Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.
Advertisement
Nature Communications volume 16, Article number: 5746 (2025)
5972
5
Metrics details
Inverted perovskite solar cell has made significant progress in recent years. Although two-step sequential deposition shows the benefits to obtain higher quality large-size perovskite crystals, the high annealing temperature, which is required to achieve phase transition, leads to the desorption of self-assembled molecules at the buried interface and induces redundant lead iodide at the top interface. Here, we propose a low temperature sequential deposition method by introduce a tailor-made 3-ethyl-1-methyl-1H-imidazol-3-ium dimethyl phosphate into lead iodide precursor solution to facilitate the sufficient reaction between lead iodide and organic salts, and lower the energy barrier from delta- to alpha-perovskite. As a result, highly crystallized and pure alpha-phase perovskite films with large grain size are fabricated, preventing the damage to buried self-assembled molecules and the formation of redundant lead iodide, which contributes to a high open circuit voltage of 1.21 V and a certified efficiency of 26.0%. The encapsulated devices show improved stability following ISOS-D-3 and ISOS-L-2 protocols.
Inverted PSCs have attracted extensive attention due to the facile preparation process, negligible hysteresis, high stability, compatibility with flexible devices and tandem devices1,2,3,4,5,6. Particularly, the efficiency of the inverted PSCs has made significant progress in recent years and surpassed the regular ones because of the development of self-assembled monolayers (SAMs), demonstrating great potential for the commercialization7,8,9,10. However, it is noteworthy that among the high-efficiency inverted PSCs (>26%) reported so far, the perovskite films are mainly processed with one-step deposition, which usually leads to small grain sizes, leaving a big room for the improvement of performance10,11,12,13.
The two-step sequential deposition method is expected to readily control the chemical reaction14,15,16, and obtain high-quality perovskite films with larger grain size compared with the one-step method (Supplementary Table 1), which had already been successfully employed in the high-efficiency regular PSCs17,18,19. Although previous works such as composition engineering, additive engineering and surface passivation have been reported to improve the sequential deposition method for inverted PSCs20,21,22,23, the certified power conversion efficiency (PCE) of the inverted PSCs based on sequential deposition (ca. 24%) still lags far behind those devices fabricated by one-step deposition, as summarized in Supplementary Table 2.
The reason behind this phenomenon is that the sequential deposition usually contains a step of phase transition from δ to α, which requires a high annealing temperature (150 °C)24,25,26. However, high-temperature annealing is not allowed for the inverted PSCs containing SAMs that are easy to desorb under high temperature and thereby increase the non-radiative recombination at the buried interface27,28,29. In addition, high-temperature annealing tends to induce the loss of the organic component, resulting in a PbI2-rich perovskite surface with high defect density, which was reported to hinder the charge transport at the top interface of inverted PSCs30,31,32.
Here, a sequential deposition method (LTSD) is proposed, which can be conducted under a low annealing temperature with complete phase transition. A tailor-made 1-ethyl-3-methylimidazolium dimethyl phosphate (EMI-DMP) is introduced in PbI2 precursor solution, which can form a porous complex with PbI2 framework to facilitate the infiltration of organic ammonium salts and promote sufficient reaction between them. Besides, EMI-DMP is found to lower the energy barrier from δ- to α-perovskite by changing the surface free energy, resulting in a complete phase transition at a low annealing temperature. As a result, highly crystallized and pure α-phase perovskite films with an average grain size of over 1.3 μm can be fabricated by LTSD while preventing the damage to buried SAMs and suppressing the formation of redundant PbI2 at the top surface. Finally, owing to the significant increase in the open-circuit voltage (VOC), the PCE of the devices significantly increases from 24.5 to 26.5% (certified 26.0%) with negligible hysteresis, delivering the current state-of-the-art performance based on sequential deposition for SAM-contained inverted PSCs (Supplementary Table 2), which can compete with the one-step deposition. The encapsulated devices show improved stability, maintaining 93.1% and 95.4% after 1000 h of ageing under 85 °C/85% RH (ISOS-D-3) and maximum power point tracking, 65 °C (ISOS-L-2), respectively.
In sequential deposition, the morphology of PbI2 has a profound impact on the subsequent reaction with the organic salts. The dense PbI2 layer will hinder the penetration of organic salts, leading to an incomplete conversion of PbI223,33. Before the investigation of the morphology, the distribution of EMI-DMP (Supplementary Fig. 1) in PbI2 film was first shown by time-of-flight secondary ion mass spectrometry (TOF-SIMS) measurement (Supplementary Fig. 2), where the featured signal of phosphorus in EMI-DMP uniformly distributes in the PbI2 film in both the vertical and horizontal directions. The top-view scanning electron microscope (SEM) images for the PbI2 films with or without EMI-DMP (shown in Fig. 1a, b) exhibit the change in PbI2 morphology from a dense layer to a porous layer after the introduction of EMI-DMP. The cross-sectional SEM measurement (Fig. 1c, d) further shows that the PbI2 film with EMI-DMP presents a fluffy shape, demonstrating a consistent morphology from top to bottom, which can facilitate the top-down diffusion of organic salts and promote the sufficient reaction between PbI2 and the organic salts. By contrast, the PbI2 without EMI-DMP shows a dense structure throughout the entire film. The height images for the two samples (Fig. 1e, f) were obtained by atomic force microscope (AFM) and the roughness increases from 29.5 nm (PbI2) to 50.9 nm (PbI2@EMI-DMP), revealing the morphological change in the PbI2 film, which is consistent with the SEM images. Figure 1g shows the X-ray diffraction (XRD) patterns of the PbI2 films with different concentrations of EMI-DMP, and Supplementary Fig. 3 shows the corresponding full width at half maxima (FWHM). As the increase in EMI-DMP, the intensity of the diffraction peak on the PbI2 crystal plane gradually decreases with an increase in FWHM from 0.35° to 0.58°, which further confirms the formation of the porous morphology for PbI2@EMI-DMP films23.
Top-view SEM images of the a PbI2 and b PbI2@EMI-DMP films. Cross-sectional SEM images of the c PbI2 and d PbI2@EMI-DMP films. AFM images of the e PbI2 and f PbI2@EMI-DMP films, the scale bar is 1 μm. g XRD pattern of the PbI2 and PbI2@EMI-DMP films. h FTIR spectra of PbI2, EMI-DMP, and PbI2@EMI-DMP samples. i XPS of Pb 4 f spectra of the PbI2 and PbI2@EMI-DMP films.
To understand the interaction between PbI2 and EMI-DMP, Fourier transform infrared (FTIR) spectroscopy was tested (Fig. 1h). The FTIR spectrum of EMI-DMP shows obvious peaks at 2925 cm−1 and 1631 cm−1, which can be attributed to the stretching vibration of C–H and C=N34,35. The peaks shift to lower wave number of 2914 cm−1 and 1625 cm−1, respectively, after the addition of EMI-DMP to PbI2. This change in the symmetric tensile vibration indicates that the density of the electron cloud of the C–H and C=N unit decreases, demonstrating the interaction between EMI-DMP and PbI2. To explore further details, X-ray photoelectron spectroscopy (XPS) was adopted. Figure 1i shows the Pb 4f XPS core spectra of the PbI2 and the PbI2@EMI-DMP films. The two peaks of the Pb 4f (143.3 and 138.4 eV) shift to lower binding energies (143.1 and 138.2 eV) after the introduction of EMI-DMP, illustrating strong interaction between EMI-DMP and Pb2+ in the PbI2 lattice, which is consistent with the FTIR results. To investigate the mechanism of the interaction, nuclear magnetic resonance analysis was performed on EMI-DMP with and without PbI2. As shown in Supplementary Fig. 4H, the imidazole rings of EMI-DMP exhibit a downfield shift of 0.05 ppm (from 7.78 and 7.70 to 7.83 and 7.75, respectively), which can be attributed to the formation of hydrogen bonds between C–H on the imidazole rings and PbI2 framework33. The formation of the porous PbI2@EMI-DMP is probably due to the hydrogen bonding between EMI-DMP and PbI2, which can disrupt the layered structure of the PbI2 film, form a cross-linked network, and create the porous structures36.
To verify the effect of annealing temperature on the sequential deposition, XRD intensity color mappings following the annealing temperature profile of the perovskite films without or with EMI-DMP are shown in Fig. 2a, b, respectively. For the perovskite film without EMI-DMP (Fig. 2a), it was found that when the annealing temperature is low (<100 °C), featured peak of PbI2 at 12.8° can be observed due to the dense morphology, which blocks the infiltration of the organic salts and leads to the incomplete conversion of PbI232,37. As the annealing temperature increases, the intensity of PbI2 gradually decreases because of the driving effect of temperature to promote the reaction. However, when the annealing temperature is over 140 °C, PbI2 appears again, which can be attributed to the escape of the organic component at the surface of the perovskite under high temperature30,38. To further confirm the location of PbI2 at low and high annealing temperatures, grazing incidence X-ray diffraction (GIXRD) measurements were performed for the perovskite films annealed at 90 °C and 150 °C. In Supplementary Fig. 5, it is clear that PbI2 locates at the buried surface of the perovskite film annealed at 90 °C, whereas it locates at the top surface of the perovskite film annealed at 150 °C. As for the δ-phase, with the increase in annealing temperature, the featured peak of δ-phase perovskite at 12.1° gradually decreases and totally disappears when the annealing temperature reaches 150 °C. However, as we discussed above, the high annealing temperature is easy to form redundant PbI2 at the top surface of the perovskite films30. In addition, the thermally unstable SAMs at the buried interface tend to desorb27,28,29. To verify the change in SAMs under high temperature, kelvin probe force microscopy (KPFM) measurements were performed and the surface potential distribution of the SAM treated with 110 °C is similar with the pristine SAM, whereas the surface potential distribution of the SAM treated with 150 °C becomes much wider (Supplementary Fig. 6), indicating the desorption of SAMs. In addition, XPS measurements were performed for the corresponding samples in Supplementary Fig. 7. Compared with pristine SAMs, the P/Ni ratio exhibits negligible change for the SAMs treated at 110 °C for 10 min. However, when the temperature increases to 150 °C, an attenuation of the phosphorus signal is observed and the P/Ni ratio decreases from 0.1518 to 0.1169, which is consistent with the KPFM results. Therefore, for the devices based on the perovskite film without EMI-DMP, serious non-radiative recombination at both interfaces will be induced.
XRD intensity color mappings following the temperature profile of the perovskite films a without and b with EMI-DMP additive. The temperature interval is 5 °C for the XRD measurements. c Free-energy calculation for the formation of α-perovskites with or without EMI-DMP additive. Top-view SEM images for d Control- and e LTSD-perovskite films. f AFM height images for Control- and LTSD-perovskite films.
The effect of EMI-DMP on the sequential deposition was then investigated in Fig. 2b. On the one hand, attributed to the porous PbI2@EMI-DMP film, the infiltration of the organic salts is facilitated and the perovskite film without PbI2 is obtained even with a low annealing temperature of 90 °C. On the other hand, the phase transition temperature is decreased and the δ-phase perovskite is eliminated with high-intensity α-phase perovskite when the temperature reaches ca. 110 °C, which will not form redundant PbI2 at the top surface. To locate the PbI2 for the perovskite film with EMI-DMP annealed at 150 °C, GIXRD was performed in Supplementary Fig. 8. The featured peak of α-perovskite at 14.1° gradually increases with larger incident angle, while the PbI2 peak at 12.8° remains unchanged, indicating that PbI2 locates at the surface of the perovskite film, which is consistent with the results observed in Supplementary Fig. 5b. These findings further confirm that high-temperature annealing leads to the formation of redundant PbI2 at the surface of perovskite films. The corresponding XRD patterns for the perovskite films annealed under different temperatures can be found in Supplementary Fig. 9. To understand the mechanism behind the decrease in phase transition temperature from 150 °C to 110 °C, the density functional theory (DFT) was used to calculate the energy difference between δ- and α-phase perovskite with and without EMI-DMP (Fig. 2c). The calculation details can be found in “Method”. As a result, the energy barrier reduces from 0.515 eV to 0.429 eV after EMI-DMP is introduced, enabling the complete phase transition under a low annealing temperature, probably due to the hydrogen bonds formed between EMI-DMP and PbI2 framework as demonstrated in Supplementary Fig. 4. To exclude other possibilities, PbI2 in dimethylformamide was spin-coated onto the substrate and the sample was then placed in a covered petri dish in glovebox filled with N2 to form the porous structure (Supplementary Fig. 10a)39. As shown in the XRD intensity color mapping (Supplementary Fig. 10b), the perovskite film without PbI2 can be obtained with a low annealing temperature due to the porous PbI2. However, the phase-transition temperature barely changes compared with the control perovskite in Fig. 2a, indicating that the porous structure may not have a strong connection with the drop in the phase-transition temperature. To further investigate the effect of hydrogen bond strength on the LTSD strategy, we used 3-butyl-1-methyl-1H-imidazol-3-ium dimethyl phosphate (BMI-DMP) and 1,3-dimethyl-1H-imidazol-3-ium dimethyl phosphate (DMI-EMP) to form weaker and stronger hydrogen bonds with the PbI2 framework, respectively, than EMI-DMP does40. According to the XRD intensity color mappings in Supplementary Fig. 11a, the weaker hydrogen bonds caused by BMI-DMP only induce a slight decrease in the phase-transition temperature from 150 °C to 130 °C, which cannot solve the problems of SAM desorption. As shown in Supplementary Fig. 11b, c when SAM is treated at 130 °C, the P/Ni ratio decreases from 0.1518 to 0.1344, indicating the desorption of SAMs. In contrast, in the case of DMI-DMP, although the stronger hydrogen bonds result in a reduction of the phase-transition temperature (Supplementary Fig. 12a), they retard the nucleation by suppressing the release of cations and anions of perovskite41,42. Combined with the low annealing temperature, the resulting perovskite films show many cracks at grain boundaries (Supplementary Fig. 12b)43. To further prove the presence of more defects in DMI-DMP-perovskite than that in EMI-DMP-perovskite, steady-state photoluminescence (PL) measurements were conducted. As shown in Supplementary Fig. 12c, DMI-DMP-perovskite film shows weaker PL intensity than EMI-DMP-perovskite film, indicating more defects and serious non-radiative recombination in DMI-EMP-perovskite. The above results indicate that the LTSD realized by introducing EMI-DMP is a promising way to obtain high-quality perovskite films with pure α-phase, prevent sacrificing SAMs, and suppress the formation of redundant PbI2.
In the ensuing discussions, Control-perovskite film refers to the sequential deposited perovskite film without EMI-DMP annealed at 150 °C, which is commonly adopted in previous works20,21,22,30. LTSD-perovskite film refers to the sequentially deposited perovskite film with EMI-DMP annealed at 110 °C. The top-view morphology of the Control- and LTSD-films was shown in Fig. 2d, e and the perovskite grains (average grain size >1.3 μm) are superior to those commonly observed in inverted PSCs fabricated by one-step deposition (Supplementary Table 1). The surface of the control film is embellished with a plethora of bright PbI2 flakes, while a neat morphology for the LTSD-perovskite film is observed, which is consistent with the XRD results. Due to the elimination of redundant PbI2 at the top surface, the roughness of the perovskite films is reduced from 31.1 nm (Control) to 26.2 nm (LTSD) as measured by AFM in Fig. 2f, which is beneficial for reducing the shunt paths at the interface of perovskite and the electron transport materials.
To probe the electronic structure of the perovskite surface, XPS was employed and the binding energy peaks of Pb 4f 7/2 and Pb 4f 5/2 located at 138.4/143.2 eV for the Control-perovskite, which shift toward lower binding energy of 138.2 eV/143.0 eV for the LTSD-perovskite, suggesting less PbI2 at the top surface (Fig. 3a)44. Meanwhile, the commonly observed Pb0 defects accompanied with PbI2 are also eliminated for LTSD-perovskite. To show the possible change in the energy level alignment, UV-vis absorption and ultraviolet photoelectron spectra (UPS) were performed. The UV-vis and UPS spectra for Control- and LTSD-perovskite films are shown in Supplementary Figs. 13 to 15. Combined with the band gap, the energy level diagram of PSCs with a structure of FTO/NiOx/Me-4PACz/Perovskite/PCBM/BCP/Cu is summarized in Supplementary Fig. 16 and it is found that the LTSD-perovskite shows a better energy level alignment with the electron transport material, which is beneficial for the charge extraction.
a XPS of Pb 4f spectra of the top surface of Control- and LTSD-perovskite films. b ToF-SIMS depth profiles of LTSD sample with a structure of FTO/LTSD-perovskite/PCBM/BCP/Cu. c XPS of Pb 4f spectra of the bottom surface of Control- and LTSD-perovskite films. d Mott-Schottky plots of the Control- and LTSD-devices, obtained under dark conditions. e tDOS of the Control- and LTSD-devices. f Dependence of the trap density on the profiling distance of the Control- and LTSD-devices measured at an AC frequency of 100 kHz. g TPV and h TPC curves of the Control- and LTSD-devices. i VOC versus light intensity plots for Control- and LTSD-devices.
To investigate the location of EMI-DMP in LTSD-perovskite film, we conducted TOF-SIMS measurements to show the species distribution. As shown in Fig. 3b, the featured signals of phosphorus are mainly located at the buried interface of the perovskite, indicating the top-to-bottom extrusion of EMI-DMP during the crystallization process45,46. To reveal the effect of EMI-DMP at the buried interface, the LTSD-perovskite was washed away to expose the FTO/NiO/SAM substrate. The KPFM result of the substrate in Supplementary Fig. 17 shows even more uniform surface potential distribution compared with the pristine SAM in Supplementary Fig. 6a, probably due to the void-filling effect of EMI-DMP. To further illustrate the voids filling effect of EMI-DMP, we employed XPS for quantitative analysis (Supplementary Fig. 18). As a result, a significant increase in the P/Ni ratio is observed for FTO/NiO/Me-4PACz exposed by removing LTSD-perovskite by DMF, compared with the fresh FTO/NiO/Me-4PACz sample (Supplementary Fig. 7a). In addition, a lift-off process was adopted to separate the perovskite films and the substrate to expose the buried surface of the perovskite47. According to Pb 4f spectra shown in Fig. 3c, the binding energy peaks of Pb 4f 7/2 and Pb 4f 5/2 shift toward lower binding energy from 138.1/142.9 eV (Control) to 137.9/142.7 eV (LTSD), indicating the passivation of the undercoordinated Pb2+ by EMI-DMP48. In addition, the I/Pb ratio for the top and bottom surfaces can be calculated for the Control- and LTSD-perovskite films. The I/Pb ratio for the Control-perovskite increases from 2.55 to 2.82 (Supplementary Figs. 19 to 21), whereas the ratio for the LTSD-perovskite film hardly changes, demonstrating that LTSD strategy optimizes the phase homogeneity of perovskite films in the vertical direction.
The above results demonstrated that the LTSD strategy eliminated the redundant PbI2 and Pb0 defects at the top surface, passivated the undercoordinated Pb2+ at the buried surface, and improved the uniformity of SAMs, which can suppress the non-radiative carrier recombination and accelerate the carrier extraction at both interfaces49. To directly show the comprehensive effect of LTSD on defects, thermal admittance spectroscopy (TAS) was carried out for the devices based on the Control- and LTSD-perovskite films and the device structure50,51. The methods for tDOS and DLCP measurements are presented in the Supplementary Note. The Mott-Schottky analysis (Fig. 3d) shows that the LTSD-device has a much higher built-in potential (Vbi) of 1.20 V compared with 1.14 V for the Control device, suggesting the reduced interfacial energy loss. The energy-dependent trap density of states (tDOS) in PSCs is shown in Fig. 3e, and reduced tDOS is observed for the LTSD-device, especially in the deep trap region (defects at surfaces). To further determine the axial spatial distribution of defects, drive-level capacitance profiling (DLCP) was carried out to visualize the dependence of the trap density on the profiling distance at an AC frequency of 100 kHz (Fig. 3f). The results indicate that the trap densities at the top and buried surfaces of the LTSD-perovskite are significantly decreased, which are consistent with the XPS results.
To validate the impact of reduced defects on the photoelectric characteristic of the perovskite film, we conducted the measurements of steady-state PL and PL mapping. As shown in Supplementary Fig. 22, the LTSD-perovskite film shows higher PL intensity than the Control-perovskite film, indicating the suppressed non-radiative recombination. PL mapping was performed in a range of 30 μm × 30 μm to show the passivation effect on a larger scale. As shown in Supplementary Fig. 23, the PL intensity for the LTSD-perovskite exhibits not only a more uniform distribution but also a higher absolute value for both top and bottom surfaces. As for the devices, a slower transient photovoltage decay (Fig. 3g), a faster transient photocurrent decay (Fig. 3h) and a smaller ideality factor (Fig. 3i) for LTSD-devices also corroborated the reduced carrier non-radiative recombination and improved charge transport.
To verify the stabilizing effect of the LTSD method on perovskite films, we monitored the Control and LTSD-perovskite films with UV-vis and steady-state PL measurements during 20 days of aging test under 1-sun illumination. The UV–vis absorption and PL intensity of Control-perovskite gradually drop, accompanied by a red shift in PL peak position, indicating a significant increase in defect density. (Supplementary Figs. 24a and 25a). By contrast, the UV-vis absorption and PL spectra of LTSD-perovskite films show no obvious change during the test due to the elimination of redundant PbI2, Pb0 and undercoordinated Pb2+ at the surfaces, which can suppress the chemical reactions with the oxygen and water and inhibit the ions migration in the devices (Supplementary Figs. 24b and 25b)52.
In order to explore the influence of LTSD strategy on the photovoltaic performance of PSCs, the devices were fabricated and the cross-section SEM image of the PSCs is shown in Supplementary Fig. 26. The concentration of EMI-DMP was optimized to be 0.6 mg/mL according to the current density–voltage (J–V) curves (Supplementary Fig. 27 and Supplementary Table 3) and the J–V curves of the champion Control- and LTSD-devices are shown in Fig. 4a. The control-device obtains a PCE of 24.5% under reverse scan with VOC, JSC, and FF of 1.16 V, 26.2 mA cm−2, and 80.8%, respectively. By contrast, the LTSD-device obtains a PCE of 26.5% under reverse scan with the VOC, JSC, and FF of 1.21 V, 26.5 mA cm−2, and 82.8%, respectively. The detailed device parameters under both scan directions can be found in Supplementary Table 4. The hysteresis for the LTSD-device is negligible due to the improved charge transfer. To isolate the impact of individual factors on device performance, we fabricated the device based on the perovskite film with EMI-DMP annealed under 150 °C and the J–V curve is shown in Supplementary Fig. 28a. Due to the negative effect of the discrete surface potential of SAM and the redundant PbI2 caused by high annealing temperature, it is observed that the corresponding VOC and fill factor is only slightly higher than that for the control-device even though EMI-DMP is introduced. To further exclude the influence of redundant PbI2, we annealed SAM at 150 °C and fabricated the device based on LTSD-perovskite. As shown in Supplementary Fig. 28b, the desorption of SAM led to an obvious decline in device performance by one percentage point compared with LTSD-device (Fig. 4a). The devices using BMI-DMP and DMI-DMP were also fabricated, exhibiting a decrease in both VOC and FF, compared with LTSD-device (Supplementary Fig. 29a–d).
a J–V curves of the Control- and LTSD-devices with an aperture area of 0.06 cm2. b Statistics of PCEs for Control- and LTSD-devices, and fifty devices were fabricated for each batch. c J–V curves of the Control- and LTSD-devices with an aperture area of 1.02 cm2. d EQE of the Control- and LTSD-devices. e Comparison of the PLQY and QFLS of the corresponding structures. f Steady-state output for the Control- and LTSD-devices. g Normalized PCEs of the encapsulated Control- and LTSD-devices aged under 85 °C/85% RH (ISOS-D-3). h MPP tracking of the encapsulated Control- and LTSD-devices (ISOS-L-2).
Fifty devices were fabricated for control- and LTSD-devices and the reproducibility is improved for LTSD-devices (22.3 ± 1.7% vs 25.6 ± 0.3%) (Fig. 4b). In addition, the enlarged devices with an aperture area of 1.02 cm2 were fabricated. As shown in Fig. 4c and Supplementary Table 5, the champion LTSD-device exhibits a PCE of 25.4%, which reduces the efficiency loss from 12.2 to 4.1%. The improved reproducibility and reduced efficiency loss can be attributed to the defects control in LTSD strategy.
To further examine the photovoltaic properties of PSCs, the typical external quantum efficiencies (EQE) of the devices are shown in Fig. 4d and the integral current densities of the Control- and LTSD-devices are calculated to be 25.77 and 25.98 mA cm−2, respectively, which are consistent with the JSC from J–V curves. We further conducted PL quantum yield (PLQY) measurement and calculated the quasi-Fermi level splitting (QFLS) for perovskite/PCBM stacks to verify the VOC difference between Control- and LTSD-devices (Fig. 4e). As a result, the PLQY is increased from 1.14 to 9.08% and the QFLS is correspondingly increased from 1.17 eV to 1.22 eV for the structure with Control- and LTSD-perovskite/PCBM stacks, respectively, which is consistent with the VOC difference in different devices. One of the LTSD-device was sent to an independent PV test laboratory for certification, and a certified PCE of 26.0% was obtained (Supplementary Fig. 30), which is the highest efficiency based on sequential deposition for SAM-contained inverted PSCs.
In Fig. 4f, the steady-state measurements were carried out to reveal the stabilized power output of the devices. The Control-device delivered a steady-state power output (SPO) of 24.1% at 0.98 V for 600 s, while an SPO of 26.2% at 1.05 V was presented for the LTSD-device. The changes in PCEs of the encapsulated devices were tracked over time under accelerated ageing conditions according to ISOS protocols. To test the damp heat stability, the encapsulated devices were continuously heated at 85°C/85% RH, following the ISOS-D-3 protocol (Fig. 4g and Supplementary Fig. 31). The Control-device went through a fast degradation and the initial efficiency dropped by 43% after 1000 h. By contrast, the LTSD-device maintained 93.1% of its initial efficiency after aging for 1000 h. Once tracked under the 1-sun illumination at the maximum power point and elevated temperature (65 °C, ISOS-L-2), the Control-device degraded to 48.1% of its initial PCE after 500 h, whereas the LTSD-device maintained 95.4% of its initial PCE after 1000 h (Fig. 4h and Supplementary Fig. 32).
Here, a low-temperature sequential deposition method is proposed, which can realize complete phase transition of perovskite without damaging SAMs or forming redundant PbI2. The resultant high-quality perovskite films with improved surfaces can accelerate the charge transport and enhance the device’s stability, contributing to the highest certified efficiency for SAM-contained inverted PSCs fabricated by sequential deposition. The LTSD strategy demonstrates the viability and potential of the sequential deposition method for obtaining efficient and stable inverted PSCs. We believe that this study can inspire attention toward the sequential deposition method and offer new opportunities for molecule design that finally contribute to the commercialization of PSCs.
Nickel acetylacetonate was purchased from Adamas. Acetonitrile, ethanol absolute, dimethylformamide (DMF), dimethyl sulfoxide (DMSO), 2-propanol (IPA) and chlorobenzene (CB) were purchased from Sigma Aldrich. (4-(3,6-dimethyl-9H-carbazole-9-yl) butyl) phosphonic acid (Me-4PACz), lead iodide (PbI2), formamidinium iodide (FAI), methylammonium bromide (MABr), 3-ethyl-1-methyl-1H-imidazol-3-ium dimethyl phosphate (EMI-DMP) and phenylethylamine chloride (PEACl) were purchased from Tokyo Chemical Industry. Methylammonium chloride (MACl) and bathocuproine (BCP) were purchased from Xi’an Yuri Solar Co., Ltd. PCBM was purchased from Luminescence Technology Corp. All the chemicals are used directly without further purification.
F-doped SnO2 (FTO) glass substrate was cleaned by sequentially washing with detergent, deionized water, absolute ethanol, acetone, and isopropanol. Before use, the FTO was cleaned with ultraviolet ozone for 20 min. NiO films were prepared by spray pyrolysis in air. 20 ml of acetonitrile solution of nickel acetylacetonate (0.04 M) was sprayed using an air nozzle onto the clean FTO glasses, which were heated to 500 °C on a hot plate at first, at a distance of about 15 cm, using clean dry air as the carrier gas, and then the coated FTO glasses were kept at 500 °C for 30 min to promote NiO crystallization53. Then the substrates were transferred to a N2-filled glove box while the substrate is still hot. After cooled to room temperature, 1.0 mg/ml Me-4PACz dissolved in ethanol absolute was deposited on the NiO films at 4000 rpm for 30 s, followed by annealing at 100 °C for 10 min. For deposition of Control perovskite layer, 1.5 M of PbI2 in DMF:DMSO (9:1) solvent was spin-coated onto substrate at 1500 rpm for 30 s, and annealed at 70 °C for 1 min, then cooled to room temperature. Then, a solution of FAI:MABr:MACl (90.0 mg:5.5 mg:9.0 mg in 1 ml IPA) was spin-coated onto the PbI2 at spin rate of 1500 rpm for 30 s, and the perovskite precursor film annealed at 150 °C for 15 min. For deposition of LTSD-perovskite layer, different concentrations (0.2–0.8 mg/ml) EMI-DMP were added into the PbI2 solution, and the annealing temperature of LTSD-perovskite films was decreased to 110 °C. For passivation layer, 1.0 mg/mL of PEACl dissolved in IPA was spin-coated on perovskite at 5000 rpm for 30 s, followed by annealing at 100 °C for 5 min. Then, 20 mg/ml of PCBM dissolved in CB was spin-coated at 1000 rpm for 30 s, followed by annealing at 70 °C for 10 min. 0.5 mg/ml of BCP dissolved in IPA was spin-coated at 6000 rpm for 30 s, followed by annealing at 70 °C for 10 min. Then, 90 nm thickness of Cu was thermally evaporated as an electrode using a shadow mask. After finishing the device fabrication, 120 nm thickness of MgF2 anti-reflection coatings were made on the glass side of the textured FTO substrate via thermal evaporation. Finally, the top of the perovskite solar cell was covered with 1.1 mm-thick coverslip glass, which was fixed with UV-curable resin (ThreeBond Group, 3013B). The solidification of the resin was conducted by UV light illumination for about 30 s.
We used the DFT as implemented in the Vienna Ab initio simulation package in all calculations. The exchange-correlation potential is described by using the generalized gradient approximation of Perdew–Burke–Ernzerhof. The projector augmented-wave method is employed to treat interactions between ion cores and valence electrons. The plane-wave cutoff energy was fixed to 450 eV. Given structural models were relaxed until the Hellmann-Feynman forces smaller than −0.02 eV/Å and the change in energy smaller than 10−5 eV was attained. Grimme’s DFT-D3 methodology was used to describe the dispersion interactions among all the atoms in adsorption models.
SEM images were measured by JSM-7800F (JEOL, China). FTIR spectra were measured by Nicolet 6700 (THERMO FISHER, USA). The XRD pattern was measured by Mini Flex 600 (Rigaku, Japan) using Cu Kα radiation. GIXRD pattern was measured by D-2422 (Rigaku, Japan) using Cu Kα radiation. The time-of-flight secondary ion mass spectroscopy was performed with an IONTOF TOF.SIMS 5-100 instrument. An Ar cluster (10 keV) was used in the sputtering and Bi3+ (30 keV) was used to probe the sample with an analysis area of 100 × 100 μm2. AFM and KPFM were measured by MFP-3D (Oxford, UK). XPS and ultraviolet photoelectron spectroscopy (UPS) were measured by AXIS Ultra DLD (China) with an Al Kα X-ray source and He Iα photon source (21.2 eV), respectively. PL mapping was measured by inVia Qontor (Renishaw, UK). PL and PL quantum yield (PLQY) were measured by FLS1000 (Edinburgh Instruments, UK). UV–vis absorption spectra were measured by Shimadzu UV 2450 spectrometry.
Current density-voltage (J–V) curves of PSCs were measured by a Keithley 2400 digital source meter under a simulated solar source of AM1.5 G (100 mW cm−2, Wacom Denso Co., Japan). The measurement was conducted forward (from −0.2 to 1.3 V) scan or reverse (from 1.3 to −0.2 V) scan. The delay time and step voltage were set as 20 ms and 20 mV, respectively. The EQE spectra were performed with the director current mode (CEP-2000BX, Bunko-Keiki).
The Mott-Schottky analysis, TAS and DLCP were performed on Zahner, Germany. For the Mott-Schottky analysis, the DC bias was scanned from 0.5 V to the 1.3 V. The amplitude of the AC bias was 20 mV. The DLCP measurements were conducted in the same DC bias range with that for the standard C–V measurement at 100 kHz. While the amplitude of the AC biases ranged from 20 to 200 mV. For the TAS measurement, the DC bias was fixed at 0 V and the amplitude of the AC bias was 20 mV. The scanning range of the AC frequency was 1–100 kHz. The ideality factors were obtained by measuring dark J–V curves under different light intensities. The transient photovoltage and transient photocurrent test were performed on an attenuated UV laser pulse (SRS NL 100 Nitrogen Laser) under 1-sun illumination. The laser wavelength is 337 nm, the repeating frequency was about 20 Hz, and the pulse width was less than 3.5 ns.
For damp-heat stability (following the testing procedures of the ISOS-D-3 protocols), the devices were put into an environment chamber with a temperature of 85 °C, humidity of RH 85% and kept in the dark. For operational stability (following the testing procedures of the ISOS-L-2 protocols), the MPP tracking was performed with a solar cell light resistance testing system (Bunkoukeiki, Japan) under a white LED light with 1 sun equivalent intensity and continuous bias voltage, and the temperature was around 65 °C. The J–V curves under forward scans were recorded every 10 h during the whole test. The maximum power points were obtained from the J–V curves and were also adjusted every 10 h.
Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.
The data that support the main findings are available in the main text and Supplementary Information. Source data are provided with this paper.
Jeng, J.-Y. et al. CH3NH3PbI3 perovskite/fullerene planar-heterojunction hybrid solar cells. Adv. Mater. 25, 3727 (2013).
Article CAS Google Scholar
Jiang, Q. et al. Towards linking lab and field lifetimes of perovskite solar cells. Nature 623, 313–318 (2023).
Article CAS Google Scholar
Lin, Y. et al. Bandgap-universal passivation enables stable perovskite solar cells with low photovoltage loss. Science 384, 767–775 (2024).
Article CAS Google Scholar
Lin, X. et al. Efficiency progress of inverted perovskite solar cells. Energy Environ. Sci. 13, 3823–3847 (2020).
Article CAS Google Scholar
Aydin, E. et al. Enhanced optoelectronic coupling for perovskite/silicon tandem solar cells. Nature 623, 732–738 (2023).
Article CAS Google Scholar
Fei, C. et al. Strong-bonding hole-transport layers reduce ultraviolet degradation of perovskite solar cells. Science 384, 1126–1134 (2024).
Article CAS Google Scholar
Liu, M. et al. Compact hole-selective self-assembled monolayers enabled by disassembling micelles in solution for efficient perovskite solar cells. Adv. Mater. 35, 2304415 (2023).
Article CAS Google Scholar
Zhao, K. et al. peri-Fused polyaromatic molecular contacts for perovskite solar cells. Nature 632, 301–306 (2024).
Article CAS Google Scholar
Zhang, S. et al. Minimizing buried interfacial defects for efficient inverted perovskite solar cells. Science 380, 404–409 (2023).
Article CAS Google Scholar
Liu, S. et al. Buried interface molecular hybrid for inverted perovskite solar cells. Nature 632, 536–542 (2024).
Article Google Scholar
Chen, H. et al. Improved charge extraction in inverted perovskite solar cells with dual-site-binding ligands. Science 384, 189–193 (2024).
Article CAS Google Scholar
Yang, Y. et al. Amidination of ligands for chemical and field-effect passivation stabilizes perovskite solar cells. Science 386, 898–902 (2024).
Article CAS Google Scholar
Dong, B. et al. Self-assembled bilayer for perovskite solar cells with improved tolerance against thermal stresses. Nat. Energy 10, 342–353 (2025).
Article CAS Google Scholar
Burschka, J. et al. Sequential deposition as a route to high-performance perovskite-sensitized solar cells. Nature 499, 316–319 (2013).
Article CAS Google Scholar
Han, Y., Xie, H., Lim, E. L. & Bi, D. Review of two-step method for lead halide perovskite solar cells. Sol. RRL 6, 2101007 (2022).
Article CAS Google Scholar
Li, H. et al. Sequential vacuum-evaporated perovskite solar cells with more than 24% efficiency. Sci. Adv. 8, eabo7422 (2022).
Article PubMed Central Google Scholar
Zai, H. et al. Wafer-scale monolayer MoS2 film integration for stable, efficient perovskite solar cells. Science 387, 186–192 (2025).
Article CAS Google Scholar
Zhao, Y. et al. Inactive (PbI2)2RbCl stabilizes perovskite films for efficient solar cells. Science 377, 531–534 (2022).
Article CAS Google Scholar
Lin, X. et al. In situ growth of graphene on both sides of a Cu-Ni alloy electrode for perovskite solar cells with improved stability. Nat. Energy 7, 520–527 (2022).
Article CAS Google Scholar
Fang, J. et al. Anion exchange promoting non-impurities enables conformable and efficient inverted perovskite solar cells. Energy Environ. Sci. 17, 7829 (2024).
Article CAS Google Scholar
Chen, P. et al. Efficient inverted perovskite solar cells via improved sequential deposition. Adv. Mater. 35, 2206345 (2023).
Article CAS Google Scholar
Wang, C. et al. Dipole molecule-mediated modulating residual PbI2 clusters in two-step-processing inverted perovskite photovoltaics. Nano Lett. 24, 15912–15920 (2024).
Article CAS Google Scholar
Sun, Q. et al. Porous lead iodide layer promotes organic amine salt diffusion to achieve high performance p-i-n flexible perovskite solar cells. Adv. Energy Mater. 13, 2301259 (2023).
Article CAS Google Scholar
Ahlawat, P. et al. A combined molecular dynamics and experimental study of two-step process enabling low-temperature formation of phase-pure α-FAPbI3. Sci. Adv. 7, eabe3326 (2021).
Article CAS PubMed Central Google Scholar
He, W. et al. Temperature-assisted crystal growth of photovoltaic α‑phase FAPbI3 thin films by sequential blade coating. ACS Appl. Mater. Interfaces 12, 55830–55837 (2020).
Article CAS Google Scholar
Bi, L. et al. Deciphering the roles of MA-based volatile additives for α-FAPbI3 to enable efficient inverted perovskite solar cells. J. Am. Chem. Soc. 145, 5920–5929 (2023).
Article CAS Google Scholar
Li, Z. et al. Stabilized hole-selective layer for high-performance inverted p-i-n perovskite solar cells. Science 382, 284–289 (2023).
Article CAS Google Scholar
Qin, Z., Chen, M., Zhang, Z., Wang, Y. & Han, L. Eliminating high-dimensional defects by upward unidirectional crystallization for efficient and stable inverted perovskite solar cells. Energy Environ. Sci. 18, 2264–2272 (2025).
Article CAS Google Scholar
Zhang, D. et al. Stable α-FAPbI3 in inverted perovskite solar cells with efficiency exceeding 22% via a self-passivation strategy. Adv. Funct. Mater. 32, 2200174 (2022).
Article CAS Google Scholar
Qu, D. et al. Phase homogeneity mediated charge-carrier balance in two-step-method halide perovskite photovoltaics. Energy Environ. Sci. 18, 1310–1319 (2025).
Article CAS Google Scholar
Ren, Y. T. et al. A spiro-OMeTAD based semiconductor composite with over 100 °C glass transition temperature for durable perovskite solar cells. Nano Energy 81, 105655 (2021).
Article CAS Google Scholar
Gao, Y., Raza, H., Zhang, Z., Chen, W. & Liu, Z. Rethinking the role of excess/residual lead iodide in perovskite solar cells. Adv. Funct. Mater. 33, 2215171 (2023).
Article CAS Google Scholar
He, J. et al. Omnidirectional diffusion of organic amine salts assisted by ordered arrays in porous lead iodide for two-step deposited large-area perovskite solar cell. Energy Environ. Sci. 16, 629–640 (2023).
Article CAS Google Scholar
Xu, Y. et al. Perovskite films regulation via hydrogen-bonded polymer network for efficient and stable perovskite solar cells. Angew. Chem. Int. Ed. 62, e202306229 (2023).
Article CAS Google Scholar
Zhao, W. et al. A special additive enables all cations and anions passivation for stable perovskite solar cells with efficiency over 23. Nanomicro Lett. 13, 169 (2021).
CAS PubMed Central Google Scholar
Lin, R.-B. & Chen, B. Hydrogen-bonded organic frameworks: chemistry and functions. Chem 8, 2114–2135 (2022).
Article CAS Google Scholar
Ko, H.-S., Lee, J.-W. & Park, N.-G. 15.76% efficiency perovskite solar cells prepared under high relative humidity: importance of PbI2 morphology in two-step deposition of CH3NH3PbI3. J. Mater. Chem. A 3, 8808–8815 (2015).
Article CAS Google Scholar
Hu, X. et al. Strain release of formamidinium-cesium perovskite with imprint-assisted organic ammonium halide compensation for efficient and stable solar cells. Nano Energy 101, 107594 (2022).
Article CAS Google Scholar
Liu, T. et al. Mesoporous PbI2 scaffold for high-performance planar heterojunction perovskite solar cells. Adv. Energy Mater. 6, 1501890 (2016).
Article Google Scholar
Dong, K., Zhang, S., Wang, D. & Yao, X. Hydrogen bonds in imidazolium ionic liquids. J. Phys. Chem. A 110, 9775–9782 (2006).
Article CAS Google Scholar
Lin, L. et al. Hydrogen bonding in perovskite solar cells. Matter 7, 38–58 (2024).
Article CAS Google Scholar
Li, M. et al. Orientated crystallization of FA-based perovskite via hydrogen-bonded polymer network for efficient and stable solar cells. Nat. Commun. 14, 573 (2023).
Article CAS PubMed Central Google Scholar
Dunlap-Shohl, W. A., Zhou, Y., Padture, N. P. & Mitzi, D. B. Synthetic approaches for halide perovskite thin films. Chem. Rev. 119, 3193–3295 (2019).
Article CAS Google Scholar
Zhidkov, I. S. et al. XPS spectra as a tool for studying photochemical and thermal degradation in APbX3 hybrid halide perovskites. Nano Energy 79, 105421 (2021).
Article CAS Google Scholar
Zheng, X. et al. Co-deposition of hole-selective contact and absorber for improving the processability of perovskite solar cells. Nat. Energy 8, 462–472 (2023).
Article CAS Google Scholar
Tan, Q. et al. Inverted perovskite solar cells using dimethylacridine-based dopants. Nature 620, 545–551 (2023).
Article CAS Google Scholar
Li, D. et al. Co-adsorbed self-assembled monolayer enables high-performance perovskite and organic solar cells. Nat. Commun. 15, 7605 (2024).
Article CAS PubMed Central Google Scholar
Mai, C.-L. et al. Donor-π-acceptor type porphyrin derivatives assisted defect passivation for efficient hybrid perovskite solar cells. Adv. Funct. Mater. 31, 2007762 (2021).
Article CAS Google Scholar
Luo, D., Su, R., Zhang, W., Gong, Q. & Zhu, R. Minimizing non-radiative recombination losses in perovskite solar cells. Nat. Rev. Mater. 5, 44–60 (2020).
Article CAS Google Scholar
Ni, Z. et al. Resolving spatial and energetic distributions of trap states in metal halide perovskite solar cells. Science 367, 1352–1358 (2020).
Article CAS Google Scholar
Fei, C. et al. Lead-chelating hole-transport layers for efficient and stable perovskite minimodules. Science 380, 823–829 (2023).
Article CAS Google Scholar
Cheng, Y. & Ding, L. Pushing commercialization of perovskite solar cells by improving their intrinsic stability. Energy Environ. Sci. 14, 3233–3255 (2021).
Article Google Scholar
Ye, F. et al. Soft-cover deposition of scaling-up uniform perovskite thin films for high cost-performance solar cells. Energy Environ. Sci. 9, 2295–2301 (2016).
Article CAS Google Scholar
Download references
The authors acknowledge financial support from the National Key R&D Program of China (No. 2021YFB3800102, L.H.), National Natural Science Foundation of China (No. U21A20171, L.H.) and Shanghai Rising-Star Program (No. 24QA2704200, Y.W.). We thank Z.B. and Y.H. for assistance with SEM measurement, X.D. for assistance with XPS and UPS measurements, Q.R. for assistance with GIXRD measurement, M.G. for assistance with AFM and KPFM measurements, and R.W. for assistance with steady PL, PL mapping and PLQY measurements from the Instrumental Analysis Center of Shanghai Jiao Tong University (China).
State Key Laboratory of Metal Matrix Composites, Shanghai Jiao Tong University, Shanghai, China
Mengjiong Chen, Zhenzhen Qin, Ziyang Zhang, Wenxiang Xiang, Yingming Liu, Chuang Tian, Siyuan Chen, Yanbo Wang & Liyuan Han
PubMed Google Scholar
PubMed Google Scholar
PubMed Google Scholar
PubMed Google Scholar
PubMed Google Scholar
PubMed Google Scholar
PubMed Google Scholar
PubMed Google Scholar
PubMed Google Scholar
M.C., Y.W. and L.H. conceived the idea. M.C. and Z.Q. designed the experiment and prepared samples. C.T. conducted PL and UV–vis characterization, Y.L. conducted stability test. Z.Z. conducted the SEM and FTIR characterization. S.C. conducted the XRD and device characterization. M.C., W.X. and Y.W. contributed to data analysis and visualization. Y.W. and L.H. directed and supervised the project. M.C. and Z.Q. wrote the first draft of the manuscript. M.C., Y.W. and L.H. revised the manuscript. All authors discussed the results and contributed to the revisions of the manuscript.
Correspondence to Yanbo Wang or Liyuan Han.
The authors declare no competing interests.
Nature Communications thanks Yongguang Tu, and the other, anonymous, reviewer(s) for their contribution to the peer review of this work. A peer review file is available.
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.
Reprints and permissions
Chen, M., Qin, Z., Zhang, Z. et al. Low-temperature sequential deposition for efficient inverted perovskite solar cells. Nat Commun 16, 5746 (2025). https://doi.org/10.1038/s41467-025-61144-y
Download citation
Received:
Accepted:
Published:
Version of record:
DOI: https://doi.org/10.1038/s41467-025-61144-y
Anyone you share the following link with will be able to read this content:
Sorry, a shareable link is not currently available for this article.
Provided by the Springer Nature SharedIt content-sharing initiative
Nature Communications (2025)
Advertisement
Nature Communications (Nat Commun)
ISSN 2041-1723 (online)
© 2025 Springer Nature Limited
Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.
