Thermal Analysis of Metal-Organic Precursors for Functional Cu:ΝiOx Hole Transporting Layer in Inverted Perovskite Solar Cells: Role of Solution Combustion Chemistry in Cu:ΝiOx Thin Films Processing

Low temperature solution combustion synthesis emerges as a facile method for the synthesis of functional metal oxides thin films for electronic applications. We study the solution combustion synthesis process of Cu:NiOx using different molar ratios (w/o, 0.1 and 1.5) of fuel acetylacetone (Acac) to oxidizer (Cu, Ni Nitrates) as a function of thermal annealing temperatures 150, 200, and 300 °C. The solution combustion synthesis process, in both thin films and bulk Cu:NiOx, is investigated. Thermal analysis studies using TGA and DTA reveal that the Cu:NiOx thin films show a more gradual mass loss while the bulk Cu:NiOx exhibits a distinct combustion process. The thin films can crystallize to Cu:NiOx at an annealing temperature of 300 °C, irrespective of the Acac/Oxidizer ratio, whereas lower annealing temperatures (150 and 200 °C) produce amorphous materials. A detail characterization study of solution combustion synthesized Cu:NiOx, including XPS, UV-Vis, AFM, and Contact angle measurements, is presented. Finally, 50 nm Cu:NiOx thin films are introduced as HTLs within the inverted perovskite solar cell device architecture. The Cu:NiOx HTL annealed at 150 and 200 °C provided PVSCs with limited functionality, whereas efficient triple-cation Cs0.04(MA0.17FA0.83)0.96 Pb(I0.83Br0.17)3-based PVSCs achieved for Cu:NiOx HTLs for annealing temperature of 300 °C.


Introduction
Perovskite solar cells (PVSCs) have witnessed significant progress related to power conversion efficiency within the last decade, climbing from 3.8% in 2009 to more than 24% in the year 2021 for single junction cells [1,2]. Some of the most promising hole transporting layers (HTLs) for inverted PVSCs are thin films of pristine or doped NiO x materials grown with various deposition methods. Thanks to its p-type semiconducting nature, high optical transmittance, enhanced electrical conductivity, and deep-lying valence band (VB) that matches well with the VB of hybrid perovskite photoactive layer, NiO x emerges as an excellent hole transporting layer (HTL) material for inverted perovskite solar cells [3][4][5][6][7][8][9][10][11][12][13][14][15][16]. Doping of NiO x with Copper (Cu) has been shown to improve the conductivity and charge collection properties, resulting in higher PCE-inverted perovskite solar cells. Jong H. Kim studied the performance of PEDOT:PSS, NiO x , and 5% Cu doped NiO x (Cu:NiO x ) HTLs Nanomaterials 2021, 11, 3074 2 of 16 relevant to the performance of inverted perovskite photovoltaic. Among them, Cu:NiO x exhibited the highest PCE (14.89%) due to having a better valence band (VB) alignment with the perovskite active layer compared to PEDOT:PSS HTL, while, in comparison to pristine NiO x , the higher PCE was ascribed to increased Cu:NiO x electrical conductivity (8.4 × 10 −4 S.cm −1 ) due to Cu doping [17]. Wei Chen et al. has reported that doping of NiO x with 5% Cu induces a slight downshift of VB from 5.16 to 5.25 eV and at the same time increases both carrier concentration and hole mobility of the Cu:NiO x HTL. The Cu:NiO x HTL-based inverted perovskite solar cells provided an increased PCE of 18.01% compared to pristine NiO x HTL-based solar cells (16.68%), retaining 95% of the PCE after 1000 h storage in air [18]. However, NiO x HTL derivatives prepared by sol-gel method usually need to be annealed at relatively high temperatures (over 400 • C) in order to achieve high crystallinity [19][20][21][22]. Sol-gel reactions are endothermic, and thus require high external thermal energy to form metal oxide lattices and remove organic residuals [23][24][25][26]. This high processing temperature increases the cost of device fabrication and also inhibits the implementation of printing manufacturing for producing next generation solutionprocessed photovoltaics on conventional flexible substrates.
Combustion synthesis methods have been reported by many researchers as a beneficial route of the synthesis of high crystalline metal oxides at lower temperatures than those used for sol-gel reactions. The combustion process's exothermic reaction provides a lower transition energy with which to form the metal oxide crystal lattices, avoiding the need for high thermal energy. Thus, the solution combustion synthesis (SCS) method is being adopted for the synthesis of metal oxides electronic thin films due to its costeffectiveness, simplicity, enhanced electronic material functionality, and relatively lower required processing annealing temperatures [27][28][29][30][31][32][33][34][35].
Over the last years, there have been many reports on the use of solution combustion synthesis of metal oxide such as Cu:CrO x and ZnO for charge transporting layers in perovskite solar cells and Amorphous Indium Gallium Zinc oxide (IGZO) for metal oxide thin film transistors applications [28,30,36,37]. We have reported the solution combustions synthesis of pristine and co-doped (Cu, Li) NiCo 2 O 4 films, applying a 300 • C annealing temperature and incorporating them as high performance HTLs in MAPbI 3 -based inverted perovskite solar cells [38,39]. We have also shown the improvements of PVSC's thermal stability based on SCS Cu:NiO x HTL by treatment of Cu:NiO x with β-alanine, showing a T80 of 1000 h under heat conditions (60 • C, N 2 ), as well as the improvement of humidity degradation resistance for SCS NiO x -based PVSC with the addition of 1% Nitrobenzene within the perovskite active layer [40,41]. Jae Woong Jung et al. have reported the solution combustion synthesis of the Cu:NiO x film at 150 • C and implemented it as HTL in MAPbI 3 perovskite solar cell. They showed that the combustion-synthesized Cu:NiO x resulted in better power conversion efficiency (PCE) for perovskite solar cells compare to devices containing a typical sol-gel synthesized Cu:NiO x HTL [42]. Other reports on the solution combustion synthesis of pristine and doped NiO x have applied a range of annealing temperatures for the fabrication of HTLs for efficient perovskite solar cells [43][44][45][46][47][48]. For example, Ziye Liu et al. reported that the fabrication of an MA 1−y FA y PbI 3−x Cl x perovskite solar cell using solution combustion synthesized NiO x , achieving a high PCE of more than 20%. The applied temperature of the NiO x solution combustion synthesis for high efficiency devices was 250 • C while, for a 150 • C annealing temperature, the devices exhibited a rapid deterioration of their PCE [46]. Ao Liu et al. demonstrated the solution combustion synthesis fabrication of optimized 5% Cu doped NiO x films for use in TFT applications, exhibiting excellent electrical performance [49]. Yi-Huan Li et al. applied a 300 • C annealing temperature for the solution combustion synthesis of Cu:NiO x , showing that it can be used as an efficient hole injection layer for the fabrication of high performing OLEDs [50]. Thus, solution combustion synthesis metal oxide thin films have emerged as a facile method for the fabrication of functional metal oxide-based charge selective contacts for electronic applications. Nevertheless, there have been reports that the combustion synthesis of thin electronic films differ from the corresponding bulk analogues, and thus a low annealing temperature combustion synthesis of thin films cannot be deduced by the bulk material behavior, even suggesting that low temperature combustion synthesis is unlikely to occur during the processing of thin film precursors [51][52][53][54][55][56]. Thus, the processing annealing conditions for the solution combustion synthesis of functional Cu:NiO x HTLs needs further investigation.
The aim of this paper is to identify the Cu:NiO x HTL annealing processing conditions and to examine the fuel to oxidizer ratio for efficient inverted PVSCs. We study in detail the Cu:NiO x thin films solution combustion synthesis process as a common metal oxides HTL that is widely used in inverted perovskite solar cells. Specifically, the reported results investigate the effect of the thermal annealing temperature (150, 200, and 300 • C) as well as the fuel (acetyl acetonate (Acac)) to oxidizer (Cu and Ni nitrates) ratio (without (w/o), 0.1 and 1.5) in the combustion synthesis process of Cu:NiO x . The study is performed at Cu:NiO x filmswith various final thicknesses (50,200, 300 nm) and for drop-casted bulk Cu:NiO x analogues (with thickness in the range of a few microns). The crystal growth process is studied by performing thermogravimetric analysis (TGA) and the crystallinity of the corresponding Cu:NiO x materials is examined by X-ray diffraction (XRD). Furthermore, we characterized the thin Cu:NiO x film properties processed under the above-mentioned conditions using XPS, contact angle, AFM, and UV-Vis spectroscopy techniques. Finally, we evaluated the impact of SCS-based Cu:NiO x thin film properties on the PCE performance of inverted PVSCs containing Cu:NiO x HTL and the triple cation perovskite [Cs 0.04 (MA 0.17 FA 0.83 ) 0.96 Pb(I 0.83 Br 0.17 ) 3 ] as a photoactive layer. Samples preparations for TGA, AFM, UV-Vis, contact angle analysis: For the thermogravimetric analysis (TGA) of the combustion synthesis behavior of the Cu:NiO x films, the samples were fabricated onto Alumina disk substrates by blade-coating in air. For the AFM, UV-Vis, and contact angle measurements, the Cu:NiO x films were fabricated by doctor blade on quartz substrates.

Materials
Device fabrication: ITO-patterned glass substrates were cleaned using an ultrasonic bath for 10 min in acetone followed by 10 min in isopropanol. The Cu:NiO x precursors were prepared and blade-coated on ITO substrates as described in the Cu:NiO x solution combustion synthesis section. For the preparations of the triple cation Cs 0.04 (MA 0.17 FA 0.83 ) 0.96 Pb(I 0.83 Br 0.17 ) 3 perovskite solutions, a previous reported method was used [57]. The perovskite films were fabricated on top of Cu:NiO x inside a N 2 atmosphere glovebox by spin-coating at 5000 rpm for 35 s and after 10 s. 300 mL of ethyl acetate were dropped onto the spinning substrate as the anti-solvent to achieve the rapid crystallization of the films. The resulting perovskite films were annealed at 100 • C for 60 min. For the electron transporting layers, a PC 60 BM film was coated on top of perovskite inside the glovebox using spin coating at 1000 rpm for 30 s from a 20 mg/mL chlorobenzene solution. To complete the devices, 7 nm of bathocuproine (BCP) was thermally evaporated followed by 80 nm of Ag. The schematic illustration of the SCS-based Cu:NiO x HTL and PVSC fabrication process and the corresponding device structure are presented in Figure 1.  Characterization: Thermogravimetric analysis (TGA) and differential thermal analysis (DTA) was performed on a Shimadzu. Samples were heated up to 400 °C in air atmosphere (200 mL min −1 flow rate) with a heating rate of 10 °C min −1 and alumina (Al2O3), a substance with the same thermal mass as the sample, was used for reference material analysis. Powder X-ray diffraction (XRD) patterns were recorded on a PANalytical X'Pert Pro X-ray diffractometer with a Ni-filtered Cu Kα source (λ = 1.5418 Å), operating at 45 kV and 40 mA. X-ray photoelectron spectroscopy (XPS) analysis conducted on a SPECS spectrometer using a Phoibos 100 1D-DLD electron analyzer and an Al Kα radiation as the energy source (1486.6 eV). Binding energy values were corrected for charging by assigning a bending energy of 284.8 eV to the C 1 s signal of adventitious carbon. For UV-Vis absorption and atomic force microscopy (AFM) measurements, the films were fabricated on quartz substrates. UV-Vis absorption measurements were performed with a Schimadzu UV-2700 UV-Vis spectrophotometer. AFM images were obtained using a Nanosurf easy scan 2 controller applying tapping mode. The thickness of the films was measured with a Veeco Dektak 150 profilometer. Contact angle (CA) measurements were performed using a KRUSS DSA 100E drop analysis system. The current density-voltage (J/V) characteristics were characterized with a Botest LIV Functionality Test System. For illumination, a calibrated Newport Solar simulator equipped with an Xe lamp was used, providing an AM1.5G spectrum at 100 mW/cm 2 , as measured by a certified oriel 91150V calibration cell. A shadow mask was attached to each device prior to measurements to accurately define 0.09 cm 2 device area.

TGA Results of Cu:NiOx (Films versus Bulk Precursors)
The synthesis behavior of the Cu:NiOx thin films and corresponding bulk mixtures onto alumina disc substrates was examined through thermogravimetric analysis (TGA) and differential thermal analysis (DTA). Figure 2a,b presents the TGA and the Characterization: Thermogravimetric analysis (TGA) and differential thermal analysis (DTA) was performed on a Shimadzu. Samples were heated up to 400 • C in air atmosphere (200 mL min −1 flow rate) with a heating rate of 10 • C min −1 and alumina (Al 2 O 3 ), a substance with the same thermal mass as the sample, was used for reference material analysis. Powder X-ray diffraction (XRD) patterns were recorded on a PANalytical X'Pert Pro X-ray diffractometer with a Ni-filtered Cu Kα source (λ = 1.5418 Å), operating at 45 kV and 40 mA. X-ray photoelectron spectroscopy (XPS) analysis conducted on a SPECS spectrometer using a Phoibos 100 1D-DLD electron analyzer and an Al Kα radiation as the energy source (1486.6 eV). Binding energy values were corrected for charging by assigning a bending energy of 284.8 eV to the C 1 s signal of adventitious carbon. For UV-Vis absorption and atomic force microscopy (AFM) measurements, the films were fabricated on quartz substrates. UV-Vis absorption measurements were performed with a Schimadzu UV-2700 UV-Vis spectrophotometer. AFM images were obtained using a Nanosurf easy scan 2 controller applying tapping mode. The thickness of the films was measured with a Veeco Dektak 150 profilometer. Contact angle (CA) measurements were performed using a KRUSS DSA 100E drop analysis system. The current density-voltage (J/V) characteristics were characterized with a Botest LIV Functionality Test System. For illumination, a calibrated Newport Solar simulator equipped with an Xe lamp was used, providing an AM1.5G spectrum at 100 mW/cm 2 , as measured by a certified oriel 91150V calibration cell. A shadow mask was attached to each device prior to measurements to accurately define 0.09 cm 2 device area.

TGA Results of Cu:NiO x (Films versus Bulk Precursors)
The synthesis behavior of the Cu:NiO x thin films and corresponding bulk mixtures onto alumina disc substrates was examined through thermogravimetric analysis (TGA) and differential thermal analysis (DTA). Figure 2a indicating the thermal instability of this precursor in the absence of any fuel additive. For the Cu:NiO x films prepared with 0.1 and 1.5 Acac/oxidizer molar ratio, the TGA profiles show similar thermal decomposition behavior, exhibiting a gradual mass loss after T > ignition temperature (T ig ). This is inconsistent with a combustion process that occurs to the combustible precursors, as is be shown below. Near 300 • C, the TGA profiles for all samples display an intense mass loss, which is associated with an exothermic peak on the DTA curve. This is attributed to the decomposition of metal complexes and the crystallization of the Cu:NiO x oxide. Thus, films present two stages of mass loss, at~130 • C and~300 • C for drying at 80 • C irrespective of Acac addition in precursor solution. The corresponding DTA results are presented in Figure 2b. For the sample w/o Acac, an negligible broad exothermic peak is observed at~130 • C and another broad exotherm peak around~300 • C. For the samples containing 0.1 and 1.5 Acac, low intensity, abrupt exothermic peaks are observed at~130 • C and broad exothermic peaks around~300 • C. Thus, the addition of Acac in the precursor induces a limited combustion process by reacting with a part of the oxidizer at~130 • C, facilitating the removal of the organic residuals.
profiles show similar thermal decomposition behavior, exhibiting a gradual mass loss after T > ignition temperature (Tig). This is inconsistent with a combustion process that occurs to the combustible precursors, as is be shown below. Near 300 °C, the TGA profiles for all samples display an intense mass loss, which is associated with an exothermic peak on the DTA curve. This is attributed to the decomposition of metal complexes and the crystallization of the Cu:NiOx oxide. Thus, films present two stages of mass loss, at ~130 °C and ~300 °C for drying at 80 °C irrespective of Acac addition in precursor solution. The corresponding DTA results are presented in Figure 2b. For the sample w/o Acac, an negligible broad exothermic peak is observed at ~130 °C and another broad exotherm peak around ~300 °C. For the samples containing 0.1 and 1.5 Acac, low intensity, abrupt exothermic peaks are observed at ~130 °C and broad exothermic peaks around ~300 °C. Thus, the addition of Acac in the precursor induces a limited combustion process by reacting with a part of the oxidizer at ~130 °C, facilitating the removal of the organic residuals.
The mass loss at ~130 °C of the sample w/o Acac suggests that the 2-methoxy ethanol, except for its role as a solvent, could also behave as a fuel. To support this, higher drying temperature (100 °C) for 1 h was applied to the film's synthesis, where most of 2-methoxy ethanol was evaporated. In this case, the TGA curve showed no mass loss and the corresponding DTA curve exhibited analogous behavior without any endothermic or exothermic reaction at 130 °C, while, for the film prepared with a Acac/oxidizer molar ratio of 0.1, a marginal mass loss occurred during the first stage of combustion (at ~130 °C). This means that the films that require non-combustive Cu:NiOx precursors require high temperatures of over 300 °C for the complete conversion of the precursors into the metal oxide lattice. Further, we compared the combustion synthesis behavior of Cu:NiOx thin films (50, 200, 300 nm) and bulk analogues (thickness range of a few microns). TGA profiles ( Figure 3a) show that, by increasing the thickness of the film, a more intense gradual mass loss occurs at ~130 °C, with the second mass loss at ~300 °C becoming less prominent, while comparatively full combustion could occur at bulk materials at ~130 °C. Thus, we can infer that the mass of the precursor has a significant impact on the complete combustion synthesis The mass loss at~130 • C of the sample w/o Acac suggests that the 2-methoxy ethanol, except for its role as a solvent, could also behave as a fuel. To support this, higher drying temperature (100 • C) for 1 h was applied to the film's synthesis, where most of 2-methoxy ethanol was evaporated. In this case, the TGA curve showed no mass loss and the corresponding DTA curve exhibited analogous behavior without any endothermic or exothermic reaction at 130 • C, while, for the film prepared with a Acac/oxidizer molar ratio of 0.1, a marginal mass loss occurred during the first stage of combustion (at~130 • C). This means that the films that require non-combustive Cu:NiO x precursors require high temperatures of over 300 • C for the complete conversion of the precursors into the metal oxide lattice.
Further, we compared the combustion synthesis behavior of Cu:NiO x thin films (50, 200, 300 nm) and bulk analogues (thickness range of a few microns). TGA profiles ( Figure 3a) show that, by increasing the thickness of the film, a more intense gradual mass loss occurs at~130 • C, with the second mass loss at~300 • C becoming less prominent, while comparatively full combustion could occur at bulk materials at~130 • C. Thus, we can infer that the mass of the precursor has a significant impact on the complete combustion synthesis reactions. The DTA results, in Figure 3c, are in accordance with the TGA profile, where the raise of thickness in Cu:NiO x precursors layers shows a more intense exothermic peak at 130 • C; this corresponds to almost complete mass loss, gradually decreasing the exothermic peak at around 300 • C. Additionally, the absence of fuel (Acac) and the impact Nanomaterials 2021, 11, 3074 6 of 16 of solvent was examined once again during the combustion process for bulk materials. As observed in TGA profile (Figure 3b), a rapid mass loss occurs at~130 • C w/o Acac for samples dried at 80 and 100 • C for 5 min, respectively. The corresponding DTA curves ( Figure 3d) show a single sharp exotherms at~130 • C that corresponds exactly to the abrupt mass loss in the TGA (Figure 3b); this process is sufficient to lead the reaction rapidly to completion for metallic Ni formation as will be shown in XRD analysis below. In contrast, in the preheated sample at 100 • C for an extended period (48 h), where most of the solvent was evaporated, combustion reaction could not occur. This sample exhibits only an intense exothermic peak of around~300 • C, which corresponds to the crystal phase formation of NiO. These observations are in agreement with previous reports that the organic solvent 2-methoxy ethanol plays a dual role of acting both as a solvent and also as a fuel in addition to Acac for the formation of the metal oxide lattices by the solution combustion synthesis [58]. reactions. The DTA results, in Figure 3c, are in accordance with the TGA profile, where the raise of thickness in Cu:NiOx precursors layers shows a more intense exothermic peak at 130 °C; this corresponds to almost complete mass loss, gradually decreasing the exothermic peak at around 300 °C. Additionally, the absence of fuel (Acac) and the impact of solvent was examined once again during the combustion process for bulk materials. As observed in TGA profile (Figure 3b), a rapid mass loss occurs at ~130 °C w/o Acac for samples dried at 80 and 100 °C for 5 min, respectively. The corresponding DTA curves ( Figure 3d) show a single sharp exotherms at ~130 °C that corresponds exactly to the abrupt mass loss in the TGA ( Figure 3b); this process is sufficient to lead the reaction rapidly to completion for metallic Ni formation as will be shown in XRD analysis below. In contrast, in the preheated sample at 100 °C for an extended period (48 h), where most of the solvent was evaporated, combustion reaction could not occur. This sample exhibits only an intense exothermic peak of around ~300 °C, which corresponds to the crystal phase formation of NiO. These observations are in agreement with previous reports that the organic solvent 2-methoxy ethanol plays a dual role of acting both as a solvent and also as a fuel in addition to Acac for the formation of the metal oxide lattices by the solution combustion synthesis [58].

XRD Results of Cu:NiO x Films and Bulk Precursors
The crystallinity of the SCS Cu:NiO x thin films (identical to the Cu:NiO x HTLs that were used within the inverted PVSCs) was examined using X-ray diffraction (XRD) analysis.

XRD Results of Cu:NiOx Films and Bulk Precursors
The crystallinity of the SCS Cu:NiOx thin films (identical to the Cu:NiOx HTLs that were used within the inverted PVSCs) was examined using X-ray diffraction (XRD) analysis. Figure 4a-c presents the XRD patterns of the films prepared using w/o, 0.1, and 1.5 Acac and annealing temperatures of 150, 200, and 300 °C . The crystal phase of NiO can be obtained for an annealing temperature of 300 °C regardless of the containing amount of Acac, while, for 150 and 200 °C , no crystal phase was detected. For the 300 °C annealing temperature, the characteristic diffraction peaks of NiO appeared at 2θ = 37.20°, 43.0°, 62.87°, and 75.20°, which can be indexed to the cubic crystal structure of NiO as (111)  The crystallinity of the as-prepared materials obtained by combustion reaction of the bulk precursors was also examined using XRD analysis ( Figure 5). Specifically, different initial molar ratios of the fuel to oxidizer (w/o, 0.1, and 1.5 Acac) at 200 °C annealing temperature of the bulk precursors were compared. The XRD results show a significant improvement in the crystallinity of the bulk Cu:NiOx compared to the corresponding thin films. Moreover, the required annealing temperature for the crystal phase formation is  The crystallinity of the as-prepared materials obtained by combustion reaction of the bulk precursors was also examined using XRD analysis ( Figure 5). Specifically, different initial molar ratios of the fuel to oxidizer (w/o, 0.1, and 1.5 Acac) at 200 • C annealing temperature of the bulk precursors were compared. The XRD results show a significant improvement in the crystallinity of the bulk Cu:NiO x compared to the corresponding thin films. Moreover, the required annealing temperature for the crystal phase formation is significantly reduced (as was also evidenced by TGA analysis) when using the solution combustion synthesis of bulk materials as compared to the thin films. In the case of samples synthesized by bulk precursors, mixed crystal phases of metallic Ni (dominant species) and metal oxide NiO (residual species) were obtained regardless of the molar ratio of fuel to oxidizer precursors. Specifically, XRD patterns revealed that intense peaks appeared at 44.0 • , 52.3 • , and 76.5 • assigned to the (111), (200), and (220) planes, respectively, of face-centered cubic Ni (JCPDS No. 87-0712) (main product). The presence of Ni phase is an indication of combustion with a rich fuel precursor; therefore, even in the precursor w/o Acac, the Ni phase is the main product implying that, again, 2-methoxy ethanol plays a dual role of acting both as a solvent and a fuel [26]. The high crystallinity of bulk materials, as evidenced by the sharper diffractions in XRD patterns, is attributed to the release of high energy in the exothermic reaction that occurred by solution combustion synthesis (SCS). Thus, in agreement with the findings from TGA measurements, the complete combustion occurs mainly due to the bulk precursor material, which has a higher mass compared to corresponding films [55].
significantly reduced (as was also evidenced by TGA analysis) when using the solution combustion synthesis of bulk materials as compared to the thin films. In the case of samples synthesized by bulk precursors, mixed crystal phases of metallic Ni (dominant species) and metal oxide NiO (residual species) were obtained regardless of the molar ratio of fuel to oxidizer precursors. Specifically, XRD patterns revealed that intense peaks appeared at 44.0°, 52.3°, and 76.5° assigned to the (111), (200), and (220) planes, respectively, of face-centered cubic Ni (JCPDS No. 87-0712) (main product). The presence of Ni phase is an indication of combustion with a rich fuel precursor; therefore, even in the precursor w/o Acac, the Ni phase is the main product implying that, again, 2-methoxy ethanol plays a dual role of acting both as a solvent and a fuel [26]. The high crystallinity of bulk materials, as evidenced by the sharper diffractions in XRD patterns, is attributed to the release of high energy in the exothermic reaction that occurred by solution combustion synthesis (SCS). Thus, in agreement with the findings from TGA measurements, the complete combustion occurs mainly due to the bulk precursor material, which has a higher mass compared to corresponding films [55].   Figure 6f), consistent with other reports [62,63]. As for the broad signals located at 940.6 and 953.9 eV (943.0 eV for the 200 • C annealed film), they are assigned to the shake-up satellite peaks of paramagnetic Cu 2+ . The incorporation of Cu 2+ ions into the NiO lattice was also verified by the Auger α parameter, that is, the kinetic energy of the Cu L 3 M 4,5 M 4,5 Auger peak plus binding energy of the Cu 2p 3/2 peak. For the 300 • C annealed film, the Auger parameter was calculated to be 1851.8 eV, which respects the existing phase of CuO [64]. Quantitative analysis from the XPS spectra also indicated that the film that annealed at 200 • C contained 6.51 wt.% CH x O y and 3.32 wt.%, with NO x containing organic compounds, while the corresponding remnants for the 300 • C annealed film was found to be 0.64 wt.% and 0.33%, respectively (see Supplementary  Table S1). The higher amount of remnants found in 200 • C annealed film suggests the incomplete combustion reaction of Cu:NiO x precursors, in agreement with the TGA results ( Figure 2). Moreover, the Cu atomic concentration (Cu doping level) in the Cu:NiO x films annealed at 200 and 300 • C was found to be 5.64% and 5.84%, respectively, which is very close to the nominal composition.

Cu:ΝiOx Thin Films Characterization
the shake-up satellite peaks of paramagnetic Cu 2+ . The incorporation of Cu 2+ ions into the NiO lattice was also verified by the Auger α parameter, that is, the kinetic energy of the Cu L3M4,5M4,5 Auger peak binding energy of the Cu 2p3/2 peak. For the 300 °C annealed film, the Auger parameter was calculated to be 1851.8 eV, which respects the existing phase of CuO [64]. Quantitative analysis from the XPS spectra also indicated that the film that annealed at 200 °C contained 6.51 wt.% CHxOy and 3.32 wt.%, with NOx containing organic compounds, while the corresponding remnants for the 300 °C annealed film was found to be 0.64 wt.% and 0.33%, respectively (see Supplementary Table S1). The higher amount of remnants found in 200 °C annealed film suggests the incomplete combustion reaction of Cu:NiOx precursors, in agreement with the TGA results ( Figure 2). Moreover, the Cu atomic concentration (Cu doping level) in the Cu:NiOx films annealed at 200 and 300 °C was found to be 5.64% and 5.84%, respectively, which is very close to the nominal composition. Furthermore, we examined the film topography of Cu:NiO x films fabricated on quartz substrates by contact angle and UV-Vis spectroscopy. Figure 7 shows the film morphology using AFM for the films synthesized from precursors containing 0.1 Acac and annealed at 150 and 300 • C, respectively. It is clearly observed that the film treated at 150 • C shows a featured structure of large particles, due to the presence of residues, with a surface roughness of 1.5 nm (Figure 7a). On the other hand, the film treated at 300 • C does not show structured features due to the small size of the Cu:NiO x particles, exhibiting a surface roughness of 0.7 nm (Figure 7b). The final thickness of the film annealed at 150 and 300 • C is~80 and~50 nm, respectively, due to the considerable amount of residue that remained in the 150 • C annealed film, as shown above through XPS analysis for films synthesized at low annealing temperatures.
Furthermore, we examined the film topography of Cu:NiOx films fabricated on quartz substrates by contact angle and UV-Vis spectroscopy. Figure 7 shows the film morphology using AFM for the films synthesized from precursors containing 0.1 Acac and annealed at 150 and 300 °C, respectively. It is clearly observed that the film treated at 150 °C shows a featured structure of large particles, due to the presence of residues, with a surface roughness of 1.5 nm (Figure 7a). On the other hand, the film treated at 300 °C does not show structured features due to the small size of the Cu:NiOx particles, exhibiting a surface roughness of 0.7 nm (Figure 7b). The final thickness of the film annealed at 150 and 300 °C is ~80 and ~50 nm, respectively, due to the considerable amount of residue that remained in the 150 °C annealed film, as shown above through XPS analysis for films synthesized at low annealing temperatures. The UV-vis absorption spectrum (Figure 8) of the films treated at 150 and 200 °C show no prominent absorption due to the amorphous nature of metal oxides, while, for 300 °C annealing film, the absorption onset at ~400 nm (~3.1 eV) and strong absorption at ~350 nm (~3.5 eV) is ascribed to the crystalline Cu:NiOx phase.  The UV-vis absorption spectrum (Figure 8) of the films treated at 150 and 200 • C show no prominent absorption due to the amorphous nature of metal oxides, while, for 300 • C annealing film, the absorption onset at~400 nm (~3.1 eV) and strong absorption at~350 nm (~3.5 eV) is ascribed to the crystalline Cu:NiO x phase. quartz substrates by contact angle and UV-Vis spectroscopy. Figure 7 shows the film morphology using AFM for the films synthesized from precursors containing 0.1 Acac and annealed at 150 and 300 °C, respectively. It is clearly observed that the film treated at 150 °C shows a featured structure of large particles, due to the presence of residues, with a surface roughness of 1.5 nm (Figure 7a). On the other hand, the film treated at 300 °C does not show structured features due to the small size of the Cu:NiOx particles, exhibiting a surface roughness of 0.7 nm (Figure 7b). The final thickness of the film annealed at 150 and 300 °C is ~80 and ~50 nm, respectively, due to the considerable amount of residue that remained in the 150 °C annealed film, as shown above through XPS analysis for films synthesized at low annealing temperatures. The UV-vis absorption spectrum (Figure 8) of the films treated at 150 and 200 °C show no prominent absorption due to the amorphous nature of metal oxides, while, for 300 °C annealing film, the absorption onset at ~400 nm (~3.1 eV) and strong absorption at ~350 nm (~3.5 eV) is ascribed to the crystalline Cu:NiOx phase.  The contact angle of water was measured on films annealed at 150, 170, and 200 • C, using a fuel to oxidizer ratio of 0 (w/o), 0.1, and 1.5 (see Supplementary Figure S2), from which the measured values are plotted in Figure 9. All the contact angles are higher than 60 • , irrespective of fuel concentration, in contrast to the contact angle of the Cu:NiO x film (0.1 molar ratio of Acac to oxidizer) annealed at 300 • C, which is substantially lower (20 • ); see Supplementary Figure S3. Thus, we infer that the remnants (see XPS analysis) in the low temperature treated film form a Cu:NiO x surface with moderate wettability, while the films annealed at 300 • C, where the surface is almost free from remnants, show an improved wettability.
(0.1 molar ratio of Acac to oxidizer) annealed at 300 °C, which is substantially lower (20°); see Supplementary Figure S3. Thus, we infer that the remnants (see XPS analysis) in the low temperature treated film form a Cu:NiOx surface with moderate wettability, while the films annealed at 300 °C, where the surface is almost free from remnants, show an improved wettability.

J-V Characterization of Cu:NiOx Films as HTLs in Planar p-i-n PVSCs
To evaluate the functionality of the different Cu:NiOx films as HTLs in solar cells, 50 nm thick Cu:NiOx films, synthesized using the previously described conditions, were implemented in inverted perovskite solar cells with structure ITO/Cu:NiOx/perovskite/PC60BM/BCP/Ag and the J-V device characteristics under 1 sun simulated light were recorded.
As it is presented in J-V curves of Figure 10a, the devices that incorporated Cu:NiOx films annealed at 150 °C and 200 °C exhibited a very limited functionality. All the devices show low Voc in the range of 0.3 V, and the generated current is below 1 mA/cm 2 . The device with a 1.5 molar ratio Acac to oxidizer that was annealed at 200 °C shows an almost linear response of the current density to the sweeping voltage, which can be attributed to partially formed metallic Ni, as can be inferred by the corresponding XRD results in Figure  4c. In contrast, the inverted perovskite solar cells which incorporate Cu:NiOx HTLs prepared from precursor solutions containing w/o, 0.1, and 1.5 Acac annealed at 300 °C, delivered higher efficiency. In Figure 10b, the J-V curves of the best performing devices under 1 sun simulated light are illustrated, and the extracted solar cell parameters of the studied devices are presented in Table 1-the average values of 12 devices for each batch are indicated in brackets. Regarding the impact of the fuel to oxidizer ratio on the devices' PCE, the devices with Cu:NiOx HTL where the precursor contained no fuel (w/o Acac) and a 0.1 Acac ratio showed similar PCE values, and the devices which incorporated Cu:NiOx film synthesized with 1.5 Acac/oxidizer ratio showed a reduced Voc and Jsc efficiency, resulting in lower PCE.

J-V Characterization of Cu:NiO x Films as HTLs in Planar p-i-n PVSCs
To evaluate the functionality of the different Cu:NiO x films as HTLs in solar cells, 50 nm thick Cu:NiO x films, synthesized using the previously described conditions, were implemented in inverted perovskite solar cells with structure ITO/Cu:NiO x /perovskite/PC 60 BM/ BCP/Ag and the J-V device characteristics under 1 sun simulated light were recorded.
As it is presented in J-V curves of Figure 10a, the devices that incorporated Cu:NiO x films annealed at 150 • C and 200 • C exhibited a very limited functionality. All the devices show low V oc in the range of 0.3 V, and the generated current is below 1 mA/cm 2 . The device with a 1.5 molar ratio Acac to oxidizer that was annealed at 200 • C shows an almost linear response of the current density to the sweeping voltage, which can be attributed to partially formed metallic Ni, as can be inferred by the corresponding XRD results in Figure 4c. In contrast, the inverted perovskite solar cells which incorporate Cu:NiO x HTLs prepared from precursor solutions containing w/o, 0.1, and 1.5 Acac annealed at 300 • C, delivered higher efficiency. In Figure 10b, the J-V curves of the best performing devices under 1 sun simulated light are illustrated, and the extracted solar cell parameters of the studied devices are presented in Table 1-the average values of 12 devices for each batch are indicated in brackets. Regarding the impact of the fuel to oxidizer ratio on the devices' PCE, the devices with Cu:NiO x HTL where the precursor contained no fuel (w/o Acac) and a 0.1 Acac ratio showed similar PCE values, and the devices which incorporated Cu:NiO x film synthesized with 1.5 Acac/oxidizer ratio showed a reduced V oc and J sc efficiency, resulting in lower PCE.
The experimental results presented within this manuscript using a triple cation Cs 0.04 (MA 0.17 FA 0.83 ) 0.96 Pb(I 0.83 Br 0.17 ) 3 perovskite formulation infer that the Cu:NiO x oxide's precursor films that were annealed at temperatures of 150 and 200 • C produce electronic films that cannot function as HTLs for efficient inverted perovskite solar cells. This is ascribed to the incomplete combustion that results in amorphous Cu:NiO x films with remnants. This result is in agreement with a previous report, where amorphous NiO x showed limited functionality as HTL when applied in organic solar cells [65]. On the other hand, as shown above, the pure crystalline phase of Cu:NiO x HTL was obtained by annealing at 300 • C for the precursors w/o and with a 0.1 ratio of Acac/oxidizer, whereas, for the 1.5 ratio, metallic Ni are likely to be present within Cu:NiO x films (as indicated within the XRD pattern in Figure 4c). The pure crystalline phases (w/o and 0.1 ratio Acac/oxidizer) of Cu:NiO x resulted to better PCEs 15.97% (average 14.48%) and 16.58% (average 14.85%), respectively, while, for the 1.5 ratio, the metallic Ni influence delivers lower PCE devices 14.90% (average 13.50%).  The experimental results presented within this manuscript using a triple cation Cs0.04(MA0.17FA0.83)0.96 Pb(I0.83Br0.17)3 perovskite formulation infer that the Cu:NiOx oxide's precursor films that were annealed at temperatures of 150 and 200 °C produce electronic films that cannot function as HTLs for efficient inverted perovskite solar cells. This is ascribed to the incomplete combustion that results in amorphous Cu:NiOx films with remnants. This result is in agreement with a previous report, where amorphous NiOx showed limited functionality as HTL when applied in organic solar cells [65]. On the other hand, as shown above, the pure crystalline phase of Cu:NiOx HTL was obtained by annealing at 300 °C for the precursors w/o and with a 0.1 ratio of Acac/oxidizer, whereas, for the 1.5 ratio, metallic Ni are likely to be present within Cu:NiOx films (as indicated within the XRD pattern in Figure 4c). The pure crystalline phases (w/o and 0.1 ratio Acac/oxidizer) of Cu:NiOx resulted to better PCEs 15.97% (average 14.48%) and 16.58% (average 14.85%), respectively, while, for the 1.5 ratio, the metallic Ni influence delivers lower PCE devices 14.90% (average 13.50%).

Discussion
In this work, we examined the solution combustion synthesis of Cu:NiOx films by using different molar ratios (w/o, 0.1, and 1.5) of fuel acetylacetone (Acac) to oxidizer (nitrates) precursors as well as various thermal processing annealing temperatures (150, 200, and 300 °C). Thermogravimetric analysis (TGA and DTA) results showed that the complete combustion process at ~150 °C can occur in bulk analogues. XRD measurements revealed that the corresponding Cu:NiOx films crystallize to NiO phase upon annealing temperature at 300 °C irrespective of Acac concentration, while, for lower annealing temperatures (150, 200 °C), no crystal phase was observed. XPS, AFM, UV-Vis spectroscopy, and contact angle measurements on the films strongly support the incomplete combustion of the Cu:NiOx thin films for annealing temperatures at 150 °C and 200 °C. XPS

Discussion
In this work, we examined the solution combustion synthesis of Cu:NiO x films by using different molar ratios (w/o, 0.1, and 1.5) of fuel acetylacetone (Acac) to oxidizer (nitrates) precursors as well as various thermal processing annealing temperatures (150, 200, and 300 • C). Thermogravimetric analysis (TGA and DTA) results showed that the complete combustion process at~150 • C can occur in bulk analogues. XRD measurements revealed that the corresponding Cu:NiO x films crystallize to NiO phase upon annealing temperature at 300 • C irrespective of Acac concentration, while, for lower annealing temperatures (150, 200 • C), no crystal phase was observed. XPS, AFM, UV-Vis spectroscopy, and contact angle measurements on the films strongly support the incomplete combustion of the Cu:NiO x thin films for annealing temperatures at 150 • C and 200 • C. XPS measurements of the Cu:NiO x film revealed the presence of a high atomic ratio of remnants for thermal annealing at 200 • C, which are remarkably reduced for films annealed at 300 • C. Surface topography images and thickness measurements via AFM and profilometer showed that the Cu:NiO x films annealed at 300 • C have a lower thickness (~50 nm) and roughness (~0.7 nm) compared to~80 nm thickness and~1.5 nm roughness for the Cu:NiO x films annealed at temperatures 150 • C due to remnants in the film. Moreover, Cu:NiO x films annealed at 300 • C have an improved hydrophilicity, showing a contact angle of 20 • , while the films annealed at 150 • C and 200 • C show angles of more than 60 • due to surface remnants. Regarding the optical absorption measurements, the 300 • C thermally annealed films exhibit a distinct absorption curve ascribed to the formed crystalline Cu:NiO x , while the lower annealing temperature films at 150 • C and 200 • C lack any strong absorption in the range of measured wavelengths due to the amorphous phase; these results are consistent with the paper reported in the XRD findings. To conclude, the presented solution combustion chemistry findings in Cu:NiO x thin films are confirmed by applying the various ratios of Acac/Oxidizer and annealing processing temperatures of SCS-based Cu:NiO x HTLs in triple cation-based Cs 0.04 (MA 0.17 FA 0.83 ) 0.96 Pb(I 0.83 Br 0.17 ) 3 inverted PVSCs. The Cu:NiO x HTLs annealed at temperatures 150 • C and 200 • C, irrespective of Acac/Oxidizer ratios, provided limited functionality in the PVSCs due to incomplete an combustion process that resulted in amorphous Cu:NiO x with remnants, as confirmed by the presented XRD and XPS measurements, respectively. The crystalline phase of Cu:NiO x HTLs and efficient inverted PVSCs performance obtained at an annealing temperature of 300 • C irrespective of the Acac/Oxidizer ratio. Following the solution combustion synthesis route that has been investigated within this manuscript, the Cu:NiO x crystalline HTLs annealed at 300 • C, with a 0.1 ratio of Acac/oxidizer resulting in 16.58% PCE for the triple cation-based Cs 0.04 (MA 0.17 FA 0.83 ) 0.96 Pb(I 0.83 Br 0.17 ) 3 inverted PVSCs.
Supplementary Materials: The following are available online at https://www.mdpi.com/article/ 10.3390/nano11113074/s1, Figure S1: XPS survey spectra of the Cu:NiO x films fabricated from precursor containing 0.1 Acac and annealed at (a) 200 • C and (b) 300 • C; Table S1: XPS calculated atomic ratios for the Cu:NiO x films fabricated form precursor containing 0.1 Acac and annealed at 200 and 300 • C; Figure S2: contact angle picture of water on films prepared using precursor containing w/o, 0.1, and 1.5 molar ratio of fuel (Acac) to oxidizer (Cu, Ni nitrates) annealed at 150, 170, and 200 • C; Figure S3: contact angle picture of water on films prepared using precursor 0.1 molar ratio of fuel (Acac) to oxidizer (Cu, Ni nitrates) annealed at 300 • C showing an angle of 20 • .