In many cases it is convenient to choose the pseudopotential to be a constant within the ion core. The parameters of the pseudopotential can be determined from the spectroscopic data for the individual atom. Results of the empirical pseudopotential energy band calculations for some elemental and compound semiconductors with diamond and zinc blende structures are shown in Figure 4.11.(1) Figure 4.12 shows the various symmetry points displayed at the zone center (Γ) and along the (100) axis (X) and (111) axis (L) inside the first Brillouin zone of a diamond lattice. The first symmetry point, Γ, is the symmetry point located at the Brillouin zone center. The conduction band minimum and the valence band maximum that located at the Γ- point in the zone center are designated as Ec and Ev, respectively. It is noted that the conduction band is defined as the lowest empty band, while the valence band is defined as the highest filled band at T = 0 K. In most semiconductors, there exists a forbidden gap between the conduction- and valence- bands, and the values of the energy band gap may vary from 0.1 to about 6.2 eV for the semiconductors. If the conduction band minimum and the valence band maximum are located at the same k- value in the first Brillouin zone, such as the Γ- point at the zone center, then the semiconductor is called the direct band gap semiconductor. Most of the III-V compound semiconductors, such GaN, GaAs, InP, InAs, and InSb, belong to this category. Direct band gap semiconductors have been widely used in photonic device applications such as laser diodes, LEDs, and photodetectors because their band structures allow for direct optical transitions. They are also widely used in high-
speed and high- frequency device applications due to the small electron effective mass and high electron mobility in these materials. If the conduction band minimum and the valence band maximum are not located at the same k- value in the first Brillouin zone, then the semiconductor is referred to as an indirect band gap semiconductor. Elemental semiconductors such as silicon and germanium belong to this category. Table 4.1 lists the energy band gaps and the effective masses of electrons and holes for the elemental and compound semiconductors.
The conduction band of a diamond or a zinc blende crystal usually consists of several subbands or satellite bands. For example, the conduction band minimum of a germanium crystal is located at the zone boundaries along the {111} axes, while for silicon it is located near the zone boundaries along the {100} axes; these are shown in Figure 4.11b and a, respectively. It is noted that the constant energy surfaces for electrons in silicon and germanium are ellipsoidal energy surfaces, while the constant energy surface near the conduction band minimum is spherical for GaAs and other III-V compound semiconductors. Figure 4.13 shows a more detailed energy band structure of GaAs calculated from the pseudopotential method.(2) The Γ- conduction band minimum is located at the zone center, the L-conduction band valleys are located at (2π/a) (1/2,1/2,1/2) along the (111) axes, and the X-conduction band valleys are located at the zone boundaries along the (100) axes. The separation between the L-valley and the Γ-band minimum is equal to 0.29 eV. The valence band maxima of the heavy- and light-hole bands are located at the Γ- point in the Brillouin zone center. Therefore, both silicon and germanium are indirect band gap semiconductors, while GaN, GaAs, InP, and InAs are direct bandgap semiconductors. For silicon, the conduction band minima consist of six ellipsoids of constant-energy surfaces along the {100} axes with the center of each ellipsoidal energy surface located about three-fourth of the distance from the zone center to the zone boundary. For germanium, the conduction band minima consist of eight ellipsoidal constant-energy surfaces along the {111} axes with the center of each ellipsoid located at the zone boundary. Thus, for germanium there are eight half- ellipsoidal conduction band valleys inside the first Brillouin zone. For GaAs, the constant-energy surface of the Γ- conduction band minimum is spherical, and is located at the zone center. The energy versus wave vector (i.e., E vs. k) relation for electrons near the bottom of the conduction band can be expressed by
()22*2cnkEkEm=+h (4.103)
for the spherical constant energy surface, and
()2222ltcltkkEkEmm⎛⎞=++⎜⎜⎝⎠h (4.104)
for the ellipsoidal constant energy surface, where ml and mt denote the longitudinal and transverse effective masses of electrons in the conduction band, respectively.
The valence bands of silicon, germanium, and GaAs crystals consist of the heavy- and light-hole bands which are degenerate at k = 0. In addition, a spin-orbit split off band is located at a few tens of meV below the top of the heavy- and light- hole bands. This can be best described by using the band structure shown in Figure 4.13 for a GaAs crystal. In this figure, it is shown that the heavy- and light-hole bands are degenerate at the top of the valence band and may be represented by a parabolic band with different curvatures. The valence band with a smaller curvature (i.e., with a larger hole effective mass) is usually referred to as the heavy-hole band, and the valence band with a larger curvature (i.e., with a smaller hole effective mass) is known as the light-hole band. The effective masses of the light- and heavy-hole bands for Si, Ge, and GaAs are also given in Table 4.1. In general, the energy versus wave vector relation (E vs. k) for the heavy- and light- hole bands near the top of the valence bands is nonparabolic and can be expressed by
()()22v*2pkskEkEm=−h (4.105)
Where s(k) is given by
()(1/222224224224xyxzyzskABCkkkkkkkkk⎡=±+++⎣ (4.106)
Note that A, B, and C in Eq.(4.106) are constants (see Problem 4.10); the plus and minus signs correspond to the heavy-hole and light-hole bands, respectively. It should be noted that the constant-energy surfaces near the top of the valence bands are warped and nonparabolic for Si, Ge, GaAs and other III-V compound semiconductors.
Another interesting and technologically important feature for the III-V semiconductors is their ability to grow the lattice-matched ternary or quaternary compound semiconductor epitaxial layers on either the GaAs or InP semi-insulating substrates (e.g., InxGa1-xP, AlxGal-xAs and InxGal–xAsyPl–y on GaAs; InxGal–xAs and InxAll–xAs on InP substrates). Using these ternary and quaternary compound semiconductors, it is possible to change many important optical, physical, and electrical properties of the III-V compound semiconductors, such as the band gap energy and electron mobility for a wide variety of applications. In addition, many novel device structures can be
fabricated using the binary/ternary superlattice and quantum well heterojunction structures (e.g., AlxGal-xAs/GaAs, InGaAs/AlGaAs). These features are extremely important for many applications in detectors, lasers, and high-speed devices using III-V compound semiconductor epitaxial layers grown by the MOCVD and MBE techniques. Figure 4.14 shows the energy band gap versus lattice constant for Si, Ge, II-VI and III-V binary compound semiconductors.(3) The solid lines denote the direct band gap materials and the dashed lines are for the indirect band gap materials. A mixture of AlP/GaP to form AlxGal–xP, AlAs/GaAs to form AlxGal-xAs, AlSb/GaSb to form AlxGal-xSb ternary compounds, and InP/GaAs/InAs to form InxGal-xAsyPl-y quaternary compound along the vertical line of Figure 4.14 yields lattice- matched epitaxial- layers grown on the GaP, GaAs, InP, and GaSb substrates, respectively. By tailoring the energy band gap of these III-V alloy systems, it is possible to produce detectors and lasers with wavelengths covering the visible to infrared spectral range. Wide band gap semiconductors such as AlN, SiC, and GaN have been widely investigated and developed in recent years, enabling the fabrication of various electronic devices for microwave, high temperature, and high power applications. Furthermore, GaN-based ternary compounds such as AlxGa1-xN and InxGa1-xN with the energy band gaps varying from 0.7 eV to 6.2 eV have been developed for UV detectors, laser diodes and LEDs. Figure 4.15(a) and (b) show the energy band gap versus alloy composition x for GaAsxP1-x and AlxGa1-xAs ternary compound semiconductors, which illustrate the band gap variation from Eg= 1.42 eV to 2.65 eV and 2.19 eV, respectively, as x varied from 1 to 0. The variation of band gap with alloy composition in an III-V ternary material system can be estimated by using an empirical formula given by:
(4.107) 2()(0)ggExEbxcx=++
Where b is a fitting parameter, and c is called the bowing parameter which could be calculated theoretically or determined experimentally. For the AlxGa1-xAs material system, the energy band gap for the Γ-, X-, and L- valleys as a function of alloying composition x can be expressed as
2()1.4251.2471.147(0.45)gExxxΓ=++− (4.108a)
2()1.900.1250.143XgExxx=++ (4.108b) (4.108c)