On July 21, 2022, the team of Professor Liu Shengzhong of the Dalian Institute of Chemical Physics of the Chinese Academy of Sciences published an article titled “Surface redox engineering of vacuum-deposited NiOx for top-performance perovskite solar cells and modules” in the journal Joule.
The surface redox engineering proposed in this work realizes the design idea of combining the vacuum-prepared charge transfer layer with the perovskite prepared in solution, and promotes the development of efficient and stable perovskite battery components.
The co-first authors of the work are Du Minyong, a doctoral student at the Dalian Institute of Chemicals, and Dr. Shuai Zhao of Chongqing University of Technology. The co-corresponding authors are Researcher Liu Shengzhong and Associate Researcher Wang Kai of Dalian Institute of Chemicals.
Background and brief results
At present, the photoelectric conversion efficiency of laboratory-size perovskite solar cells has reached more than 25%, and the preparation of large-area perovskite cells and the promotion of their industrialization process have become one of the main development directions in this field. Although one of the important advantages of perovskite batteries is that they are compatible with the solution preparation route, in the preparation process of large-area devices, only one hundred nanometers thick perovskite layer is suitable for the preparation of the solution method, while the thickness of the charge transfer layer is only a few tens of nanometers, and it is difficult to prepare a uniform and non-porous large-area charge transport layer by the solution method at this stage. In contrast, vacuum deposition technology is more controllable and more suitable for the preparation of ultra-thin large-size films. Therefore, Professor Liu Shengzhong’s team proposed a strategy to construct a large-area battery by combining a vacuum-prepared charge transfer layer and a perovskite layer prepared by solution method. However, it was found in the study that the surface of the vacuum-prepared nickel oxide hole transport layer was relatively hydrophobic, which weakened the adhesion of the perovskite precursor, and a large number of high-priced nickel ions present on the surface of nickel oxide would decompose the perovskite, forming an interfacial barrier and causing a non-capacitive lag effect, which ultimately affected the performance and stability of the device.
In response to the above problems, the team proposed a simple Surface Redox Engineering (SRE) to regulate the surface properties of electron beam evaporation nickel oxide films. The results show that SRE effectively improves the surface wettability of nickel oxide and ensures the compatibility between vacuum preparation of nickel oxide and the preparation of perovskite in solution. At the same time, the proportion of nickel ions in different valence states on the surface of nickel oxide film is reasonably adjusted, which improves the electrical performance of nickel oxide/perovskite interface and improves the interface stability. In the end, the small area trans batteries prepared on the rigid and flexible substrates achieved a photoelectric conversion efficiency of 23.4% and 21.3%, respectively, and had excellent stability. In addition, based on the earlier work, the research team successfully prepared a large-area perovskite battery module on a large-area substrate with an area of 156× 156 mm2, with an energy conversion efficiency of 18.6% and excellent stability. The surface redox engineering proposed in this work realizes the design idea of combining the vacuum-prepared charge transfer layer with the perovskite prepared in solution, and promotes the development of efficient and stable perovskite battery components.
Results and analysis
Figure 1A depicts the experimental process of SRE, in which the NiOx film first undergoes an argon plasma treatment to enhance surface wettability, a process known as NiOx’s surface oxidation engineering (SOE). It is worth noting that both EBE and SRE are performed at room temperature.
The authors used X-ray photoelectron spectroscopy (XPS) spectroscopy to elucidate changes in the surface composition of NiOx films. It is worth noting that the composition of the NiOx film is complex, in addition to the main NiO, Ni3+ is conducive to semiconductor properties, and the unstable Ni4+ will corrode the perovskite layer due to its strong oxidation properties, and -OH may trigger the deproton reaction of perovskite. Figure 1B shows the variation in the different nickel ion content. High-energy argon plasma initiates oxidation from the low-cost state of NiO and Ni(OH)2 to the high-valent state of Ni≥3+. Inferring that high-energy particle bombardment will dissociate the chemical bonds between O2- and Ni2+ in the lattice and cause atomic displacement, which will easily produce Ni2+ vacancy defects under argon plasma conditions. An Ni2+ vacancy causes the formation of two Ni3+ ions or one Ni4+ ion to maintain charge neutrality, which results in a significant increase in the composition of high-valent oxides on the surface of the film. To support this inference, the authors predicted the formation energy of Ni2+ and O2-vacancies using density functional theory (DFT) calculations. Note that unlike other oxides such as TiO2, ZnO, MoOx, WOx, and SnO2, Ni vacancies are a common defect in NiO, as evidenced by their formation energy below O vacancies under Ni-rich or O-rich conditions (Figure 1C). In addition, Ni2+ ions released by bombardment may escape into a vacuum or be adsorbed near surface defects, and then react with O2 or H2O in air to form NiO or Ni(OH)2.
The above hypothesis is very consistent with the increased O/Ni ratio during soE treatment on the surface of the NiOx film, and for the problem of reduced Ni(OH)2 content, the energy of a dozen electron volts in the argon plasma is enough to convert Ni(OH)2 to NiOOH because they have the same crystal structure. Therefore, argon plasma treatment can not only increase surface wettability, but also cause surface oxidation processes and form Ni4+ ions harmful to the perovskite layer. When the NiOx surface was further modified with Brønsted acid, the content of Ni(OH)2 and Ni4+ decreased significantly, while the content of NiO and Ni3+ remained almost unchanged. Thus, the acid can selectively eliminate the harmful Ni4+ and -OH, but retain the beneficial Ni3+ and NiO body parts. To explain the screening behavior of acids, the reaction kinetics of different valence nickel compounds and acids were tracked by ultraviolet-visible absorption spectroscopy, where NiO2, Ni2O3, NiO and Ni(OH)2 powders were used as different valence nickel compounds reacting with acids. As shown in Figure 1D, the order of chemical reactivity is Ni4+ > Ni(OH)2 >> Ni3+ > NiO. The authors also observed that NiO2 is inert in water, while it reacts violently with acids and produces a large number of bubbles. The authors reasonably infer that Ni4+ in the lattice will first be converted to free Ni4+ ions in an acid solution, and then oxidized the water to O2, accompanied by the formation of Ni≤3+. In addition, Ni(OH)2 is converted to Ni(NO3)2 in the Brønsted acidic environment.
Figure 1: The impact and mechanism of the SRE policy. (A) Schematic diagram of the SRE policy. (B) Changes in Ni species caused by SRE. (C) Under ni-rich and O-rich conditions, the formation of Ni and O vacancies can change with Fermi energy. (D) Ultraviolet-visible absorption spectrum of NiO2/Ni(OH)2/Ni2O3/NiO powder in HNO3 solution with a concentration of 3M (X axis is the wavelength). (E) state density, (F) optimized NiO surface structure, and (G) difference in average charge density of NiO surfaces with NO3-adsorbed.
Thus, Brønsted acids are able to guide the reduction reaction of NiOx after SOE treatment, selectively eliminating harmful Ni4+ and neutralizing surface hydroxyl groups. In addition, the Ni3+/Ni2+ ratios of blank group, SOE, and SRE samples were 0.87, 1.73, and 2.08, respectively. A significant increase in the Ni3+/Ni2+ ratio in NiOx necessarily means that SRE-NiOx films have enhanced p-type semiconductor characteristics. In addition, according to the DFT calculation, the adsorption energy of NO3- on the Surface of NiO is expected to be -2.90 eV, indicating that it has good energy adsorption performance. In summary, NiOx’s SRE has three functions: reduced surface tension, stabilized surfaces, and enhanced surface electrical properties.
To reveal the mechanism of effects of SRE treatment on the preparation of perovskite films on NiOx films, the authors prepared perovskite absorber layers on NiOx substrates by slit coating deposition, and all subsequent perovskite films in this study were prepared based on this method. The contact angle experiment found that the contact angle of the droplet on the untreated NiOx film was about 40.6°, which obtained a uniform perovskite layer by spin coating, but was not suitable for slit coating deposition. After SOE or SRE, the contact angle of the substrate is reduced to about 7°, indicating an increase in the wetting of the substrate, which improves the adhesion between the perovskite solution and the underlying NiOx surface, which in turn promotes 100% coverage of the perovskite precursor solution. In addition, depending on the association between the free energy (G) between heterogeneous nucleation and homogeneous nucleation, wettability will affect the crystallization of perovskite films:
where θ is the contact angle at the solid/liquid interface. The smaller the θ of the perovskite solution on SRE-NiOx, the lower the barrier of heteronucleation, which is conducive to the formation of a uniform, high-quality perovskite film. Figure 2 shows a surface SEM image of perovskite films on various substrates. The perovskite film on the original NiOx film has a large number of pinholes at the grain boundaries, which becomes a channel for charge recombination and degrades device performance. In contrast, both SOE and SRE make perovskite films dense, pinholes significantly reduced, grain size increased, and defect density reduced, which are prerequisites for enabling high-performance devices.
Figure 2: Top view SEM image of perovskite films prepared on blank, SOE, and SRE-treated NiOx films. On the right is the corresponding grain size histogram of SEM image statistics.
Based on the advantages of SRE surface treatment and the significant improvement of perovskite crystals, the authors used the slit coating method to prepare small-area trans PSCs on the electron beam NiOx film substrate. The structure is FTO/NiOx/perovskite/PCBM/BCP/Ag, as shown in the SEM cross-section of the 3A device in Figure 3A. The author systematically compares the performance of devices without any processing (Control), argon-treated devices (SOEs), and SRE-treated devices. The mean PCE of PSCs in the blank group, SOE, and SRE was 18.0±0.9%, 20.3±0.4%, and 22.7±0.4%, respectively. The continuous improvement in performance is attributed to stable surfaces after SRE treatment, better electrical performance and band matching, and better perovskite polycrystalline films. Figure 3B shows the current density-voltage (J-V) curve of the optimal SRE-PSC with a reverse-scan PCE of up to 23.4%, a VOC of 1.16 V, a JSC of 24.8 mA cm-2, and a FF of 81.4%, with negligible backlash (hi of approximately 2.8%), and a steady-state efficiency of 22.7% at the maximum power point (MPP) output at a bias voltage of 0.96 V (Figure 3C). Compared with the niOx perovskite devices currently prepared by various large-area preparation techniques, the NiOx-based perovskite devices in this study have the highest photoelectric conversion efficiency (Figure 3D). Since all functional layers are prepared at temperatures below 110 °C, the entire preparation process can be easily transferred from rigid substrates to flexible substrates. In this study, the authors used PET/ITO substrates to prepare flexible perovskite solar cellsThe device and a 90 nm antireflective layer -magnesium fluoride was prepared on the inflection surface to increase the photocurrent. Finally, the maximum PCE obtained on the flexible perovskite device was 21.3%, with a VOC of 1.09 V, a JSC of 24.5 mA cm-2, and a FF of 79.7%, and showed a small hysteresis effect (Figure 3B). In addition, the authors further characterize the external quantum efficiency (EQE) of the device, as shown in Figure 3E. The integrated current density of the EQE reaches 24.1 mA cm-2, which is almost the same as the value obtained by the J-V curve.
Based on the SRE strategy system, the authors studied the stability of electron beam NiOx trans perovskite devices. First, the authors studied the stability of unencapsulated perovskite solar cells in the laboratory atmosphere (25 °C, relative humidity <20%). As shown in Figure 3F, after aging for about 1100 hours, the SRE-PSC retains 90% of the initial PCE, while the efficiency of the blank device drops to 83% of the initial value. Subsequently, the authors continuously tested the operational stability of the perovskite device at the maximum output power point (MPP) under a standard sunlight. The study found that the T90 of the SRE-PSC has a lifespan of more than 1300 hours and exhibits excellent long-term operational stability, as shown in Figure 3G. In addition to aging in the lab, the authors further investigated degradation behavior under real outdoor conditions of extreme weather and climate change. As shown in Figure 3H, the packaged SRE device ages under outdoor conditions (Dalian, winter, January to February) and experiences a variety of weather conditions, including light/dark, low temperatures, rain, and snow. During the 41 natural daily cycles (about 1000 hours), PSCs degraded by 20%.
Figure 3: Performance of perovskite solar cells prepared by SRE strategy. (A) SEM cross-sectional image of PSC. (B) J-V curve of the best rigid/flexible device. (C) Steady-state power and current output of rigid perovskite devices. (D) PCE of NiOx-PSCs prepared using scalable techniques. (E) EQE curve. (F-H) Long-term stability of the device measured under various conditions.
The authors tested the time-resolved PL spectra of perovskite films on NiOx with or without SRE surface treatment, as shown in Figure 4A. Data analysis showed that perovskite films on SRE-NiOx showed shorter decay times (135 ns vs.360 ns) than the blank group. This suggests that SRE effectively accelerates interfacial cavity transfer kinetics. According to the dark state J-V curve of Figure 4B, the trap fill limit voltage (VTFL) of the blank group device and the SRE processor device are 0.20 V and 0.37 V, respectively. Therefore, the trap density of the SRE device is about 1.70 × 1015 cm-3, which is relatively lower than the trap density of the original device (about 3.14 × 1015 cm-3). Figure 4C shows the capacitance-to-voltage curve of the blank group and the SRE-processed PSC. The built-in electric field (Vbi) evaluated by Mott-Schottky analysis increased from 1.11 V in the blank group to 1.21 V in the SRE-PSC. The enlarged Vbi will not only accelerate the extraction of photogenerating carriers, but also promote the construction of extended depletion regions, thereby effectively inhibiting non-radiative recombination.
In addition, the authors tested electrochemical impedance spectra (EIS) of Different Surface Treatments of NiOx-based Perovskite Devices to track the effects of different surface treatments on charge transfer processes in perovskite devices. After SRE strategy treatment, perovskite devices have a larger composite resistance. Accordingly, the authors studied Rric with or without SRE surface treatment of perovskite devices under different bias voltage conditions, as shown in Figure 4D. It can be seen that under different bias voltage conditions, the PSCs treated by the SRE strategy all show a larger composite resistance than the blank group, indicating that the SRE strategy effectively inhibits the charge recombination process.
To obtain the relationship between trap density and trap depth, the authors performed a thermal conductivity spectroscopy of the device with or without SRE treatment, as shown in Figure 4E. Blank devices display relatively large tDOS values, ranging from approximately 1016 to 1020 cm-3 eV-1. After SRE processing, the defect density of the device is more than an order of magnitude lower than 0.5 eV, indicating that SRE can effectively reduce the trap density of the deep energy level at the grain boundary and interface. In addition, the authors measured the dependence of VOCs on light intensity (I) to analyze composite behavior. Figure 4F shows that voCs vary linearly with the logarithm of I. The deviation between the slope and kT/q (where k, T, and q are boltzmann constants, temperature, and elemental charge, respectively) is positively correlated with the Shockley–Read–Hall composite. As can be seen from the small slope, the SRE device significantly suppresses trap-assisted recombination, which is consistent with the increase in VOCs and FF.
Figure 4: Interaction effect of SRE on perovskite films and devices. (A) Time-resolved PL attenuation curves of perovskite films on FTO glass, blank group and SRE-NiOx films. (B) Dark current-voltage curve of pure hole devices based on blank and SRE-NiOx thin films. (C) Mott-Schottky scatterplot. (D) The relationship curve of Rrec to different bias pressures. (E) Thermal conductivity spectra of PSCs with or without SRE treatment and (F) VOC vs. variation in light intensity.
Based on the above SRE strategy and large-area slit coating deposition method, the authors prepared perovskite solar cell modules on an electron beam NiOx substrate with an area of 156×156 mm2. The large-area perovskite solar cell module in this study consists of 20 7 mm wide sub-cells in series, with a dead zone width of about 1 mm, a width of P1, P2 and P3 and a distance of adjacent lines of about 200 μm, and a geometric factor of ≥ 85.7%. As shown in Figure 5A, the PCE of the best perovskite solar cell module is 18.6% (VOC= 20.74 V; ISC=198.9 mA； FF=78.4%), which has an effective area of 174 cm2, and the I-V curves measured by reverse and forward scans almost overlap, indicating that the hysteresis effect of this component is negligible. For perovskite solar cell modules with optimal efficiency, the VOC and JSC of a single cell module calculated are 1.04 V and 22.9 mA cm-2, respectively, based on the area of their submodules (8.7 cm2). Figure 5B shows a perovskite solar cell module with a steady-state output efficiency of 18.2% at the maximum power point. In addition, the authors studied the stability of perovskite solar cell modules. In the stability study, a composite electrode structure of chromium (20 nm) and gold (130 nm) was used. As shown in Figure 5C, perovskite solar cell modules show excellent long-term stability, maintaining 80% of the initial performance after aging for 1500 hours in a dark atmosphere. In addition, perovskite solar cell modules show excellent process repeatability, as shown in Figure 5D. Among them, about 21% and about 86% of the device PCE exceeded 18.0% and 17.0%, respectively. Excellent battery module efficiency and stability result from good uniformity of the functional layer, low interface trap density, and suppressed interface compounding processes.
Figure 5: Performance of perovskite solar cell modules. (A) I-V curve of the optimal battery assembly measured in the reverse and forward scan directions. (B) Steady-state power output of the battery assembly at a fixed bias of 17.6 V. (C) Long-term stability of both modules aged in the environment. (D) PCE distribution of 14 battery modules.
The authors propose a simple SRE strategy that can not only obtain uniform perovskite films on vacuum-prepared NiOx films, but also improve the performance and stability of the interface between NiOx and perovskites. Theoretical and experimental results reveal changes in surface composition caused by redox reactions during SRE. After that, SRE significantly improved carrier transport and reduced trap-assisted compounding processes in PSCs, increasing the performance of rigid and flexible devices to 23.4% and 21.3%, respectively, while achieving long-term stability of more than 1000 hours under a variety of conditions. Notably, the SRE strategy was further extended to perovskite solar cell modules, showing excellent photovoltaic performance on substrates of 156×156 mm2, with a PCE of 18.6%. In general, the SRE strategy is conducive to the preparation of high-quality perovskite films using large-area preparation technology – slit coating method – on the vacuum preparation of NiOx films, which is conducive to promoting the commercialization process of efficient and stable perovskite solar cell modules.
This work has been funded by the Chinese Academy of Sciences Class A pilot project “Adjacent Space Science Experimental System”, the National Natural Science Foundation of China, the Innovation Fund of Dalian Institute of Chemicals and other projects. (Source: Science Network)
Related paper information:https://doi.org/10.1016/j.joule.2022.06.026