FEATURE ›
PHOTOVOLTAICS
for scaling it up in the future. With this new research, you made a solar cell the diameter of a few human hairs. What are the applications of miniaturised solar cells? The original plan was for a technology called concentrator photovoltaics. This is where you concentrate sunlight [using lenses or mirrors], so there’s the same amount of light that you would get on a large panel but in a smaller area – you could theoretically get the same power with less material. However, you also generate
way more current. Because you’ve concentrated a lot of sunlight, you’ll have high Joule losses and the solar cell will heat up. But the neat thing is that the ratio between the surface area and the volume gets larger as you miniaturise the solar cell, which means that you can theoretically dissipate the heat more easily. Over the past few years, some researchers have suggested that we can minimise these losses by shrinking solar cells down to the scale of a few hundred micrometres.
Most solar cells are made from silicon, but you used a III-V semiconductor. Why? Silicon has an indirect band gap, but most III-V semiconductors that we use for solar cells have a direct band gap, which means that their absorption is much larger. Not only that, with indirect bandgap semiconductors you can reach different sensitivities in the solar spectrum. Silicon has a band gap in the infrared, but with III-V semiconductors, you can tune the bandgap from the UV down to deep in the infrared.
Because of that, we can build
something called multi-junction collar cells, where we grow III-V semiconductors on top of each other. The uppermost subcell would be sensitive to UV and would absorb all the UV very efficiently and then, underneath that, there would be another subcell that would absorb the visible, and so on until we reach the infrared. All the upper subcells are transparent to the light that is absorbed underneath. So we absorb
24 Electro Optics May 2024
“The fact that the III-V materials are not as mature industrially as silicon was the main difficulty of this project”
light over a wide spectral region, whereas silicon is one junction so it doesn’t absorb the full spectrum of light as efficiently as the multi-junction technology.
Why were you motivated to develop a 3D framework for the solar cell? As you miniaturise the solar cell, you still need to have contact pads on the front side. There are packaging limits [on the size of electrical contacts] in the industry, of roughly 100 microns. If your solar cell is, say, 200 microns wide and you have two contact pads on each side, that is 50 microns you just lost. This means that half of the wafer would be lost to shading. In addition, multi-junction solar cell technology is extremely expensive. If we take something that costs orders of magnitude more than silicon and lose half of it, there’s a big problem. So the 2D strategy doesn’t work at a micrometric scale. You have to go 3D to mitigate losses.
How did you get around the limitations of 2D solar cell architecture? The strategy in 3D is to remove the grid on the front side, and put it on the back of the device. This way, there’s no shading. But we still need to bring the electricity generated on the front side to the back electrode. That’s where something called a ‘via’ comes in. It’s a hole that has been plasma etched and metallised so it behaves like an electric wire. In this strategy, the carriers are generated on the front side, collected by the vias, and transferred downwards through the cell where they enter the electrodes.
What were the main challenges of the project? There were a lot of challenges from the microfabrication
A graphic showing the difference between a standard solar cell and a miniaturised solar cell
standpoint. There are a wide range of III-V semiconductor materials, and all the technologies have to change slightly from one alloy to another. Silicon technology has been known for decades and we already have silicon vias, but with III-V semiconductors there are additional issues. For example, the plasma etching processes on III-V materials can be very challenging. For vias, you need something that is very narrow but very deep. This is easily done on silicon, but it’s trickier for III-V semiconductors. The fact that the III-V materials are not as mature industrially as silicon was the main difficulty of this project.
How does the performance of the solar cells you developed compare with current alternatives? It has a lot of potential to increase efficiency once it’s well optimised. The main challenge is that it will cost way more to mass-produce than the typical solar cell because of the vias. But it can reduce the shading by an order of magnitude. So, even though the cell costs more to produce, we project that the cost per wafer could go down with this technology.
Would it be possible to produce these cells at scale? This project mixed two different technologies, and the challenge
was to tie them together. There’s the III-V photovoltaic industry, which has its own process to manufacture multi- junction solar cells, and there’s the CMOS industry, which has the experience with vias and the capability to process microprocessor chips with very small dimensions. For example, in the photovoltaic industry, the III-V material growth and the metallisation of electrodes is very well established. But we also needed, via etching, processes such as chemical mechanical planarisation and atomic layer deposition, which are used in the CMOS industry. Atomic layer deposition was very important for this project because we used it to deposit a dielectric along the via to insulate it. To the best of my knowledge, atomic layer deposition is not used in the multi-junction solar cell manufacturing industry, but it is used in the CMOS industry.
What future plans do you have to develop this technology? There’s a lot of improvement to be done. When I developed the microfabrication process, I just wanted it to work. But now that we have shown that it works, why not try to improve it? Why not try to get better efficiency or more power generation? There’s a few ideas that we could investigate to achieve that. EO
www.electrooptics.com
Page 1 |
Page 2 |
Page 3 |
Page 4 |
Page 5 |
Page 6 |
Page 7 |
Page 8 |
Page 9 |
Page 10 |
Page 11 |
Page 12 |
Page 13 |
Page 14 |
Page 15 |
Page 16 |
Page 17 |
Page 18 |
Page 19 |
Page 20 |
Page 21 |
Page 22 |
Page 23 |
Page 24 |
Page 25 |
Page 26 |
Page 27 |
Page 28 |
Page 29 |
Page 30 |
Page 31 |
Page 32 |
Page 33 |
Page 34 |
Page 35 |
Page 36
