Channeling

Many targets are crystalline in nature with regular arrangement of atoms. For ions moving in certain directions, the atom rows or planes line up so that there are long-range open spaces through which the ions can travel without significant scattering. Ions are steered down these channels by glancing collisions with the atom rows or planes thus extending the final ion distribution deeper into the target. This is known as channeling.

The effect of channeling is to add a tail to the implanted distribution, which for silicon can be approximated by an exponential with decay length around 0.1 microns.

We don’t want channeling due to the following reasons:

• Modeling of channeled ion distributions is difficult and makes it difficult.

• Also channeled ion profile becomes sensitive to changes on the order of 1° in wafer tilt and beam divergence.

Channeling can be eliminated by the following methods:

• Tilt the wafer by 7° relative to the incoming beam to give the appearance of a random target, and hence avoid channeling.

• Ions can be scattered out of channels by thermal vibrations, lattice defects and foreign atoms.

• It can be reduced due to damage to the crystal caused by nuclear stopping of unchanneled ions.

• A preceding implant of silicon to amorphize the lattice structure will also reduce channeling effects.

Electronic interactions will stop the ion inside the channel. But it should be noted that because the electron density at the center of a channel is low, the electronic stopping power for a well-channeled ion would be less than that of an ion traveling randomly.

Channeling is characterized by a critical angle, Y₁,which is the maximum angle between ion and the channel for a glancing collision to occur.

Fig. 7 Schematic views of channeling (a) Ion paths through a cubic lattice, showing channeled and nonchanneled cases. (b) Back-scattering yield around a channeling direction. (c) the effect of channeling is to add a tail to the atom distribution.[2]