7a).When the devices were stored for more than two months, the PCE of the reference PSC gradually drops to 63% of its initial value, and the PCE shows a markedly rapid degradation after 35 days.By contrast, the PCE decay of the optimal PTAA/POZ-TPy-modified device remains very flat, and the device could maintain 95% of its original efficiency even after 2 months.The main reason for the raised long-term device stability of the POZ-TPy modifier can  be attributed to the increased crystallinity of the perovskite film induced by the POZ-TPy.Because the increase in the grain size of the perovskite film diminishes the grain boundary density, thereby enhancing the resistance capability of the perovskites to water and oxygen in an ambient environment.Figure 7b exhibits the maximum power point (MPP) tracking of the devices with constant simulated AM 1.5 G illumination in a nitrogen-filled glove box.The control device shows inferior performance, with 31% loss of its initial PCE after 700 h aging test, while the POZ-TPy-modified counterpart retains 91% of its original PCE under the same condition.We then performed the thermal stability test of the champion and control devices under continuous heating at 80 °C for 200 h in a nitrogen-filled glove box (Figure 7c).The efficiency of the control device drops to 74% of its original PCE, whereas the champion POZ-TPy device maintains 89% of its initial efficiency after 200 h.The enhanced thermal stability of POZ-TPy-treated devices can be attributable to the improved PTAA/perovskite interface, including better crystallinity and larger grain size of perovskites, suppression of ion migration, and reduction of surface defects.Last but not the least, we examine the performance of the newly designed POZ-TPy HTM in a larger-sized device (1.96 cm 2 ), as shown in Figure 7d.The POZ-TPy-modified PSC achieves a promising PCE of 21.17%, along with a V OC of 1.14 V, a J SC of 23.71 mA cm À2 , and an FF of 78.32%, while the PTAA counterpart delivers a PCE of 18.61%, together with a V OC of 1.11 V, a J SC of 22.98 mA cm À2 , and an FF of 72.97%.This is one of the highest efficiencies for inverted-structure PSCs with a working area of over 1.9 cm 2 .These excellent results suggest that the new POZ-TPy HTM holds great promise for future large-scale applications.Not only that, we further optimized the POZ-TPy device in combined with our previously published additive OMe-ZC3, [75] which can push ahead with the PCE soaring to 23.43%, along with a V OC of 1.16 V, a J SC of 24.37 mA cm À2 , and an FF of 82.88%.The detailed photovoltaic data were recorded in Figure S6, Supporting Information.In short, POZ-TPy exhibits striking device stability and optimized inclusion.

Conclusion
In summary, we have demonstrated a series of D-π-A type small molecules based on TPy as the acceptor moiety and benzene ring as the π-linker, which are discriminated by various donors, as HTMs for inverted PSCs.They could apparently lower the surface energy of hydrophobic PTAA, thereby improving the quality of perovskite films.Meanwhile, the POZ-TPy interface modifier not only effectively regulates the crystallization of perovskite with larger grains but also efficaciously passivates surface defects, which is beneficial to greatly reduce the nonradiative recombination of charge carriers inside perovskites and at its interface with the HTM, thereby achieving a small V OC loss of 0.12 V in the inverted PSCs.Furthermore, POZ-TPy can effectively manipulate the PTAA/perovskite interfacial energetics for a larger built-in potential and tuned a more suitable energy level alignment within PSCs, consequently decreasing the charge-extraction barrier.As a result, the POZ-TPy-modified device delivered an impressive PCE of 22.36%, along with a V OC of 1.15 V, a J SC of 23.96 mA cm À2 , and an FF of 81.16%, which not only outperform the reference PTAA devices (20.45%), but is also among one of the highest values for dopant-free inverted PSCs.This work thus delivers a simple strategy to boost the interfacial properties of PSCs and provides a new avenue to reduce the energy losses for highperforming large-area perovskite devices.

Figure 3 .
Figure 3. XPS spectra of the core-level a) Pb 4f and b) I 3 d elements of perovskite films with/without DPA-TPy, POZ-TPy, and CBZ-TPy HTMs.c) XRD patterns of perovskite films grown on varied substrates.d) Enlarged XRD patterns of (100) and (200) crystallographic planes.

Figure 5 .
Figure 5. a) Current densityÀvoltage ( J-V ) curves of the champion devices modifying PTAA with/without TPy-series treatment.b) IPCE spectra and integrated current density of the champion devices with/without modification.c) V OC versus light intensity plot of PSCs devices.d) Nyquist plot, e) Mott-Schottky plot, and f ) EQE EL -J SC curves of PSCs modifying PTAA with/without POZ-TPy.

Figure 7 .
Figure 7. a) Ambient stability of the unencapsulated PTAA and PTAA/POZ-TPy devices tested at RH: 25 AE 5% and stored in a dark room.b) Continuously illuminated with 1 sun illumination of unencapsulated devices in a nitrogen-filled glove box.c) Thermal stability of the devices tested at 80 °C in a nitrogenfilled glove box.d) J-V characterization of large-area devices of PTAA and PTAA/POZ-TPy modified perovskites based on hole-only devices.

Table 1 .
Photovoltaic parameters of champion PSCs.