Physical Characteristics of Very Low Temperature Anodic Oxides of Polycrystalline Si Films
- PDF / 459,836 Bytes
- 6 Pages / 390.24 x 621.9 pts Page_size
- 69 Downloads / 160 Views
steps. First, we performed an ozone oxidation of a p-doped 8" c-Si wafer to obtain 10 A of SiO 2. This layer enabled us to next deposit poly-Si to a thickness of 600 nm, from silane at 635'C, in-situ doped with boron at 1016 atm.cm-3. To activate the dopants we performed an anneal at 1050'C for 30 seconds and during this step, we observed that the oxygen of the thin (-10/A ) SiO 2 layer dissolved in the poly-Si layer. In this way, we obtained good electrical contact between the poly layer and the mono c-Si substrate. Aluminium was deposited on the backside of the Si substrate to improve the ohmic contact between the sample and the substrate holder in the plasma reactor. Finally, we cut the 8" wafer into 4 by 4 cm squares. For plasma oxidations, we used an ECR plasma system. It is important to note that our reactor is not in a clean room environment. The base pressure of our plasma reactor (volume 37 1) was about 10-7 Tort. During oxidation the typical microwave power used to create the oxygen plasma was 500 W at 2.45 GHz, and the oxygen pressure was fixed at 1 mTorr. During the process, the substrate was not heated, so that any increase of the substrate temperature was due only to the plasma. This temperature evolution was monitored by ellipsometry [ 1] and was always lower than 100°C. During oxidation we biased the substrate holder positively with respect to the reactor body using a continuous (dc) constant current source. When the oxide thickness changed the bias applied was increased to maintain a constant current (0.4 A). The thickness of the oxide was measured using an in-situ spectroscopic ellipsometer (SOPRA) working in the photon energy range 1.5 - 4.5 eV. During oxidation, to estimate the oxide thickness rapidly, we performed ellipsometric measurements over a smaller spectral range (3.6 - 3.7 eV), as was done for plasma oxidation of mono c-Si [ 1]. In this range, the optical indices of silicon (mono and poly crystalline) vary little with temperature and the only variable is the oxide thickness. It is important to note that in the absence of precise information, we assumed that the refractive indices of the plasma growing oxides were equal to those of the a-Si0 2. The oxides were too thin to measure by ellipsometry both the real part of the optical index and the thickness independently. For comparative purposes, we also used deposited oxide layers (plasma enhanced CVD using silane and oxygen) and thermal oxide (dry atmosphere at 900'C), on poly and mono c-Si. These were made in a clean room atmosphere. Infrared (IR) absorption spectroscopy measurements, with a 4 cml resolution, were carried out ex-situ, using a Brucker IFS 66 Fourier Transform spectrometer operating at room temperature. In order to check the macroscopic oxide homogeneity, Auger electron spectroscopy was performed using Phi model 670 spectrometer. An Ar beam was used to erode the oxide, Si and 0 atomic concentration profiles in the oxide were determined with a minimum depth resolution of 20 A. The surface roughness of the different oxides wa
Data Loading...
