Xiaoxiao
Xu
ab,
Yuhai
Sun
c,
Dingchao
He
ab,
Zheng
Liang
ab,
Guozhen
Liu
a,
Shendong
Xu
ab,
Zhaoqian
Li
ab,
Liangzheng
Zhu
*ad and
Xu
Pan
*a
aKey Laboratory of Photovoltaic and Energy Conservation Materials, Institute of Solid State Physics, Hefei Institutes of Physical Science, Chinese Academy of Sciences, Hefei, 230031, P. R. China. E-mail: shuifeng139@126.com
bUniversity of Science and Technology of China, Hefei 230026, P. R. China
cPetroleum Engineering Technology Research Institute, Shengli Oilfield Company, Sinopec, Dongying 257000, P. R. China
dAnhui Province Key Laboratory of Condensed Matter Physics at Extreme Conditions, High Magnetic Field Laboratory, Chinese Academy of Science, Hefei 230031, China
First published on 17th November 2020
In the past few years, hybrid organic–inorganic perovskite solar cells (PSCs) have shown great potential for solar energy conversion in photovoltaic applications, due to their high power-conversion efficiency and low-cost fabrication. A high-quality perovskite film with large grain size and low-defect density plays a crucial role in high-performance FAPbI3-based PSCs. To date, only a few research efforts have focused on investigating the role of modifiers in precursor solutions of perovskite materials in terms of modulating the nucleation rate to obtain high quality perovskite films. Herein, by introducing ethanedithiol (ED) and tetrabutyl ammonium perchlorate (TAP) into the precursor solution, homogeneous nucleation occurred preferentially and high-quality α-FAPbI3 perovskite films with grain size scales up to the micron-level (∼1.7 μm) were obtained. Obviously, with a decrease in the number of grain boundaries, the defect density of the final modified films was also reduced. The modified FAPbI3-based PSCs achieved a PCE of 20.15% and maintained about 92% of the initial efficiency in an air atmosphere.
The crystallization process of perovskites is a complex process which involves two steps: (i) crystal nucleation and (ii) crystal growth. The classic theory of nucleation was first explained by Gibbs (1928).24 Heterogeneous nucleation is faster than homogenous nucleation because of the lower nucleation Gibbs free energy barrier, which was found to be associated with the contact angle.15 According to the previous literature,25 thiol ligands can be used as a soft Lewis base to passivate the surface Br vacancies in perovskites. L. Ruan et al.26 reported that 1-octanethiol ligands can fill the surface defects of perovskite quantum dots and raise the stability. Furthermore, Bi et al.15 found that the ammonium group could effectively mitigate the A cation vacancy type lattice defects, while thiol functionalization (–SH) is efficient in increasing the grain size and passivating coordinatively unsaturated Pb(II) ions on the surface.
Here, we introduce precursor solution modifiers to promote a homogenous nucleation process for large-scale perovskite grains. Ethanedithiol (ED) and tetrabutyl ammonium perchlorate (TAP) are used in the perovskite precursor solution to improve the contact angle at the interface of titanium oxide (TiO2). The two available chemicals (ED and TAP) containing two special functional groups (the ammonium group and thiol group) were employed as modifiers to enlarge the particle size and reduce the defect density of the film. A complex was formed by ED and TAP and had an influence on the wetting capability of the precursor solution. By employing the strategy of modulating the contact angle of the precursor solution on the substrate, homogeneous nucleation is promoted, whereas heterogeneous nucleation is suppressed, which is conducive to the reduction of nuclei and growth of large grains.
The role of the modifiers in the precursor solution was also explored by X-ray photoelectron spectroscopy (XPS) characterization. ED contains the electron-rich atom S, while TAP contains the electron-deficient atom N. ED, TAP and their mixture were spin-coated on FTO to be investigated by XPS. As listed in Fig. S4a (ESI†), the binding energy (BE) of S 2p in the mixture is 168.73 eV, while the BE of that is 168.24 eV from ED. In comparison with pristine TAP, the BE of N 1s shifts to 402.02 eV for the mixture from 402.16 eV. These results suggest that ED interacts with TAP when mixing in the precursor solution. Since TAP contains four butyl groups, the complex of TAP and ED will increase the effect of steric hindrance, and thus the viscosity of the solution increases. We explored the viscosities of both precursor solutions, which are listed in Table S1 (ESI†). The viscosity of the W/ED+TAP precursor solution is higher than that of the control solution. It is noted that high solution viscosity causes a large surface tension force,27,28 which causes reduced grain boundary mobility. By introduction of ED and TAP, the viscosity and surface tension force of the W/ED+TAP precursor solution both increase, leading to a higher contact angle of the W/ED+TAP solution on the substrate compared to the control solution. The effect of the modifier concentration was also explored in Fig. S5 and S6 (ESI†). The results of the perovskite film fabricated using the precursor solution with pure ED or TAP are shown in Fig. S11 and S12 (ESI†). An appropriate concentration of modifiers is crucial for high-quality films and high-performance PSCs. Therefore, the appropriate concentration of ED and TAP has a positive effect on enlarging the perovskite grain size.
According to the above morphology, we speculate that the two precursor solutions with verified surface contact ability have different nucleation mechanisms (Fig. 2). The smaller contact angle of the control precursor solution on TiO2 led to a higher nucleation rate and more nuclei, which tends to heterogeneous nucleation. On the contrary, due to the larger contact angle of the W/ED+TAP solution on TiO2, homogeneous nucleation is preferred to occur.
To further explore the perovskite nucleation and growth process in the two precursor solutions, a polarizing microscope is used to observe the films during annealing at 170 °C. In Fig. 3a, the control precursor solution produces more growth sites, which continue to increase throughout the annealing process, corresponding to a high nucleation rate. For the W/ED+TAP precursor solution, fewer growth sites are produced compared with the control precursor solution. It is observed that new growth sites are hard to produce and the initial nuclei continue to grow, which tends to homogeneous nucleation. Thus, the modified precursor solution is beneficial to the formation of a large-grain perovskite.
In terms of nucleation energy, the nucleation energy barrier for heterogeneous nucleation is much lower than that of homogeneous nucleation due to the shape of the nucleus. The relationship between the free energy for heterogeneous nucleation (ΔGheterogeneous) and homogeneous nucleation (ΔGhomogeneous) can be expressed by the following formula.29
ΔGheterogeneous = ΔGhomogeneous × f(θ), |
θ denotes the contact angle of the solid/liquid interface. Generally, both heterogeneous and homogeneous nucleation occur simultaneously in the process of perovskite nucleation. According to the formula, there exists a negative correlation between the contact angle and f(θ). When f(θ) decreases, heterogeneous nucleation is less likely to occur compared with homogeneous nucleation. Consequently, homogeneous nucleation is expected to occur more easily when the solution is deposited on substrates with larger contact angles.
Fig. S7a (ESI†) shows the ultraviolet-visible (UV-vis) absorption spectra of the perovskite films. Both of the perovskite films possess a similar absorption edge (about 830 nm) in accordance with previously reported publications.30–32 The W/ED+TAP film exhibits obviously higher absorption than the control film at wavelengths of 500–800 nm. The optical bandgaps of perovskite films are determined by Tauc's equation (αhν)2 = A(hν − Eg).31 The bandgaps of the control and W/ED+TAP films are calculated to be 1.49 eV and 1.50 eV, respectively. According to the photoluminescence (PL) spectra demonstrated in Fig. S7b (ESI†), the PL emission peak of the W/ED+TAP film is located at 812 nm, while the emission peak of the control film is located at 810 nm. This result is in agreement with the optical bandgaps derived from the UV-vis spectra. The emission intensity can be affected by the morphology and crystallinity of films according to previously reported publications.32–35 In Fig. S8 (ESI†), the control film shows the reflection peaks at 11.8°, indicating that the film contains little δ-FAPbI3. Notably, the δ-FAPbI3 peaks can hardly be seen in the XRD pattern from the W/ED+TAP film. This suggests that the modifiers in the precursor solution did not change the lattice structure or induce a new phase in the perovskite layer.
In the experiment, the PSCs are fabricated with the structure of FTO/c-TiO2/m-TiO2/perovskite/spiro-MeOTAD/Au. In Fig. 4, the photovoltaic performances of the devices based on the perovskite films coming from the two different precursor solutions are investigated. The devices are measured under standard AM 1.5G illumination at an irradiance of 100 mW cm−2 (with an active area of 0.09 cm2). Fig. 4a and Table S2 (ESI†) display the current density–voltage (J–V) curves and the corresponding photovoltaic parameters. The control device based on the control film has a PCE of 17.30%, while the W/ED+TAP device fabricated from the W/ED+TAP film shows an improved PCE of 20.15%. This result may be attributed to the high-quality and large-grained perovskite films. The improved Jsc can be observed from the incident photon to current conversion efficiency (IPCE) spectra in Fig. 4b. Moreover, there is little detected hysteresis in the reverse direction and the forward direction in Fig. S10 (ESI†). For the W/ED+TAP device, the IPCE spectrum shows over 80% from 400 nm to 700 nm and it is much higher than the control device. In addition, the reproducibility and uniformity of the PSCs are investigated. The W/ED+TAP devices exhibit the best efficiency from the PCE histograms presented in Fig. 4c. The circle represents the average values among these parameters. The W/ED+TAP devices obviously display a narrower distribution than the control devices in the parameters of FF, Voc and Jsc among the 20 devices. Thus, the high-quality and large-grained perovskite films play an important role in the device photovoltaic performance. To investigate the effect of the large grains on the defect density of the perovskite films, the defect densities are evaluated by using a FTO/c-TiO2/mp-TiO2/perovskite/PCBM/Au device structure through the space-charge-limited current (SCLC) technique.36–39 In Fig. 4d and e, the following three regions can be observed: (1) the ohmic region at low bias showing a linear relationship between the current and voltage, (2) the trap-filled limit region at intermediate bias exhibiting a steep increase in current, and (3) a trap-free (SCLC) region at high bias potential. The defect density is filled with the bias increases and all the defects are filled at the trap-filled limit voltage (VTFL). The defect density can be calculated by the equation nt = (2εε0VTFL)/(eL2), where ε, ε0, e and L refer to the dielectric constants of the perovskite, the vacuum permittivity, the elementary charge and the cross-sectional thickness of the perovskite film. Fig. S9 (ESI†) shows the cross-sectional morphology and the thickness of the perovskite film. The defect density is reduced from 1.828 × 1016 cm−3 for the control perovskite film to 1.259 × 1016 cm−3 for the W/ED+TAP perovskite film, reflecting a reduction over 30%. The defect density is obviously reduced in the large-grained perovskite films, which could be attributed to the reduced GBs of the W/ED+TAP perovskite films from the modified precursor solution.
![]() | ||
Fig. 4 (a) J–V curves and (b) IPCE spectra of the PSCs. (c) Photovoltaic parameters from 20 PSCs devices. (d and e) Space-charge-limited current (SCLC) measurements from the two kinds of devices. |
Considering the fact that the humidity stability of FAPbI3 depends on the film quality of FAPbI3, the water contact angles of the control and W/ED+TAP films are tested. Fig. 5a and b show that the W/ED+TAP film achieves an improved water contact angle of 87.1°, compared to the control film (81.8°). Hence, the W/ED+TAP film has higher humidity resistance due to the high quality and large grain size. Fig. 5c exhibits the normalized PCE variation curves of the unsealed devices under 15% RH in an air atmosphere. The W/ED+TAP device can maintain 92.5% of the initial PCE after 500 hours aging time. In contrast, the control device only shows 77.6% of the initial PCE. It is supposed that enlarging the grain size of the perovskite films has a significant impact on improving the moisture resistance of photovoltaic devices.
![]() | ||
Fig. 5 The water contact angles of the (a) control and (b) W/ED+TAP perovskite films. (c) Normalized PCE variation curves of the two kinds of unsealed PSCs when exposed to 15% RH. |
Footnote |
† Electronic supplementary information (ESI) available. See DOI: 10.1039/d0tc04833j |
This journal is © The Royal Society of Chemistry 2021 |