Semiconductor thin films make solar power cheaper – news from the Enlighten Conference 2017

 

As a niche market that prioritizes power per gram over power per dollar, the space-based PV field has until now tolerated the high cost of multi-junction cells based on III–V semiconductors, since they can reach efficiencies of 30% or more and are radiation-resistant. This technology requires multiple semiconductor materials arranged in complex structures, which is why they are generally more expensive compared with other forms of PV.

Soaring demand

The massive expansion of communication networks promised by companies like OneWeb, however, demands high-volume manufacturing techniques; OneWeb alone plans to produce hundreds of spacecraft at a rate of up to 15 per week. On top of this, a huge predicted growth in the use of solar-powered unmanned aerial vehicles (UAVs) means that satellite manufacturers are no longer the only customers for lightweight, high-efficiency solar panels.

But while large, rigid solar panels can be accommodated in the vacuum of space, for UAVs the PV cells must be lightweight and able to be integrated into an aerodynamic airframe. Added to the list of requirements, therefore, is an important additional property: flexibility. As a result, PV cell producers have been encouraged to come up with innovative designs that can accommodate multiple design criteria while keeping costs down.

According to Johnson, key to meeting the needs of this expanded market is the technique of epitaxial lift-off (ELO), which allows III–V semiconductor stacks to be peeled away from their growth substrates after fabrication. This reduces the overall mass of the cell and results in flexible structures that can more easily conform to the wings and fuselages of UAVs. Meanwhile, the discarded substrates can be used again in another growth cycle, further reducing manufacturing costs.

Although ELO is still relatively expensive compared to conventional silicon processing techniques, Johnson cited a variant that promises to keep costs low. The proposed method involves the formation of a porous top layer in a monocrystalline silicon wafer by electrochemical etching. Annealing the material causes a planar void to develop at the interface between the porous and pristine silicon, allowing the upper part to be detached. Prior to lift-off, another thin layer of monocrystalline silicon can be grown on top by chemical vapour deposition (CVD), resulting in a flexible, single-crystal, 2–3 µm film that can serve as the growth substrate for other semiconductor materials.

Keeping cool

Also having applications in PV cell manufacture was a new type of plasma-enhanced CVD presented by Shashi Paul. Paul’s technique, named the Ra method, allows the fabrication of variously shaped silicon nanowires. Because the process takes place at temperatures lower than 400°C (below even 150°C for some nanowire morphologies), it is less energy-intensive than other techniques, and can make use of cheap glass or flexible plastic substrates. A continuing patent application meant that full details of the process could not be disclosed, but Paul was clear that the vapour-liquid-solid (VLS) growth mechanism is not employed.

The silicon nanowires produced by the technique exhibit excellent optical absorption across a wide range of wavelengths and incidence angles, and Paul reported that subsequent metallization layers can be applied easily. The efficiency of solar cells depends strongly on the purity and crystal structure of the materials used, which the Ra method can readily achieve. In addition, the choice of catalyst used during the growth process allows various dopants to be introduced into the nanowires, so that compositions could be achieved that yielded thin-film PV cells with efficiencies up to 6%, and with the potential to go above 15%.

Paul pointed out that, while battery-electrode performance is much less sensitive to silicon quality and purity than solar-cell performance is, stresses induced by ion migration during cycling create different challenges. The high charge capacity of silicon makes it a promising anode material for next-generation lithium-ion batteries, but only if it can be made more resilient to these stresses.

The nanowires produced by Paul’s technique, it turns out, can flex in a way that lets them accommodate ions without failing, meaning that capacity could be maintained over hundreds of cycles. Although this property of silicon nanowires has been demonstrated before, existing fabrication techniques take place at much higher temperatures, so do not enjoy all of the same advantages as Paul’s Ra method.