Doping of AlGaN Alloys

Nitride-based device structures for electronic and optoelectronic applications usually incorporate layers of AlxGal-xN, and n- and p-type doping of these alloys is typically required. Experimental results indicate that doping efficiencies in AlxGal-xN are lower than in GaN. We address the cause of these doping difficulties, based on results from first-principles density-functional-pseudopotential calculations. For n-type doping we will discuss doping with oxygen, the most common unintentional donor, and with silicon. For oxygen, a DX transition occurs which converts the shallow donor into a negatively charged deep level. We present experimental evidence that oxygen is a DX center in AlxGal-xN for x>∼0.3. For p-type doping, we find that compensation by nitrogen vacancies becomes increasingly important as the Al content is increased. We also find that the ionization energy of the Mg acceptor increases with alloy composition x. To address the limitations on p-type doping we have performed a comprehensive investigation of alternative acceptor impurities; none of the candidates exhibits characteristics that surpass those of Mg in all respects.

In this work we have applied the admittance spectroscopy technique to characterize the DX centers in AlxGa1−xAs alloys doped with silicon. Our experimental results reveal the existence of two DX centers related to silicon in AlxGa1−xAs alloys, named DX-I and DX-II centers, with thermal activation energies of 0.370 and 0.415 eV, respectively. These values are lower than those obtained by other authors using capacitance techniques. To explain this disagreement it should be noticed that capacitance techniques can be affected by the nonexponential behavior of the thermal emission transients of the DX centers in AlxGa1−xAs alloys.

We find the DX centers in Si-doped AlAs for the first time. The activation energy is measured as 0.56 eV from deep level transient spectroscopy (DLTS). The DX centers in n-AlAs exhibit a large capture energy 0.5 eV and a persistent photoconductivity. These properties are similar to those of the DX centers in AlxGa1−xAs with x∼0.3. However, the carrier concentration in the DX centers revealed by DLTS is not linearly proportional to Si donor concentration. This result is interpreted by the band structure that the DX center level lies at 30 meV above the X-conduction band (CB) minima and at 150 meV below the L-CB minima. The DX center is found not to be associated with the X-CB minima, but the L-CB minima.

Evidence for two Si-related DX like centers in Al xGa 1− xAs and GaAs

We have utilized the isothermal capacitance transient spectroscopy and modulated function waveform analysis to investigate the deep levels and DX centers in molecular beam epitaxially grown AlxGa1−xAs/GaAs Schottky diodes deposited on n+ GaAs, with Al composition ranging from x=0 to 0.5, and implanted with Si. In the concentration range x=0.19 to 0.50, two major electron trap levels (E1 and E2) were detected, which gradually changed with composition. For example, E1 changed from 0.393 to 0.339 eV and E2 changed from 0.136 to 0.287 eV. However, in pure GaAs, three major trap levels with energy E1 (hole)=0.226 eV, E2 (electron)=0.496 eV, and E3 (EL2)=0.74 eV were observed. Apparently, these levels are governed by the deep levels known as DX centers.

The field effect has been investigated on the deep levels and DX centers in molecular-beam epitaxilly-grown AlxGa1−xAs on n+ GaAs for several compositions. We have measured the isothermal and isofield capacitance transients for the major energy levels with the applied potential in the range from −0.1 to −3 V. In samples with x=0.5, two electron trap levels (E1 and E2) were detected. While E1 was strongly field dependent (changed from 0.499 to 0.287 eV) E2 was practically unchanged (0.305 eV) with respect to energy, cross section, and peak broadening. Contrary to this in pure GaAs (samples x=0), we observed three energy levels, one hole trap (E1=0.239 eV), one electron trap (E2=0.512 eV), and the usual EL2 trap (E3=0.72 eV). At this composition, however, levels E1 and E2 were absent at low reverse fields (at −1 V) and EL2 remained unchanged at 0.72 eV. We have considered the recent theories of Pool–Frenkel effect and tunneling for electron–phonon interaction to interpret the results. In particular, for AlxGa1−xAs with x=0.5, the electron–phonon coupling factor (S) was found to be about 4 and the deep centers were identified as neutral to repulsive.

Field effect studies have been performed on the deep levels in molecular beam epitaxially grown AlxGa1−xAs (with x=0.385) and InxGa1−xAs (with x=0.1) on n+-GaAs. We have measured the isothermal and isofield capacitance transients for the major energy levels with the applied electric field [here, field means normalized field, F(norm)=F(appl)×10−5] in the range −0.5 to −5 V/cm. The results of our investigation suggest that the applied field has a distinct effect on thermal emission rate, capture cross section and activation energy. In particular, the Arrhenius plots showed a nonlinear behavior that could be associated with electron–phonon interactions. Furthermore, the activation energy versus field plots indicate a linear behavior, illustrating a complex nature of the deep levels some of which, E1 in AlxGa1−xAs and E3 in InxGa1−xAs, can be characterized as DX centers.

en

The recombination (electron capture) kinetics of the ionized DX center in AlxGa1−xAs have been measured as a function of temperature and silicon doping concentration. It is shown that for x≂0.35, the silicon concentration dependence of the recombination kinetics is dominated by effects of the electron distribution in the conduction band, and is insensitive to changes in the trap characteristics. In a model kinetic calculation consistent with the data the trap is found to capture through a level 0.202 eV from the bottom of the conduction band with a width of 0.045 eV, independent of DX center concentration.

Optical ionization of DX center in AlGaAs : Se by inner-shell excitation

Intense radiation with photon energy of a few meV can induce the capture of electrons by DX centers in AlxGa1−xAs:Si.

A deep-level transient spectroscopy (DLTS) technique is reported for determining the capture cross-section activation energy directly. Conventionally, the capture activation energy is obtained from the temperature dependence of the capture cross section. Capture cross-section measurement is often very doubtful due to many intrinsic errors and is more critical for nonexponential capture kinetics. The essence of this technique is to use an emission pulse to allow the defects to emit electrons and the transient signal from capture process due to a large capture barrier was analyzed, in contrast with the emission signal in conventional DLTS. This technique has been applied for determining the capture barrier for silicon-related DX centers in AlxGa1−xAs for different AlAs mole fractions.

Localised vibrational modes (LVMs) associated with DX centers in AlxGa1−xAs and n‐type AlxGa1−xAs are worked out from a molecular model, introducing the relaxation. Preliminary studies indicate that some of the basic features observed from deep level transient spectroscopy (DLTS) can be viewed from the present calculations.

N- type AlxGa1−xN exhibits a dramatic decrease in the free-carrier concentration for x⩾0.40. Based on first-principles calculations, we propose that two effects are responsible for this behavior: (i) in the case of doping with oxygen (the most common unintentional donor), a DX transition occurs, which converts the shallow donor into a deep level; and (ii) compensation by the cation vacancy (VGa or VAl), a triple acceptor, increases with alloy composition x. For p-type doping, the calculations indicate that the doping efficiency decreases due to compensation by the nitrogen vacancy. In addition, an increase in the acceptor ionization energy is found with increasing x.

Deep-level admittance spectroscopy (DLAS) of DX centers in AlxGa1−xAs:Sn (0.2<x<0.6) reveals the presence of three levels SN1, SN2, and SN3 related to the Sn donor. While SN1 and SN3 are observed in all the samples, SN2 is prominently seen only in the indirect band-gap samples. The conventional capacitance deep-level transient spectroscopy (DLTS) is found to be unsuitable for the study of the DX center in AlxGa1−xAs:Sn with x>0.35 because of the strong freeze-out of free carriers in these samples. Even in the case of low AlAs mole fraction samples (x<0.35), the DLTS technique fails to reveal all the levels observed by DLAS and provides information only on the SN3 level.