Fluorine incorporation in preamorphized silicon

Abstract

We studied the effect of implanted fluorine on B-doped silicon formed by Si preamorphization, solid phase epitaxy (SPE) regrowth and post-SPE thermal treatments. We showed that the fluorine is an efficient diffusion inhibitor for boron, revealing the crucial importance of F implementation in the future generation devices. In samples doped with B we observed an anomalous F accumulation at the dopant implantation peak. Since the physical mechanisms driving these phenomena are not yet well understood, we investigated the effect of the presence of B and/or As on the F incorporation during the SPE process at 580 degrees C. By using As coimplantation (thus modifying the SPE rate) we demonstrated that the above mentioned increased F incorporation is due to a kinetic effect, related to the SPE rate modification by doping, while a F-B chemical bonding is refused. These data shade new light upon the mechanism responsible for B diffusion reduction by F

Implant Damage Studies with Different Implant Temperature by Spot and Ribbon Beam

“…Another interesting finding is the 2nd boron peak which was also reported other reports [1,11]. This suggests that this 2nd boron peak should not be from F-B chemical interaction [5]. This 2nd boron could still be observed in the case at the lowest implant temperature tested where no 2nd fluorine peak was observed.…”

“…With BSG at -10°C, the 2nd fluorine peak from ribbon beam implant was about the same height as that of from spot beam (Figure 1-a) but the Rs from ribbon beam is significant higher (474.89 Ohm/sq vs. 450.74 Ohm/sq). This Rs difference by spot beam and ribbon beam at low implant temperature could be related to the formation of fluorine-vacancy cluster which created an undersaturation of the interstitial concentration in the vicinity of boron profile, [6,7] judging from the shift in fluorine profile. However, the main fluorine profile shifted deeper in the depth near the a/c interface and the concentration surpassed that of the 2nd fluorine peak from ribbon beam at same depth (Figure 1-a) while Rs was significantly lower than that from ribbon beam (463.67 vs. 443.02 Ohm/sq).…”

“…Lowering the 2nd fluorine peak by lowering the implant temperature might increase the availability of silicon self-interstitial to deactivate boron in Rp region [10] and resulted in higher Rs. These fluorine-vacancy clusters created an under-saturation of the interstitial concentration in the vicinity of boron profile [5,6,7] and as driver to increase the re-activation when implant temperature further decreased to -40°C. We propose a hypothesis that the amount of interstitials at EOR reduced due to thicker amorphous layer formed during implant surpassed the lowering in blocking effect as the 2nd fluorine peak was lowered.…”