Impurity levels GaAs

In the growing process of single crystals of GaAs, the kind of impurities that dissolve easily into the crystal, and the impurity levels in the band structure for dissolved elements have to be made clear. 

A scheme showing the impurity levels of various elements in GaAs crystals is presented in Fig. 3.30. The impurity levels below the centre of the forbidden band (Fermi level of the semiconductor) act as acceptors and those above the Fermi level act as donors, except O and Se. Ge and Si function as amphoimpurities. 

Fig. 3.30 Scheme of impurity levels of various kinds of element for GaAs crystals at room temperature. Impurity levels above and below the centre of the forbidden band act as donors and acceptors, respectively. 

One consequence of the high impurity levels is the use of high dopant concentrations to control the behaviour of oxides. The dopant level is seldom below 1 in 103 moles and may be as high as 1 in 10 moles so that defects may interact with one another to a far greater extent than in the covalent semiconductors silicon, GaAs etc. 

Recent measurements [18] have extended this work to consider the manipulation of the impurity levels with a quantum well potential. GaAs/AlAs multiple quantum wells doped with Be were grown and the structural tunability of the impurity levels was demonstrated by an increase in the ls-2p absorption energy. 

Figure 3 Spin-resolved total Mn-densities of states and local Mn-densities of states on the Ga-sublattice (a) a single Mn-impurity in GaAs, and (b) (Gao.95Mno.o5)As alloy. The Fermi level coincides with the energy zero.

Calculated DOSs for the GaAs with 5% of Cr impurities without and with As-antisites are presented in Fig. 13 and they should be compared with corresponding results for GaMnAs alloy, Figs. (3) and (5). The differences can be summarized as follows (i) the Fermi energy is located within the impurity subband with a strong admixture of Cr-states. The Cr-impurity level is about 0.5 eV above the top of the valence band as... 

We depart briefly from our discussion of SI GaAs to consider an example that better illustrates some of the features of temperature-dependent Hall measurements. This example (Look et al., 1982a) involves bulk GaAs samples that have sc — F — 0-15 eV. We suppose, initially, that the impurity or defect controlling the Fermi level is a donor. Then any acceptors or donors above this energy (by a few kT more) are unoccupied and any below are occupied. Also, p n for kT eG. From Eq. (B34), Appendix B, we get... 

A nonelectronic method of measuring impurity concentrations is that of absorption spectroscopy. From Eq. (36a) it is seen that ani = avnini0, where a i is the absorption constant due to electronic transitions from level i to the conduction band. The total impurity concentration Nt can be related to ni0 by a knowledge of EF. The photon-capture cross section doping experiments or by independently measuring Nt in some sample. This process has been carried out for Cr impurity (Martin, 1979) as well as (EL2) (Martin, 1981) in GaAs. The same considerations hold for photoconductivity measurements, except that t also needs to be known, as seen from Eq. (35). 

Shown in Figure 2 are the electron donor and acceptor concentrations as a function of baking time at 800 °C for the growth of GaAs by LPE (57). The level of unintentional impurity dramatically decreases with increasing baking time. The residual acceptor and donor were attributed to C and Si, respectively. 

In order to improve the accuracy of the calculated acceptor levels in silicon and germanium, particularly for the even-parity ones, Lipari et al. [38] have used a screened point-charge impurity potential based on the wave-vector-dependent dielectric function calculated for Si, Ge, GaAs and ZnSe [65]. They make use of a phenomenological parameter a, adjusted to fit the calculated q-dependent dielectric function e(q), in this potential. The resulting potential in real space is ... 

---