Understanding of Carbon/Fluorine Co-implant Effect on Boron-Doped Junction Formed during Soak Annealing

Abstract

The formation of highly activated ultrashallow junctions is one of the main challenges for the forthcoming generation of complementary metal oxide semiconductor (CMOS) devices. Co-implantation of impurities such as carbon (C) or fluorine (F) is an attractive technique. However, junction optimization can only be achieved with a complete understanding of the underlying physical mechanisms. In this paper, the effect of C∕F co-implant on boron (B)-doped preamorphized silicon during the soak annealing is extensively studied. C∕F atoms are located in the middle range between the B∕BF2 concentration profiles and the end-of-range (EOR) defect band, with the aim of reducing the interactions of dopants with the interstitials released from EOR region. Isochronal annealing study is performed to investigate the impact of C∕F codoping on the dopant de/reactivation behavior. It is shown that transient enhanced diffusion can be reduced by both co-implant schemes. The B-doped junction formed with the C co-implant is relatively stable and dopant deactivation is inhibited, while it is presumed that F atoms form B–F complexes, which reduces the B activation level. A physical insight on the dopant-defect interactions associated with C∕F co-implant is established through the combination of diffusion and activation studies during soak annealing.

Benefits of cryo-implantation for 28 nm NMOS advanced junction formation

“…To maintain abrupt junction profiles, Ge PAI and carbon co-implants were applied before the BF 2 halo implant. The incorporated carbon helps to suppress boron TED and enhances dopant activation as carbon forms interstitial clusters and reduces the formation of extended defects [4,5].…”

Revisiting the role of strain in solid-phase epitaxial regrowth of ion-implanted silicon

“…17) Furthermore, there may be complicated interplays between the implanted dopant impurities and C. For example, enhanced dopant activation upon C + co-implant was reported. 20) On the other hand, there have also been reports showing relaxation of Si:C due to loss of substitutional carbon (C sub ), due to competition with dopant activation. 21) It is therefore important to directly measure the above-mentioned interplays in order to correctly assess the roles played by C and the dopant in the strain-included GFLS model.…”