The formation of amorphous (lysine)2PbI2
Optical and structural measurements of perovskite films reveal that the lysine reacts with residual PbI2 within FA0.85MA0.1Cs0.05PbI3 (denoted as PbI2-FACsMA) films and does not form any new crystal phases. The solid-state diffusion process of lysine into PbI2-FACsMA perovskite film is shown in Supplementary Fig. S1 (the resulting films, denoted as amo-FACsMA), which is probed using ultraviolet-visible (UV-vis) spectroscopy. There is no notable change at the absorption edge of the amo-FACsMA perovskite film (Supplementary Fig. S2). Grazing-Incidence Wide-Angle X-ray Scattering (GIWAXS) is employed to assess the types of crystal phase in these perovskite films (Fig. 1a). This characterization reveals a distinct difference between PbI2-FACsMA and amo-FACsMA films. The ring related to the PbI2 phase is observed in the PbI2-FACsMA film, while the ring associated with the PbI2 phase is absent in the amo-FACsMA film. This strong contrast indicates PbI2 disappearance without the emergence of a new crystal phase in amo-FACsMA film. To further understand this phenomenon, the crystal structure changes in amo-FACsMA film during the lysine diffusion process are monitored using X-ray diffraction (XRD) analysis (Supplementary Fig. S3). To rule out the impact of pressure and heat, the same process is conducted by employing a clean glass. The XRD pattern for PbI2-FACsMA perovskite exhibits negligible changes, while a progressive reduction of PbI2 is observed in the amo-FACsMA sample during the lysine diffusion process. Interestingly, a consistent pattern observed throughout the experiment is the disappearance of PbI2, which occurs without the formation of any new crystal structures, aligning with the results from GIWAXS.Fig. 1: Structural, morphological, and spectroscopic characterizations of amo-FACsMA films.a GIWAXS data of PbI2-FACsMA and amo-FACsMA films. Top-surface and cross-section images of (b) PbI2-FACsMA and (c) amo-FACsMA perovskite films, the dashed box represents PbI2. Scale bars are 1 μm. d XRD patterns for PbI2, Lysine, (Lysine)2PbI2, a. u., arbitrary units. e Wavelet transformations-EXAFS images of (Lysine)2PbI2. f Schematic of amorphous (lysine)2PbI2.Morphological and composition studies further indicate that the lysine reacts with the PbI2 phase and potentially forms an amorphous material, which improves the film morphologies. PbI2-FACsMA film shows substantial PbI2 residues, while amo-FACsMA film shows pinhole-free grains without PbI2 at both surface and grain boundaries (GBs) (Fig. 1b, c). High-resolution SEM images reveal evidence of solid-phase reactions for the PbI2 residues in PbI2-FACsMA film during the lysine diffusion process (Supplementary Figs. S4, S5). Time-of-Flight Secondary Ion Mass Spectrometry (TOF-SIMS) is conducted to verify the distribution of lysine within these perovskite films (Supplementary Fig. S6). In the amo-FACsMA perovskite films, lysine exhibits a vertical distribution similar to that of PbI2 in the cross-section profile of Fig. 1b, predominantly accumulated at the upper surface layer. The addition of lysine causes minimal changes in the distribution of FA+, MA+, Cs+, Pb2+, and I– (Supplementary Fig. S7). Overall, the alterations in morphology and composition, along with the absence of new crystal phases in GIWAXs and XRD, suggest the formation of an amorphous material in the surface/interface of amo-FACsMA perovskite25.Spectroscopy and structure investigations confirm that lysine can react with PbI2 and form an entirely new amorphous (lysine)2PbI2. To thoroughly characterize the product of such a reaction, different ratios of yellow PbI2 and white lysine are mixed (Supplementary Fig. S8). The blended powders are ground for 30 minutes to ensure a complete amalgamation. Subsequently, the powders transform into a light black color after a 30-minute heating process. XRD results reveal that peaks referring to PbI2/lysine can be detected at lysine to PbI2 ratios of below 2 or above 2 but not at the ratio of 2:1 (Supplementary Fig. S9). For the 2:1 ratio, there are no diffraction peaks observed in the resulting material (Fig. 1d). Combining the composition analysis result (as shown in the section of preparation of (lysine)2PbI2), we can confirm the formation of amorphous (lysine)2PbI2.X-ray photoelectron spectroscopy (XPS) is conducted to probe the changes in the elements and their chemical states in lysine, PbI2, and (lysine)2PbI2 (Supplementary Fig. S10). All core-level peaks are assigned to C 1 s, N 1 s, Pb 4 f, and I 3d. The peaks referred to Pb 4 f in (lysine)2PbI2 shift towards lower energy compared to the case of PbI2, which could be attributed to the interaction between lysine and Pb. Similarly, the I 3d peaks exhibit a shift towards lower binding energy, while the peaks of N 1 s exhibit a shift towards higher energy. Meanwhile, Fourier-transform infrared spectroscopy (FTIR) spectra indicate the shift of C = O and the absence of O-H observed in (lysine)2PbI2 (Supplementary Fig. S11). These results indicate that the adjacent amino and carboxyl groups in lysine coordinate with PbI2. The synchrotron X-ray source is further employed to probe the amorphous (lysine)2PbI2 and confirm structure information (Fig. 1e and Supplementary Fig. S12). R-space Extended X-ray Absorption Fine Structure (EXAFS) results confirm the formation of Pb-I, Pb-N, and Pb-O in (lysine)2PbI2 (Fig.1e and Supplementary Fig. S12). In R-space EXAFS data, a strong peak around 2.70 Å is assigned to the Pb-I bond, which is smaller than the distance Pb-I in PbI2 (3.17 Å)26. The peaks centering at 1.11 Å and 1.69 Å could be attributed to the formation of Pb-N/Pb-O27,28. Inspired by the crystal structure of (lysine)2NiCl229, the bond information of Pb-I/Pb-O/Pb-N, and the 6-coordinate of Pb, we firmly verify the structure of amorphous (lysine)2Pbl2 (Fig. 1f). Newly formed (lysine)2PbI2 demonstrates exceptional stability, maintaining its integrity for over 2000 hours without any observable changes (Supplementary Fig. S13).The aforementioned results indicate that lysine can react with the PbI2 phase, leading to the generation of stable amorphous (lysine)2PbI2 at the perovskite’s surface/GBs. The primary feature of amorphous materials is their isotropic nature, which exhibits significantly improved compatibility with perovskite compared to crystal passivation materials like PbI2 and lysine (Supplementary Figs. S14, S15). The influence of amorphous (lysine)2PbI2 on the performance of perovskite films is then investigated.Amorphous (lysine)2PbI2 enhanced and stabilized PerovskiteThe amorphous (lysine)2PbI2 exhibits a significantly better passivation effect compared to crystalline PbI2 OR lysine molecule passivators. The photoluminescence (PL) mapping result in Fig. 2a and Supplementary Fig. S16 reveals that both PbI2 and lysine can increase the PL intensity. However, these thin films still experience significant quenching near the GBs. In contrast, the amo-FACsMA sample displays a consistently, highest PL intensity across almost the entire surface. Moreover, amo-FACsMA shows a narrower distribution of blue-shift PL emission peak, mainly centered at 790 nm (Supplementary Fig. S17), signifying a reduction in spontaneous radiative recombination through defect/trap states in amo-FACsMA perovskite film. Meanwhile, the time-resolved PL (TRPL) spectra in Fig. 2c show that amo-FACsMA perovskite exhibits the longest PL lifetime (τ) (1.04 μs), which is far longer than PbI2/lysine passivation (Supplementary Fig. S16). Enhanced PL intensity, blue-shift and narrow PL emission peak distribution, and longer PL lifetime indicate that non-radiative recombination in amo-FACsMA films is suppressed, which can be attributed to decreased trap densities. Consistently, the Urbach energy (Supplementary Fig. S18) is significantly decreased in amo-FACsMA. The smaller Urbach energy in amo-FACsMA corresponds to a lower density of trap states30. These results indicate that amorphous (lysine)2PbI2 can notably diminish the defects in FACsMA perovskite.Fig. 2: Effect of isotropic (lysine)2PbI2 on the performance of perovskites.a PL mapping results for PbI2-FACsMA and amo-FACsMA films (area: 10 μm × 5 μm), respectively. b TRPL curves of PbI2-FACsMA and amo-FACsMA films. c Density of states of traps (tDOS) deduced from the room-temperature C–f plots for PbI2-FACsMA and amo-FACsMA. XRD results of perovskite films after (d) 55 °C heating, (e) 1 sun illumination, and (f) 55 °C heating + 1 sun illumination stability test in a glove box, the diamond represents PbI2.We have further investigated the defect physics of these samples using thermal admittance spectroscopy (TAS). All devices show typical temperature-dependent capacitance versus frequency (C–f) plots (Supplementary Fig. S19). The sub-gap energy is deduced from the temperature-dependent C–f plots. Both PbI2 and lysine passivators can decrease the trap energy to 0.49 eV and 0.47 eV, respectively, compared to the 0.73 eV trap energy of pristine perovskite (Supplementary Fig. S20). However, the amo-FACsMA-based device exhibits the lowest trap energy depth of 0.15 eV. Figure 2c and Supplementary Fig. S21 show the trap density deduced from the room-temperature C–f plots, and the amo-FACsMA sample exhibits the lowest integrated trap density of 3.2 × 1015 cm−3 eV−1. These results indicate a moderate passivation effect of crystal lysine and PbI2 passivation agent. In contrast, the weak correlation of C-f with temperatures in amo-FACsMA-based devices suppressed influence from trap states, and hence excellent passivation.In general, polycrystalline perovskites typically possess numerous defects including unsaturated dangling bonds (unsaturated Pb2+ and ion vacancies)6. The crystalline passivators like PbI2 and lysine possess the ability to passivate certain defects to a degree; however, they potentially introduce new issues, for example, lattice stress arising from alterations in ionic bond energy because of the robust interaction between crystalline passivation materials and surface/interface ions. In addition, because of the nature of crystal materials, the interface between polycrystalline perovskite and crystal passivators often hosts many defects, such as dislocations, dangling bonds, and lattice strain. These defects significantly deteriorate the PV performance. Herein, the PbI2 on the surface/interface of perovskite can serve as a seed to react with lysine, thereby forming an amorphous layer with fewer dangling bonds terminated on the surface/interface of perovskite. The interface between the crystalline and amorphous materials hosts fewer dislocations, dangling bonds, and smaller strain than the interfaces present in the polycrystalline sample (crystal-crystal, crystal-air)23,31,32. Consequently, the amorphous (lysine)2PbI2 demonstrates remarkable passivation effects, leading to a substantial decrease in trap energy depth and defect density within the amo-FACsMA sample. These outcomes are beneficial to mitigating non-radiation recombination and enhancing PV performance.In addition, the amorphous (lysine)2PbI2 significantly improves the stability of PbI2-FACsMA perovskite films. The XRD patterns of the FACsMA and amo-FACsMA perovskite films after 2000 h continuous illumination or/and heating are comparatively presented in Fig. 2d–f. The PbI2 impurity peak is distinct for the aged PbI2-FACsMA perovskite film (especially under the light and thermal coupling condition), which could be attributed to the thermal decomposition of organic cations (FA+ and MA+) in PbI2-FACsMA perovskite. Interestingly, the aged amo-FACsMA sample exhibits hardly any detectable PbI2 peak. These results suggest that the amorphous (lysine)2PbI2 can suppress perovskite de-structuring by enhancing the resistance to thermal/photo-induced decomposition of PbI2-FACsMA perovskite.In general, during the aging process of perovskites featuring FA+/MA+ cations, a significant quantity of PbI2 becomes evident, resulting from the deprotonation of FA+/MA+ cations and the subsequent formation of MA/FA, accompanied by volatilization. Herein, the NH3+ functional group in amorphous (lysine)2PbI2 can donate protons, thereby inhibiting the deprotonation of FA and MA cations. Moreover, amorphous (lysine)2PbI2 has exhibited improved compatibility and significantly reduced the lattice stress (confirmed by the GIXRD result), which favors for decrease in the escape and release of organic cation. As such, amorphous and isotropic (lysine)2PbI2 can resist the deprotonation of FA+/MA+ cations and also reduce the residual stress, therefore leading to improved stability of the thin film.Photovoltaic performances of the PSCsBenefiting from the advantages of amorphous (lysine)2PbI2, the resulting amo-FACsMA perovskite shows much enhanced device performance. Figure 3a compares the current density-voltage (J-V) characteristics of champion PSCs based on PbI2-FACsMA, and amo-FACsMA perovskites, respectively. The amo-FACsMA-based PSC exhibits an impressive PCE of 26.27% compared with 23.72% for PbI2-FACsMA. We also obtained a certified PCE of 25.94% with negligible hysteresis for amo-FACsMA -based PSC (Supplementary Fig. S22). The most striking difference is the Voc, which increases from 1.104 V in PbI2-FACsMA to 1.184 in amo-FACsMA. The EQE (Fig. 3b) is similar for both devices, with a high value of over 90% in the wavelength range of 450 ~ 700 nm. Figure 3c compares the PV parameters of FACsMA and amo-FACsMA- based PSCs for 32 devices respectively, indicating that amorphous (lysine)2PbI2 also improves the device reproducibility. In addition, the amo-FACsMA-based PSCs exhibit a smaller hysteresis (Supplementary Fig. S23 and Table S1), resulting in a stabilized output power of 26.27% (Fig. 3d). We further fabricate PSCs with an effective area of 1 cm2 based on these amo-FACsMA films. The champion amo-FACsMA device displays a PCE of 24.93%, which is far higher than that of PbI2-FACsMA devices (~ 21.38%) (Fig. 3e and Supplementary Fig. S24 and Table S2).Fig. 3: Photovoltaic and device characterization.a The J-V curves of the champion devices of PbI2-FACsMA and amo-FACsMA with an effective cell area of 0.09 cm2 in the reverse scan. b EQE spectra and integrated Jsc of PbI2-FACsMA and AMO-FACsMA-based PSCs. c The PV parameters distribution of PbI2-FACsMA and amo-FACsMA PSCs from 32 devices, respectively. d Steady-state efficiency of amo-FACsMA PSCs. e J-V characteristics of PSCs based on PbI2-FACsMA and amo-FACsMA PSCs with a 1 cm2 effective cell area in the reverse scan. f TPV curves of PbI2-FACsMA and amo-FACsMA-based devices.The significantly improved Voc in the amo-FACsMA device aligns with previous photophysical measurements on the films, i.e., the amorphous (lysine)2PbI2 effectively reduces the defect density. Further measurements on the devices also reach similar conclusions. The trap-filled limiting voltage in the space-charge limited current (SCLC) measurements decreases from 0.096 V in the FACsMA device to 0.069 V in the amo-FACsMA device (Supplementary Fig. S25), indicating reduced defects upon isotropic (lysine)2PbI2 formation. This result is also consistent with previous TAS and transient photovoltage (TPV) decay results (Fig. 3f). Compared to the FACsMA device (0.14 μs), the slower Voc decay (0.53 μs) in amo-FACsMA device indicates a longer recombination lifetime, which can contribute to a higher Voc33,34.Shelf-life and operational stability of the PSCsIn addition to improved PV performance, the amo-FACsMA device shows significantly enhanced stability. We measure the shelf life by storing the unencapsulated devices in the dark at 25 °C in an N2 glovebox. Figure 4a demonstrates a 10% decrease in the PCE of the FACsMA device after 100 days of aging, while the amo-FACsMA device exhibits minimal variation over the same aging period. We then investigate the long-term operational stability of the PSCs by aging the encapsulated devices in the ambient, using maximum power point (MPP) tracking under simulated 1-sun conditions. As shown in Fig. 4b, the amo-FACsMA-based PSCs retain over 94% of the initial PCE while the FACsMA device maintains only approximately 20% PCE after the 1000 h MPP test. Even at 85 °C, the amo-FACsMA-based PSCs also demonstrate excellent stability (Supplementary Fig. 4c).Fig. 4: Stability of the PbI2-FACsMA and amo-FACsMA PSCs.a The shelf-life stability of unencapsulated PbI2-FACsMA and amo-FACsMA PSCs. b The long-term operational stability of FACsMA and amo-FACsMA PSCs. c The long-term operational stability of PbI2-FACsMA and amo-FACsMA PSCs at 85 °C. d Cross-section images of PbI2-FACsMA and amo-FACsMA PSCs after the 500-h operational stability test at their MPP. e TOF-SIMS depth profiles for the aged devices based on PbI2-FACsMA and amo-FACsMA perovskites, Corresponding (f) 3D MA+/FA+ distribution in the aged PSCs.The enhanced stability of PSCs is attributed to the suppressed perovskite de-structuring of light-absorbing layers in amo-FACsMA PSCs. In Fig. 4d, we compare the cross-section images of the perovskite layer for the PSCs based on FACsMA and amo-FACsMA after the 500 h MPP test. Contrasting the cross-section of the initial perovskite device (Supplementary Fig. S26), the perovskite layer in the FACsMA PSCs displays a softened texture with emerged new phases. Combining our previous results on the film stability, the new phase could be assigned to PbI2. In contrast, the cross-sectional morphology of the amo-FACsMA device shows no discernible changes, which can be attributed to the stabilizing effect of isotropic (lysine)2PbI2.We further conducted an in-depth analysis of component distribution within the perovskite layer for PSCs based on the FACsMA and amo-FACsMA perovskite films. In the FACsMA device (Fig. 4e), a notable offset component distribution is observed compared with the initial devices shown in Supplementary Fig. S27. Both MA+ and FA+ cations exhibit reduced intensity within the perovskite layer and are found to migrate to the Au/BCP/C60 layer. Figure 4f illustrates the 3D MA+/FA+ distribution in the aged FACsMA PSCs, with a concentration of these cations at the top of the devices. These results indicate that FACsMA perovskite in the PSCs undergoes a perovskite de-structuring process due to the escape and release of MA/FA from the perovskite layer during the MPP test. In contrast, in the amo-FACsMA device, both MA+ and FA+ cations maintain a consistently uniform distribution before and after the stability test (Figs. 4e, f and Supplementary S27). These results are consistent with previous thin film stability results and highlight that isotropic (lysine)2PbI2 effectively suppresses the de-structuring of FACsMA perovskite in amo-FACsMA devices through stabilizing organic components, therefore leading to significantly enhanced operational stability.In summary, we have developed a new amorphous (lysine)2PbI2 enhanced and stabilized perovskites, effectively neutralizing surface/interface defects and suppressing perovskite de-structuring. Finally, the amo-FACsMA-based PSCs achieved an impressive efficiency of 26.27%, with exceptional operational stability. This work represents a breakthrough in developing highly efficient and stable perovskite materials with amorphous passivators for diverse applications, including solar cells, light-emitting diodes, and lasers.
Amorphous (lysine)2PbI2 layer enhanced perovskite photovoltaics
