Effects of Cu, K, and Guanidinium Addition to CH₃NH₃PbI₃ Perovskite Solar Cells  ^†

Abstract

Fabrication and characterization of CH₃NH₃PbI₃ perovskite solar cell with the addition of copper (Cu²⁺), and potassium (K) or guanidinium (GA) was performed. The additive effects on the photovoltaic properties, morphologies, and crystalline structures were investigated by the experimental results, electronic structures, and thermodynamic stabilities. The stability and conversion efficiency of the perovskite solar cells was improved by incorporating Cu²⁺ at the lead site, and K or GA at the organic cation, CH₃NH₃, at A-site in cubic crystal. The simultaneous addition of Cu²⁺ and K to the perovskite crystal suppressed the crystal decomposition while inhibiting desorption of MA, improving the stability of the performance.

1. Introduction

The semiconductor lead halide perovskite (LHP) attracted attention as the active layer of LEDs in the 1990s, and in the 2000s it was rapidly developed as the thin-film absorber layer of solar cells, and has been actively researched worldwide [1]. There is also a growing number of computational simulation works [2,3,4,5,6]. Although perovskite solar cells (PSC) have the advantage of being inexpensive to produce, they have serious problems of low durability and environmental pollution by Pb [7]. The durability of PSC is caused mainly by the decomposition of the perovskite crystals due to methylammonium (MA) desorption. To solve these problems, attempts to improve the interface of perovskites and to introduce additives into the perovskite layer to improve the electronic properties have been studied [8,9,10,11]. Previous studies have reported that substitution of guanidinium (GA), which is larger than MA, can inhibit MA desorption, resulting in improved performance and long-term stability [12,13,14,15]. In addition, it has been reported that the introduction of potassium (K) improves the electron transport layer (ETL) and perovskite interface state [16,17,18,19]. It has also been reported that the addition of environmentally benign transition metals other than Pb reduces toxicity and changes the electronic state to improve the performance of PSC [20,21,22,23]. Among them, the environmentally friendly transition metal Cu²⁺ has been considered as a candidate for Pb replacement, but there are few reported cases [24,25,26,27,28].

The aim of this work is to fabricate and characterize devices doped with Cu²⁺, GA, and K. The photovoltaic properties, morphologies, and crystal structure were investigated by substitution of Cu²⁺, GA, and K. The stability of the performance was measured. In addition, first-principles calculations were compared with experimental results.

2. Experimental Procedures

The preparation of perovskite solar cells is shown in Figure 1 [29,30,31,32]. The FTO (F-doped Tin Oxide) substrates were ultrasonically cleaned with acetone and methanol and dried under nitrogen gas. TiO₂ precursor solution (0.15, 0.30 M) was prepared by adding 1-butanol (1 mL) to titanium diisopropoxide bis(acetylacetonate) (Sigma–Aldrich, St. Louis, MO, USA, 0.055, 0.11 mL). The 0.15 M TiO₂ precursor solution was spin-coated onto the FTO substrate at 3000 rpm for 30 s, and then the coated substrate was annealed at 125 °C for 5 min. 0.30 M TiO₂ precursor solution was spin-coated onto the TiO₂ layer at 3000 rpm for 30 s, and then the resulting substrate was annealed at 125 °C for 5 min. This process of forming the 0.30 M precursor layer was carried out twice. The FTO substrate was then baked at 550 °C for 30 min to form a compact TiO₂ layer. To form the mesoporous TiO₂ layer, TiO₂ paste was prepared by mixing TiO₂ powder (Aerosil, Tokyo, Japan, P-25, 200 mg) and poly(ethylene glycol) (Wako Pure Chemical Corporation, Osaka, Japan, PEG #20000 20 mg) with ultrapure water (1 mL). To this solution, acetylacetone (Wako Pure Chemical Corporation, Osaka, Japan, 20 µL) and surfactant (Sigma–Aldrich, St. Louis, MO, USA, Triton X-100, 10 µL) were added, mixed for 30 min, and then allowed to stand for 24 h to remove bubbles in the solution. The TiO₂ paste was then spin-coated on the compact TiO₂ layer at 5000 rpm for 30 s. The resulting cell was heated at 125 °C for 5 min and then annealed at 550 °C for 30 min to form a mesoporous TiO₂ layer. To prepare the perovskite compound, a mixture of CH₃NH₃I (2.4 M, Tokyo Chemical Industry, Tokyo, Japan), PbCl₂ (0.8 M, Sigma–Aldrich, St. Louis, MO, USA) in DMF (Nacalai Tesque, Kyoto, Japan, 0.5 mL) solution was prepared for the standard cell. Then, these perovskite solutions were spin-coated on TiO₂ at 2000 rpm for 60 s with air blow. This process was performed three times: a solution of DPPS (Osaka Gas Chemical, OGSOL SI-30-15, Osaka, Japan, 10 mg) was prepared in chlorobenzene (0.5 mL) and dropped onto the perovskite layer during the last 15 s of the third spin coating of the perovskite precursor solution [33,34,35,36,37]. Subsequently, annealing was performed at 200 °C in air. Spiro-OMeTAD (Sigma–Aldrich, St. Louis, MO, USA, 36.1 mg) was dissolved in chlorobenzene (Wako Pure Chemical corporation, 0.5 mL). Lithium bis(trifluoromethylsulfonyl)imide (Li-TFSI, Tokyo Chemical Industry, Tokyo, Japan, 260 mg) and FK209 (Sigma–Aldrich, St. Louis, MO, USA, 188 mg) were each added to acetonitrile (Sigma–Aldrich, St. Louis, MO, USA, 0.5 mL). Immediately before membrane formation, 4-tertbutylpyridine (Sigma–Aldrich, St. Louis, MO, USA, 18 μL), the prepared solution of Li-TFSI (10 μL), and the solution of FK209 (4 μL) were mixed at 70 °C for 30 min. The spiro-OMeTAD solution was then spin-coated on the perovskite layer at 4000 rpm for 30 s. All procedures were performed in air. Finally, a gold (Au) electrode was deposited to serve as the top electrode. The layer structure of the prepared solar cell is FTO/TiO₂/perovskite/spiro-OMeTAD/Au. The prepared perovskite solar cells were stored at a temperature of 22 °C and humidity below 30%.

3. Results and Discussion

3.1. First-Principles Calculation

The electron distribution diagrams and density of states (DOS) of MAPbI₃ and Cu²⁺-doped MAPbCuI₃ were calculated by first-principles calculations [38,39,40,41,42,43,44]. MAPbCuI₃ increase in DOS due to the overlap of Cu²⁺ 3d and I 5p orbitals near the HOMO. This indicates charge transfer from I 5p orbitals through Cu²⁺ 3d orbitals. From the results, it is predicted that the short-circuit current density (J_SC) increases due to the easier hole migration. The calculation of the band gap (E_g) suggested that the simultaneous addition of Cu²⁺ and K would expand the E_g related to the open circuit voltage (V_OC). E_g is the energy gap between the highest occupied molecular orbital (HOMO) and the lowest unoccupied molecular orbital (LUMO).

The cubic and crystal models doped with Cu²⁺ and GA are shown in Figure 2. From the energy calculation, the GA-doped system had increase of energy, yielding unstable state, as compared to the standard system. These results suggest that the stability of the performance of the perovskite crystal is improved by the simultaneous addition of Cu²⁺ and K. In the other case, the perovskite crystal with addition of GA might occur the crystal decomposition with the distortion of MA.

3.2. Device Characterization

Immediately after fabrication, the current–voltage characteristic (J–V) curve and the external quantum efficiency (EQE) spectrum were measured. The performance of the Cu²⁺ 2% + K 2% doped device was inferior to that of the standard device. The surface observation of SEM and EDX analysis showed uniform morphologies and monodispersed crystal grain in the perovskite crystals containing a slight K. The GA-doped perovskite crystal exhibited a tetragonal crystal system, consistent with calculated predictions and X-ray diffraction pattern.

The stability of conversion efficiency of the devices was investigated. The performance of the devices doped with Cu²⁺ and K simultaneously maintained to be about 10.8% after 28 days, while the performance of the standard devices was reduced by about 8.5%. The optical micrographs showed grain growth with increase of size. The XRD pattern indicate the lattice constant of the crystals was decreased by doping Cu²⁺ and K.

The schematic model of atomic diffusion in the Cu²⁺, K-added perovskite crystals was discussed. The lattice constant of the perovskite crystal was considered to have decreased by the substitution of K, which was not added immediately after the device fabrication, to the desorption position of MA with the passage of time. The simultaneous addition of Cu²⁺ and K to the perovskite crystal decreased the lattice constant, yielding wide bandgap related to V_OC, as compared to those in the MAPbI₃ perovskite crystal.

4. Conclusions

The effects of the co-addition of CuCl₂, KI, and GAI to MAPbI₃ on the photovoltaic properties, microstructures, and crystal structure were investigated. The calculation of the E_g suggested that the simultaneous addition of Cu²⁺ and K would increase expansion of the E_g related to the V_OC and to improve performance. Furthermore, it was found that the devices with 2% Cu²⁺ and 2% K-doping had the best stability and performance. The addition of K in the perovskite layer promoted the uniform crystal growth with an increase of grain size while inhibiting crystal decomposition, yielding greater stability of performance.