Monovalent Copper Cation Doping Enables High-Performance CsPbIBr2-Based All-Inorganic Perovskite Solar Cells

Organic–inorganic perovskite solar cells (PSCs) have delivered the highest power conversion efficiency (PCE) of 25.7% currently, but they are unfortunately limited by several key issues, such as inferior humid and thermal stability, significantly retarding their widespread application. To tackle the instability issue, all-inorganic PSCs have attracted increasing interest due to superior structural, humid and high-temperature stability to their organic–inorganic counterparts. Nevertheless, all-inorganic PSCs with typical CsPbIBr2 perovskite as light absorbers suffer from much inferior PCEs to those of organic–inorganic PSCs. Functional doping is regarded as a simple and useful strategy to improve the PCEs of CsPbIBr2-based all-inorganic PSCs. Herein, we report a monovalent copper cation (Cu+)-doping strategy to boost the performance of CsPbIBr2-based PSCs by increasing the grain sizes and improving the CsPbIBr2 film quality, reducing the defect density, inhibiting the carrier recombination and constructing proper energy level alignment. Consequently, the device with optimized Cu+-doping concentration generates a much better PCE of 9.11% than the pristine cell (7.24%). Moreover, the Cu+ doping also remarkably enhances the humid and thermal durability of CsPbIBr2-based PSCs with suppressed hysteresis. The current study provides a simple and useful strategy to enhance the PCE and the durability of CsPbIBr2-based PSCs, which can promote the practical application of perovskite photovoltaics.


Introduction
The energy crisis and greenhouse gas emissions are the most crucial worldwide problems nowadays, which are mainly caused by the low efficiency and excessive consumption of nonrenewable fossil fuels [1,2]. Thus, the utilization of renewable energies and the development of relevant energy conversion technologies are highly urgent and essential [3][4][5][6][7]. Among various types of renewable energies/sources, solar energy has gained particular interest due to its inexhaustible and clean nature, which can be efficiently utilized by three main routes: photovoltaic cells, photocatalysis and solar-thermal power generation [1,[8][9][10]. In particular, photovoltaic cells (solar cells) have received increasing attention and achieved significant progress in the past decades because of their direct and efficient transformation of sunlight energy into electricity [11][12][13][14][15]. Nowadays, the commercial photovoltaic markets are dominated by silicon-based solar cells, showing excellent stability and high PCE of 27.6% [16]. However, silicon-based solar cells are limited by high-cost and complex fabrication procedures, and toxic by-products in the silicon manufacturing, which are harmful to the environment [17,18]. Consequently, the development of new-type solar cells with simple, cost-effective and environmentally friendly fabrication processes is highly crucial [19,20].
enhanced perovskite film quality [58][59][60]. Nowadays, the cation doping in CsPbIBr 2 is mainly focused on the utilization of divalent metal cations to substitute Pb 2+ , including Sn 2+ , Cu 2+ , Zn 2+ , Ba 2+ , Eu 2+ , etc., which effectively enhance the crystallinity and morphology of CsPbIBr 2 films [61][62][63][64][65][66]. For instance, Sn 2+ plays a crucial role in regulating the band gap and enhancing the film quality of CsPbIBr 2 , and Zhao et al. have reported that remarkably enhanced PCE (11.33%) and long-term stability were achieved by Sn 2+ doping in CsPbIBr 2based PSCs [61,64]. In our previous work, we have found that Cu 2+ doping in CsPbIBr 2 with optimized doping concentration effectively inhibited the carrier recombination by reducing the defect amount, and it enhanced the perovskite film quality by improving the crystallinity and enlarging the grain sizes, thereby remarkably improving the PCEs and humid/thermal stability of CsPbIBr 2 -based PSCs [62].
Nevertheless, it should be noted that the monovalent cation doping in the B-site of CsPbIBr 2 is much less investigated, which needs further exploration. Herein, we have employed monovalent copper cations (Cu + ) as effective B-site dopants for CsPbIBr 2 perovskites, which effectively improved the CsPbIBr 2 film quality in terms of a suppressed amount of grain boundaries, increased crystallinity and grain sizes, passivated surface defects and inhibited interfacial carrier recombination. After optimizing the Cu + doping concentration, the CsPbIBr 2 -0.50%Cu cell produced a superb PCE of 9.11%, 25.8% higher than that of the pristine cell. Furthermore, Cu + doping also remarkably improved the high-temperature and humid durability of CsPbIBr 2 -based PSCs with reduced hysteresis effect. Our current work can present some important insights for the design of high-performance all-inorganic PSCs, which may promote the widespread application of perovskite photovoltaics.

Materials and Methods
F − -doped tin oxide (FTO) glasses (2.2-mm thick and 7 Ω sq −1 ) were first cleaned by various solvents, dried by N 2 and further treated by oxygen plasma cleaning [56]. Detailed information about the fabrication of CsPbIBr 2 -based PSCs with a configuration of compact TiO 2 (c-TiO 2 )/perovskite/2,2 ,7,7 -tetrakis (N,N-di-p-methoxyphenylamine)-9,9 -spirobifluorene (Spiro-OMeTAD)/Ag can be found in our previous work [62]. As for the Cu + -doped CsPbIBr 2 -based cells, different amounts of as-prepared 0.2 M cuprous bromide/dimethyl sulfoxide (CuBr/DMSO) solution were added into the 1 M CsPbIBr 2 precursor solution to prepare various Cu + -doped CsPbIBr 2 films at fixed molar ratios of 0.25, 0.50 and 0.75%. The effective area of the cell was 0.0625 cm 2 in this work. For the thermal stability test, the carbon electrode was deposited by doctor blade for the holetransporting layer (HTL)-free CsPbIBr 2 -based PSCs and the detailed information can be found in our previous work [62].

Results and Discussion
Cu + cations with three different molar ratios of 0.25, 0.50 and 0.75% were doped into CsPbIBr 2 , which were labeled as CsPbIBr 2 , CsPbIBr 2 -0.25%Cu, CsPbIBr 2 -0.50%Cu and CsPbIBr 2 -0.75%Cu, respectively. Figure S1 displays the typical cross-sectional SEM image of as-fabricated CsPbIBr 2 -0.50%Cu cell and it was found that the 350 nm-thick perovskite film was firmly adhered on the surface of c-TiO 2 film. Based on the J−V curves of various CsPbIBr 2 cells, as depicted in Figure 1a and relevant photovoltaic parameters in Figure  S2, the Cu + doping in CsPbIBr 2 significantly boosted the PCEs of corresponding cells by increasing the fill factor (FF) and short-circuit current density (J sc ). More specifically, the unmodified cell displayed a champion PCE of 7.24% with an open-circuit voltage (V oc ) of 1.16 V, a J sc of 10.4 mA cm −2 and an FF of 0.597. In addition, the cell performance of Cu + -doped CsPbIBr 2 -based PSCs exhibited a volcano-like trend with increased Cu + doping contents, suggesting that the introduction of a suitable concentration of Cu + cations significantly enhanced the PCEs of CsPbIBr 2 cells. Particularly, the CsPbIBr 2 -0.50%Cu cell displayed the highest PCE of 9.11%, with a V oc of 1.19 V, a J sc of 11.6 mA cm −2 and a FF of 0.658 (Table S1). The CsPbIBr 2 -0.75%Cu cell exhibited a reduced PCE, which was mainly attributed to decreased FF value induced by the poor quality of CsPbIBr 2 film and higher defect concentration, which will be discussed later. Based on the EQE spectra in Figure 1b, the CsPbIBr 2 -0.50%Cu cell displayed higher EQE values than those of the pristine cell at a wavelength range of 300 to 600 nm. In addition, CsPbIBr 2 -0.50%Cu cell generated a larger integrated J sc value of 9.7 mA cm −2 than that of the pristine cell (8.1 mA cm −2 ), which was consistent with the J−V results (Figure 1a). To confirm the universality of such PCE enhancement, 20 devices for each type of PSC (without and with 0.50%Cu + substitution) were fabricated and tested with the photovoltaic parameter distributions displayed in Figure 1c. It is clear that the average PCE of the CsPbIBr 2 -0.50%Cu cell was much larger than that of the unmodified device due to the significantly improved J sc and FF values ( Figure S2). In addition, based on the maximum power point tracking (MPPT) profiles of PSCs as depicted in Figure 1d, the CsPbIBr 2 -0.50%Cu cell produced much larger stabilized PCE and J sc values than those of the pristine CsPbIBr 2 cell at the maximum power point.
It is well accepted that the PCEs of PSCs are determined by several crucial factors including the transport and recombination behavior of photogenerated carriers, defect density, sunlight absorption capability and energy level alignment between various films [67,68]. In order to elucidate the impacts of Cu + substitution on the CsPbIBr 2 film quality and the performance of CsPbIBr 2 -based PSCs, we employed XRD technique to investigate the crystallinity and the crystal structures of pristine CsPbIBr 2 and various Cu + -doped CsPbIBr 2 films. As depicted in Figure 2a, main characteristic peaks assigned to the α-phase CsPbIBr 2 structure were observed in all samples, while no obvious impurity phases were found for all investigated films, revealing that Cu + doping at different concentrations exhibited no obvious influences on the pure-phase cubic structure of CsPbIBr 2 . Based on the magnified XRD peaks in Figure 2b, the characteristic XRD peaks gradually shifted to higher angles with the increased Cu + doping amounts, implying that the Cu + cations with smaller ionic radius (0.60 Å) than that of Pb 2+ (1.19 Å) were successfully doped in the lattice of CsPbIBr 2 with a lattice contraction [69,70]. UV-vis absorption spectra were employed to evaluate the impacts of Cu + substitution on the light absorption capability of CsPbIBr 2 film. As depicted in Figure 2c, the Cu + doping effectively increased the light absorption intensity of CsPbIBr 2 film at 450-600 nm, especially for the CsPbIBr 2 -0.50%Cu film. Based on the Tauc plots of various films acquired from the UV-vis spectra in Figure S3, the band gap values of CsPbIBr 2 , CsPbIBr 2 -0.25%Cu, CsPbIBr 2 -0.50%Cu and CsPbIBr 2 -0.75%Cu were calculated to be 2.08, 2.07, 2.07 and 2.06 eV, respectively. It can be concluded that the Cu + doping can improve the light absorption capability of CsPbIBr 2 film, although the Cu + doping displayed no obvious influences on the band gap values of CsPbIBr 2 , which may be beneficial for the PCE improvement of corresponding PSCs. Based on the steady-state PL spectra of various CsPbIBr 2 films as depicted in Figure 2d, the Cu + -doped CsPbIBr 2 films exhibited much lower PL intensity than the CsPbIBr 2 film, demonstrating that the carrier Nanomaterials 2022, 12, 4317 5 of 13 recombination was effectively suppressed after introducing Cu + cations to reduce the defect amount. Moreover, CsPbIBr 2 -0.75%Cu film with excessive Cu + doping amount displayed a higher PL intensity than that of CsPbIBr 2 -0.50%Cu, implying more defects were formed in the CsPbIBr 2 -0.75%Cu film, agreeing well with the J−V results. It should be noted that the PL spectra of Cu + -doped CsPbIBr 2 films exhibited several split peaks due to the phase segregation of CsPbIBr 2 films, which may be beneficial for the PCE enhancement [62,71]. Based on the TRPL spectra as depicted in Figure S4, CsPbIBr 2 -0.50%Cu film delivered a much higher average carrier lifetime of 3.21 ns than the CsPbIBr 2 film (0.88 ns) due to the effectively suppressed defect concentration, which was favorable for the transport and separation of carriers. XPS was further employed to explore the influences of Cu + substitution on the chemical states of various ions in CsPbIBr 2 films, with results displayed in Figure S5. All the XPS peaks of CsPbIBr 2 -0.50%Cu film shifted to lower binding energies compared with the CsPbIBr 2 film due to the shrinkage of the BX 6 octahedron caused by Cu + doping induced by the changed interatomic force [58,72]. It is well accepted that the PCEs of PSCs are determined by several crucial factors including the transport and recombination behavior of photogenerated carriers, defect density, sunlight absorption capability and energy level alignment between various films [67,68]. In order to elucidate the impacts of Cu + substitution on the CsPbIBr2 film quality and the performance of CsPbIBr2-based PSCs, we employed XRD technique to investigate the crystallinity and the crystal structures of pristine CsPbIBr2 and various Cu + -doped CsPbIBr2 films. As depicted in Figure 2a, main characteristic peaks assigned to the α-phase CsPbIBr2 structure were observed in all samples, while no obvious impurity phases were found for all investigated films, revealing that Cu + doping at different concentrations exhibited no obvious influences on the pure-phase cubic structure of CsPbIBr2. Based on the magnified XRD peaks in Figure 2b, the characteristic XRD peaks gradually shifted to higher angles with the increased Cu + doping amounts, implying that the Cu + cations with smaller ionic radius (0.60 Å) than that of Pb 2+ (1.19 Å) were successfully doped in the lattice of CsPbIBr2 with a lattice contraction [69,70]. UV-vis absorption spectra were employed to evaluate the impacts of Cu + substitution on the light absorption capability of CsPbIBr2 film. As depicted in Figure 2c, the Cu + doping effec- effectively suppressed defect concentration, which was favorable for the transport and separation of carriers. XPS was further employed to explore the influences of Cu + substitution on the chemical states of various ions in CsPbIBr2 films, with results displayed in Figure S5. All the XPS peaks of CsPbIBr2-0.50%Cu film shifted to lower binding energies compared with the CsPbIBr2 film due to the shrinkage of the BX6 octahedron caused by Cu + doping induced by the changed interatomic force [58,72]. The morphology and quality of perovskite film plays a vital role in governing the performance of PSCs [54,73]. More specifically, compact perovskite films with large grain sizes and few grain boundaries are beneficial to suppress the carrier recombination [13,74]. Based on the top-view SEM images of CsPbIBr2 and various Cu + -doped CsPbIBr2 films in Figures 3a-d and S6, the pristine CsPbIBr2 film displayed an inferior morphology with abundant pinholes and small grains, while the quality of CsPbIBr2 film was significantly improved after Cu + doping in terms of reduced amount of pinholes and larger grain sizes. Particularly, the CsPbIBr2-0.50%Cu film showed a dense and uniform morphology with remarkably enlarged grain sizes and reduced amount of grain boundaries (Figure 3c,d). Nevertheless, CsPbIBr2-0.75%Cu film with excessive Cu + doping amount exhibited smaller grain sizes than those of CsPbIBr2-0.50%Cu film, as displayed in Figure  S6c,d. In addition, it was found that all elements were uniformly distributed on the CsP-bIBr2-0.50%Cu film, as depicted in Figure 3e, demonstrating homogeneous distribution The morphology and quality of perovskite film plays a vital role in governing the performance of PSCs [54,73]. More specifically, compact perovskite films with large grain sizes and few grain boundaries are beneficial to suppress the carrier recombination [13,74]. Based on the top-view SEM images of CsPbIBr 2 and various Cu + -doped CsPbIBr 2 films in Figure 3a-d and Figure S6, the pristine CsPbIBr 2 film displayed an inferior morphology with abundant pinholes and small grains, while the quality of CsPbIBr 2 film was significantly improved after Cu + doping in terms of reduced amount of pinholes and larger grain sizes. Particularly, the CsPbIBr 2 -0.50%Cu film showed a dense and uniform morphology with remarkably enlarged grain sizes and reduced amount of grain boundaries (Figure 3c,d). Nevertheless, CsPbIBr 2 -0.75%Cu film with excessive Cu + doping amount exhibited smaller grain sizes than those of CsPbIBr 2 -0.50%Cu film, as displayed in Figure S6c,d. In addition, it was found that all elements were uniformly distributed on the CsPbIBr 2 -0.50%Cu film, as depicted in Figure 3e, demonstrating homogeneous distribution of Cu + dopants. Based on the AFM images as depicted in Figure 3f,g and Figure S7, the CsPbIBr 2 -0.50%Cu film delivered a lower root-mean-square (RMS) surface roughness of 29.9 nm than the CsPbIBr 2 film (31.3 nm), benefiting the interfacial charge transfer at perovskite film/HTL interface [75,76].
The carrier recombination behavior of CsPbIBr 2 film after Cu + doping was investigated by EIS at a fixed voltage of 0.5 V under dark conditions, as depicted in Figure 4a. As can be seen, a low-frequency arc corresponding to the recombination resistance (R rec ) existed in the Nyquist plots, which was inversely proportional to the degree of carrier recombination at the interfaces between the perovskite films and TiO 2 /Spiro-OMeTAD layer [77,78]. The CsPbIBr 2 -0.50%Cu cell generated a larger R rec value than the CsPbIBr 2 cell, suggesting that the Cu + doping remarkably suppressed the interfacial carrier recombination. Based on the dark J−V curves as displayed in Figure 4b, the CsPbIBr 2 -0.50%Cu cell delivered much smaller current densities than those of the pristine CsPbIBr 2 cell, demonstrating that the photo-generated carriers were efficiently transported through charge-transporting layers of the CsPbIBr 2 -0.50%Cu cell instead of direct shunting, leading to reduced voltage/current loss and enhanced cell performance, which was attributed to the improved morphology and quality of CsPbIBr 2 -0.50%Cu films.
of Cu + dopants. Based on the AFM images as depicted in Figures 3f,g and S7, the CsP-bIBr2-0.50%Cu film delivered a lower root-mean-square (RMS) surface roughness of 29.9 nm than the CsPbIBr2 film (31.3 nm), benefiting the interfacial charge transfer at perovskite film/HTL interface [75,76]. The carrier recombination behavior of CsPbIBr2 film after Cu + doping was investigated by EIS at a fixed voltage of 0.5 V under dark conditions, as depicted in Figure 4a. As can be seen, a low-frequency arc corresponding to the recombination resistance (Rrec) existed in the Nyquist plots, which was inversely proportional to the degree of carrier recombination at the interfaces between the perovskite films and TiO2/Spiro-OMeTAD layer [77,78]. The CsPbIBr2-0.50%Cu cell generated a larger Rrec value than the CsPbIBr2 cell, suggesting that the Cu + doping remarkably suppressed the interfacial carrier recombination. Based on the dark J−V curves as displayed in Figure 4b, the CsP-bIBr2-0.50%Cu cell delivered much smaller current densities than those of the pristine CsPbIBr2 cell, demonstrating that the photo-generated carriers were efficiently transported through charge-transporting layers of the CsPbIBr2-0.50%Cu cell instead of direct shunting, leading to reduced voltage/current loss and enhanced cell performance, which was attributed to the improved morphology and quality of CsPbIBr2-0.50%Cu films.   Hole-only and electron-only PSCs with structures of FTO/PEDOT:PSS/pervoskite/ Spiro-OMeTAD/Ag and FTO/TiO 2 /pervoskite/[6,6]-phenyl-C61-butyric acid methyl ester (PCBM)/Ag were prepared to determine the trap densities of PSCs by measuring the dark J−V curves using the space charge-limiting current model [13,79]. As shown in Figure 4c,d, the traps were gradually filled until the applied voltage reached the trap-filling-limit voltage (V TFL ). The trap densities (N defects ) can be estimated by the equation of N defects = 2εε 0 V TFL eL 2 , where ε is the relative dielectric constant of CsPbIBr 2 (approximately 8); ε 0 is the vacuum dielectric constant; e represents elementary charge; and L is the thickness of CsPbIBr 2 film (350 nm, Figure S1). The V TFL values were 0.43 and 0.38 V for electron-only devices with CsPbIBr 2 and CsPbIBr 2 -0.50%Cu films, corresponding to trap densities of 3.10 and 2.75 × 10 15 cm −3 , respectively. As for hole-only devices, the hole defect densities of CsPbIBr 2 and CsPbIBr 2 -0.50%Cu films were 3.90 and 1.59 × 10 15 cm −3 based on V TFL values of 0.54 and 0.22 V, respectively. This suggested that the Cu + doping effectively reduced the trapping centers for both electrons and holes of CsPbIBr 2 film.
UPS technique was used to explore the effects of Cu + substitution on the energy level alignment of CsPbIBr 2 -based PSCs. As displayed in Figure S8 and Table S2, the valence band maximum (E VBM ) was calculated as −5.58 and −5.52 eV for CsPbIBr 2 and CsPbIBr 2 -0.50%Cu, respectively, while the corresponding conduction band minimum (E CBM ) was calculated to be −3.50 and −3.45 eV, respectively, based on the relationship of band gap = E CBM − E VBM [80]. In addition, the Fermi level of CsPbIBr 2 was slightly increased from −3.74 to −3.68 eV after the introduction of 0.50% Cu + , which was closer to the CBM position (−3.51 eV) due to the n-type nature of Cu + -doped CsPbIBr 2 [81]. The upshifted VBM position of CsPbIBr 2 -0.50%Cu effectively promoted the hole extraction from Spiro-OMeTAD to the perovskite layer. Moreover, the larger energy differences between the CBMs of CsPbIBr 2 -0.50%Cu and TiO 2 may provide a higher driving force for the electron injection from light-absorbing film to the electron-transporting layer [72]. Furthermore, the hysteresis index of the CsPbIBr 2 cell was reduced from 0.48 to 0.43 after the introduction of 0.50%Cu based on the J−V curves of corresponding PSCs under reverse and forward scan directions ( Figure S9 and Table S3), which was attributed to the improved CsPbIBr 2 film quality and promoted charge transfer, benefiting the cell stability.
Besides the device efficiency, the humid and thermal stability is another crucial factor for the development of PSCs. As depicted in Figure 5a, the CsPbIBr 2 -0.50%Cu cell retained 94% of its primary PCE after storing in humid air with a relative humidity (RH) of 15-30% at 25 • C for 400 h, much superior to the pristine device (57%) under the same conditions. Furthermore, the PCE of the CsPbIBr 2 -0.50%Cu cell maintained 80% after 600 h storage in ambient condition with a RH of 15-30%. The high-temperature air stability of PSCs was evaluated by preparing HTL-free cells (FTO/c-TiO 2 /perovskite/carbon). The photovoltaic parameters of HTL-free CsPbIBr 2 and CsPbIBr 2 -0.50%Cu cells are listed in Table S4. It was found that the CsPbIBr 2 -0.50%Cu cell delivered a superior PCE retention ratio of 97% to that of the CsPbIBr 2 device (76%) after storing in ambient condition at 85 • C for 1000 h (Figure 5b).
Cu + involved in the lattice of CsPbIBr 2 partially substituted the Pb 2+ -occupied B-site, causing a lattice contraction, and resulted in a reduced bandgap for Cu + -doped CsPbIBr 2 film. The reduced bandgap is beneficial for broadening light absorption band edge, thus facilitating the increase of photocurrent density. A previous study suggested that lattice contraction caused by doping could lead to improved heat stability for PSCs [82]. The presence of grain boundaries and pinholes might act as defect centers, trapping charge carriers, thereby reducing the PCE of PSC [83] Homogeneous and compact perovskite morphology with large grain sizes, less grain boundaries and pinholes enhanced the light capture of Cu + -doped perovskite film. The chemical bonds were enhanced after Cu + doping, leading to higher phase stability [84]. Moreover, the defect density and the nonradiative recombination of the films were reduced. contraction caused by doping could lead to improved heat stability for PSCs [82]. The presence of grain boundaries and pinholes might act as defect centers, trapping charge carriers, thereby reducing the PCE of PSC [83] Homogeneous and compact perovskite morphology with large grain sizes, less grain boundaries and pinholes enhanced the light capture of Cu + -doped perovskite film. The chemical bonds were enhanced after Cu + doping, leading to higher phase stability [84]. Moreover, the defect density and the nonradiative recombination of the films were reduced.

Conclusions
In summary, we report a monovalent Cu + cation doping approach to boost the PCEs and humid/thermal durability of CsPbIBr 2 -based all-inorganic PSCs. It was found that the introduction of a proper amount of Cu + cations into CsPbIBr 2 effectively passivated the defects, improved the perovskite film quality, suppressed the interfacial carrier recombination and enhanced the energy level alignment. Consequently, the optimized Cu + -doped cell exhibited a superb PCE of 9.11% with reduced hysteresis, which was 25.8% higher than that of pristine cell (7.24%). Moreover, the Cu + doping also remarkably improved the thermal and humid stability of CsPbIBr 2 -based PSCs. For instance, the CsPbIBr 2 -0.50%Cu device retained 97% of the primary PCE after 1000 h storage at 85 • C in humid air, while the PCE of the pristine PSC declined to 76% under the same conditions. This work provides some important insights for the fabrication of durable CsPbIBr 2 -based all-inorganic PSCs with higher PCEs, which may promote the commercialization of this technology.

Data Availability Statement:
The data that support the findings of this study are available from the corresponding author upon reasonable request.

Conflicts of Interest:
The authors declare no conflict of interest.