Measuring and testing – Speed – velocity – or acceleration – Acceleration determination utilizing inertial element
Reexamination Certificate
2000-09-07
2003-01-21
Kwok, Helen (Department: 2856)
Measuring and testing
Speed, velocity, or acceleration
Acceleration determination utilizing inertial element
C073S514380
Reexamination Certificate
active
06508124
ABSTRACT:
TECHNICAL FIELD
The present invention regards a microelectromechanical structure insensitive to mechanical stresses.
BACKGROUND OF THE INVENTION
As known, surface and epitaxial micromachining techniques allow production of microstructures within a layer that is deposited (for example a polycrystalline silicon film) or grown (for example an epitaxial layer) on sacrificial regions that are removed at the end of the manufacturing process by wet etching.
In general, the layers subject to manufacturing (deposited or grown layers) are formed at high temperatures, completely different from the operative temperatures. In addition, the various regions forming the end devices have different thermal expansion coefficients. Consequently, at the microstructure operative temperatures, residual mechanical stresses are present; in addition, in particular when the various regions are doped not uniformly, the stresses are not uniform (stress gradients); these stresses thus cause undesirable mechanical deformations of the microstructures, as described schematically hereinafter with reference to
FIGS. 1-6
.
In detail,
FIG. 1
shows in cross-section a structure
1
comprising a polycrystalline silicon bridge element
2
, formed on a monocrystalline silicon substrate
3
; a sacrificial oxice layer
4
extends between the bridge element
2
and the substrate
3
, except for two areas, where anchorage portions
5
of the bridge element
2
extend through the sacrificial oxide layer
4
, and are supported directly on the substrate
3
.
FIG. 2
shows the same structure
1
as in
FIG. 1
, in plan view.
FIGS. 3 and 4
show the structure
1
, after removal of the sacrificial oxide layer
4
, when the dimensions of the structure have been reduced (shown exaggerated in the figures, for better understanding), owing to the presence of residual stress; in particular, in
FIG. 3
, owing to the different thermal coefficient, the dimensions of the bridge element
2
are reduced (shortened) more than those of the substrate
3
; here the bridge element
2
is subjected to tensile stress, and assumes a more favorable energetic configuration. In
FIG. 4
on the other hand, the bridge element
2
undergoes a lesser reduction of dimensions than the substrate
3
; consequently, in this condition, the bridge element
2
tends to be lengthened in comparison with the substrate
3
, but, owing to the fixed anchorage portions
5
, it undergoes stress of a compressive type, causing buckling deformation.
In the case of tensile stress, the mechanical resonance frequency of bridge element
2
is shifted upwards with respect to the intrinsic value (in the absence of stress); on the other hand, in the case of compressive stress, the mechanical resonance frequency of the bridge element
2
is shifted downwards.
The average residual stress thus has the effect of modifying the resilient constant of the micromechanical structures; this modification is not reproducible, and can cause mechanical collapse of the structure (in particular in the case in FIG.
4
).
In
FIG. 5
, the projecting element
11
is formed on a monocrystalline silicon substrate
12
; a sacrificial oxide layer
13
extends between the projecting element
11
and the substrate
12
, except for an area, where an anchorage portion
14
of the projecting element
11
extends through the sacrificial oxide layer
13
, and is supported directly on the substrate
12
.
FIG. 6
shows the structure
10
of
FIG. 5
, after removal of the sacrificial oxide layer
13
. As can be seen, the release of the residual stress gradient causes the projecting element
11
to flex. In particular, indicating with &sgr;R(z) the function linking the residual stress with the coordinate z in the projecting element
11
, {overscore (&sgr;)}
R
the average residual stress &Ggr; the strain gradient, and E Young's modulus, the following is obtained:
&sgr;
R
(
Z
)={overscore (&sgr;)}
R
+&Ggr;Ez
In addition, indicating with L the length of the projecting element
11
, flexure at its free end is independent from the thickness, and is:
H=&Ggr;L
2
/2
Consequently, a positive strain gradient &Ggr; causes the projecting element
11
to bend away from the substrate
12
(upwards), whereas a negative gradient causes it to bend downwards.
In case of suspended masses, the behavior is exactly the opposite, i.e., positive stress gradients cause downward flexing, and negative stress gradients give rise to upward flexing.
In addition, the material of the package has a different coefficient of thermal expansion as compared to the material of the micromechanical structure (mono- or polycrystalline silicon). Consequently, the suspended masses may be subject to small displacements with respect to the fixed region of the micromechanical structure.
The presence of residual stress inherent to the structural material and stresses induced by the packaging material jeopardizes the performance of integrated micro-electromechanical devices.
For example, in the case of integrated micromechanical structures having a suspended mass, or seismic mass, provided with a plurality of anchorage points, the stresses inherently present in the materials or induced by packaging, by acting in different and non-uniform way on the various anchorage points, causes tension in some parts and compression in other parts, such as to modify the mutual positions of these parts and to generate non-symmetrical geometries of the structures.
For example, consider the case of an angular accelerometer with a suspended mass having an annular shape set outside the center of gravity of the suspended mass, so as to have a high moment of inertia and hence a high sensitivity. Such an accelerometer is illustrated schematically in FIG.
7
and in detail in
FIG. 8
, showing only one part thereof.
FIG. 7
shows a semiconductor material chip
20
housing an angular accelerometer
21
comprising a rotor
22
and a stator
23
. The chip
20
may moreover house circuit components (not shown) for biasing, controlling the processing signals. The angular accelerometer
21
has a barycentric axis G (defined as an axis passing through the center of gravity—not shown) coinciding with the axis of symmetry of the accelerometer. The rotor
22
(which is able to perform micrometric rotations about the barycentric axis G, in such a way that every movement of the rotor is defined by instantaneous vectors perpendicular to the barycentric axis G) comprises a suspended mass
25
having an annular shape concentric to the barycentric axis G and bearing a plurality of mobile electrodes
26
extending radially inwards from the suspended mass
25
. Each mobile electrode
26
is associated with two fixed electrodes
27
,
28
extending radially, each of which faces a different side of the respective mobile electrodes
26
. The fixed electrodes
27
,
28
, forming together the stator
23
, in practice define, together with the respective mobile electrodes
26
, a plurality of capacitive circuits; namely, all the fixed electrodes
27
, arranged on first sides (for example, on the left in the clockwise direction) of the respective mobile electrode
26
, form first capacitors with the respective mobile electrodes, whilst all the fixed electrodes
28
, arranged on second sides (for example, on the right in the clockwise direction) of the respective mobile electrodes
26
, form second capacitors with the respective mobile electrodes. The first capacitors are connected in parallel with each other and the second capacitors are also connected in parallel with each other. The first capacitor and the second capacitor associated with the same mobile electrode
26
are, instead, connected in series.
In a per se known manner, any movement of the suspended mass
25
brings about an increase in the capacitance of one of the two capacitors associated to each mobile electrode
26
and a reduction in the capacitance of the other capacitor. Consequently, by appropriately biasing the mobile electrodes
26
and the fixed electrodes
27
,
28
and by connecting them to a circuit tha
Sassolini Simone
Vigna Benedetto
Zerbini Sarah
Iannucci Robert
Jorgenson Lisa K.
Kwok Helen
Seed IP Law Group PLLC
STMicroelectronics S.r.l.
LandOfFree
Microelectromechanical structure insensitive to mechanical... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Microelectromechanical structure insensitive to mechanical..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Microelectromechanical structure insensitive to mechanical... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3030468