Specialized Diodes

By adjusting the doping levels and gradients as well as the geometry of a semiconductor crystal, we can modify the behavior of the device. This page lists a wide range of diodes whose properties have been deliberately controlled to produce specific capabilities.
Each of these specialized diodes has its own schematic symbol, shown to the right of its description below. The symbols are all specific variations on the basic diode symbol, so that the nature and function of the device is clear on a schematic diagram.


Light Emitting Diode (LED)

One of the questions semiconductor manufacturers asked themselves was, "What happens if we increase the doping levels in the silicon crystal?" Trying this gave rise, among other things, to the tunnel diode. Then they took the process even further, to the point where they skipped the silicon completely, and produced what is called a "III-V" device, named after the fact that P-type dopants are from column III of the Periodic Table (aluminum, gallium, indium) and N-type dopants are from column V (phosphorus, arsenic).
The resulting Gallium Arsenide (GaAs) crystal had the interesting property of radiating significant amounts of infrared radiation from the junction. By adding Phosphorus to the equation, they shortened the wavelength of the emitted radiation until it became visible red light. Further refinements have given us yellow and green LEDs. More recently, blue LEDs have been produced, by putting nitrogen into the crystal structure. This makes full-color flat-screen LED displays possible.
The mechanism of emitting light is interesting. The atomic structure of the LED is carefully designed so that as free electrons cross the junction from the N-type side to the P-type side, the amount of energy each electron releases as it drops into a nearby hole corresponds to the energy of a photon of some particular color. Therefore, that photon is released as a visible photon of that color.

P-I-N Diode
The p-i-n diode doesn't actually have a junction at all. Rather, the middle part of the silicon crystal is left undoped. Hence the name for this device: p-intrinsic-n, or p-i-n. Because this device has an intrinsic middle section, it has a wide forbidden zone when unbiased. However, when a forward bias is applied, current carriers from the p- and n-type ends become available and conduct current even through the intrisic center region. The end regions are heavily doped to provide more current carriers.
The p-i-n diode is highly useful as a switch for very high frequencies. They are commonly used as microwave switches and limiters.

Tunnel Diode
As we mentioned in our discussion of semiconductor physics the addition of either P-type or N-type impurities causes the Fermi level in the silicon crystal to shift towards the valence band (P-type impurities) or the conduction band (N-type impurities). The higher the doping level, the greater the shift. In the tunnel diode, the doping levels are so high that the Fermi levels in both halves of the crystal have been pushed completely out of the forbidden zone and into the valence and conduction bands.
As a result, at very low forward voltages, electrons don't have to gain energy to get over the Fermi level or into the conduction band; they can simply "tunnel through" the junction and appear at the other side. Furthermore, as the forward bias increases, the applied voltage shifts the levels apart, and gradually back to the more usual diode energy pattern. Over this applied forward voltage range, diode current actually decreases as applied voltage increases. Thus, over part of its operating range, the tunnel diode exhibits a negative resistance effect. This makes it useful in very high frequency oscillators and related circuitry.

Varactor Diode
One characteristic of any PN junction is an inherent capacitance. When the junction is reverse biased, increasing the applied voltage will cause the depletion region to widen, thus increasing the effective distance between the two "plates" of the capacitor and decreasing the effective capacitance.
By adjusting the doping gradient and junction width, we can control the capacitance range and the way capacitance changes with applied reverse voltage. A four-to-one capacitance range is no problem; a typical varactor diode (sometimes called a "varicap diode") might vary from 60 picofarads (pf) at zero bias down to 15 pf at 20 volts. Very careful manufacturing can get a capacitance range of up to ten-to-one, although this seems at present to be a practical limit.
Varactor diodes are used in electronic tuning systems, to eliminate the use of and need for moving parts.

Zener Diode
When the reverse voltage applied to a diode exceeds the capability of the diode to withstand it, one of two things will happen, yielding essentially the same result in either case. If the junction is wide, a process called avalanche breakdown occurs, whereby the current through the diode increases as much as the external circuit will permit. A narrow junction will experience Zener breakdown, which is a different mechanism but has the same effect.
The useful feature here is that the voltage across the diode remains nearly constant even with large changes in current through the diode. In addition, manufacturing techniques allow diodes to be accurately manufactured with breakdown voltages ranging from a few volts up to several hundred volts. Such diodes find wide use in electronic circuits as voltage regulators.

Schottky Barrier Diode
When we get into high-speed applications for electronic circuits, one of the problems exhibited by semiconductor devices is a phenomenon called charge storage. This term refers to the fact that both free electrons and holes tend to accumulate inside a semiconductor crystal while it is conducting, and must be removed before the semiconductor device will turn off. This is not a major problem with free electrons, as they have high mobility and will rapidly leave the semiconductor device. However, holes are another story. They must be filled more gradually by electrons jumping from bond to bond. Thus, it takes time for a semiconductor device to completely stop conducting. This problem is even worse for a transistor in saturation, since then by definition the base region has an excess of minority carriers, which tend to promote conduction even when the external drive is removed.
The solution is to design a semiconductor diode with no P-type semiconductor region, and therefore no holes as current carriers. Such a diode, known as a Schottky Barrier Diode, places a rectifying metal contact on one side on an N-type semiconductor block. For example, an aluminum contact will act as the P-type connection, without requiring a significant P-type semiconductor region.
This diode construction has two advantages in certain types of circuits. First, they can operate at very high frequencies, because they can turn off as fast as they can turn on. Second, they have a very low forward voltage drop. This is used to advantage in a number of ways, including as an addition to TTL ICs. When a Schottky diode is placed across the collector-base junction of a transistor as shown to the right, it prevents the transistor from becoming saturated, by bypassing the excess base current around the transistor. Therefore, the transistor can turn off faster, thus increasing the switching speed of the IC. The full power versions of these TTL ICs are the 74S00 series, and have switching speeds similar to ECL, and similar power requirements. The low power versions, the 74LS00 series, have switching times comparable to standard TTL, but with a much lower power requirement.
Experimentation is always in progress, and new applications are invented regularly. As new diode types come to my attention, I will add them to the list above. If you should hear of a diode type not yet in the list, please contact and let me know there. I will research the device and add it as quickly as possible. Thanks.