Finally the results were compared with Project 27 (50% Hf). All
the results between two groups are summarized below:
Project 26 (70% Hf)
|
Project 27 (50% Hf)
|
|
Type of Substrate
|
P-type
|
P-type
|
Oxide Relative
Permittivity
|
10.64
|
8.67
|
EOT (nm)
|
2.81
|
3.04
|
Doping Density (m-3)
|
4.244x1021
|
2.7x1021
|
Work Function
Difference (eV)
|
0.065
|
0.08
|
Flatband Voltage
(V)
|
0.932
|
0.53
|
Midgap Voltage (V)
|
1.367
|
1.21
|
Oxide Charge
Density (cm-3)
|
||
Flatband Condition
|
-6.660x1016
|
-3.23x1016
|
Mdigap Condition
|
-7.841x1016
|
-6.4x1016
|
It can be noticed that a higher percentage of Hf will lead
to a higher oxide relative permittivity. It could be explained by the fact that
the relative permittivity of Hf is higher than SiO2. Also, the equivalent oxide
thickness will be decreased. Nevertheless the EOT is still too large for
commercial purpose (which should be less than 2 nm). Higher percentage of Hf
will also lead to a higher doping density, which will then increase the cost of
manufacture.
Another thing should be considered is that the fixed oxide
charge density almost doubled in flatband condition. According to our supervisor,
the charge density should be controlled less than 1x1010 in order to
get a better quality of CMOS. Thus further research should be carried out to
decrease the amount of fixed oxide charge density.
没有评论:
发表评论