Diethanolamine Modified Perovskite-Substrate Interface for Realizing Efficient ESL-Free PSCs

Simplifying device layout, particularly avoiding the complex fabrication steps and multiple high-temperature treatment requirements for electron-selective layers (ESLs) have made ESL-free perovskite solar cells (PSCs) attractive. However, the poor perovskite/substrate interface and inadequate quality of solution-processed perovskite thin films induce inefficient interfacial-charge extraction, limiting the power conversion efficiency (PCEs) of ESL-free PSCs. A highly compact and homogenous perovskite thin film with large grains was formed here by inserting an interfacial monolayer of diethanolamine (DEA) molecules between the perovskite and ITO substrate. In addition, the DEA created a favorable dipole layer at the interface of perovskite and ITO substrate by molecular adsorption, which suppressed charge recombination. Comparatively, PSCs based on DEA-treated ITO substrates delivered PCEs of up to 20.77%, one of the highest among ESL-free PSCs. Additionally, this technique successfully elongates the lifespan of ESL-free PSCs as 80% of the initial PCE was maintained after 550 h under AM 1.5 G irradiation at ambient temperature.

The subsequent focus in PSCs is on the simplified fabrication technique that offers simple and inexpensive mass production in order to become economically relevant in the context of high efficiency and reasonable stabilities [22]. In this context, numerous researchers constructed ESL-free PSCs with reasonable PCEs [23]. ESL-free PSCs have a high device stability potential in addition to decent PCEs and do not need the timeconsuming manufacturing of ESLs, particularly the numerous high-temperature processes required to create TiO 2 ESL. Since it is known that TiO 2 causes the perovskite layer to decompose by creating oxygen vacancies, the removal of the TiO 2 ESL is mostly responsible

Results and Discussion
The cleaned ITO substrates were first spin-coated with DEA at 2000 rpm for 30 s, then dried at 100 • C for 10 min. X-ray photoelectron spectroscopy (XPS) was carried out to investigate the effect of DEA treatment on the surface chemistry of ITO. The XPS survey spectra of bare ITO and DEA-treated ITO are presented in Figure 1. All spectra were calibrated using the C 1s peak of 284.9 eV binding energy as a reference point. The bare ITO and DEA-treated ITO share the same peaks except at 400.4 eV, which is assigned to N 1s, showing that the DEA successfully adsorbed on the ITO substrate, as can be seen in Figure 1a,b. Figure 1c displays the high-resolution N 1s spectra of ITO/DEA, which can be divided into three peaks at 401.4, 400.4, and 399.7 eV. Sigma-state, including species containing N, can be attributed to the peaks at 401.4 eV and 400.4 eV. Additionally, it has been stated that an amine group correlates to the peak at 399.7 eV [36,37]. The shift in the binding energy, as observed in the Sn 3d (Figure 1d) signals, is indicative of a chemical interaction between DEA and the ITO substrate. This suggests a dipole interaction between ITO and DEA that is mediated by an electric field, which lowers the attractive force of the Sn nuclei and increases the Sn atom's outer-shell electron density [38].
To investigate the impact of DEA coating on the surface morphology of ITO, scanning electron microscopy (SEM) was conducted on bare ITO and DEA-treated ITO. Figure 2a,b displays the surface SEM images of ITO and ITO/DEA, respectively. As can be seen, both surfaces are compact without pinholes. However, because the DEA layer is so thin, it is hard to see the DEA-coated ITO surface clearly. Figure S1 shows that this thin layer of DEA had little influence on ITO transparency across a wide spectrum of light, indicating a great potential for high photocurrent output. The morphologies of the bare ITO and DEA-treated ITO were investigated with the help of atomic force microscopy (AFM), as shown in Figure 2c,d. The alteration in the morphology following DEA coating is not very noticeable, similar to the SEM images. However, in the DEA-treated ITO substrate, a reasonable decrease in the root-mean-square (RMS) of surface roughness from 2.8 to Nanomaterials 2023, 13, 250 4 of 14 0.9 nm is observed. The DEA coating may considerably minimize the surface defect states and promote the deposition of high-quality perovskite thin film because of the reduced roughness. We also used Kelvin probe force microscopy (KPFM) to examine bare and DEAtreated ITO, as displayed in Figure S2. This revealed that the surface potential distribution is consistent with the aforementioned surface morphologies, and the average potential of the ITO/DEA substrate is −110 mV compared to the 0.18 mV of bare ITO, which indicates desirable Fermi-level of the ITO/DEA and better band alignment with perovskite [24,39]. To investigate the impact of DEA coating on the surface morphology of ITO, scanning electron microscopy (SEM) was conducted on bare ITO and DEA-treated ITO. Figure  2a,b displays the surface SEM images of ITO and ITO/DEA, respectively. As can be seen, both surfaces are compact without pinholes. However, because the DEA layer is so thin, it is hard to see the DEA-coated ITO surface clearly. Figure S1 shows that this thin layer of DEA had little influence on ITO transparency across a wide spectrum of light, indicating a great potential for high photocurrent output. The morphologies of the bare ITO and DEA-treated ITO were investigated with the help of atomic force microscopy (AFM), as shown in Figure 2c,d. The alteration in the morphology following DEA coating is not very noticeable, similar to the SEM images. However, in the DEA-treated ITO substrate, a reasonable decrease in the root-mean-square (RMS) of surface roughness from 2.8 to 0.9 nm is observed. The DEA coating may considerably minimize the surface defect states and promote the deposition of high-quality perovskite thin film because of the reduced roughness. We also used Kelvin probe force microscopy (KPFM) to examine bare and DEA- In order to get highly efficient PSCs in ESL-free architectures, one of the main obstacles is the energy-level mismatch between ITO and the perovskite, so it is essential to investigate how DEA affects the energy levels of ITO. Figure 3a,b illustrates the outcomes of an ultraviolet photoelectron spectroscopy (UPS) experiment. For the purpose of calculating E cutoff (secondary electron edges) and E onset (Fermi edges), the intercepts of the tangents of the peaks with the extrapolated baselines were used. By estimating the difference between the incident photon energy (light source He I, 21  In order to get highly efficient PSCs in ESL-free architectures, one of the main obstacles is the energy-level mismatch between ITO and the perovskite, so it is essential to investigate how DEA affects the energy levels of ITO. Figure 3a,b illustrates the outcomes of an ultraviolet photoelectron spectroscopy (UPS) experiment. For the purpose of calculating Ecutoff (secondary electron edges) and Eonset (Fermi edges), the intercepts of the tangents of the peaks with the extrapolated baselines were used. By estimating the difference between the incident photon energy (light source He I, 21. The EF of the ITO was significantly shifted by −0.57 eV after DEA treatment, which would be advantageous for the cathode's ability to collect electrons. This is because, following DEA treatment, a favorable dipole moment formed on the surface of the ITO. As stated in previous studies, Pb-based compounds interact with -OH groups, while the amine groups in DEA tend to link with the ITO surface [34,35]. As can be seen, the DEA The E F of the ITO was significantly shifted by −0.57 eV after DEA treatment, which would be advantageous for the cathode's ability to collect electrons. This is because, following DEA treatment, a favorable dipole moment formed on the surface of the ITO. As stated in previous studies, Pb-based compounds interact with -OH groups, while the amine groups in DEA tend to link with the ITO surface [34,35]. As can be seen, the DEA modification resulted in a −0.57 eV shift in the Fermi level of ITO, which should be attributed to a chemical reaction between DEA and ITO, as supported by the FT-IR data ( Figure S4). The appearance of the N-atom signal in Figure 1 demonstrates the interaction between the N-atom of DEA and ITO. The -NH-(from the amine group) should function as an electron donor to engage in chemisorption and modify the energy levels of ITO. To examine the surface wettability of ITO treated with DEA, the water contact angle was measured ( Figure 3c). Since DEA possesses hydrophilic -OH and amine groups, the water contact angles of the ITO surface decrease from 70.1 • to 59.6 • , showing greater wetting capabilities of DEA-treated ITO. The superior wettability can decrease the nucleation barrier and aid in the formation of a high-quality perovskite layer.  The surface morphologies and cross-sectional images of the perovskite layers prepared on bare ITO and ITO/DEA substrates were taken with help of SEM, as depicted in Figure 4. It is evident that the perovskite layers that were formed on ITO/DEA have densely packed, considerably larger grains. In order to facilitate charge transfer, the perovskite layer formed on ITO/DEA substrate displays entire coverage without any grain boundary gaps. The larger perovskite grains and improved morphology on the ITO/DEA substrate can be attributed to the favorable interaction of the -OH group with Pb 2+ [35], which promotes the formation of larger grains and passivation of the perovskite's grain boundaries. A schematic representation of the DEA interaction is shown in Figure S3a, where the -OH groups of the DEA would interact with the Pb-atoms in the perovskite thin film while the N-atom of DEA tends to create a chemical connection with the Sn atom on the ITO surface. According to other researchers, the interaction between the -OH group and Pb 2+ can improve the penetration of PbI2 into a TiO2 mesoporous layer [40]. Through the direct mixing of DEA solution and PbI2 powder, we were able to further test the possible interaction between DEA and PbI2. Intriguingly, adding PbI2 to DEA has significantly altered the appearance, likely creating new products. As seen in Figure S3b, the colorless DEA solution and the yellow PbI2 powder were combined to create an entirely distinct product that was white in color. The white mixture turns into a colorless liquid after heating ( Figure S3c). To further confirm these interactions, Fourier-transform infrared (FT-IR) spectra in the wavenumber range of 600 to 4000 cm −1 were carried out, as shown in Figure S4. The spectrum of the ITO substrate shows typical broad bands below 800 cm −1 , which is assigned to the stretching vibration of In-O or Sn-O bonds [41]. The DEA-modified ITO shows broad and strong bands below 800 cm −1 . In addition to the broad and strong bands, the DEA-modified ITO displays an additional sharp and strong The surface morphologies and cross-sectional images of the perovskite layers prepared on bare ITO and ITO/DEA substrates were taken with help of SEM, as depicted in Figure 4. It is evident that the perovskite layers that were formed on ITO/DEA have densely packed, considerably larger grains. In order to facilitate charge transfer, the perovskite layer formed on ITO/DEA substrate displays entire coverage without any grain boundary gaps. The larger perovskite grains and improved morphology on the ITO/DEA substrate can be attributed to the favorable interaction of the -OH group with Pb 2+ [35], which promotes the formation of larger grains and passivation of the perovskite's grain boundaries. A schematic representation of the DEA interaction is shown in Figure S3a, where the -OH groups of the DEA would interact with the Pb-atoms in the perovskite thin film while the N-atom of DEA tends to create a chemical connection with the Sn atom on the ITO surface. According to other researchers, the interaction between the -OH group and Pb 2+ can improve the penetration of PbI 2 into a TiO 2 mesoporous layer [40]. Through the direct mixing of DEA solution and PbI 2 powder, we were able to further test the possible interaction between DEA and PbI 2 . Intriguingly, adding PbI 2 to DEA has significantly altered the appearance, likely creating new products. As seen in Figure S3b, the colorless DEA solution and the yellow PbI 2 powder were combined to create an entirely distinct product that was white in color. The white mixture turns into a colorless liquid after heating ( Figure S3c). To further confirm these interactions, Fourier-transform infrared (FT-IR) spectra in the wavenumber range of 600 to 4000 cm −1 were carried out, as shown in Figure S4. The spectrum of the ITO substrate shows typical broad bands below 800 cm −1 , which is assigned to the stretching vibration of In-O or Sn-O bonds [41]. The DEA-modified ITO shows broad and strong bands below 800 cm −1 . In addition to the broad and strong bands, the DEA-modified ITO displays an additional sharp and strong peak at 630 cm −1 , which is ascribable to the stretching vibration of the Sn-N bond. Therefore, it is reasonable to assume that the DEA and ITO substrate were linked by a Sn-N bond, exposing the -OH groups for interaction with the perovskite layer that was later added. The DEA and PbI 2 /DEA were characterized using FT-IR spectroscopy, and the variation in -OH absorption peak was monitored. Broad absorption peaks can be seen in the DEA and DEA/PbI 2 spectra between 1052 and 1458 cm −1 , which corresponds to the scissoring vibration of the -OH group, as seen in Figure S4b. The peak locations altered in the PbI 2 /DEA sample from 1052-1458 cm −1 to 1044-1448 cm −1 , indicating a connection between PbI 2 and the -OH groups of DEA. As seen in Figure S3b, the addition of PbI 2 to DEA has significantly altered the appearance, perhaps creating new compounds. These observations point to a significant interaction between DEA and PbI 2 . In perovskite deposition, the exposed -OH groups on the ITO surface would interact favorably with the PbI 2 , which is present in the perovskite precursor solution. An even and complete coverage of the perovskite layer would benefit from this advantageous interaction, ensuring good interface contact with the ITO/DEA substrate. peak at 630 cm −1 , which is ascribable to the stretching vibration of the Sn-N bond. Therefore, it is reasonable to assume that the DEA and ITO substrate were linked by a Sn-N bond, exposing the -OH groups for interaction with the perovskite layer that was later added. The DEA and PbI2/DEA were characterized using FT-IR spectroscopy, and the variation in -OH absorption peak was monitored. Broad absorption peaks can be seen in the DEA and DEA/PbI2 spectra between 1052 and 1458 cm −1 , which corresponds to the scissoring vibration of the -OH group, as seen in Figure S4b. The peak locations altered in the PbI2/DEA sample from 1052-1458 cm −1 to 1044-1448 cm −1 , indicating a connection between PbI2 and the -OH groups of DEA. As seen in Figure S3b, the addition of PbI2 to DEA has significantly altered the appearance, perhaps creating new compounds. These observations point to a significant interaction between DEA and PbI2. In perovskite deposition, the exposed -OH groups on the ITO surface would interact favorably with the PbI2, which is present in the perovskite precursor solution. An even and complete coverage of the perovskite layer would benefit from this advantageous interaction, ensuring good interface contact with the ITO/DEA substrate. The formation of high-quality perovskite layers may improve the performance of corresponding PSCs by facilitating charge transport and reducing recombination losses [42]. In addition to morphological characteristics, X-ray diffraction (XRD) and UV-vis absorption spectra were used to study how DEA deposition affected the structure and optical performance of the perovskite layers. The high-intensity peaks from the XRD pattern of the perovskite layer formed on ITO/DEA displayed slightly greater crystallinity compared to bare ITO, as shown in Figure 5a. Additionally, the perovskite layer formed on the ITO/DEA substrate exhibits slightly greater absorption capabilities throughout the entire spectral region as evidenced by UV-vis absorbance spectra in Figure 5b. Steady-state photoluminescence (PL) and time-resolved PL (TRPL) measurements were used to analyze the charge-carrier dynamics of perovskite layers grown on glass, bare ITO, and ITO/DEA substrates, as shown in Figure 5c,d. When compared to perovskite layers formed on glass and glass/ITO, the PL spectra in Figure 5c show a considerable drop in intensity for the ITO/DEA substrate, suggesting higher electron transfer to the corresponding layer [43] and lower recombination of charges [44]. A significant quenching effect for the perovskite layer deposited on the ITO/DEA substrate can be seen in the TRPL spectra given in Figure 5d, demonstrating the improved electron extraction capability The formation of high-quality perovskite layers may improve the performance of corresponding PSCs by facilitating charge transport and reducing recombination losses [42]. In addition to morphological characteristics, X-ray diffraction (XRD) and UV-vis absorption spectra were used to study how DEA deposition affected the structure and optical performance of the perovskite layers. The high-intensity peaks from the XRD pattern of the perovskite layer formed on ITO/DEA displayed slightly greater crystallinity compared to bare ITO, as shown in Figure 5a. Additionally, the perovskite layer formed on the ITO/DEA substrate exhibits slightly greater absorption capabilities throughout the entire spectral region as evidenced by UV-vis absorbance spectra in Figure 5b. Steady-state photoluminescence (PL) and time-resolved PL (TRPL) measurements were used to analyze the charge-carrier dynamics of perovskite layers grown on glass, bare ITO, and ITO/DEA substrates, as shown in Figure 5c,d. When compared to perovskite layers formed on glass and glass/ITO, the PL spectra in Figure 5c show a considerable drop in intensity for the ITO/DEA substrate, suggesting higher electron transfer to the corresponding layer [43] and lower recombination of charges [44]. A significant quenching effect for the perovskite layer deposited on the ITO/DEA substrate can be seen in the TRPL spectra given in Figure 5d, demonstrating the improved electron extraction capability from the perovskite layer to the DEA-treated substrate. The results were fitted by a biexponential decay function with the following equation.
where τ 1 and τ 2 represent decay factors that are closely related to non-radiative recombination by defects and a component of bulk recombination, respectively. Table S1 shows from the perovskite layer to the DEA-treated substrate. The results were fitted by a biexponential decay function with the following equation.
where τ1 and τ2 represent decay factors that are closely related to non-radiative recombination by defects and a component of bulk recombination, respectively. Table S1 shows decreased values for τ1 and τ2 for the perovskite layer deposited on the ITO/DEA substrate, indicating minimized recombination losses. We fabricated PSCs with and without DEA for the evaluation of photovoltaic parameters after carefully examining the morphology and substantial role of the DEA at the ITO/perovskite interface. The schematic diagram of the as-prepared device and energy levels of each layer are depicted in Figure S5. The cross-sectional SEM images, currentvoltage (J-V) characteristic curves, and photovoltaic parameters of the as-prepared PSCs are depicted in Figure 6 and Table 1 under the same scan condition. It is noteworthy that the DEA-modified PSC exhibits significantly better performance and less hysteresis in the J-V curves. The better layer quality and improved charge-carrier transport at the ITO/DEA/perovskite interface might be primarily responsible for the improved performance and decreased hysteresis, which is consistent with the dark J-V curves, as discussed below, where we observed a decrease in the density of trap states in the DEA-modified PSCs. We fabricated PSCs with and without DEA for the evaluation of photovoltaic parameters after carefully examining the morphology and substantial role of the DEA at the ITO/perovskite interface. The schematic diagram of the as-prepared device and energy levels of each layer are depicted in Figure S5. The cross-sectional SEM images, currentvoltage (J-V) characteristic curves, and photovoltaic parameters of the as-prepared PSCs are depicted in Figure 6 and Table 1, respectively. The PSC employing bare ITO displays an open-circuit voltage (V oc ) of 1.096 V, a short-circuit current density (J sc ) of 22.94 mA cm −2 , a fill factor (FF) of 74.19%, and a power conversion efficiency (PCE) of 18.65% under reverse scan condition. In comparison, the DEA-modified PSC shows improved performance of up to PCE of 20.77%, a V oc of 1.109 V, a J sc of 24.45 mA cm −2 , and an FF of 76.60% under the same scan condition. It is noteworthy that the DEA-modified PSC exhibits significantly better performance and less hysteresis in the J-V curves. The better layer quality and improved charge-carrier transport at the ITO/DEA/perovskite interface might be primarily responsible for the improved performance and decreased hysteresis, which is consistent with the dark J-V curves, as discussed below, where we observed a decrease in the density of trap states in the DEA-modified PSCs.  To further confirm the charge-carrier dynamics and charge-carrier recombination in the devices with bare ITO and ITO/DEA, the dark J-V measurement was carried out. As seen in Figure 7a, the device with DEA has a lower current density than the device with bare ITO. This shows that the DEA/perovskite interface has efficient charge transport. In order to quantitatively assess the trap density in the as-prepared devices, we also fabricated electrononly devices (ITO/Perovskite/PCBM/Au and ITO/DEA/perovskite/PCBM/Au), as shown in Figure 7b. The trap density (Ntrap) can be determined with the equation where , , VTFL, e, and L are the vacuum permittivity, relative dielectric constant, trapfield limit voltage, elementary charge, and perovskite layer thickness, respectively. Since Ntrap and VTFL are linearly related, a device with a smaller VTFL will exhibit fewer defects overall. As seen in Figure 7b, the device with DEA exhibits a significantly lower VTFL (0.25 V) than the device with bare ITO (0.33 V), indicating DEA successfully reduces the defect density. The smooth morphology with fewer grain boundaries and the high-quality crystallinity of the perovskite layer deposited on the ITO/DEA substrate are responsible for the reduced trap density. Electrochemical impedance spectroscopy (EIS) measurements were conducted in order to learn more about how DEA affects the interfacial charge transport and recombination processes in the PSCs. Figure 7c displays the Nyquist plots of the corresponding devices with a bias voltage of 0.9 V under dark conditions and a frequency range of 1 Hz to 1 MHz. At higher frequencies, the semicircle stands in for the  To further confirm the charge-carrier dynamics and charge-carrier recombination in the devices with bare ITO and ITO/DEA, the dark J-V measurement was carried out. As seen in Figure 7a, the device with DEA has a lower current density than the device with bare ITO. This shows that the DEA/perovskite interface has efficient charge transport. In order to quantitatively assess the trap density in the as-prepared devices, we also fabricated electron-only devices (ITO/Perovskite/PCBM/Au and ITO/DEA/perovskite/PCBM/Au), as shown in Figure 7b. The trap density (N trap ) can be determined with the equation where ε 0 , ε, V TFL , e, and L are the vacuum permittivity, relative dielectric constant, trapfield limit voltage, elementary charge, and perovskite layer thickness, respectively. Since N trap and V TFL are linearly related, a device with a smaller V TFL will exhibit fewer defects overall. As seen in Figure 7b, the device with DEA exhibits a significantly lower V TFL (0.25 V) than the device with bare ITO (0.33 V), indicating DEA successfully reduces the defect density. The smooth morphology with fewer grain boundaries and the high-quality crystallinity of the perovskite layer deposited on the ITO/DEA substrate are responsible for the reduced trap density. Electrochemical impedance spectroscopy (EIS) measurements were conducted in order to learn more about how DEA affects the interfacial charge transport and recombination processes in the PSCs. Figure 7c displays the Nyquist plots of the corresponding devices with a bias voltage of 0.9 V under dark conditions and a frequency range of 1 Hz to 1 MHz. At higher frequencies, the semicircle stands in for the charge transport resistance (R ct ), which originates from the substrate and perovskite contact. Because ITO/DEA made better contact with the perovskite surface, the DEA-treated PSC showed a lower R ct value (29.76 kΩ) than the ITO-based device (R ct of 54.06 kΩ), indicating more efficient electron transfer and electron extraction at the corresponding interface. Additionally, measurements of external quantum efficiency (EQE) were made in order to assess how well the J-V and EQE curves agreed. The matching J sc values integrated from the EQE curves (Figure 7d) for both devices are consistent with J-V characteristic curves.
Here, we noticed a minor discrepancy between the EQEs of the PSCs and the absorption spectra of the perovskite layers prepared on ITO and DEA-treated ITO substrates in the range of 400-500 nm. This can be attributed to the fact that the absorption spectra do not represent an external property of the PSCs but rather an internal property of the perovskites (originating from the distinctive energy-band structure) [45][46][47]. Moreover, the EQE spectrum encompasses all excitonic and/or thermal broadening effects, along with the architectural features of the PSC, and this reflects the occupied density of states as specified in the entire PSC rather than in individual perovskites.
Nanomaterials 2023, 13, x FOR PEER REVIEW 10 of 14 charge transport resistance (Rct), which originates from the substrate and perovskite contact. Because ITO/DEA made better contact with the perovskite surface, the DEA-treated PSC showed a lower Rct value (29.76 kΩ) than the ITO-based device (Rct of 54.06 kΩ), indicating more efficient electron transfer and electron extraction at the corresponding interface. Additionally, measurements of external quantum efficiency (EQE) were made in order to assess how well the J-V and EQE curves agreed. The matching Jsc values integrated from the EQE curves (Figure 7d) for both devices are consistent with J-V characteristic curves. Here, we noticed a minor discrepancy between the EQEs of the PSCs and the absorption spectra of the perovskite layers prepared on ITO and DEA-treated ITO substrates in the range of 400-500 nm. This can be attributed to the fact that the absorption spectra do not represent an external property of the PSCs but rather an internal property of the perovskites (originating from the distinctive energy-band structure) [45][46][47]. Moreover, the EQE spectrum encompasses all excitonic and/or thermal broadening effects, along with the architectural features of the PSC, and this reflects the occupied density of states as specified in the entire PSC rather than in individual perovskites. Furthermore, we acquired performance data from 9 PSCs with bare ITO and ITO/DEA, respectively, to show the reproducibility of the devices. The ITO/DEA-based PSCs exhibit improved performance and reproducibility when compared to PSCs with bare ITO, as shown in Figure S6. In addition to the high PCEs, the PSCs should have longterm stability. Figure 8a illustrates the tracking of the steady-state PCEs of the PSCs with ITO/DEA and bare ITO under AM 1.5 G at maximum power at 0.9 V bias voltage and ambient temperature. After 120 s of continuous illumination, the ITO/DEA-based device showed a stabilized PCE of 19.71%, which is considerably near the highest PCE of 20.77% Furthermore, we acquired performance data from 9 PSCs with bare ITO and ITO/DEA, respectively, to show the reproducibility of the devices. The ITO/DEA-based PSCs exhibit improved performance and reproducibility when compared to PSCs with bare ITO, as shown in Figure S6. In addition to the high PCEs, the PSCs should have long-term stability. Figure 8a illustrates the tracking of the steady-state PCEs of the PSCs with ITO/DEA and bare ITO under AM 1.5 G at maximum power at 0.9 V bias voltage and ambient temperature. After 120 s of continuous illumination, the ITO/DEA-based device showed a stabilized PCE of 19.71%, which is considerably near the highest PCE of 20.77% for the same PSC. Additionally, for 550 h of AM 1.5 G irradiation, the stabilities of the unsealed PSCs made with bare ITO and ITO/DEA substrates were evaluated every 34 h. After each stability test, the tested PSCs were kept in a glass oven at 60% humidity and 25 • C. Figure 8b demonstrates how the stability of the ITO/DEA-based PSCs exceeded that of the ITO-based devices. After 550 h, the ITO/DEA-based PSC still had 80% of the original PCE. The bare ITO-based PSCs under the same testing conditions only retained 37% of the initial PCE, demonstrating poor stability and quick degradation. The degradation of ITO-based PSCs is associated with rapid erosion because the perovskite layer formed on the bare ITO substrate has a defective surface morphology. As mentioned above, high-quality perovskite layers were created on the DEA-treated ITO substrates, with compact, large grains, and reduced grain boundaries. Starting from the grain boundaries makes it considerably easier to decompose perovskite films. Less oxygen, water, etc. will intermix within the grain boundaries of the perovskite films because there are fewer grain boundaries. The stability of PSCs made using DEA-treated ITO substrates is ultimately improved by the perovskite film with fewer grain boundaries, which effectively shields perovskite films from water penetration. Conversely, erosion occurs more rapidly in the perovskite thin films generated using the ITO substrates because these have a considerably rougher surface with smaller grain sizes and numerous grain boundaries.  Figure 8b demonstrates how the stability of the ITO/DEA-based PSCs exceeded that of the ITO-based devices. After 550 h, the ITO/DEA-based PSC still had 80% of the original PCE. The bare ITO-based PSCs under the same testing conditions only retained 37% of the initial PCE, demonstrating poor stability and quick degradation. The degradation of ITO-based PSCs is associated with rapid erosion because the perovskite layer formed on the bare ITO substrate has a defective surface morphology. As mentioned above, highquality perovskite layers were created on the DEA-treated ITO substrates, with compact, large grains, and reduced grain boundaries. Starting from the grain boundaries makes it considerably easier to decompose perovskite films. Less oxygen, water, etc. will intermix within the grain boundaries of the perovskite films because there are fewer grain boundaries. The stability of PSCs made using DEA-treated ITO substrates is ultimately improved by the perovskite film with fewer grain boundaries, which effectively shields perovskite films from water penetration. Conversely, erosion occurs more rapidly in the perovskite thin films generated using the ITO substrates because these have a considerably rougher surface with smaller grain sizes and numerous grain boundaries.

Conclusions
We used DEA to generate Sn-N and -OH-Pb chemical linkages between the perovskite layer and the ITO substrate. The interface contact, perovskite crystallinity, and film quality have all increased due to these chemical interactions, and the perovskite layer's grain boundaries and trap states have decreased. Based on the creation of interfacial dipole layers, our results imply that the addition of DEA can modify the band misfit at the ITO/perovskite contact, hence reducing the voltage deficits. Transient PL, TRPL, and IS measurements revealed that the rates of electron extraction and charge transfer have dramatically improved. The PCE of the PSCs has significantly increased as a result, rising from about 18.65% to 20.77%. Additionally, DEA modification has increased the stability of the PSCs, with 80% PCE retention even after 550 h without encapsulation. This is likely because the DEA modification improved the interfacial contact and thin-film quality of the perovskite. Our research opens up scientific pathways for the facile fabrication of ESLfree PSCs.
Supplementary Materials: The following supporting information can be downloaded at: www.mdpi.com/xxx/s1, Figure S1: Transmittance spectra of bare ITO and DEA-treated ITO substrates; Figure S2: KPFM images (a,b) and corresponding potential difference curves (c,d) of the bare

Conclusions
We used DEA to generate Sn-N and -OH-Pb chemical linkages between the perovskite layer and the ITO substrate. The interface contact, perovskite crystallinity, and film quality have all increased due to these chemical interactions, and the perovskite layer's grain boundaries and trap states have decreased. Based on the creation of interfacial dipole layers, our results imply that the addition of DEA can modify the band misfit at the ITO/perovskite contact, hence reducing the voltage deficits. Transient PL, TRPL, and IS measurements revealed that the rates of electron extraction and charge transfer have dramatically improved. The PCE of the PSCs has significantly increased as a result, rising from about 18.65% to 20.77%. Additionally, DEA modification has increased the stability of the PSCs, with 80% PCE retention even after 550 h without encapsulation. This is likely because the DEA modification improved the interfacial contact and thin-film quality of the perovskite. Our research opens up scientific pathways for the facile fabrication of ESL-free PSCs.
Supplementary Materials: The following supporting information can be downloaded at: https:// www.mdpi.com/article/10.3390/nano13020250/s1, Figure S1: Transmittance spectra of bare ITO and DEA-treated ITO substrates; Figure S2: KPFM images (a,b) and corresponding potential difference curves (c,d) of the bare ITO and DEA-modified ITO substrates; Figure S3: An strengthened link at the ITO/perovskite interface is highlighted by a coupling between the -OH group and Pb in the perovskite in this schematic example of surface modification of ITO-substrate with DEA (a), Photos of DEA liquid, PbI 2 powder and PbI 2 /DEA mixture (b), Photos of DEA liquid, PbI 2 powder and PbI2/DEA mixture under 135 • C (c); Figure S4: FT-IR spectroscopy of the ITO and ITO/DEA substrates (a), and DEA liquid and PbI2/DEA mixture (b); Figure S5: Schematic illustration of as-prepared PSC (a) and energy band diagrams of the corresponding layers (b); Figure S6: Statistical analyses of the ESLs-free PSCs prepared with bare ITO and DEA-modified ITO substrates, V oc (a), J sc (b), FF (c), and PCE (d); Table S1: TRPL data for the perovskite layers prepared on bare ITO and DEA-modified substrates.  Data Availability Statement: All the data presented in the manuscript can be obtained from the corresponding authors by reasonable request.