Ge-based substrates are being developed for applications in advanced nano-electronic devices because of their higher intrinsic carrier mobility than Si. The stability and diffusion mechanism of impurity atoms in Ge are not well known in contrast to those of Si. Systematic studies of the stable sites of 2nd to 6th row element impurity atoms in Ge crystal were undertaken with density functional theory (DFT) and compared with those in Si crystal. It was found that most of the impurity atoms in Ge were stable at substitutional sites, while transition metals in Si were stable at interstitial sites and the other impurity atoms in Si were stable at substitutional sites. Furthermore, DFT calculations were carried out to clarify the mechanism responsible for the diffusion of impurity atoms in Ge crystals. The diffusion mechanism for 3d transition metals in Ge was found to be an interstitial-substitutional diffusion mechanism, while in Si this was an interstitial diffusion mechanism. The diffusion barriers in the proposed diffusion mechanisms in Ge and Si were quantitatively verified by comparing them to the experimental values in the literature.

No abstract

The interface states between Si substrate and the gate oxide (Si/SiO 2 ) have significantly influences to device properties. The conventional measurement methods for interface states need a long turn around time evaluation period since test element group structure is necessary to be constructed. The in-line measurement in the wafer process has been expected strongly to achieve the Quick TAT analysis to increase the throughput of the process. We have studied to present a new measurement method for observing the interface states under the nondestructive condition. The specimens of the oxide, SiO 2 , on the Si wafers are prepared by the thermal oxidation of representative wafers, which surface-crystal planes have various miller indices of (100), (110), and (111). An electrode is set close to the SiO 2 surface, where the voltage and pulsed light are applied and irradiated at the same time. The photo-carriers generated at the Si surface absorption are excited to the Si upper bands and surmount the SiO 2 potential barrier. The carrier excitation could be selected by varying the wavelength of pulsed light to focus on the trapped and localized carriers' excitations from the Si forbidden band gap, excluding the Si interband transition. For the preliminary result, we demonstrate to detect the signals from the trapped carriers' excitations and their energy distribution, which is compared with the results on the distribution of the interface states density as reported in other research works. The measurement result at this time indicates the possibility toward the noncontact detection for Si/SiO 2 interface states. Figure 7. The energy distribution of the interface state density. The interface state density is normalized by 1 × 10 12 . A median and an error bar in each wafer were plotted in the graph. The error bar is variability 1σ of 6 measurement points in each wafer.Nondestructive interface state measurement by PPCM

We have developed a noncontact measurement method that enables in-line measurement and does not have any test element group (TEG) formation. In this method, the number of photocarriers excited from the interface states are counted which is called "photocarrier counting", and then the energy distribution of the interface states density (D it ) is evaluated by spectral light excitation. In our previous experiment, the method used was a preliminary contact measurement method at the oxide on top of the Si wafer. We developed, at this time, a D it measurement method as a noncontact measurement with a gap between the probes and the wafer. The shallow trench isolation (STI) sidewall has more localized interface states than the region under the gate electrode. We demonstrate the noncontact measurement of trapped carriers from interface states using wafers of three different crystal plane orientations. The demonstration will pave the way for evaluating STI sidewall interface states in future studies.

The lowest energetic configurations of metal impurities in 4throw (Sc - Zn), 5throw (Y - Cd) and 6throw (Hf - Hg) elements in Ge crystals were determined with density functional theory calculations. It was found that the substitutional site is the lowest energetic configuration for most of the calculated metals in Ge. The most stable configurations of dopant (Ga, Sb) - metal complexes in Ge crystals were also investigated. Following results were obtained. (1) For Ga dopant, 1st neighbor T-site is the most stable for metals in group 3 to 7 elements while substitutional site next to Ga atom is the most stable for metals in group 8 to 12 elements. (2) For Sb dopant, substitutional site next to Sb atom is the most stable for all calculated metals. Binding energies of the interstitial metalMiwith the substitutional dopantDswere obtained by the calculated total energies. The calculated results for Ge were compared with those for Si.