About this document
Semi-transparent solar cells divide a single absorber between daylighting and power generation. This Technical Note describes a manufacturing approach in which the thickness of a single-crystal III–V layer, set by the exfoliation depth, fixes both transmittance and output as product specifications. The full report is available in Korean. An English summary is provided on this page.
Download
An English edition is not available yet. The Korean document is below.
Summary
Ultrathin semi-transparent GaAs cells have been shown to work in the literature, but only at millimeter scale. In this approach, the ion implantation depth in a smart-cut process sets the III–V layer thickness, so transmittance, output and color coordinates are all defined by one process parameter. The note covers a manufacturing scheme that uses 1550 nm light to heat only the implant-damaged zone for exfoliation and reuses the donor wafer.
Conclusion
In a semi-transparent solar cell, transparency and output share the same absorber, so raising one lowers the other. Existing approaches leave gaps between opaque cells, cut holes or stripes into the cells, or use thin-film, organic or perovskite layers. The first two pay for transparency with active area; the third pays with long-term reliability. As the case for using windows as generating surfaces, the document estimates that a 40-story office building (40 × 40 m floor plate, 4 m floor height, 60% window-to-wall ratio) has about ten times more window area than roof area. The document starts from the opposite direction. Instead of pushing output from wide-bandgap transparent materials, it keeps only the necessary thickness of a direct-bandgap single-crystal III–V material with a high absorption coefficient. Literature values include a certified efficiency of 19.9% for a 205 nm GaAs ultrathin cell (Nature Energy, 2019) and an average visible transmittance of 27.4% for a 100 nm semi-transparent GaAs cell (Jpn. J. Appl. Phys., 2025). These demonstrations remain at millimeter-scale cells and wafer level, and the document frames the open problem as manufacturing: bringing them to square-meter architectural glass. The key step is to use the ion implantation depth in smart-cut as a product design variable. Implantation depth sets the cleave plane, the cleave plane sets the III–V layer thickness, and thickness determines absorbed light, transmittance, efficiency and color coordinates together. Changing only the exfoliation depth on the same donor wafer yields transmittance-first, balanced and output-first grades. The document labels these product specifications as development targets. The process uses 1550 nm light. Below the bandgap of the intact III–V crystal, this wavelength passes through it and is absorbed selectively in the implant-damaged zone and at the interface. Exfoliation, damage recovery and interface activation are handled by a single light source. The equipment combines a fiber laser, line-beam homogenization, a scanner or XY stage, and closed-loop monitoring with transmitted-light, reflected-light and temperature sensors, scaling to large areas without a vacuum chamber. The document presents this as a development target and a patent-based process. After exfoliation, the donor wafer surface is restored for reuse; the target is ten or more reuses per wafer, cutting the wafer cost allocated to each product to about one tenth. The finished product is architectural glass with the generating layer built into an insulated glazing unit. The stack runs from outer protective glass, an optical encapsulant, the submicron III–V layer, fine finger electrodes, a low-visibility wiring layer and an interlayer to the inner glass and an argon-filled gap, with busbars and cell interconnects placed at the glass edge and inside the frame. The document defines QoolView as designing the thickness of a single-crystal layer rather than the empty space between cells; because thickness is the specification, adding a product does not require redesigning the cell.
Topics
- Solar
