On earth, the photovoltaic cell junction for mass production tends to be just silicon, although there is a variety of material types - single crystal vs multicrystal or polycrystal, or even amorphous.

I just read that the Voyagers used indium phosphide (InP), Wikipedia says that Galium Arsenide (GaAs) are common, and I have read some satellites use triple-junction solar cells multi-layered devices used when the absolute highest efficiency is achieved.

Is there any reason - besides purchase price - that would keep all new satellites from using the highest efficiency cells? And actually, do different cells have best efficiency under different conditions encountered in spaceflight?

Pics of Juno's photovoltaics from NASA:

enter image description here

enter image description here

enter image description here


This article from 2008 was the best survey of trends in PV technology for space that I could find (requires IEEE subscription). It generalizes and essentially states that satellites in GEO have had a constant demand for more power, and that satellites in MEO or LEO have had a constant demand for higher radiation tolerance (higher end of life power). There has been continuous improvements in PV technology on both fronts, and reduced costs. Satellite manufacturers have taken advantage of these improvements, although cautiously to be sure the technology is space qualified. The paper mentions the role of multi junction cells as a big part of this, but doesn't note a particular brand or design as what the industry is converging on.

Mass, power margin, and space available on the rocket are the largest trade offs considered when choosing a solar cell technology. Flight heritage of the solar panel design is also a consideration. Lastly, for some missions it may be cheapest to just use more panels of a standard design offered by a manufacturer.

I think temperature would have the largest effect on cell operation in different missions. The solar spectrum would be unchanged (although dimmer at further differences). Radiation degradation could also be a consideration, especially for a mission like Juno, but I don't have any source comparing different materials.

Similar question answered here.

  • $\begingroup$ Can you answer: "Has the choice of photovoltaic junction types for newer spacecraft converged on one, optimal solution?" with "yes" or "no" and give some examples of PV junction choices to support your position? $\endgroup$ – uhoh Jul 22 '16 at 0:03
  • $\begingroup$ I found a paper from 2008 that talks about the trends in PV for space use and edited my answer. It agrees with my hunch that the trend is simply to use better tech as it develops and becomes reliable and cheap. $\endgroup$ – Andrew W. Jul 22 '16 at 1:14
  • $\begingroup$ So maybe Silicon might be used in low-cost situations, GaAs and InP single junction are not really used any more, and mostly it's multi-junction devices that are used now? $\endgroup$ – uhoh Jul 22 '16 at 1:20
  • 1
    $\begingroup$ Yes. Si is the cheapest beacause of it's use in semiconductors in general, low health risks, and because of the research into cheap panels for power on Earth. Multi junction cells require extra layers between the 'main' layers. Materials commonly include GaAs c-Si, InGaP, Ge, InGaAs (in no particular order). The choice depends on the wavelength of interest. Plain GaAs cells are the best single-junction cell with an efficiency somewhere in the middle and are still commonly used. $\endgroup$ – Andrew W. Jul 22 '16 at 21:42

Of course not. Until the 'perfect' photovoltiac material is determined (when would you expect that to happen? 5 years? 5 decades? 5 centuries?), we'll be forced to use less than perfect materials. Given different criteria - for instance budget, mass, life-time, light intensity (towards the Sun, or away?), the "optimum" will vary.

  • $\begingroup$ This isn't a bad default approach to take in the absence of specific expertise, but it's pretty sparse for an answer. $\endgroup$ – Nathan Tuggy Jul 21 '16 at 23:24

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.