Highly Uniform Large-Area (100 cm²) Perovskite CH₃NH₃PbI₃ Thin-Films Prepared by Single-Source Thermal Evaporation

Abstract

In this work, we report the reproducible preparation method of highly uniform large-area perovskite CH₃NH₃PbI₃ thin films by scalable single-source thermal evaporation with the area of 100 cm². The microstructural and optical properties of large-area CH₃NH₃PbI₃ thin films were investigated. The dense, uniform, smooth, high crystallinity of large-area perovskite thin film was obtained. The element ratio of Pb/I was close to the ideal stoichiometric ratio of CH₃NH₃PbI₃ thin film. These films show a favorable bandgap of 1.58 eV, long and balanced carrier-diffusion lengths. The CH₃NH₃PbI₃ thin film perovskite solar cell shows a stable efficiency of 7.73% with almost no hysteresis, indicating a single-source thermal evaporation that is suitable for a large area perovskite solar cell.

1. Introduction

Hybrid organic-inorganic halide perovskites were considered to be an attractive superstar for low-cost solar cell applications due to their high-power conversion efficiency (PCE) and straightforward fabrication process [1,2]. In the past seven years, their rapid development led to PCE being raised from 3.8% to 22.7% by device optimization, perovskite composition and improved interface engineering [3,4]. The high performance is mainly due to the excellent properties of perovskite materials, such as extinction coefficients, electron and hole bipolar transport, medium band gap, small exciton binding energy and higher carrier diffusion length, etc. [5]. Although the highest certification efficiency of perovskite solar cells has exceeded 22% [6], the device areas are still small, with most of the areas being less than 1 cm². This raises issues such as peripheral effects, measurement contacts, film uniformity, etc. [7,8,9,10]. Therefore, the development of large-area perovskite solar cells is very important for future commercialization.

Nowadays, there are many preparation methods developed for high-efficiency perovskite solar cells, such as chemical solution process and vapor-assisted solution method [11,12]. However, chemical solution methods process with imperfect surface coverage and pinholes generated in large-area perovskite thin films fabrication, which seriously influences the solar cell’s stability and efficiency [13,14,15,16]. Gu et al. [17] reported a facile seed printing method to produce a millimeter-scale perovskite single crystal film with controlled thickness and high yield through a printing process. This further proves that the perovskite single crystal film can be directly printed. Nirmal Peiris et al. [18] improved one-step deposition assisted by N₂ blow drying technique for depositing CH₃NH₃PbI₃ on TiO₂/ZrO₂/NiO screen printed electrodes, in which the low temperature treated carbon electrode method was then printed on the perovskite layer by a doctor blade. Through such methods, a significant PCE of 10.8% was achieved and the photocurrent and voltage were significantly improved. However, oxygen and moisture in the air can seriously affect the preparation of the perovskite thin film. The single source thermal evaporation method to prepare perovskite thin film under higher vacuum can eliminate the influence of oxygen and water vapor on the quality of the perovskite film. As far as we know, single-source thermal evaporation is a promising method for preparing the large-area perovskite thin films [19]. Single-source thermal evaporation is advantageous because it results in a highly uniform, pinhole-less, smooth thin films which are especially important for large-area solar cells to avoid the introduced short-circuits and benefit of the higher full fill factor and PCE [20]. In addition, single-source thermal evaporation requires no solvent and is, therefore, ideal for fabricating tandem solar cells and solar cells on flexible substrates [21].

In this work, we report the reproducible fabrication of highly uniform large-area perovskite CH₃NH₃PbI₃ thin films by scalable single-source thermal evaporation with the area of 100 cm². The microstructural and optical properties of large-area perovskite thin films were investigated.

2. Materials and Methods

2.1. CH₃NH₃PbI₃ Crystals and Powder Preparation

The 9.22 g PbI₂ (99.99%, Xi’an Polymer Light Technology, Xi’an, China) and 3.14 g CH₃NH₃I (99.5%, Xi’an Polymer Light Technology, Xi’an, China) were mixed in r-butyrolactone (30 mL, 99%, TCI Chemicals Pvt. Ltd., Shanghai, China) in a round bottom flask for 24 h with constant magnetic stirring (MYP11-2, Meiyingpu Instrument Meter Manufacturing Co., Ltd., Shanghai, China). Then, the CH₃NH₃PbI₃ perovskite precursor solution was transferred to a Petri dish and maintained at 160 °C on a hot plate until all CH₃NH₃PbI₃ perovskite precursor solutions were evaporated to give CH₃NH₃PbI₃ crystals. Finally, the CH₃NH₃PbI₃crystals were grounded into powder as the film evaporation material.

2.2. CH₃NH₃PbI₃ Thin Films Preparation

A simplified schematic diagram of single-source thermal evaporation is shown in Figure 1. Prior to deposition, the large-area (10 cm × 10 cm) K9 (BK7) glass substrate was sequentially ultrasonically cleaned with acetone, ethanol and deionized water. Before evaporation, the substrate was treated with ozone for 20 min to improve the surface activity of the substrate 0.8 g of MAPbI₃ powder was placed into the evaporation boat (tungsten boat, Sky Technology Development, Shenyang, China), and when the pressure in the vacuum chamber was pumped down to below 1.5 × 10⁻³ Pa, the current work of the evaporation boat was rapidly increased to 140 A. Subsequently, the evaporation source temperature rapidly rose to the point where the MAPbI₃ powder evaporated. The substrate did not need to be heated, and the substrate rotated at 20 r/s to ensure a uniform film formation. The entire deposition process was about 3 min, and the actual deposited perovskite film thickness was about 500 nm. Lastly, the properties of MAPbI₃ thin films were improved by annealing at 140 °C for 20 min.

2.3. Device Fabrication

The perovskite solar cells with a device structure of ITO (coated-glass)/PEDOT:PSS (CLEVIOS PVP AI4083)/MAPbI₃ active layer/PC₆₁BM (99.5%, Xi’an Polymer Light Technology, Xi’an, China)/Ag was fabricated. A PEDOT:PSS (CLEVIOS PVP AI4083) aqueous solution was spin-coated onto a cleaned ITO glass substrate to form a 50 nm thick film (4500 rpm, 30 s), and the resulting PEDOT:PSS thin film was heated for 15 min at 150 °C and then transferred into a single source thermal evaporation system. The MAPbI₃ active layer was deposited on the PEDOT:PSS thin film by single-source thermal evaporation, and then annealed in a N₂-filled glovebox (Mikrouna GmbH, Shanghai, China) at 140 °C for 20 min. The PC₆₁BM solution (20 mg/mL in chlorobenzene, Xi’an Polymer Light Technology, Xi’an, China) was spin-coated on the complete perovskite MAPbI₃ thin film (3000 rpm, 30 s). Finally, the device fabrication was accomplished by evaporating a 100 nm thick silver film as the electrode.

2.4. Characterization

The large area MAPbI₃ thin film (100 cm²) and the detected A, B, C, D and E regions are shown in Figure 2. The crystalline structures of the MAPbI₃ thin films were analyzed by X-ray diffraction technique (Ultima IV, Rigaku, Tokgo, Japan) operated at 40 kV and 40 mA with Cu Kα radiation (λ = 0.15406 nm). The surface morphology and composition of the MAPbI₃ thin films were analyzed by a SUPRA 55 scanning electron microscopy (SEM, SUPRA55, Zeiss, Oberkochen, Germany), and an energy dispersive X-ray microanalysis system (EDS, Bruker QUANTAX200, Bruker, Billerica, MA, USA). The thickness of MAPbI₃ thin films was measured by a DEKTAK XT profilometer (Bruker, Billerica, MA, USA). The Time-Resolved Photoluminescence was carried out at the Time-Resolved (FluoTime 300, PicoQuant GmbH, Shimadzu, Japan) at a peak of 780 nm. The optical transmittance properties of the MAPbI₃ thin films were obtained by spectrophotometer (Lambda 950, Perkin Elemer, Waltham, MA, USA) in the wavelength range of 300–1000 nm. The current density–voltage (J–V) curves of the MAPbI₃ thin film solar cells were recorded under simulated AM 1.5 G conditions (100 mW/cm²) with a 2400 Source Measure Unit. The illuminated area of devices was precisely set with an area of 0.12 cm².

3. Results and Discussion

Figure 3a shows the XRD patterns of A, B, C, D and E regions of the large-area perovskite MAPbI₃ thin film. It was found that the A, B, C, D and E regions of large-area perovskite MAPbI₃ thin films reveal strong characteristic diffraction peaks of (110), (220) and (330) planes of MAPbI₃, usually assigned to the tetragonal crystal structure of halide perovskite [22,23]. Also, the MAPbI₃ films in all regions did not have any impurity peak, in which the characteristic peak intensity was high, the crystallinity was high, and the preferred orientation of the film was obvious. Figure 3b–d show the typical peak intensity change, FWHM (full width of half height) variation and grain size variation of the large-area perovskite (110), (220) and (310) respectively. Surprisingly, the peak intensities and half-widths of the (110), (220) and (310) characteristic peaks in regions A, B, C, D and E were almost the same. This shows that the large-area perovskite films have uniform crystallization in all regions with almost the same grain size. In addition, there was no PbI₂ or CH₃NH₃I detected in different regions, which indicates that it only goes through sublimation for the MAPbI₃ powder to thin film during the evaporation process.

The composition of the MAPbI₃ thin film is an important factor that affects the structure, electrical and optical properties of the absorber layer. The composition and elemental ratios of A, B, C, D and E regions of large-area MAPbI₃ films were analyzed by EDS. As shown in Figure 4, all five regions have two characteristic peaks at 2.48 and 3.98 keV, which corresponds to Pb and I elements. The energy-dispersive X-ray microanalysis system determines the ratio of Pb and I elements of large-area MAPbI₃ thin films as shown in

Table 1. The composition of as-deposited films and solvent annealed films determined by EDS analysis.

Sample	Pb (at.%)	I (at.%)	Pb/I
A region	27.93	72.07	0.38
B region	27.80	72.20	0.38
C region	27.69	72.31	0.38
D region	26.81	73.19	0.36
E region	25.84	74.16	0.35

. The Pb/I value for different regions were close to the theoretical stoichiometry of 0.333 indicating the prepared MAPbI₃ films with high compositional uniformity in a large area [24]. In addition, Figure 5 shows the distribution of Pb and I elements in different regions of large-area MAPbI₃ films. It shows that the locations of the Pb and I elements were nearly uniform across the entire plane, confirming the composition uniformity of large-area MAPbI₃ thin films.

The surface morphology of regions A, B, C, D and E of large-area MAPbI₃ thin films were measured by SEM. As shown in Figure 6, all regions of evaporated MAPbI₃ thin films exhibited full surface coverage with uniform grain size on the substrate. Even when the large-area was extremely dense and uniform. Figure 6f shows the cross-section of MAPbI₃ thin films. It can be found that the larger grains distributed evenly and the grains size approaches to the film thickness of about 500 nm. As shown in

Table 2. The thickness of the MAPbI₃ layer in five areas on large-area thin film.

Region	Thickness
A	524.7 nm
B	512.5 nm
C	513.5 nm
D	510.1 nm
E	531.5 nm

, the thickness of different region films was about 520 ± 10 nm. It can be concluded that the significantly smoother and more uniform perovskite thin films prepared by the single-source evaporation is promising for large area high quality perovskite solar cells.

To investigate the influence of perovskite film quality on carrier recombination character, time-resolved photoluminescence (TRPL) was carried out by measuring the PL decay of the emission peaks at 760 nm with an excitation wavelength of 532 nm, indicating the defect states and recombination process of perovskite films, as shown in Figure 7. It shows that the time-resolved PL spectra of the A, B, C, D and E regions of large-area perovskite glass/MAPbI₃ thin film demonstrated ultrafast lifetimes (average lifetime of about 9 ns, respectively I_A = 10.6 ns, I_B = 7.4 ns, I_C = 8.2 ns, I_D = 10.4 ns, I_E = 8.4 ns) and much slower decay processes after being fitted by the double-exponential decay function I = (A₁τ₁² + A₂τ₂²)/(A₁τ₁ + A₂τ₂) where τ₁ is the faster component of trap-mediated nonradiative recombination and τ₂ is the slower component correlated to radiative recombination [25,26,27]. According to the perovskite life length and the diffusion coefficient of electrons and holes D, and according to the formula L_D² = Dt_e, the electron and the hole average diffusion lengths of 184 nm and 156 nm were deduced respectively, which is higher than CH₃NH₃PbI₃ having prepared by solution method (about 100 nm) [28]. The formation of large grains can reduce the recombination of photoelectron-hole pairs and reduce the defect state of large-area perovskite MAPbI₃ thin films, which can increase the diffusion distance of carriers [29]. These high-quality films are promising for use in large-area perovskite solar cells.

Figure 8 shows the transmission spectra of A, B, C, D and E regions of large-area perovskite MAPbI₃ thin films. The transmissivity curves of all regions almost coincide, indicating good uniformity of the large-area perovskite thin films. There was a sharp absorption edge at 780 nm indicating that the large-area perovskite MAPbI₃ thin films with good crystallinity leads to the wide range of light absorption over the entire visible region. The bandgap energy in the A, B, C, D and E regions of the large-area perovskiteMAPbI₃ thin film was calculated from:

Αh_γ = A(h_γ − E_g)ⁿ

(1)

where α is the absorption coefficient, h_γ is the photon energy, n depends on the nature of transition, $A$ is the constant, and E is the bandgap energy [30]. As shown in Figure 9, The calculated band gap of each region of large-area perovskite MAPbI₃ thin film is about 1.58 eV, which was close to the theoretical value of 1.55 eV reported by Baikie’s group [31].

Figure 10a shows the device structure of MAPbI₃ thin film solar cells with ITO coated-glass/PEDOT:PSS/MAPbI₃/PC₆₁BM/Ag and the cross-section of the solar cell device was shown in Figure 10b. It can be found that the MAPbI₃ thin film solar cell distinguishes individual PEDOT:PSS and MAPbI₃ layers clearly but obscures PCBM and Ag. The device performance was assessed using current density–voltage (J–V) sweeps under simulated sunlight. As shown in Figure 10c, the PCE of the champion cell was 7.73% with an open circuit voltage (V_OC) of 0.89 V and a short circuit current (J_SC) of 12.80 mA/cm². In particular, there was almost no photocurrent hysteresis in the J–V curve of MAPbI₃ perovskite solar cells. This result shows that the single-source thermal evaporation MAPbI₃ yields efficient solar cells with a relatively simple and planar cell architecture. The statistics of the PCE based on 20 devices are shown in Figure 10d and an average of 6.78% was obtained.

4. Conclusions

In summary, we report a facile, efficient and reproducible new method for the fabrication of large-area (10 cm × 10 cm) perovskiteMAPbI₃ thin film by single-source thermal evaporation. The dense, uniform, less porous and high crystallinity of large-area perovskite thin films can be obtained. The element ratio of Pb/I was close to the ideal stoichiometric ratio of CH₃NH₃PbI₃ thin film. In addition, there is no PbI₂ or CH₃NH₃I detected in different regions, which indicates that it only goes through sublimation for the MAPbI₃ powder to thin film during the evaporation process. These films exhibit a favorable bandgap of 1.58 eV, long and balanced carrier-diffusion lengths. The CH₃NH₃PbI₃ thin film perovskite solar cell shows a stable efficiency of 7.73% with almost no hysteresis, indicating that the single-source thermal evaporation is suitable for the large area and efficient perovskite solar cell.