1. Introduction
The development of the solar cell preparation process gradually shifted from a complex process and high energy consumption to a route of the reduced preparation process and low energy consumption. At the same time, silicon solar cells have undergone evolution resulting in traditional Al-back surface field cells (Al-BSF cells),1 PERC cells,2,3 tunnel oxide passivating contact (TOPCon),4,5 heterojunction with intrinsic thin-layer (HIT).6 Among them, the HIT solar cell is one of the most typical examples of low-temperature fabrication routes. However, it still has four problems that need to be solved: a) The a-Si:H emitters must be heavily doped utilizing flammable and toxic gas precursors such as boranes and phosphine. b) The rather narrow direct band gap of approximately 1.7 eV in a-Si:H layers results in significant parasitic absorption of UV and blue light., which requires precise thickness control. c) The required plasma-enhanced chemical vapor deposition (PECVD) results in high manufacturing costs, which impedes the spread of solar cell technology. d) Needs of low-temperature metallization process.
Dopant-free carrier-selective contacts have gained great attention to the c-Si solar cells research community recently because they can significantly reduce optical parasitic absorption and simplify the fabrication process as compared to silicon heterojunction cell architecture.7 Most of those materials are large-band-gap chemical compounds and can be divided into dopant-free hole carrier-selective contacts (DF-HCSCs) and electron carrier-selective contacts (DF-ECSCs) depending on their functions. DF-HCSCs materials usually have ultra-high work functions forming up-bending valence band which promotes hole transport. The typical examples are transition metal oxides or nitrides such as MoOx, WO3, V2Ox, CrOx, and TiNx.8-11 PEDOT:PSS as organic DF-HCSCs also have been extensively explored on silicon solar cells.12 The highest PCE of dopant-free silicon heterojunction (SHJ) solar cells is 23.83% achieved by plasma treatments1.7 nm thick of MoOx layer as the hole-selective contact.13
DF-ECSC materials have been investigated as interface modification layers in SHJ solar cells due to their very low work function. Typical examples of DF-ECSCs are alkaline-earth metal salts such as LiFx, CsF, Cs2CO3, TiOx, MgFx and rare-earth compounds like EuF3, GdF3, and CeF3.7,14-23 Moreover, polymers like branched polyethyleneimine, Triton X-100, and MgAcac are reported as DF-ECSCs.24-26 Among them, the LiFx is one of the most typical DF-ECSCs materials. This interest sparked in 2016 when an SHJ solar cell with a front a-Si:H(i)/MoO3/ITO and rear a-Si:H(i)/LiFx/Al structure achieved an efficiency of 19.42%.7
Nevertheless, The development of the DF-ECSCs technology has been hindered by the severe degradation of LiFx/Al ECSCs. Additionally, the efficiency (23.07%) of the solar cells and their stability tolerance were improved by incorporating a shielded TiO2 (1.5 nm) layer between the a-Si:H passivating layer and the low work function (WF Φ) LiFx/Al electrodes().27 However, the poor thickness-tolerant of LiFx becomes another challenge for function as a DF-ECSC in the long-term run. The thickness of the LiFx stack must be carefully controlled at 1~2 nm to ensure effective tunneling. In addition, lithium compounds materials are expensive and toxic, which harms human health. Besides, the lack of long-term stability studies is also a shortcoming for most reported electron-selective contacts.
An approach that has great potential to mitigate these mentioned disadvantages is to apply healthcare DF-ECSC materials. Strontium fluoride (SrF
2) has been widely used in toothpaste being an annexing agent in our daily life.
28 It is also reported in light-emitting devices, optical imaging, and anion conductors thanks to its low energy phonon, high ionization, and good anion conductivity.
29 Recently, it has been revealed that SrF
2 can act as a buffer layer in organic solar cells to increase efficiency as a cathode interlayer.
30 But the utilization of SrF
2 for c-Si solar cells has not been researched in the current state of research. In this contribution, we developed thickness-tolerant and highly stable DF-ECSCs for c-Si by introducing 1.5 ~ 9 nm thermal evaporated SrF
2 films. The use of X-ray photoelectron spectroscopy (XPS) indicates that the SrF
2 film, fabricated through thermal evaporation, is stoichiometric. Ultraviolet photoelectron spectroscopy (UPS) measurements indicate that the thin film exhibits an ultralow work function. The morphology and element distribution of the interface were investigated using high-resolution transmission electron microscopy (HRTEM) and energy-dispersive X-ray spectroscopy (EDX).
The optimized solar cells with 4 nm SrF2 located partially on the rear side exhibit a champion efficiency of 21.56%. Lastly, the SrF
2/Al contact is shown to be stable beyond 5000 hours in the ambient air, demonstrating that SrF
2 is a prominent DF-ECSCs layer for cost-efficiency solar cells.
2. Results and Discussion
2.1. Photoelectron Spectroscopy of the SrF2 Thin Film.
The chemical state and compositions of the deposited
SrF2 thin films on polished c-Si(n) were characterized by XPS. The surface contamination C 1s (284.8 eV) was used as a calibration standard for the binding energies of the whole input. The core level spectrums of F and Sr are illustrated in Figure 1. Figure 1A depicts the principal peak of the F 1s spectra at 684.7 eV. Two peaks are observed in the core level of the Sr 3d spectrum, as shown in Figure 1B. The Sr 3d
5/2 main peak is at 133.6 eV binding energy. The peak at a binding energy of 135.5 eV is ascribed to the Sr 3d
3/2. Both characterization results were well consistent with the previous reports of SrF
2 thin films.
31 The atomic ratio of F divided by Sr is obtained from the core-level peak areas. The relative sensitivity factor of a fraction is 1.92, demonstrating it is approximately stoichiometry.
As shown in Figure 1C,D, the SrF2 work function and valence band were characterized by ultraviolet photoelectron spectroscopy (UPS) using a He I discharge source (21.22 eV). An Au reference sample has been used to calibrate the equipment before testing. Figure 1C indicates the WF (Փ) of the fabricated SrF2 films is 2.80 eV through the secondary electron cut-off vicinity, which is lower than the reported electron selective materials LiFx, TiO2, CeF3, and reference Al (4.1 eV) in the literature.7,18,22,32 Figure 1D shows the huge valence band offset of 7.1 eV between the Fermi energy level and the valence band maximum of the SrF2 film, which can effectively block the holes’ transport. The wide band gap of SrF2 is around 7.55 eV as generally reported in kinds of literature.33,34 It can be valued that the gap between the Fermi level and the electron affinity is 0.4 eV, leading to a small conduction band offset to lightly doped c-Si(n), benefiting the electron extraction from the conduction band by thermionic emission and tunneling. Thus, by inserting the SrF2 layer, the Fermi-level pinning effect at the c-Si(n)/Al interface could be eliminated, increasing the driving power for the photo-generated carriers’ separation, and confirming the outstanding ECSCs nature of n-type c-Si.