> and finally one curious question. In all the paper researchers are
> specifying that mostly p-type silicon is used as the primary sensing
> element. Can you please tell me is it due to Farbication process or
> signal processing conditions.
>
> Thank You
> Allwyn Boustheen

This is not documented well in the literature. Here are some reasons for
using p-type piezoresistors:

P-type piezoresistors have maximum response along the <110> crystal
directions, which are parallel and perpendicular to the flat of a (100)
wafer.
This is convenient for the layout.

The piezoresistive response when a <110> p-type piezoresistor is stretched
along its length and its width is nearly equal but opposite.
This is useful in a Wheatstone bridge, in which pairs of opposite resistors
more either up or down with stress.
A Wheatstone bridge is used because it is impossible to make a single
resistor with exactly a given value,
plus it greatly reduces the effect of temperature on the output.

For certain doing levels, p-type piezoresistors also can have equal but
opposite rate of change of sensitivity and resistivity with temperature
(the resistance rises and the piezoresistance falls at the same rate).
This is important for "temperature compensation," which keeps the response
of a sensor constant with temperature.

If p-type piezoresistors are used, they are implanted and diffused into
n-type silicon.
If the n-type silicon is an epitaxial layer 15-30 um thick, on a p-type
wafer, the wafer can be electrochemically etched from the back side in KOH,
stopping on the n-type layer.
This is how most piezoresistive pressure sensors are made.

See the book "An Introduction to Microelectromechanical Systems, 2nd Ed.,"
by Maluf and Williams, Artech House, 2004, which covers much of this.

    --Kirt Williams