PART 4: SEMICONDUCTOR MANUFACTURING Doping: The semiconductive character

Related Vendors

ROHM Semiconductor GmbH

Rogers Corporation

DOWA HD Europe GmbH

A wafer in a foundry cannot “electrically” behave like a semiconductor. In fact, a material’s conductivity is a function of added impurities. Doping is the process of making semiconductors more conductive, in simple words, converting wafers into semiconductors.

What is doping?

In our guide to intrinsic and extrinsic semiconductors, the theory of doping is explained in detail. This article discusses doping with respect to semiconductor manufacturing foundries. However, this section includes the theory of doping in brief.

The process of adding impurities to a material to enhance its conductivity is known as “doping”. The added impurities are termed “dopants”. These dopants enhance the conductivity of a material by modifying its electrical band structure. Besides electrical character, doping controls the wafer’s optical, internal, and structural characteristics.

The dopant atom can either be a donor or an acceptor. The donor dopant, a pentavalent impurity, forms an n-type semiconductor. The acceptor atom, a trivalent impurity, forms a p-type semiconductor.

Trivalent impurities for P-type semiconductors (Group lll elements)

• Boron (Z=4)

• Aluminum (Z=13)

• Gallium (Z=31)

• Indium (Z=49)

Pentavalent impurities for n-type semiconductors (Group V elements)

• Phosphorus (Z=15)

• Arsenic (Z=33)

• Antimony (Z=51)

Laser welding in comparison with conventional soldering technology

Development of Cu-Cu Joining Technology by Laser Welding

Doping: Microfabrication

The process of doping in semiconductor manufacturing can be done at various stages and multiple times. Depending upon the manufacturer’s protocols, doping can be done before crystallization and after photolithography and etching processes.

Initially, dopants in small quantities are introduced at the time of the crystallization process in the wafer manufacturing procedure. Doping methods such as diffusion and ion implantation are used after chemical processes. The two main processes implemented in fabs for doping wafers, diffusion and ion implantation, are listed below:

Diffusion

One of the earliest methods used in fabs, the diffusion process forms a deep junction in the CMOS technology. The diffusion process usually starts during the chemical vapor deposition (CVD) technique or after the photolithographic process in fabs. The dopant atoms are present in the gaseous state for the diffusion process. The source of these atoms is mainly a liquid state.

Wafers are placed inside the furnace in the presence of an impurity dopant mixture gas at a high temperature. The dopant atom, known as diffusant here, moves through the crystal lattice structure of the wafer and makes covalent bonds at a high temperature between 700-1200 degrees Celsius.

The principle of the diffusant atom is to create free charge carriers inside the wafer. The doping profile in the diffusion process depends on the concentration gradient. In simple words, the diffusion process continues until thermal equilibrium. At thermal equilibrium, the concentration of diffusants is equal throughout the wafer.

The drawback of the diffusion process is the formation of parasitic characteristics. As the diffusant moves through the wafer, the crystal lattice structure is compromised. The wafer is then sent for annealing. The process involves heating the wafer to remove disturbances in the crystal structure.

E-BOOK

Semiconductors: A Comprehensive Guide

Ion Implantation

In today’s world, fabs implement the ion implantation method to control the conductivity of chips. The process has been extensively used since the late 1970s in fabs. The words “ion implantation” point out towards implantation of ions in the wafers. Unlike the diffusion method, ion implantation uses a beam to implant high-energy ions in semiconductor chips.

The ion implantation process incorporates a source, a mass analyzer, and an accelerator.

Ion implantation steps:

• The source includes dopant ions and sends them to the mass chamber.

• The mass chamber enables precise control and selection of the ions with desirable mass.

• The mass chamber sends the ions to the accelerating tube to electrically energize them.

• Finally, ions with high kinetic energy enter the crystal lattice of the substrate at a certain angle in a controlled manner.

• The ions subsequently lose energy in collisions to end the implantation process.

The doping profile depends directly upon the mass of ions and acceleration energy (in eV) to implant them. The ion implantation method fills the gaps in the diffusion process as it does not require high temperature and offers precise control of dopants. But ion collisions may damage crystal structure, similarly, making fabs carry out the process of annealing.

What after doping?

Fabs implement either diffusion or ion implantation process or both to dope semiconductors. Both processes can damage the crystal lattice structure of the wafer. In semiconductor manufacturing, annealing is a heating process to restore the distortion of the crystalline structure of the substrate due to doping damage. Annealing recrystallizes the wafer for high-quality fabrication.

"Semiconductor Manufacturing" article series

• Part 1: Introduction to semiconductor manufacturing

• Part 2: The Semiconductor “Wafers”

• Part 3: Semiconductor circuit design: key chemical procedures

• Part 4: Doping: The semiconductive character

• Part 5: Microfabrication testing: Electrical die sorting (EDS)

(ID:49812492)