The 77th JSAP Autumn Meeting, 2016

Presentation information

Oral presentation

6 Thin Films and Surfaces » 6.2 Carbon-based thin films

[14p-A26-1~16] 6.2 Carbon-based thin films

Wed. Sep 14, 2016 1:15 PM - 5:45 PM A26 (203-204)

Makoto Kasu(Saga Univ.), Toshimichi Ito(Osaka Univ.), Hitoshi Umezawa(AIST)

1:45 PM - 2:00 PM

[14p-A26-3] Study of TMB/CH4/O2/H2 plasmas for diamond doped multilayers

Alexandre FIORI1, Tokuyuki TERAJI1 (1.NIMS)

Keywords:diamond, boron doping, plasma

Sharp transitions from heavily doped (p++, [B] > 5×1020 cm−3) to lightly doped (p, [B] < 1017 cm−3) diamond films are essential to build power transistors (delta-doping) [1], or in optical systems (superlattice) [2]. Important challenges concern the accuracy and reproducibility of doping levels and deposition rates. Therefore, understanding the mechanism supporting the incorporation of boron in diamond is mandatory. In this study, we examined the effects of methane and oxygen flows on the boron incorporation, and we adjusted plasma mixtures in order to obtain sharp boron transitions in diamond multilayers.
Optical emission spectroscopy (OES) has been employed to diagnostic CVD plasmas composed of trimethylboron (TMB), methane, and oxygen, diluted into hydrogen. The comparison of Hα, Hβ, Hγ, OH*, BH*, CH*, and C2* emission lines provided a semi-quantitative characterization of the plasma composition in real time. Thus, trends between BH*/CH* peak ratio and boron concentration, and between C2/Hβ peak ratios and carbon concentration have been drawn. In addition, we identified a specific BH* super-intensity correlated with diamond in-situ etching.
We performed continuous depositions of diamond multilayers in four steps, as following, #1 deposition of a p++ layer, #2 rinsing mixture, #3 deposition of a p layer, and #4 rinsing mixture. In the case of rinsing #2 and #4, different OES features have been observed, even if the feed gas composition was the same (0.5% CH4 + 0.1% O2 to the total flow). The OES highlights the effect of the residual boron that cause in-situ etching of diamond during rinsing #2, after p++ deposition, while no etching happened in rinsing #4, after p deposition. However, we suppose that etching is responsible of larger surface roughness and worse crystallinity, which are detrimental for high quality and nano-scale multilayers. Therefore, we developed specific plasma mixtures to avoid etching during rinsing, at doping transitions, by adjusting precisely methane and oxygen flows, supported by the real time OES technique.

References:
[1] A. Fiori et al, Appl. Phys. Express 6 (2013), 045801
[2] A. Fiori et al, Appl. Phys. Lett. 105 (2014), 081109