Improvement of film quality and solar cell properties of perovskite by the addition of N-benzylhydroxylamine

Perovskite solar cells (PVSCs) can be readily manufactured through printable processes ^1–3) and can have power conversion efficiencies (PCEs) as high as 26.1%, ⁴⁾ which is comparable to that of silicon solar cells. Consequently, they are expected to be excellent power generation devices. ^5–7)

The use of perovskite layers as light absorbers in PVSCs is crucial; however, the numerous defect states in the perovskite crystal structure ⁸⁾ and the phase transitions that occur during operational conditions pose some difficulties in enhancing their reliability. ⁹⁾

The structure of perovskite crystals is represented by ABX₃ and various compositions have been investigated to stabilize its crystal structure and adjust the energy gap (Eg). ¹⁰⁾ Among these, the most widely reported compositions of PVSCs that exhibit a high PCE feature methylammonium (MA⁺) as the A-site ion. ^11,12) However, during the formation of the perovskite layers, heat treatment can readily remove MA⁺ from the crystal structure. ^13,14) Therefore, compositions in which the A-site cation is replaced by a high molecular weight formamidinium (FA⁺) have been introduced to enhance stability. ¹⁵⁾ In addition, since FAPbI₃ readily forms a photo-inactive δ-phase with a large Eg at room temperature, a method to stabilize the crystal structure by introducing bromide (Br⁻) at the X site has been implemented. ¹⁶⁾ However, the introduction of Br⁻ causes a blue shift in Eg, which is undesirable from the perspective of the Shockley–Queisser limit. ¹⁷⁾

Furthermore, the easy formation of lead iodide (PbI₂), which has a low optical absorption coefficient, via ion migration and desorption from the perovskite surface poses a challenge in achieving high reliability. ¹⁸⁾

Several other methods, including the addition of diverse amine compounds and organic cations, such as phenethylammonium, ^19,20) guanidine, ²¹⁾ and triazoles, ²²⁾ have been reported to function as defect passivation agents for the perovskite crystals. These compounds improve the film quality of the perovskite and device performance by passivating surface defects and enhancing the charge transfer characteristics, photostability, and trapping state. Additionally, a low-dimensional perovskite layer can be used to improve the operational stability. ²³⁾ However, the introduction of amine materials with large-size molecules may cause other issues, such as increased insulation, and adversely impact the band gap and crystal structure.

This study aims to improve the stability of the perovskite layers without affecting the Eg and crystal structure of the perovskite by introducing N-benzylhydroxylamine (N-BzHoA) as an additive, because the hydroxyl group of N-BzHoA exhibits stronger hydrogen bonding properties than the amino group. In our previous study, we investigated the effects of methoxyamine and hydroxylamine additives with alkyl substituents on the additive concentration and its effect on the perovskite films. The results show that an additive concentration of 1% improved the perovskite crystal structure and trap states density, resulting in improved PCE and lifetime. ²⁴⁾ Furthermore, the benzyl group adsorbs iodine and prevent halogen ion desorption. ²⁵⁾

In this study, the effect of the additive was verified by introducing 1% of the N-BzHoA into the FAPbI₃ precursor solution. The properties and stability of the perovskite thin film were evaluated and the corresponding inverted PVSC was fabricated to evaluate its photovoltaic properties and reliability. The fabricated device structure was composed of glass/ITO/[2-(3,6-Dimethoxy-9H-carbazol-9-yl)ethyl]phosphonic acid (MeO-2PACz)/perovskite/C₆₀/bathocuproine (BCP)/Ag. MeO-2PACz (TCI, >98% purity) was dissolved in 2-propanol (IPA) at 2 mM. The perovskite precursor solution (FAPbI₃) was adjusted to 1.48 M in DMF:DMSO (4:1, volume ratio), followed by stirring overnight. In separate glass vials, 1.48 M of potassium iodide (KI, Sigma-Aldrich, purity: 99%) and N-BzHoA (Sigma-Aldrich, purity: 99%) were prepared and stirred in the same manner. KI was introduced to suppress crystal defects. ²⁶⁾ The target devices were prepared by adding 5% and 1% (molar ratios) of KI and N-BzHoA, respectively, to the perovskite precursor solution.

ITO-coated glass substrates were treated with ultraviolet (UV) ozone for 20 min before use. MeO-2PACz was spin-coated onto the ITO substrate at a rate of 3000 rpm for 30 s and then annealed at 100 °C for 10 min on a hot-plate in a nitrogen-purged glove box. A 0.45 μm polytetrafluoroethylene (PTFE) filter was used to filter the perovskite precursor solution, and the filtered solution was then spin-coated onto the MeO-2PACz layer (1000 rpm, 10 s, followed by 6000 rpm, 30 s). After the first 15 s of spin-coating, chlorobenzene was dropped as an anti-solvent. C₆₀, BCP, and Ag were formed through vacuum evaporation, with film thicknesses of 20 nm, 8 nm, and 70 nm, respectively. The devices were encapsulated using glass caps and UV-curable resin before the measurements.

We fabricated and characterized single thin films and cells both with and without introducing N-BzHoA as an additive into the perovskite precursor solution. Figure 1 shows the device structure of the inverted perovskite solar cell ²⁷⁾ and the molecular structure of N-BzHoA, which was investigated as an additive. Additionally, Fig. S1 shows the energy diagram of the fabricated device. UV–Vis absorption spectroscopy, X-ray diffraction (XRD), Fourier transform IR spectroscopy (FTIR), scanning electron microscopy (SEM), and atomic-force microscopy (AFM) were used for single film evaluation. For device characterization, current density-voltage (J-V) characteristics were measured to evaluate the effects of the N-BzHoA on the PCE. The stability of the PVSC samples was evaluated under continuous 1-sun illumination with solar simulated light at room temperature.

Zoom In Zoom Out Reset image size

The UV–Vis absorption spectra are shown in Fig. 2, wherein it is evident that the absorption intensity of the film increased with the addition of N-BzHoA. However, the position of the absorption-edge wavelength remained unchanged between the control and the N-BzHoA-added films. In Fig. S2, Eg is confirmed by the Tauc plot based on the UV–Vis absorption results, with a value of 1.53 eV for both the control and N-BzHoA-added films. These results indicate that the introduction of N-BzHoA does not influence the Eg of the perovskite itself.

Zoom In Zoom Out Reset image size

XRD measurements were conducted to investigate the effect of N-BzHoA on the crystal structure of perovskite. Figure 3(a) shows that the control film has a small δ-phase-derived peak and a large PbI₂-derived peak, indicating that the α-phase perovskite cannot be preferentially formed. As the δ-phase perovskite and PbI₂ have a large Eg and are less photoactive, ²⁸⁾ a process that can preferentially form black, photoactive α-phase perovskite films is desirable. By introducing N-BzHoA, the peaks originating from PbI₂ and the δ-phase perovskite were suppressed, while those originating from the α-phase perovskite were enhanced significantly. These results suggest that the addition of N-BzHoA contributes to the formation of an ideal perovskite crystal structure.

Zoom In Zoom Out Reset image size

Figure 3(b) shows the SEM images of the surfaces of the control and N-BzHoA-added perovskite films. In the control film, numerous small white grains that appear to be PbI₂ are visible, which is consistent with the XRD results. In contrast, the amount of PbI₂ is dramatically reduced on the surface of the film treated with N-BzHoA, and the grain size of the perovskite is increased, which is consistent with the UV–Vis absorption spectra. These results suggest that the incorporation of N-BzHoA contributes to the formation of an ideal perovskite crystal structure and is effective in improving its photovoltaic properties. ²⁹⁾ In addition, the AFM results in Fig. 3(c) show a comparison of the roughness of the perovskite surface. The perovskite surface should be smooth to ensure a uniform contact with the electron transport layer and more efficient carrier transport. ³⁰⁾ The surface roughness of the control film is large (RMS = 38 nm), whereas that of the N-BzHoA-added film exhibits a modified surface with an RMS of 16.4 nm. These results indicate that the introduction of N-BzHoA into the perovskite precursor solution is an effective method for achieving both a large grain size and uniform surface state.

Figure 4(a) and Table I present the photovoltaic characteristics of the inverted PVSCs fabricated with the control and N-BzHoA-added precursor solutions. The control device exhibits a PCE of 14.81%, whereas that with N-BzHoA addition exhibits improved properties, with a J_sc of 21.55 mA cm⁻², a V_oc of 1.06 V, a fill factor of 0.76, and a PCE of 17.49%. Furthermore, the device with N-BzHoA exhibits significantly reduced hysteresis in the current density-voltage characteristics between scan directions.

Zoom In Zoom Out Reset image size

Figure 4(b) shows the results of the lifetime evaluation under continuous light exposure. Although the PCE of the control device decayed to approximately 30% of its initial value after 100 h, the N-BzHoA-added device retained a PCE of more than 89%. We believe this is because the introduction of benzyl groups increased the hydrophobicity of the perovskite film, which prevented the ingress of moisture and increased stability. These results emphasize that the introduction of N-BzHoA is an effective method for improving the stability of inverted PVSCs.

Figure S3 shows the thermal stabilities obtained through XRD measurements, before and after heating (200 °C, 1 h). It can be confirmed that, compared to the control film, the addition of 1% N-BzHoA prevented the occurrence of PbI₂-derived peaks, resulting in improved thermal stability.

To further verify the versatility of N-BzHoA, we investigated its synergistic effect with methylammonium chloride (MACl), which is used to increase perovskite grain size. ³¹⁾ As shown in Fig. 4(c), the simultaneous addition of MACl was attempted, and PCE was found to be limited in devices in which only MACl was added to FAPbI₃. In contrast, devices in which both MACl and N-BzHoA were added to the perovskite precursor solution exhibited a significantly improved PCE of 17.01%. The SEM images in Fig. S4 show that the perovskite film with only MACl has voids and white grains that appear to be PbI₂, whereas that with both MACl and N-BzHoA has uniform and larger grain sizes. These results suggest that N-BzHoA is a versatile additive.

Additionally, a hole-only device (ITO/MeO-2PACz/perovskite/PTAA/Ag) was evaluated, as shown in Fig. 4(d), and the defect density of the perovskite film was calculated as follows:

$$\begin{eqnarray}&&{{\rm{N}}}_{{\rm{defects}}}=\frac{2\varepsilon {\varepsilon }_{0}{{\rm{V}}}_{{\rm{TFL}}}}{{\rm{q}}{{\rm{L}}}^{2}}\end{eqnarray} \tag{ 1 }$$

where N_defects, ε, ε₀, V_TFL, q, and L correspond to the trap density, relative permittivity, vacuum permittivity, trap filling limit voltage, elementary charge, and perovskite layer thickness, respectively. The V_TFL value, which was obtained at the intersection of the TFL and ohmic regions, of the control device was 0.73 V and the defect state density was 2.36 × 10¹⁶ cm⁻³. In contrast, the N-BzHoA-modified device had a V_TFL value of 0.38 V and decreased defect state density of 1.23 × 10¹⁶ cm⁻³. ^32,33) These results indicate that the introduction of N-BzHoA as an additive contributes to the increase in grain size and the formation of a preferential α-phase crystal structure, and is thus effective in reducing the trap state density.

Figure 5 shows the results of the FTIR analysis, which helps to elucidate the mechanism by which the introduction of N-BzHoA contributes to the enhanced PVSC performance. The following peaks can be observed: 3340 cm⁻¹ (N–H and O–H stretching vibrations), 1720 cm⁻¹ (=NH), and 1530 cm⁻¹ (C=C). ³⁴⁾ The introduction of N-BzHoA into the precursor solution shifted each peak to a lower wavenumber region. These results suggest that the formation of OH⋯I or NH⋯I hydrogen bonds between N-BzHoA and perovskite can passivate the defect state, contributing to its enhanced PCE and crystallinity. ³⁵⁾

Zoom In Zoom Out Reset image size

In conclusion, the introduction of a small amount of N-BzHoA as an additive into the FAPbI₃ perovskite precursor solution preferentially formed the α-phase as well as promoted an increase in grain size and decreased surface roughness. The fabricated device exhibited improved trapping states as well as enhanced V_oc and FF values, resulting in a PCE of 17.49%. The synergistic effect observed when combining N-BzHoA with MACl demonstrates its versatility as an additive in PVSCs. This study's findings also provide avenues for broader applications and advancements in perovskite-based solar technologies, with N-BzHoA potentially contributing to the development of more stable and robust PVSCs.

Acknowledgments