We formed phosphorous(P)-ion-implanted n-BaSi₂ films on p-Si(111) substrates and demonstrated solar-cell functionality of the n-BaSi₂/p-Si heterojunction under AM1.5 illumination. The BaSi₂ films were grown by molecular beam epitaxy, followed by implantation of P ions to the BaSi₂ films using PF₃ gas at an energy of 10 keV and a dose of 1 × 10¹⁴ cm⁻². Subsequent postannealing was conducted at 500°C in Ar for different durations (t = 30–480 s) to activate the P atoms. The diffusion coefficient for P atoms in BaSi₂ was evaluated from the depth profiles of P atoms by secondary-ion mass spectrometry. The activation energies of lattice and grain boundary diffusion were found to be 1.1 ± 0.6 and 2.5 ± 0.6 eV, respectively. From the analysis of Raman and photoluminescence spectra, the ion implantation damage was recovered by the postannealing. For one treated sample with t = 120 s, the internal quantum efficiency reached 67% at a wavelength of 870 nm. This is the highest ever achieved for n-BaSi₂/p-Si heterojunction solar cells. Ion implantation is thus applicable to BaSi₂ films grown by any other method. This achievement thereby opens a new route for the formation of BaSi₂ solar cells.

We performed advanced opto-electrical simulations on thin-film BaSi₂ solar cells. First, absorption spectra of BaSi₂-pn homojunction solar cells on Si substrate were calculated based on flat and/or pyramidally-textured surfaces, wherein 20-nm-thick n⁺-BaSi₂ was the topmost electron transport layer. By changing the front surface structure from flat to texture, the reflectance decreased in the wavelength (λ) range 700–1200 nm and the photocurrent density (J_ph) delivered by the photogenerated carriers in the 500-nm thick p-BaSi₂ layer increased by 1.2 mA/cm². Simulations revealed that the key factor inhibiting light absorption in the p-BaSi₂ layer was parasitic absorption in the n⁺-BaSi₂ and in the c-Si substrate. To solve these optical issues, we propose a new device structure, Al-doped n⁺-ZnO (AZO, 50 nm)/i-ZnO (20 nm)/p-BaSi₂ (500 nm) heterojunction solar cell (HJSC). In this device structure, the parasitic absorption reduced drastically, and J_ph reached 30.23 mA/cm². Furthermore, by replacing the Si substrate with a glass substrate, the light trapping worked more effectively, and the absorber layer thickness required for J_ph to saturate was reduced to 1 μm, yielding 32.06 mA/cm². Based on these simulation results, we manufactured n⁺-AZO/p-BaSi₂ HJSC. The internal quantum efficiency exceeded 30% at λ = 600 nm, meaning that we demonstrated the operation of n⁺-AZO/p-BaSi₂ HJSC for the first time. We investigated origins of small efficiencies compared to those simulated, and found that the passivation of defects in the p-BaSi₂ layer and the reduction of carrier recombination at the i-ZnO/p-BaSi₂ interface would significantly improve the solar cell performance.

Barium di-silicide (BaSi₂) is a very promising absorber material for high-efficiency thin-film solar cells, due to its suitable bandgap, high light absorption coefficient, and long minority-carrier lifetime. In this study, we compare the nanostructure, layer composition, and point defects of BaSi₂ thin films deposited by Radio Frequency (RF) sputtering, Thermal Evaporation (TE), and Molecular Beam Epitaxy (MBE), using Doppler Broadening Positron Annihilation Spectroscopy (DB-PAS) depth profiling, Raman spectroscopy, and x-ray diffraction. Our DB-PAS study on thermally annealed RF-sputter deposited and on TE-deposited BaSi₂ layers, in a comparison with high quality BaSi₂ films produced by MBE, points to the presence of vacancy-oxygen complexes and Si or Ba mono-vacancies, respectively, in the (poly)crystalline BaSi₂ films. The degree of near-surface oxidation increases, going from MBE and TE to the industrially applicable RF-sputtered deposition synthesis. The use of a-Si capping layers on the thermally annealed RF-sputtered BaSi₂ films leads to a clear reduction in sub-surface oxidation and improves the quality of the BaSi₂ films, as judged from DB-PAS.