IMPATT-Diode
The IMPATT diode is a high-frequency semiconductor device of microelectronics, which is part of a diode to the electronic components. The name is derived from the English term Impact Ionization Avalanche Transit Time diode in German it is called avalanche transit-time diode ( LLD). The main representatives of the IMPATT diode family, the Read diode, the one-sided abrupt pn junction, the double-sided double drift diode, hi -lo and lo -hi -lo- diode and PIN diode. Other runtime diodes are the BARITT diode DOVETT diode and the TRAPATT diode.
The IMPATT diode is one of the most efficient semiconductor sources of microwave generation up to 100 GHz, and thus one of the most powerful sources of microwave energy in the solid state physics. It is used in microelectronic circuits in which high-frequency energy sources are needed. These are among other applications in telecommunications, for example, for stations in the millimeter-wave communication such as radar for civilian air and ground use or control of missiles in the military and similar applications. Benefits of IMPATT diodes are that they are cheap to produce, have a low power consumption, reliable, can deliver high performance and are lightweight. The disadvantage is the high noise, the sensitivity of the element under working conditions and high reactances. The reactances are strongly dependent on the amplitude of oscillation and therefore must be considered in the circuit design with very great care, so as not to upset or even to burn through the diode.
- 4.1 Small Signal Analysis
- 6.1 forms
Principle of operation
IMPATT diodes are caused by impact ionization and the use of transit -time characteristics of the electrons in the semiconductor structure for the manufacture of a dynamic effective negative differential resistance at microwave frequencies. Often this resistance is also referred to by the English term negative dynamic resistance or the appropriate abbreviation NDR. A negative resistance is a source of energy in the form of a current or voltage source in general. The negative resistance comes about by means of two delay times, which in turn cause a delay of the high frequency current relative to the high-frequency voltage. Is referred to in this context, a negative phase shift between the high-frequency current and the high-frequency voltage. The delay time is determined by the caused by the avalanche breakdown " avalanche time ," the second shift arises from propagation delays through the diode in the n pip structure or p nin structure in the drift region about. If the two delay times together form a half-period, is produced a negative electrical resistance at the corresponding frequency.
History
The first experimental observation of an IMPATT oscillation by Johnston, deLoach and Cohen took place in 1965. These were operated in a blocking polarization in the avalanche breakdown region and the microwave range silicon diode.
The physicist William B. Shockley summed up this negative resistance because of its structural simplicity as early as 1954 in the eye. The great advantage is that it is, in contrast to the transistors, which have three connections to be a device with two terminals. In 1958, Read brought a proposal to develop a high-frequency semiconductor diode, which should consist of a relatively high resistance of an avalanche zone at one end and a drift region. The drift zone serves as a transit time range for the generated charge carriers. CA Lee, RL Batdorf, W. Wiegman and G. Kaminsky were the first documented this oscillation. Misawa, guilds, and Hines developed this, the small-signal theory which underpins that a negative resistance with IMPATT properties of diode barrier layers or semiconductor-metal contacts is to obtain, regardless of the doping profile.
Stationary case
Be considered in the following text, the field distribution, the breakdown voltage, and space charge effects in stationary conditions. In the first image above, the doping profile and the field distribution of an idealized Read diode can be seen. The Ionisationsintegrand is given by:
? p? n herein and the respective ionization rates of electrons and holes. And W is the depletion region width. Between them is the drift zone.
The avalanche condition is given by:
Due to the strong dependence of the alphas from the electric field one can find that the avalanche zone is limited greatly locally. That is, the multiplication process is carried out in a very narrow region near the highest electric field intensity between 0 and Xa.
Special cases of the read diode of the single- abrupt p -n junction, and the PIN diode also known as Misawa diode. A further embodiment, the double-sided abrupt P -N - pn junction. The avalanche zone is located at silicon in the vicinity of the depletion zone center. For? N? P and are very different in silicon. In GaP the alphas are almost the same and you can make the following simplification
Because of the avalanche zone is almost symmetrical to the depletion zone center.
Breakdown voltage
The general methods for the determination of the breakdown voltage in abrupt transitions can also be for symmetrical double-sided abrupt transition as applied for example p - pn -n junctions. It is calculated as
In this equation, the maximum field being at the point x = 0 Where x is shifted depending on the material of the symmetry plane due to the different alphas. The maximum field at x = 0, when the doping is known to be read from a chart. Then I can be the breakdown voltage with the help of the above equation to calculate. The reverse voltage at breakthrough is UB Ue. In this equation u is the diffusion potential ( built-in voltage of the pn junction ), which is given by 2 ⋅ ( K ⋅ T / q) ⋅ ln ( NB / ni). Symmetrical for abrupt transitions, the diffusion voltage in practice is negligible.
For the read diode and the Hi-Lo - diode, the breakdown voltage and depletion layer width is given by the following equations
The epitaxial layer is crucial in the Read diode for the width of the depletion zone. For the calculation, the same tables can be re- used as the maximum field of the abrupt pn junction. Provided that the n-or p- region is given, these values meet almost completely ( with a deviation of about one percent ), as well at the read diode and the hi -lo diode. However, under the proviso that xA is smaller than b.
The Durchbruchspanung a lo -hi -lo- diode with a very Smallen Q "nugget " is given by
In this equation Q is the number of impurities per square centimeter in the " clumping ". The maximum field can be calculated from the field-dependent Ionisationskoeffizienten.
Regions
The avalanche region of an ideal pin diode extends across the entire intrinsic layer. However, this region is limited to the read diode and the pn-junctions in a very narrow region near the metallurgical junction. A good approximation for Xa can be obtained by means of the avalanche breakdown condition. With increasing distance of x from the metallurgical junction of the contribution to the integral decreases, so that it can be assumed at 95 % of a meaningful contribution. For the Read diode is calculated from xA
Analogously, for the hi -lo diode to one side and the double-sided abrupt junction transitions (for the case that they are operated in the breakdown voltage ), the following equation is used
Effects
In operating conditions, the high current densities during avalanche break, which cause a significant temperature increase at the transition and space charge carrier effects have to be considered.
The ionization rates of electrons and holes decrease with increasing temperature. Consequently, the breakdown voltage with increasing temperatures. When the DC power is increasing ( the product of the reverse voltage and the reverse current) to increase both the temperature at the junction and the breakdown voltage. To avoid that it comes through extreme heat rise to strongly confined places to total failure of the device, IMPATT diodes must have a suitable heat dissipation.
Space charge effects due to the carrier generation, charge carriers that cause fluctuations in the electric field in the depletion region. This results in the change of the differential DC resistance. For abrupt transitions increases this, and for pin diodes he takes off.
Dynamic case
In the following the injection phase and the transit time of an ideal component is considered. That is to say that the current at the site is injected with a phase of φ and that move the injected charge carriers with a saturation velocity of the drift region. The amount of AC line current density at the point is equal to the total exchange current density with phase shift.
The entire alternating current in the drift region is composed of the sum of the conduction current and displacement current. Herein is the complex alternating field.
The complex alternating field is obtained from the above two equations.
By integration we obtain the impedance Z.
In this equation, C is the capacitance per area. And is the transit angle and is calculated as follows. For the real and the imaginary part we obtain the following expressions from the above equation
Small-signal analysis
The first considerations for small-signal analysis were prepared by William T. Read. M. guilds and MF Hines developed these theories further. The following simplifications were made here: and saturation velocities of electrons and holes are equal.
According to literature, the current density is the avalanche breakdown and the total exchange current density. The current density in avalanche breakdown is the exchange current density of the particles in the avalanche region. Assuming a thin avalanche region, there are no delays of the current density up to the entry into the drift region. To the second assumption on which the current density in avalanche breakdown propagating as continuous wave ( in which only the phase changes ), the drift rate calculated to
Large-signal analysis
The picture shows a Read- diode in the large-signal work area.
Production
For the manufacture of IMPATT diodes, there are generally three techniques: epitaxy, diffusion and ion implantation. In this case, combinations of these three techniques are available. For all three steps, it is necessary that the doping is performed with high accuracy.
Epitaxy generally means an ordered crystal growth. Here, a new layer is transferred ( epitaxial growth ), which assumes the given atom ordering on an existing substrate. Frequently, especially the molecular beam epitaxy ( MBE) is used. In MBE, the thickness of the doping and the layer can be determined in almost atomic scales. It is therefore used particularly for diodes in the millimeter range and even smaller. It is not only the most accurate but also the most expensive method. A further possibility consists in the diffusion. Since the diffusion of solid substances with one another is very lengthy, one tries to speed up this process using different methods. Energy must be added in general, this can be done using laser, ultraviolet radiation, or by vapor deposition. In ion implantation, impurities are ionized and accelerated electromagnetic penetration into the substrate.
The diode produced is finally fixed in a microwave housing. Here is the diffused side or the metal electrode in contact with a copper or diamond heat conductor, so the heat generated during the pn junction can be diverted quickly.
Molding
There many different forms have been developed. A relatively simple structure, the p -n- n . It is produced either by double epitaxial process or by diffusion to an epitaxial layer. This is the n substrate reduce the series resistance. In order to reduce losses and to preserve uniformity that may occur due to the skin effect, the substrate is only a few microns in size. The thickness of the epitaxial layer must also be controlled: When breakthrough is no epitaxial layer remain which makes the component unusable.
Another method is the Schottky diode, with a semiconductor-metal contact which is rectifying. Thus, the structure is n -n- metal. The structure is similar to the first structure, but has several advantages: Thus, the maximum field occurs at the metal- semiconductor interface, the generated heat can be rapidly conducted away from the metal contact. The component may also have the shape of a truncated cone. Here, when the maximum field is shifted from outside to inside, the breakdown takes place inside the device. Because the diode can be produced at relatively low temperatures, it is possible to preserve the original high-quality epitaxial layer. A disadvantage is however, that the metal electrode, electrons and holes can be attacked with a high energy, so there is no long-term durability.
The hi -lo diode commonly used has the structure N in metal. It is a modified read diode in which the p layer was replaced by a metal contact - there is thus at the same time, a Schottky diode. Since it is mainly a majority carrier device, the minority carrier storage effect, is prevented. This gives the diode has a higher efficiency, especially those of microwaves. Here, a strict control of doping profiles is necessary so that a specific frequency can be set. Using a self-limiting anodic Ätzungsmethode the highly doped layer can be thin or the surface can be made low doped. As a result, the breakdown occurs at a desired voltage and a desired frequency is produced at the same time. Most Schottky barriers have a large lock and vice versa a low saturation current. The disadvantage is that gallium arsenide react at operating temperature with platinum, whereby the pn junction is shifted. This changes the breakdown voltage, the power decreases. Applying a small amount of platinum ( from 200 to 500 angstroms ) on the epitalen surface, followed by a tungsten or tantalum layer, reduces the reaction.
A fourth possibility is to build according to the scheme n -np -p , a double drift diode. It is produced by ion implantation and is useful for diodes, which are used for millimeter-wave generation. Both the output power and the impedance per unit area by approximately doubled, so this structure leads to higher efficiency.