Solution-Processable NiOx:PMMA Hole Transport Layer for Efficient and Stable Inverted Organic Solar Cells

For organic solar cells (OSCs), nickel oxide (NiOx) is a potential candidate as the hole transport layer (HTL) material. However, due to the interfacial wettability mismatch, developing solution-based fabrication methods of the NiOx HTL is challenging for OSCs with inverted device structures. In this work, by using N, N-dimethylformamide (DMF) to dissolve poly(methyl methacrylate) (PMMA), the polymer is successfully incorporated into the NiOx nanoparticle (NP) dispersions to modify the solution-processable HTL of the inverted OSCs. Benefiting from the improvements of electrical and surface properties, the inverted PM6:Y6 OSCs based on the PMMA-doped NiOx NP HTL achieves an enhanced power conversion efficiency of 15.11% as well as improved performance stability in ambient conditions. The results demonstrated a viable approach to realize efficient and stable inverted OSCs by tuning the solution-processable HTL.


Introduction
In recent years, organic solar cells (OSCs) based on non-fullerene acceptors (NFAs) have continued to set new efficiency records by virtue of developments in novel organic photovoltaic materials as well as innovations in device-processing techniques [1][2][3][4]. For the future commercialization of NFA-based OSCs, in addition to realizing the highest possible device efficiency, it is also crucial to achieve highly stable OSCs via facile largescale, roll-to-roll compatible fabrication processes [5][6][7][8]. A modern single-junction OSC device typically consists of a bulk-heterojunction (BHJ) active layer, top metal electrodes, bottom transparent electrodes, and charge transport layers between the active layer and the electrodes. For the conventional-structured devices, metals of low work function (WF), namely aluminum (Al), calcium (Ca), or barium (Ba), are used as the cathodes on the top. In the inverted device structure, the direction of charge carrier transport is reversed. Therefore, silver (Ag) or gold (Au) with high WF are utilized as the top anodes. As Ag and Au are stable metals under ambient conditions, OSCs with the inverted structure are generally considered to possess higher stability than the conventional-structured devices; hence, they are more favorable for the mass production of practical OSCs [9,10].
Charge transport layers are introduced to improve the interfacial contacts between the active layer and electrodes and to enhance the electron/hole extraction efficiencies. For inverted OSCs, the electron transport layer (ETL) is fabricated on the transparent metal oxide cathode that can withstand high temperatures. In contrast, the hole transport layer (HTL) of the inverted device has to be deposited atop the BHJ active layer, which is vulnerable to excessive heat, moisture, and chemicals [11][12][13][14]. For such reasons, the materials and fabrication processes of the HTL in inverted OSCs demand careful selections to minimize the impacts on the BHJ active layer. The conductive polymer poly (3,4ethylenedioxythiophene):poly (styrene sulfonic acid) (PEDOT:PSS) dispersed in water is a inverted OSCs exhibited an optimized power conversion efficiency (PCE) of 15.11%. Additionally, for the device with a PMMA-doped NiOx HTL, the PCE could maintain 75% of its initial value under ambient conditions for an extended period of 30 d.

Materials and Methods
The PM6 donor and Y6 acceptor were purchased from Solarmer Materials, Inc., Beijing, China. The PMMA (Mw~350,000) was purchased from Sigma-Aldrich (Shanghai) Trading Co., Ltd., Shanghai, China. All other chemical reagents were ordered from TCI (Shanghai) Development Co., Ltd., Shanghai, China. The molecular structures of PM6, Y6, and PMMA are displayed in Figure 1a, and the device structure of the OSC is depicted in Figure 1b. To obtain NiOx NP dispersions, nickel chloride hexahydrate (NiCl2·6H2O, 2 g, 8.4 mmol) was added to 100 mL ethanol, then stirred at room temperature to obtain a clear green solution. Next, sodium hydroxide (NaOH) solution (5 mol/L) was added to the green solution until the pH value reached 10. The obtained turbid green mixture was centrifuged, and the remaining precipitation was rinsed twice with deionized water and ethanol to remove any soluble impurities. The rinsed product was dried overnight at 120 °C to obtain nickel(II) hydroxide (Ni(OH)2) powder. The Ni(OH)2 powder was calcined at 300 °C for 2 h to obtain NiOx NPs in black powder. Finally, the NiOx NPs were re-dispersed in the mixture of deionized water, isopropanol, and n-butanol (4:15:1 by volume) via ultrasonic treatment. The NiOx NP concentration was 10 mg/mL. The PM6:Y6 BHJ blend solution was prepared by stirring the chloroform-dissolved donor and acceptor blend (1:1.2 by weight) with 1-chloronaphthalene as a solvent additive (0.5% by volume) for 24 h, and the blend concentration was 20 mg/mL. Sol-gel precursor of zinc oxide (ZnO) was prepared by dissolving zinc acetate dihydrate (Zn(CH3COO)2·2H2O, 0.1 g) with ethanolamine (NH2CH2CH2OH, 0.029 mL) in 2-methoxy ethanol (CH3OCH2CH2OH, 1.0 mL) after vigorously stirring for 12 h. To obtain the PMMA-doped NiOx NP dispersions, the powder PMMA was dissolved firstly in DMF, and the PMMA concentration was 10 mg/mL. The as-prepared NiOx NP dispersions were subsequently mixed with various contents of PMMA solution (1%, 3%, or 5% v/v) and stirred for 1 h until the mixture was uniformly distributed. To obtain NiO x NP dispersions, nickel chloride hexahydrate (NiCl 2 ·6H 2 O, 2 g, 8.4 mmol) was added to 100 mL ethanol, then stirred at room temperature to obtain a clear green solution. Next, sodium hydroxide (NaOH) solution (5 mol/L) was added to the green solution until the pH value reached 10. The obtained turbid green mixture was centrifuged, and the remaining precipitation was rinsed twice with deionized water and ethanol to remove any soluble impurities. The rinsed product was dried overnight at 120 • C to obtain nickel(II) hydroxide (Ni(OH) 2 ) powder. The Ni(OH) 2 powder was calcined at 300 • C for 2 h to obtain NiO x NPs in black powder. Finally, the NiO x NPs were re-dispersed in the mixture of deionized water, isopropanol, and n-butanol (4:15:1 by volume) via ultrasonic treatment. The NiO x NP concentration was 10 mg/mL. The PM6:Y6 BHJ blend solution was prepared by stirring the chloroform-dissolved donor and acceptor blend (1:1.2 by weight) with 1-chloronaphthalene as a solvent additive (0.5% by volume) for 24 h, and the blend concentration was 20 mg/mL. Solgel precursor of zinc oxide (ZnO) was prepared by dissolving zinc acetate dihydrate (Zn(CH 3 COO) 2 ·2H 2 O, 0.1 g) with ethanolamine (NH 2 CH 2 CH 2 OH, 0.029 mL) in 2-methoxy ethanol (CH 3 OCH 2 CH 2 OH, 1.0 mL) after vigorously stirring for 12 h. To obtain the PMMAdoped NiO x NP dispersions, the powder PMMA was dissolved firstly in DMF, and the PMMA concentration was 10 mg/mL. The as-prepared NiO x NP dispersions were subsequently mixed with various contents of PMMA solution (1%, 3%, or 5% v/v) and stirred for 1 h until the mixture was uniformly distributed.
The inverted OSC devices were fabricated with the structure of indium tin oxide(ITO)/ ETL/PM6:Y6/HTLs/Ag. The pre-cleaned ITO/glass substrates were treated with UVozone for 10 min to improve their wettability with the ETL material. The ZnO precursor was subsequently spin-coated at 5000 rpm for 40 s on the ITO/glass substrate and then annealed at 150 • C for 30 min in the air to form the ETL. In a glove box of pure nitrogen atmosphere, the BHJ blend solution was spin-coated atop the ETL at 2000 rpm for 1 min to fabricate the active layer (100 nm) and then annealed at 100 • C for 10 min. Next, the HTLs (50 nm) were fabricated by spin-coating the undoped or PMMA-doped NiO x NP dispersions at 1000 rpm for 30 s, followed by the 5000 rpm spinning for 5 s to remove the redundant mixed solution for NP dispersing. After drying in the nitrogen-filled glove box for 30 min, the Ag electrodes (100 nm) were sequentially deposited under a high vacuum of 1 × 10 −5 Pa.
An LED light source (VeraSol-2, Oriel Instruments, WA, USA) was used as the sunlight simulator. Current density-voltage (J-V) measurements of the OSCs were performed with a semiconductor characterization system (4200A, Keithley, CA, USA). The external quantum efficiency (EQE) studies were performed using a quantum efficiency analysis system (Solar Cell Scan 100, Zolix, Beijing, China). The transmittance spectra of the films were measured using a UV-vis spectrophotometer (UV-3600, Shimadzu, Kyoto, Japan). Ultraviolet photoelectron spectroscopy (UPS) measurements were performed using a photoelectron spectrometer (PHI 5000 VersaProbe III, ULVAC, Chigasaki, Japan). The morphologies of different films were surveyed using an atomic force microscope (MultiMode 8, Bruker, MA, USA) and a field emission scanning electron microscope (GeminiSEM 300, ZEISS, Jena, Germany). The contact angle measurements were performed using Dataphysics OCA200. For the device stability study, all the OSCs were stored without extra encapsulation in the air at 20 • C and 40% relative humidity.

Results and Discussion
To evaluate the photovoltaic performance of inverted PM6:Y6 OSCs with PMMAdoped NiO x NP HTLs, the J-V curves were measured under AM 1.5G condition. As presented in Figure 2a and Table 1, OSCs with undoped NiO x HTLs exhibit a decent average PCE of 12.86 ± 0.15% with an open circuit voltage (V OC ) of 0.852 V ± 0.007 V, a short-current density (J SC ) of 22.58 mA/cm 2 ± 0.37 mA/cm 2 , and a fill factor (FF) of 66.71% ± 0.52%. The maximum PCE of the NiO x HTL device is 13.06%. After the doping of PMMA, the devices fabricated with NiO x :PMMA HTLs achieve higher average and maximum PCEs compared to the control device. The improvements mainly benefit from the enhanced J SC and FF values. By introducing the NiO x NP HTL with 3% of PMMA dopant, the OSC device yields the highest average PCE of 14.95 ± 0.12% and maximum PCE of 15.11%, accompanied by optimized J SC of 24.92 mA/cm 2 ± 0.13 mA/cm 2 and FF of 70.33% ± 0.45%. As the PMMA content increases to 5%, the J SC and the FF exhibit noticeable drops, resulting in declined average and maximum PCEs (14.11 ± 0.10% and 14.25%, respectively), which can be attributed to the less favorable electrical properties of the HTL caused by the excessively doped insulating polymer [31]. It can be anticipated that doping an appropriate amount of PMMA may improve the surface morphology and the deposition quality of the HTL, thus achieving superior device parameters. To further investigate the enhanced device performance, EQE spectra of OSCs with different HTLs are presented in Figure 2b. After doping PMMA into the NiOx NP HTL, the photoresponse of the modified devices received notable improvements over a broad wavelength range from 300 to 900 nm. The improvements can be ascribed to the more efficient charge carrier transportation and collection. The integrated current density values from the EQE spectra are 22.95 mA/cm 2 , 24.55 mA/cm 2 , 25.22 mA/cm 2 , and 23.97 mA/cm 2 for devices fabricated with the undoped NiOx HTL and PMMA:NiOx HTLs of different dopant ratios, respectively. The results align with the JSC values recorded from the J-V measurements, indicating their good reliability. The hole mobility (µh) of the OSCs with different HTLs was evaluated using the space-charge-limited current (SCLC) method [34]. The hole-only devices were fabricated with the structure of ITO/PEDOT:PSS/PM6:Y6/HTLs/ Ag, and the J-V characteristics in dark conditions are provided in Figure 3a. The device built with the NiOx:PMMA HTL provided a higher µh of 4.81 × 10 −4 cm 2 /Vs than the device of the undoped HTL (µh = 3.48 ×  To further investigate the enhanced device performance, EQE spectra of OSCs with different HTLs are presented in Figure 2b. After doping PMMA into the NiO x NP HTL, the photoresponse of the modified devices received notable improvements over a broad wavelength range from 300 to 900 nm. The improvements can be ascribed to the more efficient charge carrier transportation and collection. The integrated current density values from the EQE spectra are 22.95 mA/cm 2 , 24.55 mA/cm 2 , 25.22 mA/cm 2 , and 23.97 mA/cm 2 for devices fabricated with the undoped NiO x HTL and PMMA:NiO x HTLs of different dopant ratios, respectively. The results align with the J SC values recorded from the J-V measurements, indicating their good reliability. The hole mobility (µ h ) of the OSCs with different HTLs was evaluated using the space-charge-limited current (SCLC) method [34]. The hole-only devices were fabricated with the structure of ITO/PEDOT:PSS/PM6:Y6/HTLs/ Ag, and the J-V characteristics in dark conditions are provided in Figure 3a. The device built with the NiO x :PMMA HTL provided a higher µ h of 4.81 × 10 −4 cm 2 /Vs than the device of the undoped HTL (µ h = 3.48 × 10 −4 cm 2 /Vs). The enhanced hole mobility indicates that PMMA doping can effectively promote the hole transportation of the NiO x NP HTL. As presented in Figure 4a,b, the behaviors of charge carrier recombination were further explored by measuring the dependence of JSC or VOC on the various incident light intensities (Pin). The degree of bimolecular recombination was qualitatively analyzed by employing the following equation [36]: Under ideal conditions, when the value of α reaches 1, it represents the ideal condition that before recombination occurs, all the free charge carriers are collected at the electrodes. As shown in Figure 4a, the α values of the devices based on the NiOx HTL and NiOx:PMMA HTL are 0.956 and 0.976. The slightly higher α value indicates less bimolecular recombination in the PMMA-doped HTL-based OSC device. To gain an in-depth understanding of the exciton dissociation behaviors of the OSC devices, the photocurrent density (J ph ) dependence on the effective voltage (V eff ) was plotted for devices fabricated with NiO x or NiO x :PMMA HTLs (Figure 3b) [35]. J ph is defined as the difference between illuminated current densities (J L ) and dark current densities (J D ). V eff is obtained by the difference between V 0 and V a , where V 0 is the voltage when J ph = 0, and V a is the applied bias voltage. The saturated current density (J sat ) can be obtained at a high V eff when all excitons are assumed to be dissociated into free charge carriers, and the electrons/holes are collected by the corresponding electrodes. At V eff = 3 V, the undoped device has a J sat of 23.56 mA/cm 2 , while the device of the NiO x :PMMA HTL exhibits a J sat of 25.80 mA/cm 2 . Under short-circuit conditions or under maximum power output conditions, respectively, the exciton dissociation probability (P diss ) and the charge collection probability (P coll ) are defined by J ph /J sat . The OSC device fabricated with the undoped HTL has a P diss of 93.82% and a P coll of 85.66%, while the PMMA-doped HTL-based device has a P diss of 96.71% and a P coll of 87.91%. The result reveals that by introducing the PMMA-doped NiO x NP HTL, the OSC device achieves more efficient exciton dissociation as well as charge collection, which are the leading causes for the improved J SC .
As presented in Figure 4a,b, the behaviors of charge carrier recombination were further explored by measuring the dependence of J SC or V OC on the various incident light intensities (P in ). The degree of bimolecular recombination was qualitatively analyzed by employing the following equation [36]: As presented in Figure 4a,b, the behaviors of charge carrier recombination were further explored by measuring the dependence of JSC or VOC on the various incident light intensities (Pin). The degree of bimolecular recombination was qualitatively analyzed by employing the following equation [36]: Under ideal conditions, when the value of α reaches 1, it represents the ideal condition that before recombination occurs, all the free charge carriers are collected at the electrodes. As shown in Figure 4a  The situation of trap-assisted recombination was qualitatively evaluated using the following equation [37]: k is the Boltzmann constant,  is the thermodynamic temperature, and q is the elementary charge. An n value close to 1 suggests trap-assisted recombination occurs less frequently in OSC devices. As depicted in Figure 4b, the n value of the OSC device using the undoped NiOx NP HTL is 1.461. For the device with a modified HTL, the n value drops significantly to 1.176. The suppressed trap-assisted recombination can account for the enhanced FF in the PMMA-doped device.
To study the effect of PMMA doping on energy levels, UPS measurements of pristine and doped NiOx HTLs were conducted. The binding energies of the cutoff region (Ecutoff) Under ideal conditions, when the value of α reaches 1, it represents the ideal condition that before recombination occurs, all the free charge carriers are collected at the electrodes. As shown in Figure 4a, the α values of the devices based on the NiO x HTL and NiO x :PMMA HTL are 0.956 and 0.976. The slightly higher α value indicates less bimolecular recombination in the PMMA-doped HTL-based OSC device.
The situation of trap-assisted recombination was qualitatively evaluated using the following equation [37]: k is the Boltzmann constant, T is the thermodynamic temperature, and q is the elementary charge. An n value close to 1 suggests trap-assisted recombination occurs less frequently in OSC devices. As depicted in Figure 4b, the n value of the OSC device using the undoped NiO x NP HTL is 1.461. For the device with a modified HTL, the n value drops significantly to 1.176. The suppressed trap-assisted recombination can account for the enhanced FF in the PMMA-doped device.
To study the effect of PMMA doping on energy levels, UPS measurements of pristine and doped NiO x HTLs were conducted. The binding energies of the cutoff region (E cutoff ) and the onset region (E onset ) can be extracted from the UPS spectra in Figure 5a. The band gap energy (E g ) was estimated by the Tauc plot method from the transmittance spectra in Figure S1. The E g values for undoped NiO x and PMMA-doped NiO x were 3.74 eV and 3.77 eV, respectively. With the excitation energy value (hν) of 21.22 eV (He I), the WF was calculated by subtracting the E cutoff from the hν. The pristine NiO x HTL had a WF of 5.07 eV. For the NiO x :PMMA HTL, the WF value increased to 5.16 eV. Combining the results of E g , WF, and E onset , the valence band maximum (VBM) and conduction band minimum (CBM) can be calculated. For the pristine and modified HTLs, the VBM levels were 1.72 eV and 1.71 eV, and the corresponding CBM levels were 5.46 eV and 5.48 eV, respectively. Given the above results, the energy level diagram of the OSCs is plotted in Figure 5b. Rather than the recombinational hole transport of other metal oxides (e.g., MoO x ) that occurs at the interface between the active layer and HTL, NiO x exhibits intrinsic p-type hole transport, whereby the photogenerated holes from the donor are directly transported to the anode through the valence band of the HTL [29]. Therefore, the energy level matching between the WF of the NiO x HTL and the HOMO of the donor is an important factor for achieving desirable device performance [38]. The deeper WF and the shifted energy level alignments of the PMMA-doped NiO x HTL help to optimize the hole extraction and collection efficiencies, thus contributing to the improved J SC and FF for the modified OSC device. Figure S1. The Eg values for undoped NiOx and PMMA-doped NiOx were 3.74 eV and 3.77 eV, respectively. With the excitation energy value (hν) of 21.22 eV (He I), the WF was calculated by subtracting the Ecutoff from the hν. The pristine NiOx HTL had a WF of 5.07 eV. For the NiOx:PMMA HTL, the WF value increased to 5.16 eV. Combining the results of Eg, WF, and Eonset, the valence band maximum (VBM) and conduction band minimum (CBM) can be calculated. For the pristine and modified HTLs, the VBM levels were 1.72 eV and 1.71 eV, and the corresponding CBM levels were 5.46 eV and 5.48 eV, respectively. Given the above results, the energy level diagram of the OSCs is plotted in Figure 5b. Rather than the recombinational hole transport of other metal oxides (e.g., MoOx) that occurs at the interface between the active layer and HTL, NiOx exhibits intrinsic p-type hole transport, whereby the photogenerated holes from the donor are directly transported to the anode through the valence band of the HTL [29]. Therefore, the energy level matching between the WF of the NiOx HTL and the HOMO of the donor is an important factor for achieving desirable device performance [38]. The deeper WF and the shifted energy level alignments of the PMMA-doped NiOx HTL help to optimize the hole extraction and collection efficiencies, thus contributing to the improved JSC and FF for the modified OSC device. To reveal the morphological evolutions of the NiOx NP HTL after the modification of PMMA, the surface morphologies of the different HTLs were characterized by atomic force microscopy (AFM) and scanning electron microscopy (SEM). The AFM height images of the bare NiOx HTL and the PMMA-doped NiOx HTL are shown in Figure 6a,b. The bare NiOx HTL has a relatively coarse surface, accompanied by a root mean square (RMS) roughness value of 2.793 nm. By contrast, the PMMA-doped HTL has a smoother surface with a significantly decreased RMS value of 1.603 nm. From the top-view SEM images in Figure 6c,d, visible pinholes can be found for the unmodified NiOx HTL. After the doping of PMMA, the long-chain polymer with high molecular weight could fill the gaps between the NiOx NPs, resulting in an HTL film of more continuously organized NPs with fewer pinholes [39,40]. The AFM and SEM results suggest improved surface smoothness and uniformity of the NiOx:PMMA HTL film, which could be beneficial for promoting charge transportation and restricting charge recombination. The pinhole-less and denser morphology of the PMMA-doped HTL could also help to improve device stability by resisting the penetration of water, oxygen, and the top metal electrode [33]. To reveal the morphological evolutions of the NiO x NP HTL after the modification of PMMA, the surface morphologies of the different HTLs were characterized by atomic force microscopy (AFM) and scanning electron microscopy (SEM). The AFM height images of the bare NiO x HTL and the PMMA-doped NiO x HTL are shown in Figure 6a,b. The bare NiO x HTL has a relatively coarse surface, accompanied by a root mean square (RMS) roughness value of 2.793 nm. By contrast, the PMMA-doped HTL has a smoother surface with a significantly decreased RMS value of 1.603 nm. From the top-view SEM images in Figure 6c,d, visible pinholes can be found for the unmodified NiO x HTL. After the doping of PMMA, the long-chain polymer with high molecular weight could fill the gaps between the NiO x NPs, resulting in an HTL film of more continuously organized NPs with fewer pinholes [39,40]. The AFM and SEM results suggest improved surface smoothness and uniformity of the NiO x :PMMA HTL film, which could be beneficial for promoting charge transportation and restricting charge recombination. The pinhole-less and denser morphology of the PMMA-doped HTL could also help to improve device stability by resisting the penetration of water, oxygen, and the top metal electrode [33].
Contact angle (CA) measurements were carried out to collect additional evidence of improvements brought about by doping the PMMA. The CA images of undoped or PMMA-doped NiO x NP dispersions on the PM6:Y6 BHJ film are provided in Figure 7a,b. Owing to the incorporation of the polymer, the modified NiO x NP dispersions have a reduced CA from 22.17 • ± 0.17 • to 18.45 • ± 0.14 • , indicating its improved wettability on the BHJ. This could benefit the deposition quality of the HTL film, further boosting the device performance of OSCs [29]. The CA images of distilled water on the NiO x and NiO x :PMMA HTL films are displayed in Figure 7c Contact angle (CA) measurements were carried out to collect additional evidence of improvements brought about by doping the PMMA. The CA images of undoped or PMMA-doped NiOx NP dispersions on the PM6:Y6 BHJ film are provided in Figure 7a,b. Owing to the incorporation of the polymer, the modified NiOx NP dispersions have a reduced CA from 22.17° ± 0.17° to 18.45° ± 0.14°, indicating its improved wettability on the BHJ. This could benefit the deposition quality of the HTL film, further boosting the device performance of OSCs [29]. The CA images of distilled water on the NiOx and NiOx:PMMA HTL films are displayed in Figure 7c,d. Since the CA of water on the NiOx:PMMA film increased from 29.33° ± 0.19° to 35.10° ± 0.23° over the pristine film, the enhanced hydrophobicity of the HTL could provide better protection for the active layer from water in ambient conditions. Owing to the incorporation of the polymer, the modified NiOx NP dispersions have a reduced CA from 22.17° ± 0.17° to 18.45° ± 0.14°, indicating its improved wettability on the BHJ. This could benefit the deposition quality of the HTL film, further boosting the device performance of OSCs [29]. The CA images of distilled water on the NiOx and NiOx:PMMA HTL films are displayed in Figure 7c,d. Since the CA of water on the NiOx:PMMA film increased from 29.33° ± 0.19° to 35.10° ± 0.23° over the pristine film, the enhanced hydrophobicity of the HTL could provide better protection for the active layer from water in ambient conditions. Device stability is a crucial factor for the commercial applications of OSCs. To investigate the long-term effects of PMMA doping, the OSCs with undoped or PMMA-doped NiOx HTLs were stored without extra encapsulation in ambient conditions for 30 d. During this period, J-V measurements were performed daily to study the evolutions of device Device stability is a crucial factor for the commercial applications of OSCs. To investigate the long-term effects of PMMA doping, the OSCs with undoped or PMMA-doped NiO x HTLs were stored without extra encapsulation in ambient conditions for 30 d. During this period, J-V measurements were performed daily to study the evolutions of device performance parameters. As seen in Figure 8, after being stored in air for 30 d, both the devices with undoped or PMMA-doped HTLs were able to maintain over 95% of their initial V OC values. However, the undoped device suffered notable degradations in J SC and FF. Consequently, its PCE decreased to 51% of the initial values. In comparison, the J SC and FF of the PMMA-doped device declined less significantly, resulting in a PCE of 75% of the initial value. The improved device performance stability in ambient conditions can be attributed to the enhanced moisture/oxygen-blocking effects of the NiO x :PMMA HTL, which is in accordance with the results of morphology and CA characterizations. Device stability is a crucial factor for the commercial applications of OSCs. To investigate the long-term effects of PMMA doping, the OSCs with undoped or PMMA-doped NiOx HTLs were stored without extra encapsulation in ambient conditions for 30 d. During this period, J-V measurements were performed daily to study the evolutions of device performance parameters. As seen in Figure 8, after being stored in air for 30 d, both the devices with undoped or PMMA-doped HTLs were able to maintain over 95% of their initial VOC values. However, the undoped device suffered notable degradations in JSC and FF. Consequently, its PCE decreased to 51% of the initial values. In comparison, the JSC and FF of the PMMA-doped device declined less significantly, resulting in a PCE of 75% of the initial value. The improved device performance stability in ambient conditions can be attributed to the enhanced moisture/oxygen-blocking effects of the NiOx:PMMA HTL, which is in accordance with the results of morphology and CA characterizations.

Conclusions
In this study, a facile modification strategy of a solution-processable NiO x NP HTL was realized for the inverted-structured OSCs. By doping the NiO x NP dispersions with a commercially available polymer, namely PMMA, the PM6:Y6 device with the modified HTL exhibited an average PCE of 14.95 ± 0.12% and a maximum PCE of 15.11%, which was a significant improvement from the PCEs of the undoped-NiO x -HTL-based device. With more favorable electrical and surface properties of the NiO x :PMMA HTL, efficient charge transportation as well as suppressed charge recombination were simultaneously accomplished. The PMMA-doped HTL also helped the OSC device to maintain 75% of its initial PCE value after aging in ambient conditions for 30 d. Overall, this work proposed a novel solution-processable, polymer-assisted metal oxide nanocomposite HTL for the future mass production of inverted OSCs with excellent device performance and prolonged device shelf life [41].
Funding: This work was financially supported by the National Science Foundation of China (U21A20492 and 62275041), and the Sichuan Science and Technology Program (2022YFH0081, 2022YFG0012, and 2022YFG0013). This work was also sponsored by the Sichuan Province Key Laboratory of Display Science and Technology and the Qiantang Science and Technology Innovation Center.