August 13th, 2020
Category High-Efficiency Solar Cells
With today’s record efficiency of 44.0 % under a concentration of 942 suns  III-V solar cells have not reached their full potential. The multi-junction concept has the perspective of obtaining efficiencies above 45 % and maybe even 50 % under concentration. New designs for cells with more than three junctions require the use of thin-film cells in some form [27, 59, 60] either as a free standing thin-film cell or in combination with cells grown on different wafer materials through a layer transfer process.
A basic method for obtaining a thin-film cell by etching/polishing of the wafer after growth is widely used [1, 28, 29], but this is an expensive method because the wafer, which has a large contribution in the total cell cost, is lost...Read More
One of the aims of ELO is to allow multiple reuse of the wafer after lift-off to reduce cell cost. Ideally, a minimum of wafer re-preparation between consecutive lift-offs is required. However, exposure to the HF solution increases the surface roughness of the wafer on a nanometer scale. The roughness is about 0.3 nm for new so-called epi-ready wafers and between 2 and 4 nm for wafers that have been subjected to a 20 % HF solution for 21 h. In addition during ELO reaction residues are left on the surface , mainly elemental arsenic. During storage of the etched wafer under ambient conditions, from this arsenic As2O3 particles are formed...Read More
Wafer-based cells have a full area back contact. A thin-film cell allows access to the backside of the device. This makes it possible to apply a metal grid on front and back side. In this design the cell can be used as a bifacial cell  where illumination takes place from both top and bottom side using a relatively cheap mirror setup.
Fig. 21.10 Examples of thin-film cell designs: semitransparent InGaP cell (left) and bifacial GaAs cell (right)
Such a cell can also be applied as semitransparent cell for mechanical stacking on a lower bandgap cell. An example is a thin-film InGaP cell combined with a Si cell  or an InGaAsP/InGaAs tandem cell grown on InP combined with a GaAs cell . In Fig. 21.10 an InGaP and GaAs semitransparent cell are shown...Read More
For space solar cells high efficiency, high radiation resistance, a low weight, and flexibility are the desired features. Because of their highest power output wafer – based III-V cells have been dominant over other cell types since the late 1990s. Wafer-based cells are mounted on a rigid honeycomb structure to prevent the wafers from breaking, resulting in relatively high weight. A thin-film cell on a flexible host substrate does not require a rigid panel assembly. This introduces new low weight options for panel design. In the past, other attempts were made for thin – film space cells which used a-Si or CIGS materials...Read More
The first good quality thin-film III-V cell, a 4 cm2 GaAs cell with a 23.3 % efficiency, was made by Kopin in 1990  at a time when the best wafer-based GaAs cell performance was 25.1 %. The thin-film was not lifted off chemically like in ELO but mechanically using the CLEFT (Cleavage of Lateral Epitaxial Films for Transfer) technique . The cell was used in a mechanical tandem in combination with CIS bottom cell. The total efficiency was 25.8 % which made it the best mechanically stacked tandem cell at that time intended for use in space. No progress in CLEFT was reported after 1990.
Fig. 21.8 Flexible ELO thin-film cells with a metal foil backing
Conventional triple junction cells grown on Ge have their limitations. They are inflexible, brittle, and relatively heavy. Therefore wafer-based cells require some kind of structural support to prevent damage. Thin-film III-V cells produced with the ELO technique offer new opportunities for device design, based on the fact that the thin-film carrier can be selected on its material properties rather than crystal growth demands. Figure 21.8 shows an example of flexible thin-film cells with a metal backing. Substrate reuse and thus lower cell cost has been the main driver for the development of ELO thin-film cells, but apart from this there are many interesting new concepts and applications which are not possible for wafer-based cells.Read More
Alternative approaches for ELO are directed to the production of microchips using a transfer-printing technique to peel and print a large number of small thin-film structures onto glass or plastic . The layer stack for this method is identical to other ELO methods: a device structure that is grown over an AlAs release layer, only in this approach the AlAs layer is much thicker (1 ^m) than generally applied for the full area lift-off. In this method not the entire wafer area is lifted off, but small area device structures. To do this, before lift-off the material is separated into square blocks by vertical etching through the device structure to enable exposure of the release layer to HF by immersion...Read More
For the application of ELO the process needs to have a sufficiently high etch rate. It is and always will be a relatively slow process. With an etch rate of less than 1 mm/h it initially took more than a day to lift-off a 2 in. diameter thin-film
structure. Therefore an important goal was to speed up the process to a time scale of hours for lifting of a 4 in. thin-film, which is in the same order as the deposition time of the solar cell layer structures by MOCVD...Read More