...

HOME

INTRODUCTION

MECHANISMS

DESIGN

STATUS

Electroabsorption modulators (EAMs) are just one type of optical modulators. Currently, the optical modulators that dominate the market are the so-called Mach-Zehnder modulators (MZMs) based on either LiNbO3 [see an example by Lucent (data sheet: 10 Gb/s version 188 kB, 40 Gb/s version 260 kB) ] or GaAs. It uses the refractive index change as a function of the applied electric field (the electro-optic effect). Problems of MZMs include: big size and the difficulty of integration with the light source, the laser diodes.

The EAMs are based on the electroabsorption effect: They control the light intensity by changing the absorption coefficient as a function of the electric field. Because the absorption coefficient can be changed quite drastically by applying the electric field in some semiconductors, the EAMs are very effective and the size can be quite small. (EAMs are typically around 200 um long while the MZMs are centimeters long. The final packaged devices are of course bigger than these.) Another advantage of EAMs is the ease of integration with the light source, as the laser diode and the EAMs are based on the very similar material and structure.

There are two different physical mechanisms that are currently used: the Franz-Keldysh effect (FKE) and the quantum-confined Stark effect (QCSE). The FKE refers to the phenomenon that occurs when we apply the electric field to the bulk semiconductors (typically thicker than a tenth of microns). To have an efficient optical modulator, we want minimum absorption when there is no electric field and large absorption when we apply the electric field. To achieve this, we use the material whose bandgap energy is larger than the light energy that we use. (For 1.3 or 1.55 um, the most commonly used material system that has this range of bandgap is InP-based.) When we apply the electric field to the semiconductor, the energy band is tilted. This makes the effective bandgap smaller at the non-zero electric field than the bandgap at the zero electric field. This can be physically described by the virtual transition of electrons from the valence band to the conduction band and the tunneling through the triangular barrier. As the degree of tilt is related to the amount of the applied electric field, the more electric field we apply, the more transition occurs, thereby increasing the absorption coefficient. (Absorption in semiconductor happens as we know by transitions of electrons from the valence band to the conduction band.) The amount of absorption coefficient change is as large as a few hundred inverse centimeters as the electric field is changed from zero to hundreds of kV/cm in FKE.

The QCSE happens in the quantum wells as its name indicates. A quantum well can be made in semiconductor by growing two different materials in such a way that the energy band forms a well due to the band discontinuity (sometimes called band offset). One example is InGaAs and InP, which have the different band line-up. So if InP is grown and then InGaAs and InP, they form a quantum well if the thickness of InGaAs is in the order of 100 angstroms (several tens of atomic layers). For the QCSE to be useful for the optical modulator, the wells in the conduction band and in the valence band should spatially conicide as in the case of InGaAs and InP (this is called type I heterostructure). More than one quantum well are commonly used to enhance this effect. When more than one quantum well are grown, the barrier material should be thick enough not to cause coupling between adjacent quantum wells because coupling between adjacent wells broadens the quantized energy levels inside each well (so called superlattice effect). When the barrier layers are thick enough so that the coupling between adjacent well is small, this group of quantum wells are called mutiple quatum wells (MQWs).

When light is incident on the MQWs, the light is absorbed if its energy is equal to the one that is required to lift an electron from the valence band to the conduction band. After this, the electron in the conduction band and the hole created in the valence band form an exciton, a hydrogen-like quasi-particle bound by the Coulomb interaction. The existence of the exciton at room temperature is one of the biggest characteristics that distinguishes the quantum wells from the its bulk counterpart, and the main reason why we use MQWs to make an optical modulator: The absorption peak by this exciton is very sharp at zero electric field. Here also we have to make the quantum well not absorb at zero electric field by carefully designing the quantum well thickness with the given materials. This means that the exciton peak is at the shorter wavelength than the light wavelength. When the electric field is applied, the exction peak becomes broadend and the position itself to the longer wavelength. This makes the absorption coefficient becomes very high and as a result, the output transmitted light will be very small: the operating principle of the EAM. Compared with the amount of absorption coefficient change with the FKE , the QCSE has about 5 ~ 10 times bigger change. This makes the EAM made of the MQW materials very attractive.