Transistor Structure: Bipolar Transistor Fabrication

Essential details of the structure of typical bipolar transistors with explanations to help understand how they work.

Home » Electronic components » this page

Transistor Tutorial Includes:
Transistor basics Gain: Hfe, hfe & Beta Transistor specifications Transistor and diode numbering codes Choosing replacement transistors

Bipolar transistor fabrication and their structure has come a long way since the first transistors were made.

Today’s transistors are made using sophisticated processes and the transistor structure enables them to have very high levels of performance.

The original transistor made by Bardeen, Brattain and Shockley consisted of two very closely spaced contacts on a germanium base. The structure of this transistor consisted of two point contacts on a germanium base area.

Today the transistors are made in a variety of ways and have a variety of different structures. They may be diffused, epitaxially grown or they may use a mesa construction.

Bipolar transistor structure: basics

In essence a transistor consists of an area of either p type of n type semiconductor sandwiched between regions of oppositely doped silicon. As such devices can be either a p-n-p or an n-p-n configuration.

There are three connections, namely the emitter, base, and the collector. The base is the one in the centre and it is bounded by the emitter and collector. Of the two outer two the collector is often made larger as this is where most of the heat is dissipated.

Transistor basic structure and circuit symbols

The base derives its name from the first point contact transistors where the centre connection also formed the mechanical "base" for the structure. It is essential that this region should be as thin if high levels of current gain are to be achieved. Often it may only be about 1 um across.

The emitter is where the current carriers are 'emitted', and the collector is where they are 'collected'.

Point contact transistor structure

The very earliest transistors used a point contact structure. This transistor structure was easy to manufacture using very low technology equipment, but was not reliable.

As the name indicates, this bipolar transistor structure uses wires to make a point contact on the semiconductor material.

Diagram of the structure of a point contact bipolar transistor showing the base semiconductor and the two wires forming point contacts onto the semiconductor. — Point contact transistor structure

Alloyed junction structure

Another transistor structure that was widely employed in the early days of transistors was the alloyed junction.

Diagram of the structure of an alloyed junction bipolar transistor structure showing the base semiconductor and the two p+ areas forming the emitter and collector. — Alloyed junction transistor structure

The alloy junction transistor structure used a germanium crystal as the base for the whole structure, as well as being the base connection. The emitter and collector alloy beads were then fused on opposite sides. There were several types of improved alloy-junction transistors developed over the years that they were manufactured.

The alloy-junction transistors became obsolete in the early 1960s, with the introduction of the planar transistor which could be mass-produced easily while alloy-junction transistors had to be made individually.

Diffusion transistors

Unlike the previous transistor structures where the contacts were added to the semiconductor crystal externally, the diffusion transistor enabled the various regions of the transistor to be created by diffusing dopants into the semiconductor crystal to give regions of the required characteristics, p-type, n-type, p+, n+, etc.

The earliest diffusion transistors used a form of diffused-base transistor structure. These transistors still had alloy emitters and they even sometimes possessed alloy collectors like the earlier alloy-junction transistors. Only the base was diffused into the substrate, although sometimes the substrate formed the collector.

Planar transistor structure

The planar transistor structure was developed at Fairchild Semiconductor in 1959 and it represented a major breakthrough in technology. Not only did it enable bipolar transistors to be manufactured more easily, but it also set in place the basis of future integrated circuit technology as well.

The planar transistor structure also includes a passivation layer on the external areas of the crystal. This protects the junction edges from contamination, and allows much less expensive plastic packaging to be used without risking degradation transistor performance as a result of contamination entering the crystal lattice, especially in the regions around the junctions.

Surprisingly, the first planar transistors had inferior performance levels when compared to their alloy junction relations, but the diffused planar structure transistors could be mass produced and as a result they cost much less, making them a very attractive option. However the early difficulties have been overcome and planar transistors offer very high levels of performance.

Simplified diagram of the structure of a planar junction bipolar transistor structure showing the base semiconductor and the areas forming the emitter, base and collector. — Simplified planar transistor structure

It is worth noting that the collector area has a larger volume than the emitter. Although in many respects the two terminals could be exchanged, the collector is where the most power is dissipated and therefore it is made to have the larger volume.

It can also be seen that in this transistor the current flow takes place in a vertical plane in the diagram.

Further differences exist in the doping levels used in the transistor structure as well. The emitter doping is generally higher than the base doping to provide a high injection efficiency. Also the collector doping is lower than the base doping.

A common approach to form the emitter and base junctions is to use a process known as a double diffusion technique. Using the double diffusion technique the base area diffusion is undertaken first to provide the larger base area. Then the smaller emitter area is diffused with a higher dopant level to provide the shallower more heavily doped emitter.

Lateral planar transistor structure

In some cases it can be advantageous to have a lateral transistor structure.

Diagram of the structure of a lateral planar junction bipolar transistor structure noting that current flow is in the horizontal plane on the diagram. — Lateral planar transistor structure

It can be seen from this diagram that the current flow is in the horizontal plane rather than the vertical plane. This format has advantages in some applications, but requires more diffusion processes and is therefore more complex and hence more expensive. As such it is only used when performance and characteristics demand it.

Transistor doping & doping profiles

Whatever the construction method is used for a transistor, the base layer is kept very thin. It is typically less than 1µm for high current gain.

For most bipolar junction transistors, the current flow is in a vertical plane although a lateral structure is possible if needed.

In terms of doping levels within the transistor structure, the emitter doping is generally the highest. This provides the highest injection efficiency. Next is the base doping. Finally the collector is given the lowest doping level so that the neutral base function is a weak function of the collector base voltage.

The typical doping profile for a transistor structure shows the different regions for the transistor with their doping levels.

Typical doping profile for standard silicon transistor (NPN)

As can be seen from the transistor structure diagram, the emitter doping is very much higher than ether the base of collector regions.

Although it is possible for transistors to operate with emitter and collector reversed, the performance will be inferior. As a result of the incorrect doping levels. Also the collector is made larger and able to have its heat conducted away more easily as this is the region of the transistor structure where most heat is dissipated.