Your browser is not supported. Please update it.

9 August 2019

GaAs and III-V to make solar comeback on new HVPE process?

We have a variety of improvements coming out of solar over the coming years which are lined up to continue revolutionizing the technology and drive down prices, and the US National Renewable Energy Laboratory (NREL) has come out with yet another, which promises to dramatically lower cost of panels.

This is more about the process of how they are built and involves a new growth technique, called dynamic hydride vapor-phase epitaxy (D-HVPE) which will open up the use of substances which exhibit higher photo voltaic effect referred to as III-V materials due to their position in the periodic table. This includes (Gallium Arsenide) GaAs and Indium Gallium Phosphte (GaInP), which have better electronic and optical properties, with higher electron mobilities than Silicon.

So this process could be important for both building VLSI chipsets as well as making solar panels.

Currently these materials are only used when cost is not an issue but best performance is, for instance for electronics on satellites.

They are also used in power amplifiers in cell phones and in light emitting diodes (LEDs) for general illumination applications with nitride-based III-V.

But it is not the materials costs that is prohibitive, it is how they are made. The D-HVPE process was reported in the journal Crystals last year, and this has been followed up by a paper in the current edition of Nature,

The NREL researchers have refined the D-HVPE process to produce solar cells more than 20 times faster than the process now commonly used which is called metalorganic vapor-phase epitaxy (MOVPE). This allows the base layer to be grown in about 23 seconds, instead of the traditional 8 minutes using MOVPE. Given that many layers have to be grown, this results in a rate of about 15 micrometers an hour.

MOVPE rapidly deposits atoms layer-by-layer on a semiconductor wafer and an older process – hydride vapor phase epitaxy (HVPE) –  which was dumped in favor of MOVPE in the 1970s was re-tried.

This will open up a huge number of markets where these devices would be useful. There are university departments all over the world trying the alternative process of making solar panels based on cheap, freely available metals like Copper, whereas this is a process that remains with the best materials, but makes the process faster and therefore cheaper.

It would certainly change the financial equation if Electric Vehicles could simply have solar panels on their roof, which would be good enough to fill or at least half fill their batteries, in key sun-drenched parts of the world, or have mobile phones where their screens could recharge them. It would put far less onus on the grid environment to deliver every piece of electricity anywhere.

We have also heard a lot lately about Perovskite development approaching 28% efficiency, and carbon nano-tubes allowing heat to be converted into light, and fed back into photo-voltaic absorption, capable of achieving 80% efficiency. All of these promise orders of magnitude reductions in cost per KWh and that’s without the relatively simple process of bifacial panels, which are driving the current round of improvements.

The title of the paper is “Gallium arsenide solar cells grown at rates exceeding 300 µm h−1 by hydride vapor phase epitaxy.”

NREL scientists say they have already used the D-HVPE process to make solar cells with a 25% efficiency, that 27% is in sight and 29% is theoretically possible with just a few “technical hurdles” to clear.

This was all achieved by developing a two chamber reaction which moves the substrate from one chamber, where the first layer of chemicals is deposited, to the second chamber and another layer.

A number of researchers in the past continued to pursue III-V class materials in solar until very recently, and this change may bring that research back alive once more.

D-HVPE for GaInP has been shown at rates up to 206 micrometers an hour, some 50 times faster than MOVPE growth rates. GaInP is used as a passivating layer in gallium arsenide (GaAs) solar cells and can also be used as a light-absorbing layer in multijunction solar cells.

Funding for the research came from the Energy Department’s Solar Energy Technologies Office and the Advanced Research Projects Agency-Energy. Let’s see if anyone adopts the technology any time in the next few years.