Impact of Si precursor mixing on the low temperature growth kinetics of Si and SiGe

Abstract

I have quantified, in a 300 mm ASM Epsilon 3200 tool, the impact of adding SiH4 or SiH2Cl2 to Si2H6 on the low temperature growth kinetics of Si and SiGe. Increasing the amount of Si atoms by adding the formers to the latter resulted in Si growth rates (GR) reductions (i) at 20 Torr, in the 500 °C–575 °C range and (ii) at 525 °C, in the 20–90 Torr range. Those GR decreases were more pronounced with SiH2Cl2 than with SiH4. I have otherwise investigated the 525 °C, 20 Torr growth kinetics of SiGe with germane (GeH4) as the Ge source and various mixes of Si precursors. The GR associated with a SiH2Cl2 + GeH4 chemistry was lower than the Si2H6 + SiH2Cl2 + GeH4 GR, which was itself lower than the Si2H6 + GeH4 one. Meanwhile, the Ge content was much higher with SiH2Cl2 + GeH4 than with Si2H6 + SiH2Cl2 + GeH4. The later was itself ∼1% to ∼4% higher than the Si2H6 + GeH4 one depending on the GeH4 flow. As for pure Si, Si2H6 governed the SiGe growth kinetics, with SiH2Cl2 acting as if it was a source of HCl (GR reduction and slight Ge content increase). The situation was more complex when adding SiH4 to Si2H6 + GeH4. The Si2H6 + SiH4 + GeH4 GR was always lower than the Si2H6 + GeH4 one. Meanwhile, the SiH4 + GeH4 GR was either lower or higher than the other two depending on the GeH4 flow. The Ge content was significantly higher with SiH4 + GeH4 than with Si2H6 + SiH4 + GeH4 (27%–53% range ⇔ 9%–32% range). The latter was itself ∼2% lower than the Si2H6 + GeH4 one. As before, Si2H6 governed the SiGe growth kinetics, with SiH4 slightly reducing both the GR and the Ge content, which is unusual. Plotting the SiGe GR as functions of the Ge concentration showed that the addition of SiH4 or SiH2Cl2 molecules did not dramatically change the SiGe growth kinetics, which was predominantly governed at 525 °C, 20 Torr by Si2H6.