Resumen de Characterization of alternative carrier selective materials and their application to heterojunction solar cells
Luis Guillermo Gerling Sarabia
Crystalline silicon (c-Si) solar cells can be considered a highly industrialized and mature product with a record conversion efficiency of 26.6%, not far from the practical limit of 29.4% (for single p/n junction devices). Accordingly, current research and development are addressing some remaining efficiency and cost limitations, including the reduction of (1) carrier recombination in highly doped materials, (2) parasitic absorption by narrow band gap films and (3) high temperature energy-intensive processing (especially critical for wafer thicknesses below 100 µm).
In parallel, thin-film PV (e.g. organics and perovskites) have introduced a large number of dopant-free, hole- or electron-selective materials with optoelectronic properties that are comparable or superior to standard p- and n-doped layers in c-Si. Consequently, this thesis work explores novel heterojunctions between c-Si and these carrier-selective contact materials, putting special emphasis on TMO thin films whose wide energy band gap (>3 eV), surface passivation and large work function (>5 eV) characteristics permit their utilization as transparent/passivating/hole-selective front contacts in n-type c-Si (n-Si) solar cells.
To this purpose, a comparative study among three thermally evaporated TMOs (V2O5, MoO3 and WO3) allowed correlating their chemical composition with thin film conductivity, optical transmittance, passivation potential and contact resistance on n-Si substrates. The variation of these properties with film thickness, air exposure or temperature annealings was also studied.
Overall, V2Ox outperformed the other oxides by obtaining higher implied open-circuit voltages and lower contact resistances, translating into higher selectivities.
Next, a thorough study of the TMO/c-Si interface was performed by electron microscopy, secondary ion-mass spectrometry and x-ray photoelectron spectroscopy, identifying two separate contributions to the observed passivation: (1) a chemical component, as evidenced by a thin SiOx interlayer naturally-grown by chemical reaction during TMO evaporation; and (2) a ¿field-effect¿ component, a result of a strong inversion (p+) of the n-Si surface, induced by the large work function difference between both materials. Considering all this, an energy band diagram for the TMO/SiOx/n-Si heterojunction was proposed, reflecting the possible physicochemical mechanisms behind c-Si passivation and carrier transport.
Then, the characterized TMO/n-Si heterojunctions were implemented as front hole contacts in complete solar cell devices, using thin TMO films (15 nm) contacted by an indium-tin oxide (ITO) anti-reflection/conductive electrode and a silver finger grid. As rear electron contacts, n-type a-SiCx:H thin films (20 nm) were used in localized (laser-doped) and full-area configurations, the former contacted by titanium/aluminum while the latter by ITO/silver electrodes. The best performance solar cells were obtained for V2Ox/n-Si heterojunctions, characterized by an open-circuit voltage (VOC) close to 660 mV and a maximum conversion efficiency of 16.5%. Additional characterization confirmed the good quality of the induced p+/n-Si junction, with ideality factors close to 1 and built-in potentials above 700 mV. Moreover, a photocurrent gain of ~1 mA/cm2 (300¿550 nm wavelength range) was directly attributed to the difference in energy band gaps between TMOs (>2.5 eV) and the a-SiCx:H reference (~1.7 eV). On a sideline, hole-selective contacts based on PEDOT:PSS polymer solutions were also characterized, resulting in a moderate conversion efficiency of 11.6% in ITO-free devices.
Finally, it is worth emphasizing the high degree of innovation in this thesis project, reporting for the first time the properties of these alternative contact materials in the context of c-Si photovoltaics and contributing to a more generic understanding of solar cell operation and design.
