H. C. Gatos

Electron mobility and free‐carrier absorption in InP; determination of the compensation ratio

Theoretical and experimental studies of the electron mobility and the free‐carrier absorption of n‐type InP were carried out in the temperature range 77–300 °K. All major scattering processes and screening effects were taken into consideration. It was found that the experimental dependence of electron mobility and free‐carrier absorption on temperature and/or on carrier concentration can be consistently explained only when the effect of compensation is quantitatively taken into account. Convenient procedures are presented for the determination of the compensation ratio from the values of electron mobility and from the free‐carrier absorption coefficient. The high contribution of optical‐phonon scattering in InP limits the applicability of the free‐carrier absorption approach to electron concentration n≳1017 cm−3. Electron mobility, however, can be reliably employed for the determination of the compensation ratio for n≳1017 cm−3 at 300 °K and n≳1015 cm−3 at 77 °K.

Origin of the 0.82‐eV electron trap in GaAs and its annihilation by shallow donors

The concentration of the major electron trap (0.82 eV below the conduction band) in GaAs (Bridgman grown) was found to increase with increasing As pressure during growth. It was further found that (for a given As pressure) the concentration of this trap decreased with increasing concentration of shallow donor dopants (Si, Se, and Te). Donor concentrations above a threshold of about 1017 cm−3 led to the rapid elimination of the trap. On the basis of these findings, the 0.82‐eV trap was attributed to the antisite defect AsGa formed during the postgrowth cooling of the crystals.

Crystallographic Polarity in the II‐VI Compounds

Crystallographic polarity associated with the zinc‐blende and the wurtzite structures has been studied in the following II‐VI sulfides, selenides and tellurides; ZnS, ZnTe, CdS, CdSe, CdTe, HgSe, and HgTe. The CdS and CdSe have the wurtzite structure and the others the zinc‐blende structure. Consistent with theoretical calculations, the geometric structure factor for x‐ray scattering was found different in opposite directions along the polar axis. Consequently, the two types of surfaces perpendicular to the polar axis were identified. The etching behavior of these surfaces was correlated with the x‐ray results so that simple etching tests were developed for the identification of the {111} (zinc blende) or the {00.1} (wurtzite) surfaces.

Application of scanning electron microscopy to determination of surface recombination velocity: GaAs

A method is reported for the determination of the surface recombination velocity by scanning electron microscopy; this method is based on an established relationship between the effective diffusion length of the minority carriers, the penetration depth of the electron beam, and the surface recombination velocity. Values of surface recombination velocity, up to about 2.5×106 cm/sec were determined in n‐type GaAs with a bulk minority‐carrier lifetime of the order of 10−8–10−10 sec; in GaAs, with carrier concentrations exceeding 1018 cm−3, recombination velocity of about 3×106 cm/sec represents a saturation value.

Quantitative study of the charge transfer in chemisorption; oxygen chemisorption on ZnO

A new experimental approach is presented for studying the charge‐transfer process involved in the chemisorption on polar semiconductors. This approach utilizes the surface piezoelectric effect, contact potential difference measurements, and surface photovoltage spectroscopy. From the study of oxygen adsorption on ZnO it was found that the rate of electron transfer varies exponentially with the surface barrier height and is proportional to the oxygen pressure (from 10−3 to 20 Torr). Furthermore, it was found that the charge transfer is characterized by a thermal activation energy of about 0.72 eV. At room temperature this activation energy constitutes the most significant rate‐limiting factor and is largely responsible for the extremely slow rate of chemisorption. A model for chemisorption was developed in which the thermal activation is treated as an intermediate nonelectronic step involving metastable activated surface states. Upon capturing electrons from the bulk these states become stable surface stat...

Identification of oxygen‐related midgap level in GaAs

An oxygen‐related deep level ELO was identified in GaAs employing Bridgman‐grown crystals with controlled oxygen doping. The activation energy (825±5 meV) of ELO is almost the same as that of the dominant midgap level: EL2 (815±2 meV). This fact impedes the identification of ELO by standard deep level transient spectroscopy. However, we found that the electron capture cross section of ELO is about four times greater than that of EL2. This characteristic served as the basis for the separation and quantitative investigation of ELO employing detailed capacitance transient measurements in conjunction with reference measurements on crystals grown without oxygen doping and containing only EL2.

Native hole trap in bulk GaAs and its association with the double‐charge state of the arsenic antisite defect

We have identified a dominant hole trap in p‐type bulk GaAs employing deep level transient and photocapacitance spectroscopies. The trap is present at a concentration up to about 4×1016 cm−3, and it has two charge states with energies 0.54±0.02 and 0.77±0.02 eV above the top of the valence band (at 77 K). From the upper level the trap can be photoexcited to a persistent metastable state just as the dominant midgap level, EL2. Impurity analysis and the photoionization characteristics rule out association of the trap with impurities Fe, Cu, or Mn. Taking into consideration theoretical results, it appears most likely that the two charge states of the trap are the single and double donor levels of the arsenic antisite AsGa defect.

Electronic characteristics of “real” CdS surfaces

Abstract Photovoltage spectroscopy (including photovoltage inversion and photovoltage quenching) was employed to determine the electronic characteristics of “real” (basal and prismatic) surfaces of CdS. In room atmosphere surface states with the following positions were found in the cadmium surfaces: Ec-Et≅0.05 eV, 0.4 eV, 0.8 eV, and Ev-Et≅0.83 eV; the same surface states were present in the sulfur surfaces, with the exception of those at Ec-Et≅0.4eV. In the prismatic and unetched basal surfaces, states at Ec-Et≅l.l eV were found in addition to all of those found on the cadmium surfaces. The surface states at Ec-Et≅0.8 eV were eliminated in high vacuum from all surfaces studied; they are apparently associated with adsorbed ambient molecules (most likely oxygen). After illumination with white light in ultrahigh vacuum (10−11 torr) only the states atEv-Et≅ ≅0.83 eV were detected in all surfaces; the states at Ec-Et≅0.05 eV were present in the sulfur surfaces and prismatic surfaces and the states at Ec-Et = 1.1 eV in the prismatic surfaces.

Electron mobility in n‐type GaAs at 77 K: Determination of the compensation ratio

Electron‐mobility values in n‐type GaAs at 77 K were computed utilizing a variational procedure and taking into account all major scattering mechanisms. Mobility values are tabulated as a function of electron concentrations, providing a convenient means for the determination of the compensation ratio, i.e., the determination of the total concentration of ionized impurities.

Passivation of the dominant deep level (EL2) in GaAs by hydrogen

We showed that hydrogen incorporated into single crystals of GaAs (by exposure of the crystals to hydrogen plasma) renders the major deep donor level (EL2) located at 0.82 eV below the conduction band at room temperature electronically inert. We attribute this passivation process to the interaction of hydrogen with the unsaturated bonds of the antisite AsGa defect (believed to be responsible for EL2) leading to the formation of stable As–H bonds.