The Electrical Behavior of Implanted Bismuth in Silicon

Abstract

The electrical behavior of 20–50-keV Bi implanted layers in silicon has been evaluated using Hall effect and sheet resistivity measurements. Implants of greater than 1014/cm2 performed at room temperature show a large increase in the number of carriers/cm2 for short anneals at ≈600°C that is associated with the reordering of the lattice. Although Rutherford scattering measurements (performed at Chalk River Nuclear Laboratories) indicate that ≈80% of the Bi atoms are substitutional, only ≈20% are effective electrically. The peak value of the carrier density distribution is ten times the corresponding maximum solid solubility.

Quadrupole shift of nuclear magnetic resonance of donors in silicon at low magnetic field

“…The three samples were individually annealed after the ion implantation, at 650 C for 30 min in an evacuated quartz tube. This process, designed to maximize the number of donor-readout center (D-R) pairs [47], led to an activation efficiency [48][49][50][51] 60%, resulting in a Bi donor concentration less than 1.1 10 18 cm −3 , and a 121 Sb and a Si: 123 Sb concentration less than 4×10 16 cm −3 (in all cases below the metal-insulator transitions [52,53]).…”

Hyperfine clock transitions of bismuth donors in silicon detected by spin-dependent recombination

“…These conditions yielded a maximum bismuth concentration of 1.8 × 10 18 cm −3 (above the solubility limit [25]) in the depth of 90 to 150 nm from the surface. The post-implantation annealing, performed at 650 • C for 30 min in an evacuated quartz tube, led to an activation efficiency [7,[26][27][28] below 60%, resulting in the Bi donor concentration less than 1.1 × 10 18 cm −3 (below the metal-insulator transition [29]). This process was designed to maximize the number of D-R pairs, instead of fully activating all the implanted Bi atoms [30].…”

The use of aluminium ion implantation for power semiconductor devices

“…The main problem associated with aluminium in silicon is its low electrical activity after the annealing (see, e.g., Ref. [3]). Sadana et al [4] and, in a very recent paper, Chang et al [2] explain this by the fact that the concentration of A1 exceeds its solubility limit at the anneal temperature and precipitates form (there is no agreement with the nature of the precipitates: Sadana et al find aluminium precipitates whereas Chang et al claim the precipitates to be aluminium oxide particles).…”