The Detailed Dependence of Implanted Phosphorus Profiles in (100) Single-Crystal Si on Key Implant Parameters
- PDF / 378,745 Bytes
- 6 Pages / 414.72 x 648 pts Page_size
- 5 Downloads / 187 Views
715 Mat. Res. Soc. Symp. Proc. Vol. 396 © 1996 Materials Research Society
over the surface of a wafer has been accurately characterized for various nominal tilt and rotation angles for the Eaton NV6200AV. We have used the variation of the implant angles over the wafer to advantage by choosing the nominal tilt and rotation angles for an implant (having a specific energy and dose) such that the practical range of the angle space, i.e. tilt angles of 0'-10o and rotation angles of 0' - 450 is covered. Due to the 4-fold rotational symmetry of silicon about the direction and the mirror symmetry about the direction, a range of rotation angle from 00 to 450 is sufficient to sample the crystal structure at all rotation angles (0' to 3600). The number of wafers required to be able to cover the useful region of the angle space for a particular dose/energy combination was reduced to six in this manner. Commercially available silicon wafers having nominal (100) orientations can have a small crystal cut error. Therma-Wave maps [2] were made on the two end wafers in each batch in order to measure and correct for the crystal cut error. The steps discussed above, in addition to the highest possible degree of control of the wafer alignment during the implant, are aimed at providing accurate experimental data for model development. Over 400 implanted samples were chosen for SIMS (Secondary Ion Mass Spectrometry) analysis. Samples were chosen for energies of 15, 30, 50, 80, 120, and 180 keV, doses of lx10 13, lxl014 , 5x1014 , 2x10 15 , and 8x10 15 cm- 2, and various combinations of tilt and rotation angles. The samples were sputtered using Cs+ primary ion bombardment at energies of 5keV for the low energy implants (80keV). Such a low SIMS measurement energy was used in order to avoid knock-on effects that can distort the measured profile. At higher implant energies, these knock-on effects have a negligible effect on the profile, and so the higher SIMS measurement energy can be used. The beam current varied from 0.075 1RA for the lowest energy implants to 0.7gtA for the highest energy implants. The rastered area varied from a 175tim square to a 400 gtm square, and the corresponding measurement area was 30m to 180am in diameter. The sputtering rate varied from 13 Angstrom/sec to 42 Angstrom/sec. High mass resolution conditions were used to distinguish 31p- from the interference of the molecular species 3°SiH- at 31 amu. The accuracy of the measured concentrations is believed to be better than 50%. The conversion of sputtering time to depth is based on the measurement of the analytical crater depth with a Tencor stylus profilometer and is accurate to within 10%. ANALYSIS OF PROFILES Observations of the profile dependencies will be described first; then the discussion and interpretation of these dependencies will be given. As expected, the profiles become deeper as the energy increases, as shown in Figure 1. An example of the dose dependence, reflecting the damage accumulation with dose, is shown in Figure 2. Clearly the channeling tai
Data Loading...
