Introduction

Solar energy is a green clean energy source with a wide range of applications. Organic–inorganic hybrid perovskite solar cells have drawn great attention thanks to their high photoelectric conversion efficiency, simple manufacturing, and low cost1,2,3. The initial efficiency of perovskite solar cells was only 3.8%, but recently, the efficiency of single-junction perovskite solar cells has reached a high record to 24.2%4 and the perovskite/silicon tandem solar cell incresed to 25.2%5.

The device structures, recombination mechanisms, interface engineering, and material synthesis should be studied to further improve the efficiency. The key to improving the efficiency of solar cells is to improve the quality of perovskite film6. Solvent annealing is reportedly as an effective way to improve the crystallinity of some organic semiconductors, but this method’s effectiveness in inorganic semiconductors has not yet been confirmed. The current study shows that solvent annealing can be applied to increase the crystallinity and grain size of the perovskite films. Huang et al. found that the introduction of DMF/DMSO solvent vapor during the growth of CH3NH3PbI3 crystals can effectively improve the crystallinity and grain size of the film, passivate the film defects, and improve the device performance7. Liu et al. reported the introduction an anti-solvent vapor (e.g., alcohol vapor) to replace DMF vapor during the annealing procedure, which can improve the growth of perovskite crystals and increase the grain size of the perovskite MAPbI3 crystals, thus high crystallinity and pinhole-free MAPbI3 film could be obtained8. Zhang et al. created a different solvent atmosphere to anneal the perovskite film, the results show that the perovskite crystal quality was significantly improved when annealing in a poor mixed solvent [IPA: DMF = 100:1 (v/v)]9.

In addition to efficiency, the environmental stability and light stability under operating conditions are other key factors in photovoltaic and other optoelectronic applications10. Compared with its 3D counterpart, the Ruddlesden–Popper phase layered 2D perovskite thin film shows good stability but low efficiency11,12. The poor efficiency could be attributed to the inhibition of the out-of-plane charge transport of organic cations, which acted like insulating spacers between conductive inorganic plates13. Recently, attempts to utilize two-dimensional layered hybrid compounds in perovskite films have achieved breakthrough results. Smith I.C. et al. reported a layered (PEA)2(CH3NH3)2Pb3I10 perovskite light absorber for solar cell applications. The solar cell has an interesting open circuit voltage of 1.18 V and a photoelectric conversion efficiency of 4.73%. Moreover, this absorber was relatively stable up to 46 days in air with 52% relative humidity14,15. Mitzi D.B. explored the photovoltaic-related properties of 2D MA2Pb (SCN)2I2 perovskite, which can be used as an absorber layer for the top cell of a tandem solar cell16. Although superior device performance has not yet been achieved, this 2D layered mixtures have been demonstrated as effective new perovskite film with adjustable photoelectric properties and enhanced air stability17,18. In contrast to 3D perovskite, 2D perovskite [CH3(CH2)3NH3)2(CH3NH3)n-1PbnI3n+1 (BA)2(MA)n-1PbnI3n+1, n = 1, 2, 3, 4, …, ∞] have better optoelectronic property tunability because of their greater degree of freedom in quantum mechanics and chemistry, and, more importantly, higher environmental stability. Therefore, the development of 2D perovskite thin films will directly aid in improving the stability of perovskite solar cells.

Based on vacuum preparation method, the dual-source or single-source thermal evaporation methods, are also available to deposit perovskite thin films19,20,21. However, the dual-source thermal evaporation requires precise simultaneous control of the evaporation source of organic and inorganic materials, but the effective control of the film compounding process is very difficult. The easy deviation from the stoichiometric ratio directly leads to a decrease in film quality and repeatability22. To our knowledge, single-source thermal evaporation is a effective method for preparing large-area, high-efficiency perovskite solar cells23. In this study, 2D perovskite (BA)2(MA)3Pb4I13 thin film was prepared by single-source thermal evaporation, and the effects of solvent annealing on the microstructural and optoelectronic properties of the thin film were investigated.

Results

Figure 1a shows the crystal structure of the (BA)2(MA)3Pb4I13 powder and the thin films including the as-deposited and solvent-annealed thin films. The prepared (BA)2(MA)3Pb4I13 powders have characteristic diffraction peaks of 2D perovskite. However, the as-deposited thin film shows broad peaks, which indicates low crystallinity. After the solvent annealing, the stronger characteristic diffraction peaks of the (060), (080), (111), (131), and (222) planes usually refer to the 2D (BA)2(MA)3Pb4I13 perovskite crystal structure. These results indicate that the perovskite crystallinity is increased, with fewer low-dimensional defects and/or larger perovskite grain sizes, and less scattering of internal grain boundaries (Fig. 1b)24,25,26,27. Figure 1c displays the FWHM of the 2D perovskite (060), (080), and (111) peaks. The FWHM of the solvent-annealed thin film is significantly smaller, indicating better crystallization28. Based on the Debye-Scherrer formula, D = Kλ/(β cosθ), (D is the grain size of crystals, K is a constant, λis the wavelength of the X-ray, βis the FWHM, and θ is the diffraction angle29), the grain sizes of the as-prepared and solvent-annealed thin films were calculated and the results are shown in Fig. 1d. After solvent annealing, the 2D perovskite grain size becomes markedly larger, suggesting that the solvent annealing can improve the crystallinity of the thin film, which might lead to a higher efficiency for device.

Figure 1
figure 1

XRD patterns of the (a) (BA)2(MA)3Pb4I13thin films prepared by single-source thermal evaporation, (b) changes in the typical peak intensities, (c) FWHM changes, and (c) change in grain size.

Full size image

The composition of the (BA)2(MA)3Pb4I13 film is an important factor affecting the structural, electrical, and optical properties of the light-absorber. Figure 2 and Table 1 show the composition of the (BA)2(MA)3Pb4I13 thin films measured by EDS. Two typical peaks located at 2.48 and 3.98 keV, corresponding to the Pb and I elements. The atomic ratio of Pb to I of the as-deposited thin film is approximately 0.392. It decreases to 0.365 for the solvent annealed films, which is much close to the stoichiometry of the (BA)2(MA)3Pb4I13 film, indicating the formation of pure-phase 2D perovskite thin films. Figure 3 displays the elemental distribution of the thin film after solvent annealing, and shows that the Pb and I have uniform distribution in the entire plane without element enrichment or deficiency.

Figure 2
figure 2

The EDS spectral line patterns of the (a) as-deposited and (b) solvent-annealed films.

Full size image
Table 1 Compositions of the as-deposited and solvent-annealed films determined by EDS analysis.
Full size table
Figure 3
figure 3

EDS-MAP of the (BA)2(MA)3Pb4I13 thin films (a) before and (b) after solvent annealing; distribution of (c) I and (d) Pb elements.

Full size image

Figure 4 show the morphology of as-deposited and solvent-annealed (BA)2(MA)3Pb4I13 thin films. Figure 4a illustrates that the as-deposited thin film exhibits complete surface coverage but with small grains on the substrate. The cross-section in the inset of Fig. 4a shows no distinct grains which may easily lead to poor reproducibility and photocurrent hysteresis of the 2D perovskite solar cells30,31,32,33. Solvent vapor of γ-butyrolactone introduced during the annealing of the 2D perovskite causes the recrystallization of (BA)2(MA)3Pb4I13. Precise control of recrystallization can improve the quality of perovskite film34,35,36. After the γ-butyrolactone vapor annealing treatment (Fig. 4b), the (BA)2(MA)3Pb4I13 thin film shows denser and larger grain distribution, and the defects are significantly reduced. Therefore, more photogenerated charges can successfully reach the electrode instead of recombining in the grain boundary.

Figure 4
figure 4

Morphology of the (a) as-deposited and (b) solvent-annealed film (b). Cross-sectional scanning electron micrographs of the (a) as-deposited and (d) solvent-annealed films.

Full size image

Time-resolved PL (TRPL) decay measurements were performed to study the influence of the charge transfer process in the (BA)2(MA)3Pb4I13 thin film. Figure 5 displays the patterns and the lifetimes of the charge carriers in the thin films were estimated by fitting the data37. The average decay time (τave) of the (BA)2(MA)3Pb4I13 films were calculated according to the formula, τa = (A1τ12 + A2τ22)/(A1τ1 + A2τ2), and the charge carrier life time extracted from the as-deposited (BA)2(MA)3Pb4I13thin films is 1.34 ns38. Under γ-butyrolactone solvent annealing, the TRPL lifetimes are increased to 26.29 ns, which is in very good agreement with previously reported values39. Using the formula \({{L}_{D}}^{2}={\rm{D}}{\tau }_{s}\), in which the fluorescence lifetime of perovskites and the diffusion coefficients of electrons and holes D, the electron and hole diffusion lengths are deduced to be 314 nm and 266 nm, respectively40. The longer lifetime (τs) indicates an increased charge-diffusion length (LD) of the (BA)2(MA)3Pb4I13 thin films because of the better crystallization in the solvent-annealed thin films. This characteristic reduces the recombination of photoelectron–hole pairs.

Figure 5
figure 5

Time-resolved photoluminescence lifetime of the as-deposited and solvent-annealed (BA)2(MA)3Pb4I13 thin films.

Full size image

The optical transmittance properties were obtained by a UV/visible/near-IR spectrophotometer in the wavelength range of 300–1000 nm. Figure 6 shows the transmittance spectra for the as-deposited and solvent-annealed (BA)2(MA)3Pb4I13 thin films prepared by single-source thermal evaporation. As shown in the previous SEM image, more defect states due to smaller grains, the absorption edge is clearly moving toward the IR region after solvent annealing, indicating wide range of light absorption caused by enhanced crystallinity. The absorption range of the as-deposited thin film was lower than that of the annealed thin film due to the improvement of the film’s crystallinity as we mentioned above. Compared with perovskites with multiple nano-grains, the solvent-annealed perovskite film has fewer grain boundaries, this facilitates a greater range of light absorption by the absorbing layer. The band-gap energy can be calculated as \(\alpha h\nu ={\rm{A}}{(h\nu -{E}_{g})}^{n}\), where α is the absorption coefficient, hv is the photon energy, A is the constant, n depends on the nature of transition, and E is the band-gap energy41. Figure 7 shows that the band gap of the as-deposited thin film is 2.40 eV and decreases to 1.89 eV after solvent annealing, which is close to the theoretical value42. Hydrogen bond exists in perovskite, the presence of hydrogen bonds may affect the optical band gap of (BA)2(MA)3Pb4I13, Similarly, Filip et al.43,44 have experimentally shown perovskite tunable optical bandgaps.

Figure 6
figure 6

Optical transmittance spectra of the as-deposited and solvent-annealed (BA)2(MA)3Pb4I13 thin films.

Full size image
Figure 7
figure 7

Estimation of the optical band gap of the as-deposited and solvent-annealed (BA)2(MA)3Pb4I13 thin films.

Full size image

The perovskite solar cells with a device structure of ITO/PEDOT: PSS/2D perovskite (BA)2(MA)3Pb4I13/PC61BM/Ag (Fig. 8a) were fabricated. PEDOT: PSS and PCBM were the hole and electron transport layers, respectively. Figure 8b shows the J-V curves of the (BA)2(MA)3Pb4I13 perovskite solar cells based on the as-deposited and solvent-annealed (BA)2(MA)3Pb4I13 thin film. The Jsc, Voc, FF, PCE, Rs, and Rsh of the corresponding devices are summarized in Table 2. In the 2D perovskite solar cell, the (BA)2(MA)3Pb4I13 thin films without γ-butyrolactone vapor treatment are presented as the as-deposited PSCs, while the films treated with γ-butyrolactone are presented as the solvent-annealed PSCs. The as-deposited PSCs exhibit Jsc of 6.51 mA/cm2, VOC of 0.85 V, FF of 45.82%, Rs of 729.93 Ω, and Rsh of 616.78 Ω. These characteristics result in a low PCE of 2.54%. Compared with the as-deposited PSCs, when the γ-butyrolactone solvent vapor is introduced during annealing, the performance of the 2D perovskite solar cell is significantly enhanced. Jsc substantially increases to 10.98 mA/cm2, Voc increases to 0.95 V, Rs is reduced to 26.55 Ω, and PCE increases to 4.67%. The γ-butyrolactone vapor during annealing can lead to enhanced crystallinity and larger grain size of (BA)2(MA)3Pb4I13, passivate defects, and improve device performance. Hence, the solvent annealing produces a high quality non-porous 2D perovskite film with a high purity phase, complete surface coverage, and good crystallinity. These characteristics can suppress internal recombination and leakage currents and promote photoelectric conversion of 2D perovskite solar cells.

Figure 8
figure 8

Device structure of (a) 2D perovskite solar cell; (b) J-V characteristics before and after solvent annealing.

Full size image
Table 2 Photovoltaic performances and fitting parameters used for the impedance spectra of the (BA)2(MA)3Pb4I13-based perovskite solar cells before and after solvent annealing.
Full size table

Conclusions

The effect of solvent annealing on grain growth is investigated to enhance the photovoltaic-device efficiency of 2D perovskite (BA)2(MA)3Pb4I13 thin film prepared by single-source thermal evaporation. Solvent annealing can effectively enhance the crystallization of (BA)2(MA)3Pb4I13 thin film with denser and larger crystal grains. The element ratio of Pb/I is close to the ideal stoichiometric ratio. The films show a favorable band gap of 1.896 eV and long electron and hole diffusion lengths of 314 nm and 266 nm, respectively. The performance of the (BA)2(MA)3Pb4I13 perovskite solar cell is significantly enhanced, that is, Jsc remarkably increases to 10.98 mA/cm2, Voc increases to 0.95 V, and Rs is reduced to 26.55 Ω. The solvent-annealed (BA)2(MA)3Pb4I13thin-film solar cell prepared by single-source thermal evaporation shows an efficiency of 4.67%. Thus, the proposed method is promising for preparing large-area and efficient 2D perovskite solar cells.

Methods

(BA)2(MA)3Pb4I13 crystal and powder preparation

PbI2 (7.38 g, 99.99%, Xi’an Polymer Light Technology), MAI (1.91 g, 99.5%, Xi’an Polymer Light Technology), and BAI (1.61 g, 99.5%, Xi’an Polymer Light Technology) were mixed in γ-butyrolactone (150 ml, 99%, TCI) in the beaker for 24 h with constant magnetic stirring. The 2D perovskite solution was then transferred onto a glass culture dish and maintained at 150 °C on a hot plate. Until all 2D perovskite solutions were evaporated, (BA)2(MA)3Pb4I13 crystals can be obtained as shown in Fig. 9. Then, the prepared (BA)2(MA)3Pb4I13 crystals were ground into powders as the film evaporation material.

Figure 9
figure 9

Schematic of the fabrication procedure including crystal preparation, single-source thermal evaporation, and solvent annealing.

Full size image

(BA)2(MA)3Pb4I13 thin-film preparation

Figure 9 shows the fabrication of the crystals, including the (BA)2(MA)3Pb4I13 crystals preparation, single-source thermal evaporation, and solvent annealing. Prior to deposition, the ITO glass substrate was cleaned, and 0.8 g of (BA)2(MA)3Pb4I13 perovskite powder was weighed. The powder was placed in the evaporation boat. The distance from the evaporation source to the substrate was 25 cm, and the substrate speed was 40 rpm. Once the chamber pressure was pumped down to below 1 × 10−3 Pa, the working current of the evaporation source was rapidly raised to 150 A, and then the film was deposited. Until the powder was completely evaporated, the as-deposited 2D perovskite (BA)2(MA)3Pb4I13 thin films has a thickness of approximately 400 nm. Solvent annealing was then performed under a γ-butyrolactone atmosphere (40 μl) at 120 °C for 20 min.

Device fabrication

The perovskite solar cells have a device structure of ITO/PEDOT: PSS/(BA)2(MA)3Pb4I13/PC61BM/Ag was prepared. An aqueous solution of PEDOT-PSS (CLEVIOS PVP AI4083) was spin-coated onto ITO glass substrate to form a 50-nm thick thin film (4500 rpm for 40 s). The obtained PEDOT-PSS film was placed on a hot plate at 160 °C for 20 minutes and then transferred to a single source evaporation deposition system. The (BA)2(MA)3Pb4I13 absorber layer was deposited by single-source thermal evaporation and then solvent annealed in a N2-filled glove-box for 20 min at 120 °C. The PC61BM solution (20 mg/ml in chlorobenzene) was spin-coated on the (BA)2(MA)3Pb4I13 thin film at 3000 rpm for 30 s. Lastly, 90 nm-thick Ag cathode was prepared by thermal evaporation in a vacuum of approximately 3.0 × 10−4 Pa.

Characterization

The crystalline structure of the (BA)2(MA)3Pb4I13 thin films and powder were analyzed by X-ray diffractometer (Ultima IV). The composition and surface morphology of perovskite film and powder were analyzed by energy-dispersive X-ray microanalysis system (Bruker QUANTAX 200, Bruker, Billerica, MA, USA) and a SUPRA 55 scanning electron microscope, respectively. The time-resolved PL was recorded using the steady-state spectroscopy and time-resolved (Fluo Time 300, Pico Quant GmbH). The thickness of the (BA)2(MA)3Pb4I13 thin films were measured by a DEKTAK XT profilometer (Bruker, Billerica, M A, USA). The optical transmittance properties were obtained by a UV/visible/near-IR spectrophotometer (Lambda 950, PerkinElmer). The J-V curves of the 2D perovskite solar cells were recorded in simulated AM 1.5 G conditions (100 mW/cm2) with a Keithley 2400 Source Measure Unit.