Calculations and measurements of contact resistance of semi-transparent Ni/Pd contacts to p-GaN

Calculations of specific contact resistance as a function of doping and barrier height were performed for p-type GaN. These calculations took into account two valence bands, each with different effective masses, and show that at low doping, the heavy hole band accounts for most of the conduction, whereas at heavier doping, the light hole band dominates conduction. These calculations also indicate the barrier height for typical contacts to p-GaN is between 0.75 eV and 1 eV. Specific contact resistance measurements were made for oxidized Ni/Au, Pd, and oxidized Ni/Pd ohmic contact metal schemes to p-GaN. The Ni/Pd contact had the lowest specific contact resistance, 6×10⁻⁴ Ω cm². Auger sputter depth profile analysis showed some Ni diffused away from the GaN surface to the contact surface with the bulk of the Pd located in between two areas of Ni. Both Ni and Pd interdiffused with the GaN at the semiconductor surface. The majority of the oxygen observed was with the Ni as NiO. Angle-resolved-x-ray photoelectron spectroscopy (AR-XPS) analyses showed the formation of predominantly NiO and PdO species, with higher Ni and Pd oxides at the contact surface.     

Influence of arsenic on the atomic structure of the Si(112) surface

The surface science techniques of low-energy electron diffraction (LEED), x-ray photoelectron spectroscopy (XPS), and scanning tunneling microscopy (STM) have been used to characterize the clean Si(112) surface and the influence of an As monolayer on the properties and structure of the surface. In agreement with previous studies, the clean surface is found by both LEEd and atomically resolved STM images to be unstable with respect to faceting into other stable planes. Procedures for in-situ deposition of As onto clean Si surfaces were devised and XPS results show that approximately one monolayer of As can be deposited free of any contamination. The As/Si(112) surface is characterized by a sharper LEED pattern than for the clean surface and by STM images characterized by long rows along the direction with a regular width of 1.9 nm. This is consistent with a doubling of the periodicity in the direction of the bulk-terminated unit cell. This implies that As yields a stable but reconstructed Si(112) surface.     

Reaction chemistry and resulting surface structure of HgCdTe etched in CH4/H2 and H2 ECR plasmas

We report on several new aspects of etching of Hg_1−xCd_xTe (x = 0.22), HgTe, and CdTe in CH₄/H₂/Ar plasmas generated by an electron cyclotron resonance plasma source. Using a residual gas analyzer, we have identified elemental Hg, TeH₂, Te(CH₃)₂, and Cd(CH₃)₂ as the primary reaction products escaping from a HgCdTe surface during the plasma exposure. We have also demonstrated that a bias is not needed to etch HgCdTe at moderate temperatures (30-40°C), as previously suggested by other researchers. We have also developed a technique that avoids the formation of hydrocarbon polymer films on a HgCdTe sample during etching. Moreover, we have examined by x-ray photoelectron spectroscopy analysis and ellipsometry the surface condition of HgCdTe resulting from etching with this technique at zero bias. After exposure to the CH₄/H₂Ar plasma (or to a H₂/Ar plasma only), the HgCdTe samples exhibited a depletion of the HgTe component in the near surface region (increase of the x-value). The depletion covered a range from virtually x = 1 after H₂/Ar (10:2 in sccm) etching to values 0.4 < x < 0.5 after CH₄/H₂Ar (7:7:2 in seem) etching. Exposures to the plasmas were found to result in surface roughening of HgCdTe, however, plasmas rich in H₂ were observed to cause significantly rougher surfaces than plasmas with small H₂/CH₄ ratios. This difference in the resulting surface condition is attributed solely to chemical effects since the respective ion energies are considered to be below the damage threshold for HgCdTe in both cases. We also investigated the etching of HgTe and CdTe single crystals. The etch rate of HgTe was found to be over one order of magnitude higher than that of CdTe under similar conditions. This large difference in etch rates is assumed to be responsible for the observed preferential etching of the HgTe component indicated by the HgTe depletion of the HgCdTe surface region.