In order to limit the global warming below 2°C or 1.5°C, the world needs to reach a state where CO2 emissions achieve a net zero balance. This requires the decarbonization of the energy system, which is a major source of CO2 emission. Photovoltaic technology, which could directly convert solar energy into electricity, plays an important role in achieving the carbon neutralization. In recent years, organic-inorganic hybrid perovskite materials have gained tremendous development in the fields of photovoltaics, detectors, and light-emitting diodes, benefited from their unique optoelectronic properties and low-cost fabrication processes. The low resource consumption and low temperature solution processability make perovskite solar cells a serious contender for next-generation photovoltaics with simultaneous high-performance, low-cost and low CO2 footprint. Currently, single-junction perovskite solar cells (PSCs) have reached a certified power conversion efficiency (PCE) of 25.7%, approaching the Shockley–Queisser limit of single-junction solar cells. Stacking two perovskites with a complementary bandgap in tandem device provides a new and effective route to surpass the efficiency limit of single-junction PSCs.
A typical all-perovskite tandem solar cell consists a wide-bandgap (wide-Eg, 1.7-1.9 eV) top subcell and a low-bandgap (low-Eg, 1.1-1.3 eV) bottom subcell, which are usually connected by interconnection layers or mechanical stacking, forming the two-terminal (2-T) and four-terminal (4-T) configurations, respectively. We previously reported the combination of lead (Pb) and tin (Sn) precursors in forming mixed Sn-Pb low-Eg perovskites due to the bowling effect [J. Am. Chem. Soc. 138, 12360-12363 (2016)]. Such a mixed Sn-Pb perovskite, i.e., (FASnI3)0.6(MAPbI3)0.4, has a Eg as low as 1.25 eV, acting as a promising bottom subcell for all-perovskite tandem solar cells. Subsequently, we regulated the crystallization process of the low-Eg perovskite film via varying precursor concentration, and found that 1.6 M precursor delivered the optimal thickness of 620 nm. The final low-Eg PSCs achieved the first certified 17% efficiency for Sn-Pb low-Eg PSCs, and we further constructed 4-T all-perovskite tandem solar cells employing a semitransparent 1.57 eV perovskite top cell [Nat. Energy 2, 17018 (2017)]. Later, we incorporated lead chloride in the Sn-Pb precursor, and found that proper amount of chlorine enlarges the grain size, reduces electric disorder, and prolongs the carrier lifetime, finally leading to highly improved efficiency to 18.4% with a thick low-Eg absorber of 750 nm. The thicker perovskite layer could harvest more photons and enhance the spectral response at the long wavelength NIR region. We then fabricated 2-T all-perovskite tandem solar cells with chlorine-doped low-Eg perovskite as bottom subcell absorber, a 1.75 eV wide-Eg perovskite as top subcell absorber, and ultrathin Ag/MoOx/ITO as an interconnection layer, and consequently obtained the PCE of 21% [Nat. Energy 3, 1093-1100 (2018)].
High-quality perovskite absorber layer is the essential prerequisite for efficient PSCs, especially for tandem solar cells with multilayers. Currently, the most popular fabrication process of perovskite films is based on solution method, i.e., that is depositing perovskite precursor solutions on the conductive substrates and then annealing for final crystallized absorber layers. It is found that the solvents used for dissolving precursors always have coordination with lead halides by forming adducts in the wet film, and then evaporated out through ion-change with organic cations during annealing. The annealing process also plays vital roles in determining the crystallinity and quality of the final absorber films. In the commonly reported normal annealing process, which directly anneals the wet film with film side upward at certain temperatures, solvents are releasing only at vertical direction, giving no effect on surface morphology with relatively small grains and texture features. Moreover, compared with the solvent annealing process via excessive solvent vapor during grain growth, we found that large amount of solvents could dissolve the as-formed grains and then lead to relatively enlarged grain size in the perovskite films. However, this solvent annealing process is not compatible with the Sn-based compositions, since excessive solvents would accelerate the crystal dissolving, causing the appearance of pinholes. So far, universal annealing methods are rarely reported to effectively improve the film quality of different perovskites with various compositions.
We expect a universal annealing process that can simultaneously enhance crystallinity and keep homogenous film morphology for most kinds of perovskite materials. Inspired by the solvent annealing process and single-crystal growth method that cause grain recrystallization with external solvents in confined spaces, we adapted these methods by precisely controlling the amount of coordinating solvents and the releasing properties. Herein, we reported a close space annealing (CSA) strategy to regulate the crystallization process of perovskite absorber layers by slowing down the solvent releasing process in the intermediate film, and made the residual solvents involved into the grain growth process during annealing. This CSA method is compatible with perovskite films with various compositions and bandgaps. In such a CSA process, the coordinated solvents in the intermediate-phase perovskite can be accurately controlled by preheating the as-casted film with certain temperature and time. Then, the film is reversed back with the film side downward to the solvent permeable covers, and the residual solvents could slowly volatilize out, merge adjacent grains together and flatten the surface. As a result, the slowed solvent releasing process leads to larger grains and smoother surface of the perovskite film.
Schematic illustration of close space annealing process and possible grains merging dynamics.
It is worth mentioning that the releasing rates of coordinated solvents are different for various permeable covers, which are soda-lime glass, filter paper, and printing paper in our study. Two evaporation directions are illustrated, i.e., vertical and horizontal directions, of the solvents in the intermediate phase perovskite film, which could lead to obviously different morphology of the final perovskite films. When the vertical releasing rate is low, solvents would slowly escape out in a timely way and then pass along the film surface during annealing. This leads to recrystallization of grains and smooth surface. In the cases of papers-based CSA processes, the solvents could release around the edges (horizontal direction) and by permeation through the membrane, thus the grains are gently enlarged and the texture feature is kept to some extent. The permeated solvents could be further released at the rough surface of the hotplate, but does not contribute to the morphological changes of perovskite film. Note that the overall horizontal solvent releasing rate in printing paper-based CSA process is relatively slower than that of the filter paper-based one, leading to smaller grain size and more distinguishable surface texture of corresponding perovskite film compared to the latter one. Therefore, the vertical and horizontal solvent releasing should be balanced to facilitate the growth of high-quality films with both large grains and textured surface.
This CSA method is universal and feasible for both wide-Eg and low-Eg perovskites. The resulting PCEs of 1.25 eV low-Eg (FASnI3)0.6(MAPbI3)0.4 and 1.75 eV wide-Eg FA0.8Cs0.2Pb(I0.7Br0.3)3 PSCs are 21.51% and 18.58%, respectively. We further fabricated the all-perovskite tandem solar cells employing the CSA-processed wide-Eg and low-Eg perovskite layers, and obtained the PCEs of 25.15% and 25.05% for 4-T and 2-T all-perovskite tandem solar cells, respectively. Moreover, the improved film quality has positive effects on the device stability. One unencapsulated 2-T all-perovskite tandem solar cell retained 90% of its original PCE after continuous illumination for about 450 h when measured in a glovebox.
Our work offers a promising and effective approach to achieve efficient single-junction cells and multi-junction tandem solar cells.
This work titled “A universal close-space annealing strategy towards high-quality perovskite absorbers enabling efficient all-perovskite tandem solar cells” was published in the latest volume of Nature Energy journal (DOI: https://doi.org/10.1038/s41560-022-01076-9).