Transient Enhanced Diffusion for Ultra Low Energy Boron, Phosphorus, and Arsenic Implantation in Silicon

  • PDF / 411,050 Bytes
  • 6 Pages / 420.48 x 639 pts Page_size
  • 18 Downloads / 291 Views

DOWNLOAD

REPORT


compared to the equivalent BF 2 implant. However, its physical mechanisms are still not clear and may be attributed to two possible factors: (1) the presence of high hydrogen content accompanying the boron doping in silicon, and (2) formation of more stable defects from big cluster bombardment, which hamper or delay the release of interstitials from them. In this work, TED has been studied in terms of implant energy, species, and preconditioning implant. TED is observed for all three implant species, B, P, and As, with diffusion enhancements of 2.4-5.5. Boron diffusion enhancement decreases from 4.3 to 2.4 as implant energy drops from 2 to 0.25 keV. The study also shows that preamorphization by Ge implant to a depth of 250 A can completely eliminate the channeling tail of a 2 keV B implant. Although higher boron diffusion enhancement results from the Ge implant, a shallower profile is still preserved as compared with the one without Ge implant. In an attempt to understand the reduced TED from decaborane implant, a hydrogen implant is applied to samples prior to boron implant to simulate a hydrogen doped environment. The results show negligible difference between the samples with and without hydrogen pre-doping. EXPERIMENT Ultra low energy boron, phosphorus, and arsenic ions were implanted into n- or p-type bare silicon samples at 0.25-2 keV to doses of (2-10)x1014 at./cm 2. All implants were performed with a combination of 7-degree tilt and a 22-degree rotation with respect to the (110) plane to reduce the channeling effects. For 0.25 keV boron implant, the incident beam was decelerated from either 2 or 4 keV in order to maximize the beam current and thus enhance the implant throughput. Some samples also received preamorphization implant of 20 keV Ge to 2x104 at./cm 2 or pre-doping implant of 0.5 or 3 keV hydrogen to >7x 10'4 at./cm 2 prior to 2 keV B implant. Following B, P, and As implants, some implanted wafers were subsequently annealed through rapid thermal anneal (RTA) at 1000-1050 'C for 10-15 sec in a nitrogen ambient. Nominal ramp rates of 35-50 'C/sec were used. Sheet resistance of annealed samples was measured with four point probe. Implant dopant profiles in as-implanted and annealed samples were determined using secondary ion mass spectroscopy (SIMS) at Charles Evans. Oxygen-leak techniques were used to detect low energy boron profiles with enhanced depth resolutions. RESULTS AND DISCUSSION Figure l(a) shows two SIMS profiles of boron implanted at 250 eV to a dose of 2x1014 at./cm2, obtained as boron ions were decelerated from 2 and 4 keV, respectively. These two profiles are almost identical from the surface peak to a concentration level of 3x 10 18 at./cm 3, but they begin to deviate from each other in the low concentration tail region. Two different depths of 250 and 400 A are obtained from the profiles at the lxlO17 at./cm 3 concentration level. The longer tail for deceleration from 4 keV is due to beam energy contamination from 4 keV B neutrals, a component that cannot be decelerated to 0.25 keV

Data Loading...