Experimental methods
Films Deposition and Characterization: n-type c-Si wafers (100) with a thickness of 170 µm and resistivity of 1–3 Ω cm were cleaned by the Radio Corporation of America (RCA) procedure and then dipped in HF (1-2%) solution. Prior to deposition of ZnTiO3, thin SiO2 passivation layers were grown by thermal treatment under 300 °C for 30 min in air atmosphere. Following this, ZnTiO3 films were thermally evaporated at the rate of 0.1 Å s-1 from high-purity (99.9%) ZnTiO3 powder, the thickness of ZnTiO3were monitored using a crystal oscillator, and then metallized with a thermally evaporated Al electrode. Contact resistivity (ρ c) of c-Si/SiO2/ZnTiO3/Al annealed at different temperatures was measured using the transfer length method (TLM), and the TLM pattern has the length of 20 mm and the pad spacing of 0.5, 1, 1.5, 2, 3, and 4 mm, respectively. The thickness of ZnTiO3 was varied and the thickness of Al was 150 nm. The ρ c was extracted from current-voltage (IV ) curves obtained from a Keithley 2400 source meter. To examine the thermal stability of the contacts, the samples were placed on a hot plate at different temperatures for 30 minutes in air ambient. As a comparison, a TiO2 film with a thickness of 4 nm was also deposited as an electron-selective layer by atomic layer deposition (ALD), using titanium tetrakis (dimethylamide), and H2O as precursors at 150 °C, and the sample was also annealed under the same condition. Two ZnTiO3 samples were fabricated for the chemical composition and work function (WF ) measurement by X-ray photoelectron spectroscopy (XPS) (ThermoFischer, ESCALAB Xi+) and ultraviolet photoelectron spectroscopy (UPS) (Kratos AXIS Ultra DLD). They were prepared as following: 50 nm thick ZnTiO3 layer and 100 nm thick Al were deposited on n-type polished c-Si wafers sequentially. Only one sample was annealed on a hot plate at 300 °C and the other one did not receive annealing. Then the Al layers of the two samples were removed using a dilute hydrochloric acid solution at room temperature, and then were rinsed in deionized water. Prior to XPS and UPS measurements, the samples were further etched by Ar ions with the etching depth of 5 nm.
Solar Cells Fabrication and Characterization: The c-Si wafers were n-type with a thickness of 170 µm and resistivity of 1–3 Ω cm. After texturing to form an array of random pyramids in KOH solution, standard RCA cleaning and dilute HF dip were performed. Subsequently, p+ emitter was formed by boron diffusion. Then the front surfaces were passivated with Al2O3/SiNx stack layers. The front Ag grids were fabricated by screen-printing paste and firing. For the electron-selective contacts, the rear surfaces of the wafers were cleaned by diluted HF (1-2%) solution and rinsed with deionized water. Then a SiO2 layer was thermally grown on the rear surface at 300 °C for 30 min in air atmosphere, following with thermal evaporation of ZnTiO3/Al. To examine the thermal stability of the devices, the post-annealing treatment was performed on a hot plate under different temperatures for 30 minutes in air ambient. The current density-voltage (J -V ) characteristics of the solar cells were investigated under the illumination of AM1.5G (100 mW cm-2, 25 ℃) using Newport 92250 A-1000 solar cellI -V tester and Keithley 2400 source meter.
The cross-sectional analysis of the electron-selective contacts was prepared using the focused ion beam (FIB) lift-out technique by Thermalfisher scois 2 FIB system. Then high-angle-annular dark-field scanning transmission electron microscopy (HAADF-STEM) images, energy-dispersive X-Ray spectroscopy (EDX) mapping and line scan were used to characterize the c-Si/SiO2/ZnTiO3/Al contacts before and after annealing using Thermalfisher Titan Themis Z system. For the passivation effect measurements, external quantum efficiency (EQE) spectra, photoluminescence (PL) mapping and effective minority carrier lifetimes (τ eff) were examined. The EQE of the solar cells with different post-annealing temperatures were tested via Crowntech QTEST HIFINITY 5. PL images of both an as-fabricated solar cell and a 300 °C post-annealed solar cell were obtained from a FL-B01 system with an 808 nm Laser injection. Effective carrier lifetime of solar cells with different post-annealing temperatures was extracted by Suns-V OCmeasurements using Sinton Instruments WCT-120.
Results and Discussion
The contact property of ZnTiO3 on n-type c-Si was investigated by measuring the contact resistivityρ c of c-Si/SiO2/ZnTiO3/Al stacked layers, using the TLM model, as shown in Figure 1a. Note that the contact resistivity is one of the most important parameters to evaluate the carrier selectivity. Figure 1b shows the dependence ofρ con annealing temperature after metallization. The thicknesses of ZnTiO3 are 1, 3 and 5 nm, respectively. It can be seen that all of the c-Si/SiO2/ZnTiO3/Al contacts possess ultralow 𝜌c (less than 5 mΩ cm2) before annealing. The 𝜌c of the c-Si/SiO2/ZnTiO3/Al contacts with ZnTiO3 thickness of 1 nm is unchanged when the post-annealing temperature is ≤ 250 °C, and then it starts to increase slightly, and further increases sharply when the annealing temperature is > 350 °C. For the electron-selective contact with ZnTiO3 thickness of 3 nm, its 𝜌c shows a slight increase until the annealing temperature of 350°C and increases drastically when the annealing temperature is > 400 °C. For the electron-selective contact with ZnTiO3 thickness of 5 nm, the 𝜌c decreases with annealing temperature (≤ 350 °C) and starts to increase sharply at the temperature of 400 °C. Obviously, the SiO2/ZnTiO3/Al stacks can form highly thermal stable ohmic contacts with n-type c-Si when the temperature is ≤ 350 °C. Moreover, the thicker ZnTiO3, the better heat resistance. Figure 1c compares the thermal stability of Ohmic contacts between our ZnTiO3-based stacks and other well-known Ohmic contacts.18, 30, 39 LiF- and SiO2-based contacts are stable within a relatively small annealing temperature range (<200 °C). TiO2-based contacts have improved thermal stability. In contrast, ZnTiO3-based contacts have the best thermal stability and possess the lowest 𝜌c over the largest annealing temperature range.
The influence of post-metallization annealing temperature on the ZnTiO3-based layers was also investigated in device level. The solar cell structure is depicted in Figure 1d. SiO2/ZnTiO3 stacks are utilized as passivating, electron-selective contacts. The thickness of ZnTiO3 is between 1 and 5 nm. The solar cells were annealed on a hotplate in air ambient for 20 min with temperature range from 25-400 °C. The corresponding cell parameters (V OC,J SC, FF and PCE) at each annealing temperature step are presented in Figure 1e-h. As can be seen, theV OCimproves from 624mV to 640 mV as the annealing temperature increases from 25 °C to 250 °C for the ZnTiO3 thickness of 1 nm. When the annealing temperature is further increased, theV OC drops obviously. For the solar cells with thicker ZnTiO3 (3 nm, 5 nm), although their initial V OC (about 600 mV) are lower than that of the cells with 1 nm ZnTiO3, they are improved more drastically, and reach around 650 mV. Moreover, it is obvious that the thicker the ZnTiO3 film, the higher the optimal annealing temperature. Similar behaviors can also be observed from J SC. The FF of cells with the ZnTiO3 thickness of 1 nm and 3 nm have negligible change within the annealing temperature of 350 °C, while the FFof the cells with 5-nm-thick ZnTiO3 is increased by 2% in absolute value when the annealing temperature increases from 25 °C to 300 °C. Therefore, the FF results are consistent with the contact resistivity results. As the comprehensive result ofV OC, J SC and FF , the PCE of the ZnTiO3/c-Si heterojunction solar cells increase remarkably with post-annealing temperature. The optimal temperature is either 250 or 300 °C depending on the thickness of ZnTiO3.