Performance Comparison between the Nanoporous NiO x Layer and NiO x Thin Film for Inverted Perovskite Solar Cells with Long-Term Stability.

The development of hole-transport layers (HTLs) that elevate charge extraction, improve perovskite crystallinity, and decrease interfacial recombination is extremely important for enhancing the performance of inverted perovskite solar cells (PSCs). In this work, the nanoporous nickel oxide (NiO x ) layer as well as NiO x thin film was prepared via chemical bath deposition as the HTL. The sponge-like structure of the nanoporous NiO x helps to grow a pinhole-free perovskite film with a larger grain size compared to the NiO x thin film. The downshifted valence band of the nanoporous NiO x HTL can improve hole extraction from the perovskite absorbing layer. The device based on the nanoporous NiO x layer showed the highest efficiency of 13.43% and negligible hysteresis that was better than the one using the NiO x thin film as the HTL. Moreover, the PSCs sustained 80% of their initial efficiency after 50 days of storage. This study provides a powerful strategy to design PSCs with high efficiency and long-term stability for future production.

Introduction

Perovskite solar cells (n>an class="Chemical">PSCs) have made an impressive progress with maximum power conversion efficiency (PCE) from 3.8 to 25.5% within a decade[1,2] due to high absorption in the visible region,[3] long carrier diffusion length,[4] high carrier mobility,[5] low exciton binding energy,[6] and tunable band gaps by exchanging composition.[7,8] Single cationic perovskite materials such as methylammonium lead iodide (MAPbI3) and formamidinium lead iodide (FAPbI3) have been proposed as the absorbing layer in PSCs with PCE values up to 20.8 and 18.94%, respectively.[9,10] Compared with single cationic perovskites, multiple-cation perovskite materials solve the disadvantage of instability and achieve higher device performance. Saliba et al. incorporated rubidium cations into PSCs to reveal an optimized open-circuit voltage (VOC) of 1180 mV, a short-circuit current density (JSC) of 22.8 mA/cm2, a fill factor (FF) of 81%, and a certified PCE of 21.8%.[11] Furthermore, the device retained 95% of its initial performance after 500 h of aging at 85 °C in a nitrogen-filled glovebox. Jeon et al. demonstrated PSCs using (FAPbI3)0.85(MAPbBr3)0.15 as the active layer, confirming a certified PCE of 23.2% with a slight hysteresis behavior.[49] The device maintained 92.6% of its initial PCE value after 310 h of continuous illumination and almost 95% for more than 500 h of thermal annealing at 60 °C. Saliba et al. reported cesium-containing triple cation perovskite Cs(FA0.17MA0.83)1–Pb(I0.83Br0.17)3 as the light absorber.[12] The PCE value of the optimized PSC dropped from 21.17 to ∼18% after aging for 250 h under constant illumination in a nitrogen atmosphere. Apparently, the adoption of multiple-cation perovskite materials is a good choice for commercialized production in the near future.

PSCs have been extensively developed in two difn>an class="Chemical">ferent device configurations, that is, regular and inverted types. The regular PSC has a device structure of anode/electron-transport layer (ETL)/perovskite/hole-transport layer (HTL)/cathode since it is derived from the first perovskite-related literature.[1] However, the n–i–p configuration usually encounters a serious drawback of hysteresis, which causes efficiency drop and instability of devices.[13−15] Another problem is device instability which arises from the morphological deformation of (2,2′,7,7′-tetrakis(N,N-di-p-methoxyphenylamine)-9,9′-spirobifluorene) (spiro-MeOTAD) as the HTL.[16,17] To avoid the above problems, inverted PSCs with the p–i–n configuration have been developed because of their simple device structure, free of high-temperature processing, and little hysteresis effect.[18−20] Besides, inverted PSCs can be combined with traditional solar cells such as silicon or copper indium gallium selenide solar cells to construct tandem devices with high efficiency.[21,22] In the inverted PSC structure, poly(3,4-ethylenedioxythiophene):poly(styrene sulfonate) (PEDOT:PSS) has been widely used as the HTL owing to the simple fabrication process and hole extraction ability from the perovskite layer.[23,24] However, organic PEDOT:PSS may corrode and deteriorate transparent conductive oxide electrodes, such as fluorine-doped tin oxide (FTO) and indium tin oxide, due to its acidic nature.[25]

Apart from organic materials such as PEDOT:n>an class="Chemical">PSS, inorganic materials can also serve as the HTL in inverted p–i–n devices, which possess better thermal stability, higher hole mobility, and lower expense compared with traditional organic hole-transport materials.[26] It is believed that the development of inorganic hole-transport materials plays an important role in the commercialization of PSCs with low cost, high efficiency, and long-term stability. Among miscellaneous metal oxide materials, nickel oxide (NiO) is particularly attractive as it belongs to p-type semiconductors with high hole mobility, excellent chemical stability, and good transmittance in the visible range.[26,27] The matched energy level alignment between NiO and the perovskite facilitates hole extraction and electron blocking.[28,29] Many approaches have been proposed to prepare NiO HTLs for inverted PSCs, including sol–gel deposition,[30] nanoparticle dispersion,[31] spray pyrolysis,[32] and atomic layer deposition.[33] Moreover, an additional passivation layer can be introduced between NiO and the perovskite layer to reduce pinhole defects on the NiO surface and increase the crystallinity of the perovskite. Wang et al. utilized polystyrene as a passivation medium to modify the surface of the NiO film and increase the perovskite quality.[34] Highly efficient devices with a larger perovskite grain size, fewer interfacial defects, and suppressed charge recombination were achieved. A very high PCE of 19.99% and a VOC of 1.149 V were obtained, while no hysteresis was observed. Chen et al. adopted a conjugated poly(bithiophene imide) (PBTI) for the passivation of grain boundaries in inverted planar PSCs.[35] The incorporation of PBTI between NiO and the perovskite resulted in lower defect density, reduced charge recombination, and high efficiency of devices. An optimized PCE of 20.67% was obtained with the PBTI treatment.

In this research, we investigated the formation mechanism and characterization of nanoporous NiO for the fabrication of inverted n>an class="Chemical">PSCs. Nanoporous NiO has been reported to assist the deposition of a pinhole-free perovskite film with larger grain size.[36] Different precursors including nickel sulfate heptahydrate (NiSO4·7H2O), nickel acetate tetrahydrate (Ni(CH3COO)2·4H2O), and nickel chloride (NiCl2) have been adopted to examine the formation condition of nanoporous NiO layers. Transmission spectroscopy and ultraviolet photoelectron spectroscopy (UPS) were carried out to explore the transmittance in the visible range and energy levels of nanoporous NiO, respectively. From the literature survey, we notice that UPS has not been applied to investigate the band structure of nanoporous NiO so far. As for the fabrication of inverted PSCs, [6,6]-phenyl-C61-butyric acid methyl ester (PC61BM) doped with tetrabutylammonium tetrafluoroborate (TBABF4) and titanium (diisopropoxide) bis(2,4-pentanedionate) (TIPD) were chosen as ETLs. The device with the configuration of FTO/NiO/perovskite/PC61BM/TIPD/Ag was fabricated and evaluated. Both NiO thin film and nanoporous NiO layer were used as the HTL for comparison.

Experimental Section

Materials

FTO-coated glass substrates were purchased from Ruilong Opn>toelectronics Technology Co., Ltd. n>an class="Chemical">Potassium persulfate (K2S2O8, purity 98%) was purchased from Showa. NiSO4·7H2O (purity 98%) and aqueous ammonia (NH3(aq), 25–28 wt %) were bought from Acros and Sigma-Aldrich, respectively. High-purity perovskite precursors including lead iodide (PbI2, purity 99.9985%), lead bromide (PbBr2, purity 99.99%), and methylammonium bromide (MABr, purity 99.5%) were purchased from Luminescence Technology Corp., Taiwan. Formamidinium iodide (FAI, purity 98%) was bought from STAREK Scientific Co., Ltd. Cesium iodide (CsI, purity 99.9%) was bought from Alfa Aesar. PC61BM (purity 99%) was purchased from Solenne B.V., the Netherlands. Other chemicals and solvents were purchased from Alfa Aesar or Acros and used without further purification.

Preparation of the NiO Thin Film

The NiO thin film was spn>in-cast from its precursor solution on the n>an class="Chemical">FTO-coated glass substrate, which was cleaned stepwise in detergent, deionized water, acetone, and IPA under ultrasonication for 30 min each, followed by UV–ozone exposure for 25 min. To prepare the 0.1 M NiO precursor solution, 2.62 mg of NiSO4·7H2O was dissolved in 10 mL of methanol at 80 °C with stirring in a sealed glass vial overnight. The NiO precursor solution was then spin-coated on the FTO substrate at 1500 rpm for 30 s, followed by drying at 80 °C for 10 min. After drying, the substrate was moved into a high-temperature oven for calcination at 450 °C for 1 h to obtain the NiO thin film.

Preparation of the Nanoporous NiO Layer

The nanoporous NiO layer was prepared according to the previous literature with some modified parameters.[36] The pre-cleaned n>an class="Chemical">FTO substrate was taped in a Petri dish with FTO face upward. To prepare the NiO growth solution, 8 mL of 0.1 M NiSO4·7H2O was added to a solution of 6 mL of 0.025 M K2S2O8 in DI water and 2 mL of aqueous ammonia (25–28 wt %). The mixed growth solution was shaken and immediately poured into the Petri dish. The reaction was carried out for 5 min and the substrate was taken out, cleaned with deionized water to remove loose black particles, and further dried at 100 °C for 1 h. After drying, the substrate was moved into a high-temperature oven for calcination at 450 °C for 1 h to obtain the nanoporous NiO layer.

Device Fabrication

After the preparation of the NiO thin film or nanoporous n>an class="Chemical">NiO layer, the absorbing perovskite Cs0.05FA0.81MA0.14Pb(Br0.15I0.85)3 layer was deposited by spin coating. For the perovskite solution used in this research, a mixture of CsI (17.5 mg), FAI (190.2 mg), MABr (21.8 mg), PbI2 (548.6 mg), and PbBr2 (77.1 mg) was dissolved in a mixed solvent (1 mL) consisting of N,N-dimethylformamide and dimethyl sulfoxide with a 4:1 volume ratio at 70 °C for 1 h with stirring. The perovskite film was deposited on top of the NiO thin film or nanoporous NiO layer with a two-step spin-coating process in a nitrogen-filled glovebox. The first step was 1000 rpm for 10 s with a ramp-up rate of 200 rpm/s, and the second step was 5000 rpm for 20 s with a ramp-up rate of 1000 rpm/s. Chlorobenzene (300 μL) was dropped onto the substrate 10 s before the end of the second spin-coating step. The substrate was then annealed at 100 °C for 1 h. The PC61BM solution (20 mg/mL) containing 2 wt % of TBABF4 in chlorobenzene was spin-coated on the formed perovskite layer at 3000 rpm for 30 s, followed by heating at 100 °C for 10 min. Afterward, 0.1 wt % of TIPD solution in isopropanol was spin-coated on the PC61BM layer at 5000 rpm for 30 s. Finally, 100 nm of the Ag electrode was deposited by thermal evaporation at a base pressure of 7 × 10–6 Torr. The active area of each device is 4 mm2.

Characterization Methods

The cross-sectional and top-view scanning electron microscopy (SEM) micrographs of samples were investigated with an ultrahigh-resolution ZEISS AURIGA Crossbeam scanning electron microscope. The morpn>hology and roughness of samn>an class="Chemical">ples were examined with a Bruker Innova atomic force microscopy (AFM) with the tapping mode. The UPS measurements for the NiO thin film and nanoporous NiO layers were performed on a Thermo VG-Scientific/Sigma Probe spectrometer. A He I (hν = 21.22 eV) discharge lamp was used as the excitation source. X-ray photoelectron spectroscopy (XPS) measurements were conducted by a Thermo K-Alpha X-ray photoelectron spectrometer for elemental composition analysis of samples. The steady-state photoluminescence (PL) spectra of the perovskites on different substrates were measured using a Princeton Instruments Acton 2150 spectrophotometer. A KIMMON KOHA He–Cd laser with double excitation wavelengths at 325/442 nm was utilized as the light source. The transmission spectra were recorded with the same spectrophotometer using a xenon lamp (ABET Technologies LS 150) as the light source. To perform time-resolved PL (TR-PL) measurements, a 473 nm pulsed laser (Omicron) was utilized as an excitation light source. The TR-PL signals were recorded by a time-correlated single-photon counting module (PicoQuant MultiHarp 150 4N) combined with a photomultiplier tube through an Andor Kymera 328i spectrometer. The apparatus was assembled by LiveStrong Optoelectronics Co., Ltd. from Taiwan. X-ray diffraction (XRD) patterns and crystallinity of samples were measured by a Rigaku D/MAX2500 X-ray diffractometer. The current density–voltage (J–V) characteristics of the PSCs were measured using a Keithley 2400 SourceMeter under AM 1.5G simulated sunlight exposure (Yamashita Denso YSS-150A equipped with a xenon short arc lamp, 1000 W) at 100 mW/cm2 under an ambient environment. The scan rate for J–V measurements was 25 mV/s. The external quantum efficiency (EQE) measurements were performed on an assembled apparatus in the laboratory, comprising a solar simulator (ABET Technologies LS 150, USHID UXL-150MO), a monochromator (Prince Instruments Acton 2150), and a Keithley 2400 SourceMeter.

Results and Discussion

Characterization of the NiO Thin Film and Nanoporous Layers

The NiO thin film was prepn>ared via the sol–gel process and its formation has been discussed previously.[37,38] Meanwhile, the nanopn>orous n>an class="Chemical">NiO layer was prepared by chemical bath deposition.[39−41] The involved chemical reactions are listed as follows.

The first step (1) describes the reaction between the starting material NiSO4 and n>an class="Chemical">ammonia in water to form nickel hydroxide Ni(OH)2. Then, the second step (2) reveals the formation of the species NiO(OH) by reacting Ni(OH)2 with S2O82– from K2S2O8, as indicated in the part of Section . Obviously, in this step, the persulfate salt induces Ni2+ oxidation to generate Ni3+ species NiO(OH). Afterward, the as-deposited precursor film containing Ni(OH)2 and NiOOH was thermally converted to NiO by calcination. The advantages of this approach include simple processing, usage of DI water as solvent, and short reaction time within 10 min. Figure a, c shows the top-view and cross-sectional SEM images of the NiO thin film, respectively. It can be seen that a thin and dense NiO layer is deposited on the FTO surface that possesses many grain boundaries. The thickness of the NiO thin film was estimated to be ca. 30 nm. Figure b,d reveals the top-view and cross-sectional SEM images of the nanoporous NiO layer, respectively. Sponge-like nanostructures and interconnecting networks were clearly observed, exhibiting higher surface area for perovskite filling and crystallization. The formed nanoporous NiO layer has a high degree of porosity with a pore size of 100–300 nm and a wall thickness of 20–30 nm. Furthermore, nanoporous NiO flakes were vertically aligned on the FTO substrate with a thickness ranging from 100 to 120 nm. The cross-sectional SEM image of the nanoporous NiO layer can be seen in Figure S1 in the Supporting Information, revealing a flake-like morphology. The surface morphology of our nanoporous NiO layer is similar to those in the previous literature.[36,39] Apart from the SEM observation, AFM experiments were also carried out to investigate the morphology and average roughness (Ra) of the prepared samples. Figure e,f shows the topographic AFM images of the NiO thin film and nanoporous layer, respectively. Many NiO nanospheres were formed on the FTO substrate, with a low Ra value of 11.1 nm. In Figure f, highly porous flakes were aligned on the FTO surface, which are consistent with the SEM observation. The Ra value of the nanoporous NiO layer is estimated to be 13.3 nm, which is slightly larger than that of the NiO thin film due to its sponge-like interconnecting networks.

Figure 1

Top-view and cross-sectional SEM images of the (a,c) NiO thin films and (b,d) nanoporous NiO layer deposited on FTO substrates; AFM topographic images of the (e) NiO thin film and (f) nanoporous NiO layer.

Top-view and cross-sectional SEM images of the (a,c) NiO thin films and (b,d) nanopn>orous n>an class="Chemical">NiO layer deposited on FTO substrates; AFM topographic images of the (e) NiO thin film and (f) nanoporous NiO layer.

Figure a shows the transmission spectra of the NiO thin film and nanoporous layer from 300 to 750 nm. It is seen that the nanoporous n>an class="Chemical">NiO layer has a higher transmittance of 60–90% in the range of 300–450 nm and an even higher transmittance of 90–95% in the range of 450–750 nm, as compared with the NiO thin film. This is beneficial for incident photons to enter devices and to be absorbed by the absorbing layer. Figure b reveals the absorption spectra of the NiO thin film and nanoporous layer in the visible range. The optical band gaps (Eg) of the NiO thin film and nanoporous layer were estimated from their absorption edges around 350 nm to give 3.48 and 3.5 eV, respectively, which is similar to the previous literature.[42] From transmission and absorption measurements, we conclude that light is easier to pass through the nanoporous NiO layer owing to its porous structure, as compared with the NiO thin film.

Figure 2

(a) Transmission and (b) absorption spectra of the NiO thin film and nanoporous NiO layer.

(a) Transmission and (b) absorption spectra of the pan class="Chemical">NiO thin film and nanopn>orous n>an class="Chemical">NiO layer.

The UPS spn>ectra of the n>an class="Chemical">NiO thin film and nanoporous layer were measured to examine the change of the energy levels, as shown in Figure . The work function (φw) is derived by subtracting the binding energy cutoff in the high binding energy region (around 14.1 eV) from the He I photon energy (21.22 eV).[23] Since φw is defined as the energy difference between the EF and the vacuum level, the EF values of the NiO thin film and nanoporous layer were determined to be −7.05 and −7.1 eV, respectively, from Figure a. Furthermore, the binding energy cutoffs in the low binding energy region around −1.9 eV indicate the energy difference between the EF and the valence band (VB) level.[43] Therefore, the VB levels of the NiO thin film and nanoporous layer were calculated to be −5.16 and −5.23 eV, respectively, from Figure b. By combining VB levels from UPS experiments and Eg values from optical measurements, the conduction band (CB) levels of the NiO thin film and nanoporous layer were calculated to be −1.68 and −1.73 eV, respectively. The above results demonstrate that the energy levels of NiO can be slightly altered by different nanostructures. The downshifted VB level of the nanoporous NiO layer is matched better with the perovskite absorbing layer than the NiO thin film, which can improve the hole extraction from perovskite to NiO HTL.

Figure 3

UPS spectra of the NiO film and nanoporous NiO layer at (a) high and (b) low binding energy regions.

UPS spn>ectra of the n>an class="Chemical">NiO film and nanoporous NiO layer at (a) high and (b) low binding energy regions.

The XRD patterns of the NiO thin film and nanopn>orous layer on the n>an class="Chemical">FTO substrates are shown in Figure . The diffraction signals of the NiO thin film are found at 2θ = 37.0, 43.2, and 62.8°, corresponding to the (111), (200), and (220) planes, respectively.[44] According to the XRD patterns, the prepared NiO is well consistent with the cubic phase.[45] Furthermore, the three diffraction signals of the nanoporous NiO layer are also observed at similar 2θ positions, confirming that the crystalline structure of NiO is not affected by different nanostructures.

Figure 4

XRD patterns of the NiO thin film and nanoporous NiO layer deposited on FTO substrates.

XRD patterns of the NiO thin film and nanopn>orous n>an class="Chemical">NiO layer deposited on FTO substrates.

To further probe Ni3+ and n>an class="Chemical">Ni2+ components in the NiO thin film and nanoporous layer, XPS experiments were performed and the corresponding Ni 2p3/2 and O 1s signals are displayed in Figure . The multicomponent band can be deconvoluted into four different states at 854.0 (Ni2+), 855.8 (Ni3+), 861.2 (Ni2+ satellite), and 864.1 eV (Ni3+ satellite) that have been reported in the previous literature.[29,30] The received Ni3+/Ni2+ ratios were calculated from Figure a,b to be 1.23 and 1.25 for the NiO thin film and nanoporous layer, respectively. The slightly increased Ni3+/Ni2+ ratio means that the nanoporous structure has better hole-transport ability than the thin film. Figure c,d shows the O 1s spectra of the NiO thin film and nanoporous layer, being fitted with two states at around 529.3 eV (O2– from NiO) and 531.3 eV (O2– from Ni2O3).[36] The ratios of the obtained O element from NiO and Ni2O3 were calculated to be 0.424 and 0.454 for the NiO thin film and nanoporous layer. Similarly, it shows that the NiO nanoporous layer has a higher proportion of Ni3+ to contribute to hole conductivity compared to the thin-film state.

Figure 5

XPS spectra of Ni 2p3/2 and O 1s elements in the (a,c) nanoporous NiO layer and (b,d) NiO thin film.

XPS spn>ectra of Ni 2pn>3/2 and O 1s elements in the (a,c) nanopn>orous n>an class="Chemical">NiO layer and (b,d) NiO thin film.

Figure a,b shows the cross-sectional SEM image of the perovskite deposited on the n>an class="Chemical">NiO thin film and nanoporous NiO layer, respectively. The thickness of the perovskite layer on both HTLs was estimated to be 550 nm. It is evident that the large and intact perovskite crystals with a submicron grain size up to 500 nm were formed when deposited on the nanoporous NiO layer, as revealed in Figure b. In contrast, the perovskite crystals deposited on the NiO thin film looked smaller with an average grain size of 300 nm, as shown in Figure a. The perovskite films with a large grain size are beneficial for reducing charge recombination, allowing effective carrier extraction and transport from the perovskite absorbing layer to the corresponding charge-transport layers.[34,46] The top-view SEM images of the perovskite deposited on the NiO thin film and nanoporous NiO layer are displayed in Figure c,d, respectively. No pinholes could be found for both perovskite films. The grain size of perovskite crystals on the NiO thin film is estimated to be in the range of 100–200 nm in Figure c, while larger perovskite crystals with a grain size of 100–350 nm were observed on the nanoporous NiO layer in Figure d. The highly porous surface of the nanoporous NiO layer provides nucleation sites for the perovskite growth to achieve a larger grain size compared with the NiO thin film. Our results prove that the nanoporous NiO layer can perform as a template to obtain a high-quality perovskite film.

Figure 6

Cross-sectional SEM images of PSCs based on the (a) NiO thin film and (b) nanoporous NiO layer as the HTL; top-view SEM images of the perovskite deposited on the (c) NiO thin film and (d) nanoporous NiO layer.

Cross-sectional SEM images of PSCs based on the (a) n>an class="Chemical">NiO thin film and (b) nanoporous NiO layer as the HTL; top-view SEM images of the perovskite deposited on the (c) NiO thin film and (d) nanoporous NiO layer.

The XRD patterns of the perovskite deposited on the n>an class="Chemical">NiO thin film and nanoporous NiO layer are shown in Figure . The composition of the deposited perovskite layer was Cs0.05FA0.81MA0.14Pb(Br0.15I0.85)3, as mentioned in the Experimental Section. Intense diffraction peaks at 2θ = 14.04, 19.92, 24.52, 28.36, 31.76, 34.96, 40.52, and 43.16° are observed, which correspond to (001), (011), (111), (002), (012), (112), (022), and (003) planes. The location of these diffraction peaks of the perovskite is consistent with the previous reports.[47] Moreover, the intensity of diffraction peaks of the perovskite on the nanoporous NiO is found to be higher than that on the NiO thin film, implying better crystallization and morphology of the perovskite film. Here again, the nanoporous NiO layer can serve as a better template for the perovskite growth to obtain higher crystallinity.

Figure 7

XRD patterns of the perovskite films on the NiO thin film and nanoporous NiO layer.

XRD patterns of the perovskite films on the n>an class="Chemical">NiO thin film and nanoporous NiO layer.

The steady-state PL spn>ectra of the n>an class="Chemical">perovskite on the glass, NiO thin film, and nanoporous NiO layer are revealed in Figure a. The PL emission of the perovskite Cs0.05FA0.81MA0.14Pb(Br0.15I0.85)3 is located at 755 nm that is similar to the previous literature.[11] It is seen that the perovskite deposited on the glass has the highest PL intensity, while the one on the nanoporous NiO layer owns the lowest PL emission. The reduced PL emission can be attributed to the nanoporous structure that helps to extract carriers more efficiently from the perovskite layer, indicative of the decreased charge recombination and improved JSC value for device fabrication.[45] The TR-PL measurements of the perovskite on the glass substrate, NiO thin film, and nanoporous NiO layer were carried out and the corresponding results are revealed in Figure b. It is clearly seen that the perovskite deposited on the nanoporous NiO layer possesses a faster PL decay curve compared with that on the NiO thin film, implying that more effective hole–electron separation can be accomplished. The PL decay curves were fitted using a biexponential model and the average lifetime (τavg) was estimated from the fitted curve data using the equation τavg = ∑Aτ2/∑Aτ,[27,31] where A is a constant and τ is the lifetime. The perovskite Cs0.05(MA0.85FA0.15)0.95Pb(Br0.15I0.85)3 on the glass substrate had a τavg of 121.47 ns, while the perovskite deposited on the NiO thin film and nanoporous NiO layer showed shorter τavg values of 34.3 and 26.52 ns, respectively. This indicates more effective charge extraction by the nanoporous NiO HTL from the perovskite active layer as compared with that by the NiO thin film.

Figure 8

(a) PL emission spectra and (b) TR-PL decay curves of the perovskite on the glass substrate, NiO thin film, and nanoporous NiO layer.

(a) PL emission spn>ectra and (b) TR-n>an class="Chemical">PL decay curves of the perovskite on the glass substrate, NiO thin film, and nanoporous NiO layer.

The p–i–n device structure of the inverted PSC based on the nanoporous n>an class="Chemical">NiO HTL is shown in Figure a, revealing a sandwiched FTO/nanoporous NiO/Cs0.05(MA0.85FA0.15)0.95Pb(Br0.15I0.85)3/PCBM/TIPD/Ag architecture. The energy level diagram of the whole device is depicted in Figure b. The VB and CB levels of NiO have been discussed in the previous part, while the energy levels of the remaining components were referred to our previous report.[43] In our device architecture, electrons can be smoothly transported from the perovskite absorbing layer to the Ag electrode through PCBM/TIPD, while holes are migrated stepwise from the perovskite layer to nanoporous NiO and collected at the FTO electrode. Moreover, the deeper VB level of the nanoporous NiO HTL would result in a higher VOC value. The J–V curves of the devices under AM 1.5 G illumination are shown in Figure c, and the measured parameters including JSC, VOC, FF, PCE, series resistance (RS), and shunt resistance (RSh) are summarized in Table . The optimized device based on the nanoporous NiO showed a VOC of 1.02 V, a JSC of 18.9 mA/cm2, a FF of 70%, and a PCE of 13.43% in the reverse scan, which is significantly higher than the one based on the NiO thin film (VOC = 1 V, JSC = 16 mA/cm2, FF = 66%, and PCE = 10.53%). The statistical distribution of 20 individual devices for all photovoltaic parameters is depicted in Figure S3 in the Supporting Information. It can be seen that our devices possessed good reproducibility and PSCs based on the nanoporous NiO layer showed relatively higher photovoltaic parameters. The improved device performance is mainly ascribed to the increased JSC value. In order to investigate the reason to the increased JSC, hole-only devices with the configuration of FTO/NiO thin film or nanoporous NiO layer/Ag were fabricated and their current–voltage characteristics were measured and are displayed in Figure S2 in the Supporting Information. It reveals that the nanoporous NiO layer has a higher conductivity than the NiO thin film, indicative of the enhanced hole-transport ability. The hole mobility (μh) was inferred from the space-charge limited current equation J = (9/8)εε0μh(V2/L3). The μh values of the nanoporous NiO layer and the NiO thin film are calculated to be 2.41 × 10–3 and 2.03 × 10–3 cm2/V s, respectively, which are close to the previous report.[43] The increased JSC value can be ascribed to the enhanced hole mobility. The reduced charge recombination, as discussed in the PL and TR-PL part, is also responsible for the increased JSC value. We also notice that VOC of the device based on the nanoporous NiO layer is higher than that on the NiO thin film, which has been predicted from the downshifted VB levels. Furthermore, the PSC using the nanoporous NiO HTL exhibited negligible hysteresis compared with that using the NiO thin film as the HTL. The eliminated hysteresis is realized due to the higher crystallinity and reduced surface defect of the perovskite, which has been reported in the previous study.[48]Figure d shows the EQE spectra and integrated current densities of devices as a function of wavelength using the nanoporous NiO layer and NiO thin film as the HTL. The results demonstrate that the device based on the nanoporous NiO has a higher photon-to-electron conversion capability from 350 to 700 nm compared to that based on the NiO thin film. The integrated J values for the devices based on the nanoporous NiO layer and NiO thin film were calculated to be 18.2 and 15.6 mA/cm2, respectively, which are similar to the JSC values from Table . As shown in Figure e, both the resulting PSCs using the nanoporous NiO layer and NiO thin film as the HTL maintained 80% of their initial efficiency over a period of 50 days stored in the nitrogen glovebox and measured in ambient air at 25 °C, revealing long-term stability for future production.

Figure 9

(a) Device structure based on the nanoporous NiO, (b) energy level diagram, (c) J–V characteristics, (d) EQE spectra and integrated current density, and (e) normalized PCE evolution of devices based on the NiO thin film and nanoporous NiO layer.

Table 1

Device Performance of Inverted PSCs Based on Different NiO HTLs

HTL	scan direction	JSC (mA/cm2)	VOC (V)	FF (%)	best PCE (%)	avg PCEa (%)	RS (Ω·cm2)	RSh (kΩ·cm2)
NiOx thin film	forward	16.9	0.98	62	9.87	9.36	278	9834
	reverse	16.0	1	66	10.53	9.97	252	10,573
nanoporous NiOx	forward	18.9	1.02	69	13.26	12.54	213	13,160
	reverse	18.9	1.02	70	13.43	12.73	194	15,657

Data were obtained from 20 devices.

(a) Device structure based on the nanoporous NiO, (b) energy level diagram, (c) J–V characteristics, (d) EQE spn>ectra and integrated current density, and (e) normalized PCE evolution of devices based on the NiO thin film and nanoporous NiO layer.

Conclusions

In this research, we compared two different nanostructures of n>an class="Chemical">NiO applying in inverted PSCs. The sponge-like nanostructures and interconnecting networks of NiO can provide nucleation sites for the perovskite growth to achieve a larger grain size compared with the NiO thin film. The higher crystallinity and pinhole-free perovskite could generate carriers more efficiently, while effective extraction of carriers by the nanoporous NiO HTL and PCBM/TIPD ETL was achieved to reduce charge recombination and improve JSC of devices. The optimized PSC based on the nanoporous NiO layer exhibited a high PCE of 13.43%, negligible hysteresis, and excellent device stability for 50 days of storage. To date, our results provide a simple and effective approach to achieve a high-quality perovskite absorbing layer on the nanoporous NiO layer for future application in photovoltaics.