The zener effect is the dominant breakdown effect in low voltage zener diodes. Let's take some time to learn more about it.
Zener diodes are diodes with a known breakdown voltage, which is called the zener voltage. There are two effects, that cause this breakdown: the zener effect and the avalanche breakdown. The zener effect is the dominant breakdown effect at voltages below 5 V. Compared with the avalanche breakdown, it is noticeable by it less steep I-V curve. Today, we want to take a closer look at it.
The zener effect is named after Clarence M. Zener, who was an American physicist researching on breakdown effects in different materials. In 1934, he published a theory predicting the properties of the zener breakdown. The first zener diode was built in 1950 by Bell Labs and named after him. But what was it that Zener predicted? And how does the zener effect work in semiconductors? Are you ready for a little deep dive into semiconductor physics?
Disclaimer: I'm not a physicist and semiconductors are an extremely complicated topic. I cannot guarantee for correctness of my explanations. For the ease of understanding some details might be left out or presented in an over simplified manner. If you think, that parts of this article need correction, contact me at firstname.lastname@example.org.
To understand how the zener effect in zener diodes work, we first need to recap, what happens on a p-n junction. I already described this in lengths in a dedicated article on the p-n junction, however, to understand the zener effect, we should take another look at the p-n junction – this time with the help of the band model.
If you followed up on my past articles, the following image might be familiar to you. It shows the process of electron excitation and relaxation in semiconductors in the band model. I'll use this image to recap the most essential parts about the band model for you.
The band model is based on the quantum mechanic observations, that electrons are not located at fixed positions in an atom. Instead, their positions have to be described probabilistically. However, this does not mean that the electrons are scattered totally random around the nucleus. Around the nucleus there are different energy bands in which the electrons are likely located. In between these bands there are gaps: the so-called forbidden zones or band gaps. These gaps described a region in which electrons are very unlikely to be found. An electron cannot simply move into another band. It needs enough energy to jump across the band gap into the next energy band. This happens at random when electrons collide with another, e.g., due to internal thermal motion, and as well under the external influences like colliding photons.
The most interesting energy bands in the context of electronics are the valence band and the conduction band. The conduction band is the first band in which electrons can freely move between atoms, which makes it possible for a current to flow. Inside the valence band, the electrons are captured and can only move, if another atom next to them is missing an electron. In these case they can fill up the empty spot, whereby this spot moves to the atom the electron originated from. The places with missing electrons are called electron holes or defect electrons and allow for conduction within the valence band.
The Y-axis of the band model describes the energy level. For electrons, it is favorable to move to a lower energy level. Electron holes shift towards a higher energy level.
Now, with the basics covered, let's have a look at the p-n junction :
As we know, diodes make use of doped semiconductor material. In the p-type region, the electron acceptors, facilitate an easy creation of electron holes in the valence band. In the n-type region the electron donors provide electrons which can be easily excited into the conduction band. The consequence is, that there are many free electrons in the n-type material and many free holes in the p-type material, through which electric charges can be transported. These are the so-called majority charge carriers.
The electrons diffuse through the junction into the p-type material, where they recombine with free holes. Equivalently, the holes in the valence band of the p-type material diffuse to the n-type region and recombine with the electrons there. This process also causes a charge build-up at the junction: the p-type side becomes negatively charged and the n-type side positively. In the band model, the electric field around the junction is visually represented by bent energy bands. The force of electric field opposes the diffusion process. It hinders the majority charge carriers from passing through the junction. The electrons would now have to move uphill, which is energetically unfavorable. Without an external electric field, both forces are at equilibrium at an internal electric potential of around 0.7 V. A depletion region is formed, in which there are only very few free majority charge carriers.
If a forward voltage is applied to the diode, the external electric field supports the diffusion process. The depletion region shrinks. Once the internal potential of 0.7 V, which is equal to what we know as the forward voltage, is overcome, the current can freely flow through the junction.
To explore this in the band model, we need to integrate the external voltage into the visualization of the energy bands. The rule for this is to add the external potential difference to the internal one. The electric potential on the n-type side, to which the negative pole is connected, is now higher than the one on the p-side side, to which the positive pole is connected. The electrons can flow downhill.
Now for the interesting part: what happens if the diode is reverse biased? Again, we add the potential difference of the externally applied voltage to the internal one. We end up with the following situation:
There is now a big potential barrier between both sides of the junction. Normally, no current can flow through it. The external electric field enforces the drift process and thus the width of the depletion region. The exception to this is a small reverse leakage current, that is caused by minority charge carriers. The band model allows to easily see what is meant by reverse leakage current. If there happens to be a free electron inside the conduction band of the p-type material, it gets accelerated by the applied electric field and moves through the junction. However, free electrons inside the conduction band of the p-type material and equivalently free holes inside the valence band of the n-type material, are too rare to create a noteworthy current flow.
So much for the working principle of classic diodes. Things change, when the zener effect comes into play. Due to their high doping levels, zener diodes have a very narrow depletion region. In it there are high electric field forces. The energy bands are steeply bent and there are now regions in the p-type material's valence band, that have the same energy level as the n-type material's conduction band. This enables a quantum mechanic effect called tunneling effect. With a certain probability electrons can tunnel through a potential barrier to a place with the same energy level at the other side. In case of zener diodes this is what is known as zener effect: electrons tunnel from the valence band of the p-type material into the conduction band of the n-type material.
Note, that Zener himself did not discover the tunnel effect, he already knew about it. What he achieved was to describe at what rate electrons would tunnel through the junction and to provide a mathematical model for it. With it, he showed that this effect kicks in suddenly once a certain voltage is exceeded. He also examined under which conditions it would occur, namely at high field strengths and a narrow band gap, like it is present in a zener diode.
Zener didn't develop his theory specifically for diodes, but for dielectric insulators. Regarding the zener effect, they have the similar properties as a semiconductor junction, except that for most insulators the zener effect is only of theoretical, and not of practical relevance. For most materials as well as for other types of semiconductors, the avalanche breakdown - which we will cover next - is of much higher importance. If you are interested in more details, Zener's publication is available in the archive of the Royal Society.