Effects of Copper Substitution in Methylammonium-Based Perovskite Solar Cells ^†

Abstract

The addition of copper bromide to the perovskite precursor solutions increased the conversion efficiencies of the devices. On the other hand, the short-circuit current densities decreased with an increase in the added amounts of copper (Cu). From first-principles calculations, the partial substitution of lead with Cu resulted in the formation of a Cu d orbital energy level in the forbidden band, which worked as a recombination center, causing the generated carriers to disappear. Experiments and calculations show the effects of Cu substitution on the electronic structures and the ability of the addition of Cu compounds to further improve the device performance.

1. Introduction

Perovskite materials with excellent photovoltaic properties have been studied using a wide variety of approaches, including first-principles calculations [1,2,3,4,5], machine learning [6,7,8,9,10], and device characteristics simulations [11,12,13,14,15], in addition to experiments [16,17,18,19,20,21,22]. Although the conversion efficiency and stability of perovskite solar cells are gradually improving, most perovskite materials with excellent performance contain toxic Pb. In order to reduce the perovskite solar cells’ toxicity for commercialization, alternative elements to Pb are being investigated [23,24,25,26,27,28,29]. In previous studies investigating the effect of Cu compound addition in methylammonium (MA)- or Cs-based perovskites, it was reported that the additions of small amounts of Cu to the perovskite precursor solution increased the grain size and improved the film quality, which contributed to improved device properties [30,31,32,33,34]. Furthermore, the combination of Cu with alkali metals or organic cations more stable than MA improved the conversion efficiency and stability of perovskite solar cells [35,36,37,38].

In this study, the effects of the addition of Cu compounds to the perovskite precursor solution and the substitution of Pb with Cu on the device properties and electronic structure were investigated [39]. The amount of Cu added was varied in the range of 0, 1, 2, 3, and 12.5%, and the current–voltage characterization and X-ray diffraction measurements were performed. In addition, first-principles calculations were performed to determine the effect of Cu substitution in methylammonium-based perovskite crystals from the band structure and partial density of states.

2. Device Fabrication and Computational Conditions

The structure of the fabricated perovskite solar cell is fluorine-doped tin oxide glass/compact TiO₂/mesoporous TiO₂/CH₃NH₃PbI₃/spiro-OMeTAD/Au. CuBr₂ was used as the Cu compound to be added to the perovskite precursor solution, with 0% Cu as the standard device and different Cu additions in the range of 1, 2, 3, and 12.5%. The details of the experimental methods are described in the previous papers [35].

A 2 × 2 × 2 supercell was built, and one of the eight B-sites was substituted with Cu to produce a structural model of MAPb_0.875Cu_0.125I₃. The energy gap and carrier effective mass were calculated from the band structure, and the information on orbitals was obtained from the partial density of states (pDOS). The details of the calculation method are described in the previous papers [39].

3. Results and Discussion

Figure 1 shows the cell parameters as a function of the amount of Cu compound added to the perovskite precursor solution.

Table 1. Device properties of perovskite solar cells.

Cu2+	JSC	VOC	FF	RS	RSh	η	ηave	Eg
(%)	(mA cm−2)	(V)		(Ω cm2)	(Ω cm2)	(%)	(%)	(eV)
0	21.6	0.879	0.657	4.04	1664	12.5	10.4	1.55
1	20.9	0.898	0.681	3.49	1526	12.8	11.6	1.55
2	20.4	0.919	0.701	3.03	1507	13.1	11.7	1.56
3	20.2	0.892	0.638	4.85	1213	11.5	10.5	1.56
12.5	14.3	0.870	0.591	5.07	621	7.33	6.07	1.60

shows the values of the cell parameters obtained from the current–voltage characterization. The highest open circuit voltage (V_OC) and fill factor (FF) were obtained when 2% Cu was added, and the conversion efficiency was higher than that of standard devices. The short-circuit current density (J_SC) tended to decrease with an increase in the Cu addition, and the conversion efficiency decreased compared to the standard device when the Cu content exceeded 2%.

Table 2. Parameters obtained from X-ray diffraction measurements.

Cu2+	I100/I210	Lattice Parameter	Crystallite Size
(%)		(Å)	(Å)
0	3.4	6.267(0)	394
1	3.7	6.269(1)	549
2	4.6	6.268(0)	663
3	4.2	6.268(1)	548
12.5	2.2	6.243(0)	625

shows the results of the X-ray diffraction pattern analysis. The crystallite size increased with the addition of Cu to the precursor solution, with 2% Cu showing the highest conversion efficiency and the highest perovskite (100) plane orientation. The increase in the lattice parameter with the addition of small amounts of Cu was attributed to the lattice distortion due to Cu substitution. For 12.5% Cu, the ionic radius of Cu is smaller than that of Pb; thus, the lattice shrank as more Cu was incorporated into the perovskite crystals.

First-principles calculations were performed to investigate the cause of the decrease in J_SC with an increase in the Cu addition. Figure 2 shows the calculated band structure and density of partial states.

Table 3. Parameters obtained from first-principles calculations. The * means effective mass.

Model	Total Energy	Energy Gap	me*/m0	mh*/m0	mh*/m0	Contribution of Transition	Oscillator Strength
	(eV cell−1)	(eV)	Pb	I	Cu	(%)
MAPbI3	−3495	1.480	0.229	0.208	−	16.7	0.0168
Cu 12.5%	−3363	1.438	0.240	0.239	0.355	62.2	0.0287

shows the values of the parameters obtained from the first-principles calculations. From Figure 2c, the valence and conduction bands are composed of I p and Pb p orbitals, respectively. When Pb is partially substituted by Cu, the band gap decreases due to the lower energy of the Pb p orbital. Furthermore, an energy level of the Cu d orbital was formed in the forbidden band. Considering the experimental results, the energy level of the Cu d orbital worked as a defect level, causing the loss of the generated carriers and the reduction in the J_SC. In the absence of Cu, electrons are excited from the I p orbital to the Pb p orbital. In the presence of Cu, electrons can be excited from the I p orbital to the Cu d or Pb p orbital, resulting in two patterns of excitation processes. Despite the increase in the excitation probability, the decrease in device properties with an increase in the Cu addition means that it is difficult to extract electrons excited from the I p orbital and trapped in the Cu d orbital as charge carriers. Experimental and computational results indicate that the Cu substitution has a negative effect on the electronic structure and that the defect levels formed in the forbidden bands reduce the J_SC. Therefore, the enhancement of device properties by the addition of Cu compounds to the precursor solution is attributed to improvements in the microstructure of the perovskite film, such as an increase in crystallite size and perovskite crystal (100) plane orientation.

4. Conclusions

The addition of 2% Cu to the perovskite precursor solution improved the V_OC and FF and enhanced the conversion efficiency of the device. The calculated band structure and density of partial states indicate that the substitution of Pb for Cu has a negative effect on the electronic structure, with the formation of Cu d orbital energy levels in the forbidden band causing carrier recombination and lowering of the J_SC. Therefore, it is difficult to adopt Cu as a replacement element for Pb, but the addition of small amounts of Cu compounds to the perovskite precursor solution was shown to contribute to the further enhancement of device properties.