Semiconductor Light Emitters Feel the Force

Everyday electronic devices such as transistors, microprocessors and mobile phones, rely upon currents of electrons flowing through a semiconductor in response to an applied voltage. Now our experiments have revealed that the optical excitations responsible for light generation in semiconductors, can be moved around in a similar fashion using applied voltages. The discovery, which has been reported in the research journal Science, resulted from collaboration between scientists at High Field Magnet Laboratory (University of Nijmegen, the Netherlands), Toshiba Research Europe Limited (Cambridge, UK), and University of Cambridge.

We have found that the motion of certain complexes of electrons and holes (called excitons) can be controlled by an applied voltage. Excitons play an important role in determining the optical properties of semiconductors, such as light absorption and emission, which are of uppermost importance for electro-optic applications. The research demonstrates for the first time that charged excitons drift when subjected to the electric field produced by the applied voltage.

Charged excitons on the moveLight absorbed by a semiconductor promotes electrons out of a full energy band into a higher energy, empty band. Usually the negatively charged electron liberated by light is strongly attracted to the positively charged hole it vacated in the full energy band. This process leads to the formation of an exciton. An exciton can be regarded as a temporarily storage of the energy of the absorbed photon. When eventually the electron falls back into the hole (exciton recombination) the photon is re-emitted. The net charge of an exciton is equal to zero, and therefore the exciton is unaffected by an applied voltage. However, by attaching an extra electron to the photo-created electron-hole pair, thereby giving it net negative charge, the three-particle complex can be made to drift towards a positive electrode. The place where the emitted photon leaves the semiconductor therefore can be chosen by the application of an electric field.

These results show that charged excitons drift in an electric field in a similar way to electrons. It suggests we may be able to control light emission in semiconductors in previously unimagined ways.

Physical background

This new phenomenon was demonstrated using the transistor structure shown in the figure, which allows control of the concentration of excess electrons under the gate region by applying a voltage between the gate and drain contacts. The active layer of the transistor consists of an ultrathin (30 nanometer) layer of galliumarsenide, just beneath the surface of the sample.

Figure 1: A focused laser beam illuminates a transistor, that contains a 30 nanometer thin galliumarsenide active layer, just beneath the sample surface. The laser light excites electrons from a filled to an empty energy band, leaving behind an empty place (a hole) in the filled band. Eventually the electron will fall back into the hole by emitting a photon. As long as that process did not occur the negatively charged electron pairs up with the positively charged hole, due to their mutual attraction. The resulting particle is called an exciton, which has no net charge. After introducing excess electrons to the active layer of the transistor, by applying a gate voltage, each photo-excited exciton captures an additional electron to form a negatively charged exciton or in short trion. Application of a voltage between the source and drain of the transistor results in an eletcric field along the active layer, which drives the trions towards the collector. As a result, the spatial distribution of the light emission intensity, as measured by a special optical microscope, becomes strongly asymmetric. The negatively charged exciton emits it photon at a position different from the point where it was created. The position where the light is emitted can thus be controlled by application of an electric field.

An incident laser excites electron-hole pairs within the active layer of the semiconductor under the gate region. The electron and hole bind to form a complex called an exciton, which can be considered as the semiconductor analogon of the hydrogen atom, i.e. a negatively charged exciton bound by a positive nucleus (the hole). Since the exciton as a whole has no net charge, it is insensitive to an electric field created by a voltage applied between the source and drain of the transistor. Thus light emission is restricted to the same area of the device that is excited by the laser.

However, strikingly different behaviour is observed after introducing excess electrons to the active layer. In this case the photo-excited electron-hole pairs bind to an excess electron, to form a three-particle exciton, called a trion or negatively charged exciton. This semiconductor complex can be regarded as the analogue of the negative hydrogen ion, H^-. Now we have found, that due to their net negative charge, trions are able to drift in an applied electric field, travelling over distances as long as several micrometers. Thus upon applying a voltage between the source and drain of the transistor, the light emissive area, as seen through a special optical microscope, is skewed away from the incident laser spot (see figure). The trions drift in the opposite sense to the field because of their negative charge. Reversing the polarity of the applied field causes the excitons to drift in the opposite direction. This is the first time that excitons have been manipulated in this way by applied voltages.

Figure 2: The optical setup is a specially designed, 5 meter long, microscope, consisting of a microscope objective placed in a liquid Helium bath (4.2 K), just above the sample, and an ocular which focuses the light on the entrance slit of a monochromator. The exciting laser beam is guided through a spatial filter to ensure that the beam is gaussian and, via the second lens, to set the divergence of the beam. The 40 X microscope objective focuses the laser to a 1.2 mm-diameter spot on the sample surface, and collects the luminescence. The luminescence (PL) is guided through the monochromator and detected by the CCD camera. By this method the image on the CCD camera is spectrally resolved along its horizontal axis and spatially resolved along its vertical axis.

Further Information

• Fabio Pulizzi, phone: +44 (0)115 9515195, email: Fabio.Pulizzi (at) nottingham ac uk
• Peter Christianen, phone: +31-24-3652245, email: P.Christianen (at) science ru nl
• Jan Kees Maan, phone: +31-24-3653422, email: JC.Maan (at) science ru nl
• Andrew Shields, Toshiba Research Europe Ltd, Cambridge, UK, email: andrew.shields (at) crl toshiba co uk