A Critical Regime for Amorphization of Ion Implanted Silicon
- PDF / 1,133,309 Bytes
- 6 Pages / 414.72 x 648 pts Page_size
- 6 Downloads / 204 Views
417 Mat. Res. Soc. Symp. Proc. Vol. 321. @1994 Materials Research Society
The present work examines damage accumulation in silicon and the effects that ion flux, fluence and species have upon it, in the regime where dynamic defect annealing plays a significant role. The effects of pre-existing damage on the amorphization process will be used to illustrate that amorphization is nucleation limited in this regime. EXPERIMENTAL PROCEDURE Implantations involving rare gas ions were undertaken using a 200 keV Whickham ion implanter, located at the Royal Melbourne Institute of Technology, while all other implantations were performed on the 1.7 MeV NEC 5SDH tandem accelerator located at the Australian National University. In all cases the bombarded wafer was thermally coupled to a copper block with conducting paint and clips. The block was either heated or cooled to maintain the silicon samples at the required temperature which ranged from 00 to 340' C. A constant ion flux was maintained throughout each irradiation. Rutherford backscattering channeling (RBS-C) and cross-section transmission electron microscopy (TEM) were used to analyze the resultant damage. The number of displaced atoms (Nd) was calculated using the computer program Nd. This program [6], based on work by Ziegler [7], fits a dechanneling component from RBS-C spectra before extracting Nd over a given range. The depth of material used to determine Nd was constant for each ion type but varied between ion species, due to the difference in their implantation ranges. TRIM [8] was used to calculate the energy deposition and implanted ion profiles. RESULTS AND DISCUSSION Figure 1 displays Nd resultant from an 80 keV Si bombardment at a series of temperatures 2 and ion fluences. The data shown is for irradiations ranging from 2 to 10 x 1015 ions/cm 2 which were all delivered at a rate of 2.7 X 1013 ions/cm /s. The data for each fluence have been fitted with the same function as obtained for the 1 x 1016 ions/cm 2 series. Examining the data for each ion fluence it becomes clear that three distinct regions exist. These are as indicated on the 1 x 1016 ions/cm 2 data set. In region I the wafer is totally amorphized over the irradiated region (confirmed by TEM) and no temperature dependence is observed in the residual damage structure. Hence Nd values in this region appear similar for all implant doses. This region reflects a dominance of defect production and amorphization over annealing. At higher temperatures another region (III) exists, for a specific fluence, where temperature also has little effect on the final value of Nd. In contrast to region I, dynamic defect annealing is dominant and the sample remains totally crystalline, although, TEM results (figure 2a) have revealed a band of interstitially-based extended-defects. Despite dynamic defect annealing dominating in this regime the fact that Nd is non-zero implies the annealing is not complete. In region II (the critical regime) the extreme sensitivity of Nd to even slight changes in temperature suggests that
Data Loading...
