Modelling of Rapid Thermal and Furnace Oxide Growth
- PDF / 1,306,514 Bytes
- 6 Pages / 417.6 x 639 pts Page_size
- 102 Downloads / 204 Views
612
Two ellipsometers were used to measure the oxide thicknesses. A helium neon laser based system set at 700 was used for a rapid determination of film thickness and for uniformity tests, and a UV ellipsometer set at 75 was used for more accurate determination of very thin oxide layers. Excellent agreement between these two systems was obtained. Good agreement was also obtained between these techniques and cross sectional TEM for very thin oxides and the Nanometrics Nanospec/AFT Micro-area Gauge for thick oxides. For furnace oxidations the wafer was slowly pushed into a conventional diffusion furnace and left for fifteen minutes in a nitrogen atmosphere to reach the set temperature. The ambient was then changed to dry oxygen for the duration of the oxidation which ranged from a few minutes to several days. Finally the ambient was changed back to nitrogen and the wafer was slowly withdrawn. This anneal cycle is shown schematically in Fig. I. For halogen lamp oxidations in the Heatpulse 210-T it was not possible to switch gases rapidly enough to mimic the procedure used for the furnace oxidations. For this reason an oxygen ambient was used at all times. The anneal cycle used is shown in Fig. 2. An anneal at 4000C was carried out to aid reproducibility of the subsequent temperature peak. The ramp rate was in excess of 200 C/second and therefore very little oxidation occurred before the oxidation temoerature was reached. Temp. Temp.
St sacs
020 8 sas pro-heat
S15
mins pre-heat
t rains oxidation
Time
oidation
Tim.
Data Loading...
