Articles you may be interested inPhotoluminescence and secondary ion mass spectrometry investigation of unintentional doping in epitaxial germanium thin films grown on III-V compound by metal-organic chemical vapor deposition

This paper reports on depth profiles of Mn implants, after annealing, obtained by secondary ion mass spectrometry. Two types of samples implanted with manganese were studied: unintentionally doped n-type (5–10×1015 e cm−3), and n-type S-doped (n=5×1018 e cm−3) substrates. In unintentionally doped substrates it is found that Mn exhibits a well-defined two-species diffusion front as other acceptors: Be, Zn, or Cd. On the contrary in S-doped substrates it does not move. The study of the correlations between the movement of the residual impurities and the implanted Mn atoms, or Zn in a comparative sample, has led us to propose a model based on an interstitial-substitutional reaction involving the impurity sites and taking place in the bulk of the semiconductor.

No abstract

Some (AlGa)Sb layers grown by molecular beam epitaxy were found unexpectedly to be n type. Characterizing these layers by secondary ion mass spectrometry it was observed that they contained selenium (Se). Systematic analyses showed that selenium was present in all our antimonide layers, even those which were p type, at concentrations between 6×1014 and 3×1017 cm−3. The thermodynamical study of this contaminant incorporation led us to conclude that it comes from the solid antimony used for growths. That was confirmed by spark source mass spectrometry investigations. During growths, selenium behaves as other elements of group VI, sulphur for example, and this behavior is described by a simple kinetic model. It incorporates more easily at low substrate temperatures and saturates at levels depending on the antimony (Sb4) flux. At higher substrate temperatures, the incorporation is balanced by desorption according to an activation energy of 3.2 eV. As a consequence, selenium constitutes an accurate thermal probe to follow substrate temperature variations during growths. As could be expected, sulphur was also found to contaminate (AlGa)Sb films at levels of the order of a few 1015 cm−3. However, oxygen was not detected, probably being lower than 1016 cm−3 the detection limit of the analytical technique. This residual doping of antimonides by chalcogens presumably compensates their natural p-type doping, but relationships with mobilities have not been evidenced yet.

InP device degradations have often been related to the presence in the active layer of gold atoms having migrated from the contacts. We have studied gold thermal diffusion in InP in the temperature range 400–700 °C, using secondary ion mass spectrometry (SIMS). We have pointed out SIMS artifacts and how to avoid them. We have found small values for the diffusion coefficient: 2×10−12 cm2/s at 550 °C. By deep-level transient spectroscopy measurements, gold appeared to behave as a shallow donor, with a level situated at 0.55 eV from the conduction band. Our conclusion is that gold thermal migration from the contact is not the mechanism responsible for the device degradation.