Young's modulus and residual stress of pnc-Si deposited with different substrate biases
Last week, I reported on the the morphology of pnc-Si deposited with different substrate biases. We hypothesized that the reason for this was due to the different initial film stress of the as-deposited a-Si. Unfortunately, the pressure-displacement method does not allow us to measure amorphous films since it requires the membrane to be nearly flat at zero pressure (under zero/tensile stress). I went ahead and measured the crystallized films to find whether or not there was a difference in the Young’s modulus or residual stress in the pnc-Si state. Here is a table summarizing the results for a film deposited with 00, 25 and 50W substrate bias.
| wafer | bias (W) | E (GPa) | σ (MPa) | Porosity (%) | Eadj (GPa) | σadj (MPa) |
| 662 | 00 | 91 | 64 | 4.1 | 95 | 67 |
| 664 | 25 | 100 | 42 | 5.8 | 106 | 45 |
| 665 | 50 | 116 | -21 | ~0 | 116 | -21 |
The Eadj and σadj are the adjusted values based on the membrane porosity of Young’s modulus and residual stress, respectively. Harrop et al. and Kovacs et al. suggested that bulk E and σ values could be calculated from their perforated counterpart by a (1-porosity) factor.
From these results, I have concluded that the crystallized films have very similar material properties and whatever difference in the initial stress state was eliminated in the phase transition.
To interrogate the amorphous film stress state we will need to employ a wafer bow measurement using a profilometer (the same method that JP used in his senior design project).