Charge Asymmetric Resonance Tunneling (CART) LED

Commonly a single quantum well or multiple quantum wells are used as the active layer in LEDs. For fabrication of a highly efficient device the number of carriers recombined inside the active layer should be maximized and the number of carriers recombined outside the active layer should be minimized. This needs optimization of capture rates for electrons and holes into the active layer.

In polar III-V and II-VI semiconductors the most effective channel of the carrier capture in quantum well is by emission of polar optical phonons. The corresponding carrier capture rate can be roughly estimated as the quantum well width divided by product of the carrier thermal velocity over polar phonon emission time. Thus, the capture rate depends on the quantum well parameters and the carrier masses. As a rule in III-V and II-VI semiconductors the electron effective masses are much lighter and corresponding thermal velocities are higher then those for holes. For this reason, for the narrow quantum well, which provides optimal carrier confinement and maximal optical matrix element, a part of the electrons are not captured in the active layer and recombine outside of it. This reduces the efficiency of LED devices.

General scheme of Charge Asymmetric Resonance Tunneling (CART) LEDFig. 1 General scheme of CART LED

To solve this problem our team had suggested a LED structure based on a system of two wells with Charge Asymmetric Resonance Tunneling (CART) which allows to enhance the number of the electrons captured into the active layer with quantum well [1]. The phenomenon of the charge asymmetric resonance tunneling uses the quantum mechanical effect of strong exponential dependence of the potential barrier tunnel penetrability on the mass of the tunneling particle. This phenomenon can be used in LED design for all semiconductors that have heavy hole masses significantly higher than electron masses. Such semiconductors are most of III-V and II-VI Group semiconductors.

General scheme of CART LED under applied forward bias is shown in Fig. 1.

The system consists of an emitter of the electrons, an emitter of holes and an active layer. The hole emitter is coupled with active layer in such a way that holes can be freely supply into the active layer without a barrier. The electron emitter is coupled to the active layer via a barrier. The barrier design uses the charge asymmetric resonance tunneling phenomenon which allows to make the barrier transparent for electrons and blocking for holes. The phenomenon of the charge asymmetric resonance tunneling uses the quantum mechanical effect of strong exponential dependence of the potential barrier tunnel penetrability on the mass of the tunneling particle. This phenomenon can be used in LED design for all semiconductors that have heavy hole masses significantly higher than electron masses. To achieve the CART phenomenon the electron energy level positions in active layer should be fitted to the electron energy level position in the emitter by the adjusting material composition and sizes of the emitter and the active layer. The width of the tunnel barrier should be chosen in range 5- 50Ǻ to make it transparent for electrons and not transparent for holes. At the same time the hole energy level positions in active layer should be chosen lower than the energy level positions for holes in the electron emitter. This forbids the hole penetration without thermal activation into the electron emitter. It is important that even the small amount of thermally activated holes have no possibility to tunnel into the electron emitter for the chosen barrier width because of their heavy mass.

Advantages of this design are:

  • increase of capture efficiency of the electrons into the active layer due to direct resonance tunneling of the electrons from the electron emitter into the active well

  • suppression of the electron leakage into the hole emitter

  • elimination of the parasitic light generated outside the active layer

  • the electron emitter acts also as a good current spreading layer

References

[1] Y.T.Rebane, Y.G.Shreter and W.N.Wang, phys.stat.sol.(a) vol 180, pp 121-126 (2000)