Source: Phase Transformations and Complex Properties – University of Cambridge

Solid State Diffusion in Semiconductor Processing

Solid state diffusion is a fundamental process used in semiconductor manufacturing to introduce dopant atoms into materials like silicon. This process plays a crucial role in the production of devices such as solar cells.

Introduction to Solid State Diffusion

Heating a wafer at high temperatures in the presence of dopant atoms causes these atoms to be incorporated into the top surface of the wafer. In silicon solar cell processing, boron is commonly used to create a p-type base, while phosphorus is used for the n-type emitter layer.

Calculation of Diffusion Profiles

The diffusion process can be described by Fick’s law, where the flux density (j) is proportional to the product of the diffusion coefficient (D), concentration volume (N), and distance (x). Diffusion profiles can be calculated for specific cases, such as diffusions from unlimited and limited sources.

Diffusion from an Unlimited Source

Diffusions from unlimited sources typically result in shallow junctions with high surface concentrations of dopant atoms. The diffusion process is often described by the complementary error function, with parameters such as impurity concentration at the surface (N0), diffusivity (D), depth (x), and time (t) influencing the profile.

Diffusion from a Limited Source

Diffusions from limited sources involve a two-step process: a short pre-deposition followed by a longer drive-in at higher temperatures to create a deep, lightly doped emitter. The final dopant profile is often Gaussian in shape and can be described using mathematical equations.

While simplified models are useful, second-order effects can lead to deviations from these ideal profiles. Computer simulations are commonly employed to account for these complexities in real-world semiconductor processing.

Source: MDPI