Skip to main content

Epitaxy and 3 Important Techniques for Fabrication of Heterostructures

Molecular Beam Epitaxy (MBE)
Level 3 (up to Physics B.Sc.)
Level 3 requires the basics of vector calculus, differential and integral calculus. Suitable for undergraduates and high school students.
Updated by Alexander Fufaev on
Table of contents
  1. What are the epitaxial growth models?
  2. Technique #1: Molecular Beam Epitaxy (MBE) Here the method 'Molecular beam epitaxy' and its advantages and disadvantages are explained.
  3. Technique #2: Metal Organic Chemical Vapor Deposition (MOCVD) Here the process 'Metal-Organic Chemical Vapour Deposition' and its advantages and disadvantages are explained.
  4. Technique #3: Liquid phase epitaxy (LPE) Here, the 'Liquid Phase Epitaxy' method and its advantages and disadvantages are explained.

Structures obtained by epitaxy (e.g. semiconductor GaAs) are used in LEDs, solar cells and laser diodes.

  • Homoepitaxy - is when the atomic layers deposited on the substrate consist of same atoms as the substrate itself.

  • Heteroepitaxy - is when the atomic layers deposited on the substrate consist of different atoms than the substrate.

What are the epitaxial growth models?

The epitaxial growth models theoretically describe how the atoms attach to the substrate. There are three important growth models, classified according to the bonding between the layer atom and the substrate.

  • Volmer-Weber growth: Here, the bond between the atoms of the layer is stronger than the bond to the substrate. As a result, (three-dimensional) islands are formed on the substrate.

  • Frank van der Merve growth: Here, the bonding between the layer atoms is weaker than between the substrate and the layer atom, leading to the formation of monolayers.

  • Stranski-Krastanov growth: Initially, the bonding between layer atoms to the substrate is stronger than the bonding between layer atoms to each other, which is why first a few monolayers form on the substrate. After a few monolayers, the bonding between the individual layer atoms predominates, so that thereafter rather (three-dimensional) islands are formed. In this way, individual quantum dots can be produced.

There are three main techniques for producing epitaxially grown structures. They differ not only in the procedure, but also in the growth rate (how quickly structures are formed), production costs, etc.

Technique #1: Molecular Beam Epitaxy (MBE)

Illustration : Basic setup of an MBE machine.

In this process, the materials (for the layers) are evaporated in the effusion cells. Individual atoms escape through a very small opening in the effusion cell and land on the substrate. The shutters in front of the effusion cells allow individual atom beams to be blocked, so that material layers can be varied as a result. In addition, the substrate is heated to ensure better diffusion (distribution) on the substrate.

Advantages of the technique:

  • Temperature for heating the substrate can be set independently of the temperature in the effusion cells. This allows growth at lower temperatures, which leads to better control of growth. In addition, thinner layers can be grown as a result.

  • Almost all materials can be vaporized in the effusion cells, i.e. more material combinations are possible.

  • Since growth is relatively slow, it can be observed in real time using electron microscopy (RHEED).

Disadvantages of the technique:

  • The size of the substrate is limited to a dozen centimeters because the atomic beams from the effusion cells can no longer be directed to fall uniformly on the substrate for larger substrates.

  • Fluctuating growth rates if the amount of gas in the effusion cells becomes smaller.

Technique #2: Metal Organic Chemical Vapor Deposition (MOCVD)

Illustration : Basic setup of an MOCVD machine.

A carrier gas consists of atoms that are supposed to dock to the substrate and are in combination with organic substances. This carrier gas flows into a reaction chamber laminarly (without turbulence). For example, galium (Ga) forms a compound with three \(\text{CH}_3\) molecules and arsenic (As) with one \(\text{H}_3\) molecule. These molecules are blown into the reaction chamber where a vacuum (\(0.1\, \mathrm{bar}\) to \(1 \, \mathrm{bar}\)) is established. The reaction chamber is heated so that the incoming molecules diffuse to the substrate and decompose there - due to high temperature. While galium and arsenic accumulate on the substrate, the residual gases (organic compounds) are removed from the reaction chamber.

The growth rate depends on the temperature of the substrate. At low temperatures, the chemical reaction (decomposition) is slow, although there are enough molecules on the substrate surface. At higher temperatures, almost all molecules decompose on the substrate, but their amount on the substrate is limited by diffusion to the substrate surface. In this temperature regime, the growth rate is independent of the temperature. At too high temperatures, the atoms detach from the substrate (desorption), which in turn leads to a decrease in the growth rate.

Advantages of the technique:

  • Mass production possible, since the size of the reaction chamber can theoretically be any size.

  • Higher and especially reproducible growth rate than molecular beam epitaxy, due to higher temperatures.

Disadvantages of the technique:

  • Lower quality sample than with molecular beam epitaxy, because at higher temperatures growth becomes more difficult to control.

  • Dangerous because of toxic arsine and highly explosive.

Technique #3: Liquid phase epitaxy (LPE)

Illustration : Basic setup of an LPE machine.

The saturated melt is brought into contact with the substrate under a hydrogen atmosphere. It is passed over the substrate (or the other way around). The substrate is then cooled, reducing the solubility limit of the material in the melt so that a pure layer of the material is formed on the substrate as a result.

Advantages of the technique:

  • High growth rate

  • Technique is cost-effective

Disadvantages of the technique:

  • No very precise control of the layer thickness. The method is therefore more suitable for thicker layers (> 10 nm).

In the next lesson, we'll look at how photolithography can be used to build integrated circuits.