RESUMEN
We introduce cracked film lithography (CFL) as a way to reduce the cost of III-V photovoltaics (PV). We spin-coat nanoparticle suspensions onto GaAs thin-film device stacks. The suspensions dry in seconds, forming crack networks that we use as templates through which to electroplate the solar cells' front metal grids. For the first time, we show that heating the crack template allows it to flow and refill cracks, which decreases crack footprint and improves final grid transmittance. We demonstrate 24.7%-efficient single-junction GaAs solar cells using vacuum-free CFL grids. These devices are only 1.7% (absolute) less efficient than the baseline grids patterned by photolithography with the loss mostly resulting from the reduced transparency of the CFL pattern. Additional optimization could decrease this difference. Initial cost modeling suggests that CFL is more scalable than photolithography: In particular, CFL's lower materials and equipment costs could greatly reduce the levelized cost of electricity of III-V PV at scale, a potential step toward terrestrial deployment.
RESUMEN
Cracked film lithography (CFL) is an emerging method for patterning transparent conductive metal grids. CFL can be vacuum- and Ag-free, and it forms more durable grids than nanowire approaches. In spite of CFL's promising transmittance/grid sheet resistance/wire spacing tradeoffs, previous solar cell demonstrations have had relatively low performance. This work introduces macroscopic nonuniformities in the grids to improve the short-circuit current density/fill factor tradeoff in small area Cu(In,Ga)Se2 cells. The performance of optimized baseline grids is matched by CFL grids with microscopic openings and macroscopic patterns, culminating in a 19.3% efficient cell. Simulations show that uniform CFL grids are enhanced by patterning because it leads to better balance among shadowing, grid resistance, and transparent conductive oxide resistance losses. Thin-film module efficiency calculations are performed to highlight the performance gains that metal grids can enable by eliminating the transparent conductive oxide losses and widening monoliths. Adding the patterned CFL grids demonstrated in this work to CIGS modules is predicted to reach 0.7% higher efficiency (absolute) than screen-printed grids.
RESUMEN
The fundamentals of using cracked film lithography (CFL) to fabricate metal grids for transparent contacts in solar cells were studied. The underlying physics of drying-induced cracks were well-predicted by an empirical correlation relating crack spacing to capillary pressure. CFL is primarily controlled by varying the crack template thickness, which establishes a three-way tradeoff between the areal density of cracks, crack width, and spacing between cracks, which in turn determine final grid transmittance, grid sheet resistance, and the semiconductor resistance for a given solar cell. Since CFL uses a lift-off process, an additional constraint is that the metal thickness must be less than 1/3 of the crack template thickness. The transmittance/grid sheet resistance/wire spacing tradeoffs measured in this work were used to calculate solar cell performance: CFL-patterned grids should outperform screen-printed grids for narrow cells (0.5-2 cm wide) and/or cells with high semiconductor sheet resistance (≥100 Ω/sq), making CFL attractive for monolithically integrated thin-film photovoltaic modules.
RESUMEN
The chemical structure of the Zn(O,S)/Cu(In,Ga)Se2 interface in high-efficiency photovoltaic devices is investigated using X-ray photoelectron and Auger electron spectroscopy, as well as soft X-ray emission spectroscopy. We find that the Ga/(Ga+In) ratio at the absorber surface does not change with the formation of the Zn(O,S)/Cu(In,Ga)Se2 interface. Furthermore, we find evidence for Zn in multiple bonding environments, including ZnS, ZnO, Zn(OH)2, and ZnSe. We also observe dehydrogenation of the Zn(O,S) buffer layer after Ar+ ion treatment. Similar to high-efficiency CdS/Cu(In,Ga)Se2 devices, intermixing occurs at the interface, with diffusion of Se into the buffer, and the formation of S-In and/or S-Ga bonds at or close to the interface.
