Semiconductors play an important role in solving the planet’s energy issues. There are two distinct, but related, phenomena: the conversion of (sun) light in electricity and the conversion of electrical power in visible light. The first conversion is known as the photo-voltaic (PV) technology and the second conversion is the one that is used by Light Emitting Diodes (LED’s) used in Solid State Lighting applications. Both conversions enjoy considerable interest from scientists, governments, energy companies as well as citizens. Clear is that both energy conversions can contribute substantially in solving the availability and distribution of energy around the planet.
A key factor for the successful acceptance (at least in terms of economically feasibility) of both PV and LED’s is the efficiency of these two types of energy conversions since this will directly impact cost per Watt or cost per Lumen. Indeed, the question arises are there fundamental limitations to these energy conversions? For PV cells it has been reported that the upper efficiency of on silicon based cells will run at about 30%. For LED’s there has not been reported so far a fundamental barrier that would keep the LED away from 100% efficiency (however, the fact that the device heats up during operation hints already to a less than 100% efficient light conversion).
On the efficiency of PV cells I will come back in a future contribution, for now I would like to focus on the efficiency of a LED. A LED is typically constructed from a classical p-n junction but in the LED case the p and n material are separated by what is called an active zone that can be either doped or intrinsic. The semiconductor material must be a direct band gap semiconductor in order to have sufficient conversion efficiency[i]. By putting the LED in a forward bias the electrons and holes that arrive in the active zone can recombine in two different ways:
- Radiative recombination. It is this recombination that fuels the light emission from the LED.
- Several other non-radiative recombination processes occur as well. These reduce the amount of holes and electrons available for light emission.
There are other loss (non-radiative) mechanisms operating (such as absorption of the photons by the semiconductor) that further reduce the light generation efficiency.
Recently an article in the Journal of Applied Physics[ii] appeared that gives good insight in the different factors that influence the power-light conversion efficiency. An important factor is the so-called wall plug efficiency, defined as follows:
Wall Plug Efficiency = emission power/electrical power
a pretty straightforward definition. In the article all the different recombination and loss mechanisms are mathematically described and then put together in one model for the LED. This model can then calculate the behavior (and thus wall plug efficiency) of the LED device in terms of operating conditions (temperature, current, voltage), material properties (semiconductor material such as GaN or GaAS and doping) and LED structure (thickness of the different layerings, metal contacts and lay out of the active layer). This is of great help when optimizing the LED device for conversion efficiency.
Let me summarize a few important conclusions from the article:
- There is not a fundamental reason why the power-light conversion cannot be 100%. Even stronger, the conversion can be more than 100% (see next point for explanation)! However, the high efficiencies may not always in a practical operating window (for instance at the current densities the LED needs to run because of a certain required light output per surface area semiconductor).
- The energy of the photon may come not only because from the band gap energy difference but phonons (thermal energy from the lattice) may contribute as well. In that case the LED can act as a heat pump: the device cools actually and can in that way extract heat from the environment and achieve efficiency better than 100% (using the above wall plug efficiency definition).
- Further improvements will be possible to increase the light output of LED’s.
Thus, we can expect to see in the coming years more developments coming to improve the Solid State Light technology and this will be a very valuable contribution to our energy strategy.
[i] See for an explanation: http://en.wikipedia.org/wiki/Direct_and_indirect_band_gaps
[ii] O. Heikkila, J. Oksanen, J. Tulki, Ultimate limit and temperature dependency of light-emitting diode efficiency, Journal of Applied Physics 105, 093119 (2009)
© Copyright John Schmitz 2010