Surface Passivation of Sputtered NiO x Using a SAM Interface Layer to Enhance the Performance of Perovskite Solar Cells.

Sputtered NiO x (sp-NiO x ) is a preferred hole transporting material for perovskite solar cells because of its hole mobility, ease of manufacturability, good stability, and suitable Fermi level for hole extraction. However, uncontrolled defects in sp-NiO x can limit the efficiency of solar cells fabricated with this hole transporting layer. An interfacial layer has been proposed to modify the sp-NiO x /perovskite interface, which can contribute to improving the crystallinity of the perovskite film. Herein, a 2-(3,6-dimethoxy-9H-carbazol-9-yl)ethyl]phosphonic acid (MeO-2PACz) self-assembled monolayer was used to modify an sp-NiO x surface. We found that the MeO-2PACz interlayer improves the quality of the perovskite film due to an enlarged domain size, reduced charge recombination at the sp-NiO x /perovskite interface, and passivation of the defects in sp-NiO x surfaces. In addition, the band tail states are also reduced, as indicated by photothermal deflection spectroscopy, which thus indicates a reduction in defect levels. The overall outcome is an improvement in the device efficiency from 11.9% to 17.2% due to the modified sp-NiO x /perovskite interface, with an active area of 1 cm2 (certified efficiency of 16.25%). On the basis of these results, the interfacial engineering of the electronic properties of sp-NiO x /MeO-2PACz/perovskite is discussed in relation to the improved device performance.

Introduction

Perovskite solar cells (PSCs) offer several advantages because of their ease of fabrication, low cost, and ability to produce transparent, flexible devices with good quality on a laminated output.[1,2] The device performance of PSCs with inverted structure is influenced by the hole transport layer (HTL). Some semiconductor materials that are used as hole transport materials (HTMs) in PSCs have attracted the attention of researchers; these materials include PEDOT:PSS, CuO, graphene oxide, V2O5, PbS, and PTAA (poly(bis(4-phenyl)(2,4,6-trimethylphenyl)amine)).[3−7] The main reasons for preferring these inorganic HTMs over organic HTMs are the higher hole transporting ability, higher stability, and lower cost of manufacturing of the first.

Notably, NiO has been successfully applied as a wide bandgap (3.6–4.0 eV) p-type semiconductor material in an inverted structure.[8−10] The preference of NiO is based on its intrinsic properties, including suitable work function and adequate charge carrier mobility, which can sufficiently match the energy level of perovskites by adjustment of the O2– or Ni2+ concentration.[11−14] Another advantage of NiO is that multiple methods are applicable and available for its deposition,[15−17] whereby sputtering offers suitable control of the composition of NiO, allowing for roll-to-roll fabrication.[18,19] Perovskite based on sputtered NiO (sp-NiO) can show an operational stability as high as 4000 h.[20] According to its stoichiometry,[21] NiO is a Mott–Hubbard insulator. However, nonstoichiometric composites (such as NiOOH and Ni2O3, and Ni3+ species) can be induced by the oxidation of NiO, which significantly improves the p-type conductivity.[19]

Pristine NiO has low conductivity, which may degrade hole extraction by aggravating charge carrier recombination.[22] Consequently, methods to increase the conductivity of NiO have emerged for treating NiO.[11,23−25] Notably, nickel vacancies dominate the p-type conduction in nontreated NiO.[26] Moreover, the internal p-type conductivity is limited because the Ni vacancies in untreated NiO have a large ionization energy. Consequently, extrinsic treatments such as dopants, which contain shallow acceptor levels, are preferred,[27] helping to increase the conductivity of NiO and, thus, yielding enhanced PSCs.[24,28−30]

In addition, the charge carrier transfer is significantly affected by the interface defects/traps occurring between the NiO and perovskite layers because charge extraction takes place at the interface, which suffers from charge recombination.[31] These defects may be minimized by introducing a layer between perovskite and NiO. Moreover, the introduced interface layer improves energy level matching; hence, this has become a preferred method for realizing further improvement of PSC performance.[32] Furthermore, self-assembled monolayers (SAMs) form layers by the self-assembly of surfactant molecules at surfaces. SAMs are capable of being physisorbed or chemisorbed onto a number of surfaces and forming extremely thin layers when they are applied using solution-processed techniques, such as dip coating, slot-die coating, blade, spraying, and spin coating.[33] Surface modification using SAM has been widely applied in PSCs.[34−36] Consequently, SAM-based HTL for inverted perovskite solar cells have been used[37−39] and helped in achieving 21% efficiency for a single-junction device.[40]

Bai et al.[41] modified a NiO crystal film surface using a small molecule (diethanolamine). As a result, the chemical reaction rate of the conversion of PbI2 to MAPbI3 (MA+ = CH3NH3+) was slowed, creating a better interface and film quality. Zhang et al.[42] focused on reducing trap-assisted recombination between the NiO and perovskite layer by introducing ferrocenedicarboxylic acid (FDA) to modify NiO. Hence, the modifications yielded improved power conversion efficiencies (PCEs) to 18.2% with improvement of the crystallization of the perovskite layer, hole transport, and collection abilities. Furthermore, the resulting PSCs were stable, which concurs with another modification that was performed using ferrocenedicarboxylic acid and PTAA. This modification attained better perovskite films with efficient hole extraction.[43] To passivate the NiO surface, Wang et al.[32] investigated a series of para-substituted benzoic acid (R-BA) SAM layers on NP NiO. As result, the devices that included SAM had PCEs of 18.4% and were less affected by trap-assisted recombination, minimized energy offset between NiO NP and perovskite, and changed the surface wettability.

Recently, it has been shown that modification of NiO is the key challenge to improve the open-circuit voltage (Voc) by reducing the defects of NiO.[44−46] In addition, to reduce the surface defect and hydroxyl group presented on NiO, Mann et al. used 3-(triethoxysilyl)propylamine (TSPA) as the SAM between NiOx and perovskite. They found that SAM passivated the surface of NiO and reduced the recombination of the charge.[47]

A new generation of SAM, [2-(3,6-dimethoxy-9H-carbazol-9-yl)ethyl]phosphonic acid (MeO-2PACz), was developed for the first time by Al-Ashouri et al. as a hole-selective contact with intrinsic scalability, ease of processing, low cost, and free of dopants. Another improvement entails enabling highly efficient p–i–n PSCs and a record-efficiency monolithic perovskite/CIGSe tandem device.[48] Furthermore, selecting SAMs for application in perovskite devices is an important factor, whereby ([2-(9H-carbazol-9-yl)ethyl]phosphonic acid) (2PACz) and ([2-(3,6-dimethoxy-9H-carbazol-9-yl)ethyl]phosphonic acid) (MeO-2PACz) can create an interface that is energetically well aligned with the perovskite absorber with minimal nonradiative recombination.[48] After this work, they further designed a tandem perovskite solar cell using the same SAM (MeO-2PACz) and found that a fast hole extraction was linked to a low ideality factor.[49] They also investigated the ITO surface coverage using MeO-2PACz on the top surface and introduced NiO as the intermediate layer.[40] Lastly, Sun et al. developed a method to enhance the interaction between MeO-2PACz and ITO using a sputtered NiOx layer with triple-cation perovskite devices. The result shows that NiO passivates the ITO and prevents direct contact between perovskite and ITO, which contributes to improving the PCE.[50] Herein, we developed a similar strategy using sp-NiO by coating MeO-2PACz on top of sp-NiO to study the surface defect of NiO and perovskite interfaces using different analytical techniques with the commonly used perovskite methylammonium lead iodide (MAPbI3). Despite these remarkable contributions, the application of MeO-2PACz, by modulating the crystal contact or layer between NiO/perovskite, needs more investigation. Therefore, the current study involves a systematic investigation of the effect of MeO-2PACz on a NiO layer. Second, an in-depth analysis of the defect at NiO and perovskite layers is conducted, whereby a thin layer of MeO-2PACz is inserted, as illustrated in Figure .

Figure 1

Schematic representation of the device structure and chemical structure of MeO-2PACz.

The surface functional groups of NiO are composed of hydroxylated groups[51,52] that react with phosphonic groups.[41,53] Hence, the three main goals of using a SAM include the following: First, a SAM can synchronously contribute to improving the crystallinity and stability of the perovskite layer. Second, a SAM should be able to passivate the defects of the interface. Third, a SAM can be applied to have superior interface contact properties compared to the plain NiO layer. The current work shows that MeO-2PACz reduces the defects in sp-NiO and enhances the quality of the perovskite film by enlarging the domain size, increasing the charge attraction efficiency, and reducing charge recombination. In this case, PSCs with a modified sp-NiO/perovskite interface yielded PCEs of 17.2% with an active area of 1 cm2 (certified efficiency of 16.25%, Figure S3). Therefore, the outcome proves that it is feasible to treat the sp-NiO/perovskite interface because the performance of PSCs with a sp-NiO HTL is significantly improved.

Experimental Section

Structure of the Device

The device containing the SAM interface had the following structure: ITO-coated glass/sp-NiO/MeO-2PACz/perovskite and (CH3NH3PbI3)/PC61BM/aluminum-doped zinc oxide (AZO)/Ag, which is illustrated in Figure . In this representation, sp-NiO acts as the HTL layer, which blocks electrons, and PC61BM/AZO acts as the electron transport layer.

Hole Transport Layer Deposition

A radiofrequency sputtering method was applied to deposit the 20 nm NiO hole transporting layer onto indium tin oxide (ITO) glass (10 ± 2 ohm/square, Ra < 2.6 nm) at room temperature. The radio frequency (RF) sputtered equipment was obtained from Sanyu Electron Co., Ltd., Tokyo, Japan, SVC-700 RF II NA. Commercially available 99.9% pure NiO (Kojundo Chemical Lab. Co., Ltd., Saitama, Japan) was used as the target, and the sputtering was done at 3.5 Pa argon pressure. The sputtering chamber was evacuated to reach less than 2 × 10–3 Pa before the deposition.

Self-Assembly Monolayer Deposition

The glass substrate with sputtered NiO was brought inside a glovebox. SAM preparation utilized MeO-2PACz (TCI, >98.0%) (6 mg) dissolved in 18 mL of ethanol. Extra care was taken to prevent particle formation, whereby all solutions were prepared inside a glovebox and filtration was carried out using 0.22 μm syringe filters. The next step was to spin coat the MeO-2PACz over sp-NiO HTL at 3000 rpm for approximately 30 s. Annealing was performed for 10 min on a hot plate (100 °C). The perovskite film deposition entailed two steps, starting with spinning.

Perovskite Layer Deposition

Perovskite layer deposition began with the preparation of the perovskite precursor solutions, whereby 5-AVAI (5-aminovaleric acid hydroiodide) (6.3 mg) and PbI2 (Kanto Chemical, 98% purity) (1260 mg) were dissolved in DMF–DMSO (2.85–0.15 mL). Methylammonium chloride (MACl) (Wako Chemicals, battery grade]) (5 mg) and methylammonium iodide (MAI) (Wako Chemicals) (95 mg) were dissolved in 2 mL of ethanol and then stirred overnight at 300 rpm/70 °C. Subsequently, spin coating was performed by preparing perovskite using a two-step interdiffusion method; more details can be found in our previous work.[54] The perovskite films were deposited in two steps, starting with spin coating PbI2 solution onto MeO-2PACz (3000 rpm, 30 s), annealing for 3 min on a hot plate (100 °C) and spin coating the MAI solution (4000 rpm, 30 s) as a second step. The films obtained were then annealed under a methylammonium chloride (MACl) vapor environment to improve the perovskite film.[55]

Electron-Transport Layer and Metal Electrode Deposition

Preparation of the metal electrode and electron transport layer began by dissolving phenyl-C61-butyric acid methyl ester (PC61BM) in 20 mg/mL of chlorobenzene, which was followed by spin coating over the previously prepared perovskite films (1000 rpm, 7 s, then 3000 rpm for 30 s). Thereafter, the samples were annealed at 100 °C for 15 min. To spin coat aluminum-doped zinc oxide (AZO) (Avantama AG (N-21X)), the AZO solution was placed onto the PC61BM layer (1500 rpm, 5.5 s, then 4000 rpm, 20 s), which was followed by annealing on a hot plate (100 °C, 10 min). The cells were completed by thermally evaporating 150 nm thick Ag electrodes. The final step of the process involved encapsulation with cavity glasses, which were sealed using ultraviolet curable resins, including UVRESIN XNR5516Z, NagaseChemteX, Japan.

Characterization

X-ray diffraction (XRD) patterns were obtained using a Bruker D8 advanced X-ray diffractometer, Cu Kα radiation, λ = 1.54050 Å, with a scan rate set at 3° min–1. A field emission scanning electron microscope (Hitachi S-4800) was used to capture a morphological image of the films, and the accelerating voltage was set to 5 kV. An ultraviolet–visible–near-infrared spectrometer (7200, V-JASCO) was used to measure the absorption and transmittance spectra. Moreover, a spectrofluorometer (FP8500, JASCO) was used to obtain the photoluminescence (PL) spectra. The time-resolved photoluminescence (TRPL) was measured with a fluorescence lifetime spectrometer (Quantaurus-τ from Hamamatsu-Photonics K.K.) equipped with ∼405 nm laser diode (typical peak power of 400 mW) at 200 kHz repetition rate. Another characterization step entailed carrying out X-ray photoelectron spectroscopy using a Versa Probe II (ULVAC-PHI, Japan). The current density–voltage (J–V) curves were measured under 1 sun using an AM 1.5G spectral filter with an active area of 1 cm2, which was calibrated by a silicon reference cell. A spectrometer (SM-250IQE, Bunkokeiki, Japan) was used to characterize the incident monochromatic photon-to-electron conversion efficiency (IPCE) spectra, calibrated by a silicon reference cell. Photothermal deflection spectroscopy (PDS) was used for different treatments of the NiO surface to characterize the defect levels present in the perovskite/NiO and NiO/ITO bandgaps. PDS is a suitable method because it enables the detection of the defect levels for all charge states within the measured sample depth; it detects the charge states in accordance with the absorption coefficient.[56−58] In this case, monochromatic light was used to irradiate the sample surface. The light source was a halogen lamp modulated at a chopping frequency ranging between 7 and 11 Hz with incident light at a normal angle to the sample surface. A filter was fixed and used to facilitate the measurement over a wavelength range from 700 to 1200 nm. The previous pumping light was concentrated using a cylindrical lens. The dimensions of the lens were 1 × 10 mm2. A semiconductor laser (660 nm) was then used to probe the sample surface in parallel with the light. During probing, the laser was deflected based on the thermal energy produced by the combination of the electrons activated by the pump light. During the process, there was a need to enhance the deflection of the laser probe. This was achieved by conventionally dipping the PDS sample into a fluorinert (FCI) solution characterized by a high coefficient-of-temperature dependence of the refractive index, δn = δT, where T and n represent the temperature and refractive index, respectively.

Finally, photoelectron spectroscopy was applied using an ultrahigh vacuum (UHV). X-ray photoelectron spectroscopy (XPS) with a nonmonochromatic source was measured (Al Kα; 1486.6 eV, spot size; 10—300 mm) at a pass energy of 10 eV. XPS was used to determine the composition and chemical state of the components for depths of up to 10 nm.

Results and Discussion

Understanding the Effect of MeO-2PACz (SAM) on the NiOx Layer

A correlative investigation was conducted into the impact of the MeO-2PACz (SAM) interface on NiO and its influence on the performance of a solar cell device. Hence, the focus is to understand the properties and crystallization of the NiO film, as well as the interaction between NiO and MeO-2PACz.

Structural and Optical Characterization of NiO

Treatment of NiO with MeO-2PACz was investigated usingscanning electron microscopy (SEM), as shown in Figure a,b. The surface morphology of the pristine NiO film and treated NiO films were investigated, whereby minimal changes were observed. In this case, the results depict the same domain size and good coverage. Notably, the SEM images do not show a significant difference, implying that with or without the MeO-2PACz layer the morphological properties are retained.

Figure 2

(a, b) SEM images of NiO before and after treatment with MeO-2PACz SAM, (c) XRD patterns for NiO and (d) XRD patterns for NiO/MeO-2PACz SAM.

The transmittance of NiO and NiO/MeO-2PACz was investigated, as shown in Figure S1, using UV–vis transmittance spectra. The same NiO substrate was measured before and after MeO-2PACz treatment to show the minimal change by the introduction of the SAM.

The crystallinity of the film was studied by XRD, and the results are presented in Figure c,d in which the evolution of the XRD patterns obtained for the NiO film and the field after being modified with MeO-2PACz are shown. The dominant (111), (200), and (220) peaks fit well to crystalline NiO.[59] Notably, in the two films, there are no significant changes, which means that the SAM of MeO-2PACz should not change the crystal structure of NiOx.

Surface Analysis for NiO

XPS was conducted to investigate the properties of NiO. In this case, the chemical components of the pristine NiO film and NiO/MeO-2PACz were analyzed. The characteristic peaks for Ni 2p3/2 and O 1s are presented in Figure . Furthermore, decomposition of the XPS spectrum shows that the Ni 2p spectrum can be fitted by two oxidation states, namely, Ni2+ and Ni3+. The other two peaks are related to satellite peaks. Ni2+ is found at 853.7 and 854.1 eV for the pristine NiO and modified NiO surfaces, respectively, as shown in Figure a,c. This peak corresponds to NiO6 octahedral bonding in the cubic rock salt NiO structure.[60] Another peak was observed at 855.3 eV for pristine NiO and at 855.7 eV for modified NiO, which can be attributed to Ni3+ comprising NiOOH[61] and Ni2O3.[59,62] In previous studies, this peak has been assigned to the O vacancy in NiO.[15] The broad peaks observed at 860.6 and 864.4 eV are assigned to the shakeup processes (satellites) for NiO.[61,63] After treatment of the NiO film, a significant decrease occurs in the integrated area of overall Ni peaks, as summarized in Table . Notably, a shift of ∼0.5 eV was observed for the dominant peaks in Ni 2p3/2 for pristine and treated NiO, implying that electron transfer occurs. This is similar to a previous result,[64] which shows a shift in the core component of NiO after treatment. In conclusion, the surface passivation cannot be supported by the XPS results. In addition, a high-resolution XP spectrum for the C 1s and N 1s is shown in the Figure S2. The position for C 1s peaks for pristine NiO is the first peak was obtained at 284.9 eV and can be assigned to the C–C bond, whereas the second peak was at 286.4 and can be identified as representing the C–O–C bond. A third peak is found at the position of 288.3 eV, which is related to C=O. Moreover, the NiO with MeO-2PACz layer has three C 1s peaks which are 284.9, 286.4, and 288.6 eV. Those peaks are related to C–C, C–O–C, and C=O, respectively. The N 1s spectrum was found in the NiOx with MeO-2PACz layer at 400.1 eV as shown in Figure S2.

Figure 3

XPS surface spectra: (a) Ni 2p3/2 for pristine NiO, (b) O 1s for pristine NiOx, (c) Ni 2p3/2 for NiO treated with MeO-2PACz, and (d) O 1s for NiO treated with MeO-2PACz.

Table 1

Summary of the Relative Intensity of the Ni 2p, O 1s, and P 2p Peaks Showing the Area under the Curve for the Pristine NiOx and NiO/MeO-2PACz Samples Obtained from XPS

		control (NiOx)		NiOx/MeO-2PACz
elements	peaks	position (eV)	integrated area (%)	position (eV)	integrated area (%)
Ni 2p	Ni2+ (NiO)	853.7	2.6	854.1	1.8
	Ni3+ (NiOOH), Ni2O3	855.3	10.8	855.7	7.8
O 1s	NiO (Ni2+)	529.2	21.8	529.6	18.2
	Ni2O3 (Ni3+)	530.9	13.1	531.3	11.7
	NiOOH	531.7	6.5	532.1	5.9
	OH	533.0	1.4	533.2	4.3
P 2p	2P3/2			133.1	1.2
	2P1/2			133.9	0.6

The binding energy in the O 1s spectra was resolved into four main oxygen states, as shown in Figure b,d. The peaks at 529.2 and 529.6 eV are attributed to NiO or Ni2+62 for pristine NiO and NiO/MeO-2PACz. Moreover, the peaks at 530.9 eV for pristine NiO and 531.3 eV for NiO/MeO-2PACz are assigned to O-bonded Ni2O3 or Ni3+, as reported previously.[62] In addition, an increased amount of Ni2O3 has been reported to contribute to an increased work function (WF).[65] The peaks at 531.7 and 532.1 eV are related to NiOOH.[66−68] The higher binding energy at 533 eV is attributed to hydroxyl (OH) groups,[69] and the intensity of this peak increases after the treatment, which is probably because of the hydroxyl group present in the MeO-2PACz structure.

Evidence that the NiO surface was covered with MiO-2PACz after the treatment is indicated by the P 2p peak observed for the two phosphates and metaphosphate, as shown by the XPS spectra in Figure S2. Because the main P 2p binding energy is 133.1 eV, this peak is attributed to phosphate groups, which indicates a phosphorus binding state.[62]

Understanding the Role of MeO-2PACz (SAM) on a Perovskite Film

Understanding the impact of MeO-2PACz (SAM) on the performance of perovskite devices requires investigating the morphology and optical properties of perovskite films. Additionally, the defect level at the NiO/MeO-2PACz interface and the effect of the treatment on the energy level are investigated.

Effect of Surface Modification on the Morphology of Perovskite Film and the Optical Properties of Perovskite Film after Treatment of NiO

SEM analysis was carried out to gain insights into the perovskite morphology on NiO and NiO/MeO-2PACz, as shown in Figure a–d. A significant difference in perovskite domain size when deposited onto the MeO-2PACz layer is obtained from the quantified domain size. Based on the surface SEM images, an increase in domain size is observed because of the MeO-2PACz interface layer. Hence, treating NiO with MeO-2PACz influences the perovskite’s bulk properties with uniform perovskite crystallinity because of the passivation of the surface defect in NiO, leading to a slight enhancement in domain size; here, suppression of recombination is expected. SEM images of the cross-section of samples with a layered structure ITO/NiO/MeO-2PACz/perovskite indicates no significant morphological difference for the perovskite.

Figure 4

SEM images of perovskite (PVK) films: (a) top surface SEM without treatment, (b) top surface SEM with MeO-2PACz interface, (c) cross-sectional scanning image without treatment, (d) cross-sectional scanning image with a MeO-2PACz interface.

Steady-state PL and TRPL were performed to investigate the photophysical properties of perovskite films with and without the MeO-2PACz interface layer. Figure a shows the PL results obtained for the perovskite films. The PL intensities are increased after MeO-2PACz treatment, which indicates a significant suppression of recombination in the perovskite layer.[70] Further measurements included TRPL analysis, which aimed to understand the recombination lifetime in perovskites. The results are illustrated in Figure b, which shows the control perovskite device and modified film with a MeO-2PACz layer. A single wavelength of 402 nm was applied as an excitation source. Figure b shows the faster decay reflecting the Förster charge injection into the MeO-2PACz-treated interface. Effective transfer of the charge carrier was attained for the modified sample according to the TRPL results. A higher steady-state PL indicates slightly better surface passivation; thus, the faster TRPL decay likely stems from charge transfer and not higher nonradiative recombination.

Figure 5

Steady-state PL for perovskite films with and without treatment (a), TRPL measured for perovskite films with and without treatment (b), and XRD patterns for perovskite films on pristine NiO (c) and XRD for modified NiO(d).

The PL and TRPL results indicate the role of the MeO-2PACz layer in surface passivation. Hence, these results explain the enhancement of the Jsc for MeO-2PACz-modified PSCs, which will be discussed later.

The XRD results for perovskite films with and without the MeO-2PACz layer are presented in Figure c,d, which indicates the effect of MeO-2PACz on the surface. In this case, the diffraction peaks for the perovskite layer with the MeO-2PACz treatment are similar to the control film without a MeO-2PACz layer. However, a new diffraction peak appears at 12°, which is attributed to PbI2, which indicates the presence of unreacted PbI2 content in the perovskite film. Introducing the MeO-2PACz interface layer between perovskite/NiO results in the total disappearance of PbI2 crystals, which can be explained by their conversion in the perovskite phase.

The XRD results obtained lead to the conclusion that NiO without treatment affects the perovskite film because of the improper perovskite crystallization on NiO without a MeO-2PACz underlayer, hence resulting in the appearance of residual PbI2 that appears unclear. However, the presence of H2O and OH groups and interstitial oxygen on NiO or the presence of Ni3+ in NiO can cause improper perovskite crystallization.[15,18,71,72] On the other hand, the device containing the MeO-2PACz interface layer shows a better passivation effect than perovskite/NiO. Thus, the MeO-2PACz layer was utilized to passivate the defects in the perovskite layer.

Studying the Defect Level at the Interface

To further support our claim that the MeO-2PACz layer can passivate the NiO surface, PDS was applied to further investigate the pristine NiO and NiO/perovskites; the results obtained are compared with MeO-2PACz-treated NiO, as shown in Figure a,b. The intensity of the PDS signal measured for NiO without treatment is higher than that of the treated samples. Moreover, the slope of the spectrum depicts the structural order of the surface. A quantitative description of disorder is indicated by the Urbach energy, which is described as the inverse slope of the PDS signal. PDS signals are observed to range from 4 to 3.2 eV, as shown in Figure a, and from 1.65 eV to ∼1.7 eV, as shown in Figure b, which indicate a reduction of the defect level at the NiOx surfaces and NiOx/perovskite interface, respectively. Conversely, the Urbach energy shows a decline in value with treated NiO, which indicates a reduction in the defect level.[73,74] We can conclude that the MeO-2PACz layer can passivate the surface of NiO, improving the crystalline quality and decreasing the number of defects.

Figure 6

PDS spectra and the Urbach energy as determined from the inverse slope of the PDS signals for (a) NiOx and (b) perovskite.

Effect of Surface Modification on Photovoltaic Properties

The effects of surface modification were investigated, as shown in Figure a,b. Figure a represents the current J–V, a characteristic control for a perovskite device and a modified device with a MeO-2PACz SAM layer. Figure b shows the external quantum efficiency (EQE) spectra for the control perovskite device and modified device with the SAM layer.

Figure 7

(a) Representative current density–voltage (J–V) characteristics for the control perovskite device and the device modified using the MeO-2PACz interface layer. (b) EQE spectra for the control perovskite device and the device modified with a MeO-2PACz interface layer. Integrated Jsc for the MeO-2PACz treated and untreated devices are 17.7 and 16.3 mA/cm2, respectively.

The results are summarized in Table , indicating the average device parameters for PSCs, whereby the data for the control device and device modified with the MeO-2PACz interface layer are presented. The average results are derived from measurements for eight perovskite solar cell devices for each condition (Figure S6).

Table 2

Device Parameters for Perovskite Solar Cells Including the Control Devices and Devices Modified Using a SAM Layer Interfacea

device	Jsc (mA/cm2)	Voc (V)	FF (%)	Rs (Ω·cm2)	Rsh (Ω·cm2)	η (%)
perovskite/NiOx	17.3 ± 0.68	1.0 ± 0.06	0.7 ± 0.03	8.7 ± 2.08	1027.5 ± 334.78	11.9 ± 0.74
perovskite/MeO-2PACz/NiOx	20.1 ± 0.18	1.11 ± 0.01	0.8 ± 0.01	4.8 ± 0.45	4810.7 ± 437.98	17.2 ± 0.03

The results are derived from the eight perovskite solar cells for each condition.

The effect of film quality on carrier recombination was studied by comparing the photovoltaic performance of the fabricated devices with and without the MeO-2PACz treatment. The current J–V characteristics with and without MeO-2PACz treatment were analyzed, as shown in Figure a. Notably, the PCEs for the devices with and without MeO-2PACz treatment were determined to be 17.2% and 11.9%, respectively. The devices were fabricated under the same conditions, and the results are summarized in Table . The results indicate that there is a significant enhancement of Voc, Jsc, FF, and Rsh obtained for the MeO-2PACz-treated device when compared with the untreated device. The certified data for the MeO-2PACz-treated device are shown in Figure S3 with efficiency of 16.25%. The preliminary stability testing over 100 h was also conducted under maximum power point tracking (MPPT) conditions. The MeO-2PACz-treated and untreated devices revealed similar performance as shown in Figure S4, showing almost no reduction in the power conversion efficiencies.

The EQE was measured to study the efficiency of photocurrent conversion. The results show that the EQE covers the entire visible range from 300 to 800 nm for the treated devices and control device, as presented in Figure b. The results confirm that the MeO-2PACz-treated device has a higher EQE than the control device, and bandgap energies for both devices are about 1.55 eV that which is typical value for MAPI perovskites.

An analysis of the internal quantum energy (IQE) measurements for the devices was conducted, and the results are presented in Figure S5. The values obtained correspond to the ratio of the carrier charge, which is collected by the solar cell, to the number of photons absorbed under illumination. The results show that the charge carriers generated and collected for solar cell operation are significantly higher when the perovskite device is treated with MeO-2PACz.

Electronic Properties and Energy Level for Perovskite Films

To determine the mechanism behind this improvement, the electronic structures of NiO with and without a MeO-2PACz interlayer were investigated using ultraviolet photoelectron spectroscopy (UPS). The UPS results for NiO and NiO/MeO-2PACz, as measured under a −10 V bias for the valence band (VB), are shown in Figure a, and the secondary electrons are shown in Figure b. Additionally, the WF for pristine NiO and NiO covered with MeO-2PACz was determined to be 4.9 and 5.5 eV, respectively.

Figure 8

UPS spectra for NiO and NiO/MeO-2PACz, as measured under a bias of −10 V. (a) Valence band spectra for NiO and NiO/MeO-2PACz. (b) Work function (φ) is calculated using the equation φ = hν; the secondary electron (SE) is used as the cut off. (c) Energy level diagram for pristine NiOx (left) and NiO/MeO-2PACz (right).

The obtained VB values for perovskite on NiO and NiO/MeO-2PACz are close to the Femi level, which improves the device performance by facilitating charge transfer between perovskites and NiO, as shown in Figure c. The values of the WF and IE for the perovskite layer were taken from literature.[25] Before MeO-2PACz treatment, an energy gap of 0.1 eV between the VBM of perovskite and the VBM of NiO can be observed, which can lead to the formation of a hole trap at the interface for charge carriers with insufficient kinetic energy to overcome the gap. After NiO was treated with MeO-2PACz, the VBM of perovskite exceeded the VBM of NiO. Thus, the energy gap is compensated due to an energy shift, resulting in a proper energy level alignment for hole transport over the interface. The energetic gap between the VBM of perovskite and NiO changes from +0.1 eV to −0.3 eV. More details for the extraction of these values can be found in Table S3.

Conclusions

In the current study, we modified a hole transport layer (NiO) with a MeO-2PACz (SAM) layer. Inserting the MeO-2PACz interface between a NiO hole transport layer and perovskite layer helped decrease defects by passivation of NiO. On the basis of the results obtained, the MeO-2PACz on NiO interface significantly improves the defect level and device performance. This improvement in performance can be attributed to a reduction in charge recombination, increase in extraction efficiency, and enhancement of the perovskite film quality with its large domain size. The analysis further indicates that, generally, the introduction of MeO-2PACz results in the passivation of NiO surface defects, resulting in an enhancement of the crystallization. In summary, interface modification leads to various positive effects, including better interfacial contact, better energy level alignment, enhanced crystallization, and an increase in the power conversion efficiency. Thus, our results offer a promising mechanism for improving the performance of inorganic carrier transport layers in PSCs, which can potentially be extended to other combinations of inorganic semiconductors and functional organic molecule dopants in the future.