ABSTRACT
Schottky diodes and FET gates share common processing challenges. Schottky contacts will be discussed first because the basic discussion is applicable to both. FET gate processing will be discussed in more detail later. 12.1 Schottky DiodesSchottky diodes perform functions in silicon that cannot be achieved by p-n junction devices. The devices have a low turn-on voltage and are fast, being majority carrier (electrons for n-type Si) devices with a short recovery time. In the case of GaAs and other III-V compounds, Schottky diodes and field-effect transistor (FET) gates fulfill the needs for functions that otherwise could not be accomplished. Diffusion processes are not needed and p-n junctions are avoided and lack of a good gate oxide technology is circumvented. The process associated with Schottky diodes and gates is very simple and amenable to submicron gate formation, thus making high-frequency operation easier to achieve. The basic physics of metal-semiconductor junctions was discussed briefly in Chapter 2. A few results of importance are repeated here. The I-V curve of a metal-semiconductor junction showing rectifying or diode behavior, with current density in the forward direction, has the exponential form given by
J ≈ exp qV nkT
Ï Ì Ó
¸ ˝ ˛
(12.1) The reverse current and the breakdown behavior are discussed later. 12.1.1 Depletion WidthIn an n-type semiconductor, when a voltage is applied to a Schottky contact, electrons are depleted out of the depletion region and the applied voltage controls the depletion width (the depth under the metal). The negative charge of the electrons is balanced by the positive charge of the ionized dopant atoms. The depletion width can be derived by Poisson’s equation:d V
dx
2 = – re( )x s
(12.2)where es is related to the dielectric constant k and permittivity of free space by es = ke0. If the charge density is a constant Nd V dx
2 = qNes (12.3) Integrating twice and using proper boundary conditions V x qNw x xw( )= - ÊËÁ ˆ¯˜ÈÎÍÍ ˘˚˙˙22
(12.4)
V(w) = V = qNw2 2e
(12.5)or w2 = 2esV qN
(12.6) Therefore depletion width w = √{2 . (esV)/qN}. The applied voltage V is related to the gate voltage Vg by V = Vbi – Vg, where Vbi is the built-in voltage with no external voltage applied (Vbi values for some common metals on GaAs are listed in Table 2.1). The width of the depletion region goes down with increasing doping and is very narrow for doping levels over 1E18/cm3.
