Micromechanical component i.e. micro-mirror, has positioning element i.e. mirror plate, adjustable with respect to base substrate by bending spring mechanism based on different internal stresses of subunits

The invention relates to a micromechanical component. Furthermore, the invention relates to a production method for such a micromechanical component.

State of the art

Examples of an electrostatic drive with two comb electrodes are known from the prior art. By applying a voltage between the two comb electrodes, the rotatably arranged actuator comb electrode can be rotated relative to the fixedly mounted stator comb electrode about a predetermined axis of rotation. By turning the actuator comb electrode, a coupled actuator, such as a micromirror, can be adjusted. It is known that the torque exerted on the actuator comb electrode increases at a constant voltage as the maximum distance of the electrode fingers from the axis of rotation increases.

The <patcit><text>US 2005/0117235 A1</text></patcit>, the <patcit><text>US 2004/0263938 A1</text></patcit> and the <patcit><text>US Pat. No. 6,806,992 B2</text></patcit> describe micromirrors in which the actuator electrode fingers are at least partially attached to the outer sides of a frame with a mirror plate coupled thereto. The associated stator electrode fingers lie in a plane adjacent to the plane of the actuator electrode fingers. However, arranging the actuator electrode fingers and the stator electrode fingers above and / or below in an adjacent plane requires an adjustment process that is hardly satisfactory to realize.

Another problem of conventional micromirrors is the reflection of the incident light beam at the light incident window. In this case, a part of the light is often reflected by the light incident window on the image plane of the deflected by the mirror plate light beam. This often leads to annoying light points on the image plane.

The <patcit><text>EP 1 688 776 A1</text></patcit> therefore describes a housing for a micromirror with a tilted cover glass. However, the production of such a micromirror requires a relatively high effort and is therefore relatively expensive. In addition, the packaging of the micromirror housing with the inclined cover glass is relatively expensive.

The pamphlets <patcit><text>US 2005/0100269 A1</text></patcit> and <patcit><text>US 2005/0238282 A1</text></patcit> show a micromirror with a spring arrangement for tilting, wherein a rotation axis outside the mirror and is disposed away from it. The publication<patcit><text>DE 601 12 236 T2</text></patcit> shows a radiation detector with a thermal displacement element with a two-piece spring assembly for displacing a reflection plate, wherein also a rotation axis outside the reflection plate and is arranged away from it. The publication<patcit><text>US 2002/0036265 A1</text></patcit> also shows such an arrangement. The publication<patcit><text>US 2005/0231065 A1</text></patcit> shows a deflectable micromirror with comb drive, which is suspended rotatably about a central axis. The publication<patcit><text>US 2004/0114259 A1</text></patcit> shows a micromirror with a spring arrangement for displacement, wherein the micromirror is arranged parallel to the base substrate.

Disclosure of the invention

The invention provides a micromechanical component having the features of claim 1 and a manufacturing method for a micromechanical component having the features of claim 9.

The present invention is based on the finding that an actuating element, which is fastened to a base substrate via a spring device, can be adjusted in a preferred starting position without an external force by forming the spring device from at least two sub-units with different residual stresses in such a way that one after or Bending the spring device executed simultaneously with the production of the micromechanical component due to the different residual stresses, the control element in the preferred starting position.

In a preferred embodiment, the actuating element can be adjusted via the bending of the spring device due to the different residual stresses of the first subunit and the second subunit to an initial position in which it is inclined by an initial inclination angle not equal to 0 ° and not equal to 180 ° to the base substrate , In this way, the adjusting element can be adjusted to a desired angle of inclination to the base substrate and / or a light incident window arranged fixedly thereto, without having to carry out a further complicated adjustment. For example, the initial inclination angle can be chosen such that a light beam reflected at the light incidence window does not fall on an image plane of the micromechanical component designed as a micromirror.

In a further preferred embodiment with an electrostatic and / or magnetic drive, the electrostatic and / or magnetic drive is designed to counteract a spring force of the spring device and to adjust the actuating element with respect to the base substrate from the starting position into at least one end position. For example, theelectrostatic and / or magnetic drive a fixedly arranged on the base substrate stator electrode unit and an adjustable means of the spring means connected to the base substrate actuator electrode unit. Preferably, the actuator-electrode unit may be coupled to the actuator. The adjustment of the two electrode units to one another is omitted in this case. By way of example, at least one of the two electrode units is produced from an electrode material layer and subsequently turned out of the plane of the electrode material layer by means of the at least one spring device about an axis of rotation. In this way, the unwound electrode unit can be easily arranged at a certain inclination angle to the other electrode unit.

In particular, the actuator electrode unit and the adjusting element can be arranged on a frame. The entire frame of the micromechanical component can be rotated about an axis of rotation via the at least one spring device. The deflection of the frame determines the deflection of the mirror by means of the quasi-statically operated electrostatic drive. The frame is deflected so far that the deflection angle is above an optical angle of 15 ° required for a good image quality. The deflection angle of the frame is therefore always greater than zero. Because the frame always has angles greater than zero, the reflection on the glass pane is pushed out of the picture. The at least one spring device is designed in this way

In addition to avoiding reflection on the glass surface, forming the actuator and the actuator electrode assembly on the frame provides sufficient flexural rigidity to withstand accelerations up to 2500 g. The in-plane rigidity of the electrostatic drive is thus sufficient to avoid pull-in.

In addition, a comparatively large torque is achieved by the comparatively large distance of the electrode units from the axis of rotation even with a small applied voltage. The electrostatic drive is thus easy to operate with an operating voltage below 80 volts. This makes it possible to realize the electronic control in BCD technology.

An inclined arrangement of the two electrode units relative to one another requires a comparatively small total area for the two electrode units, the adjusting element and the two spring devices. Thus, a scanner can be realized whose area is smaller than 4 × 6 mm.

For example, the first residual stress of the first subunit and / or the second residual stress of the second subunit are an extrinsic residual stress and / or an intrinsic residual stress.

In an advantageous development, the spring device comprises a first spring and a second spring, which are arranged in two sections, wherein the spring device is bent so that the first spring has a curvature on a base substrate facing the underside of the first spring and the second spring has a curvature has on a side facing away from the base substrate top of the second spring. This brings the advantage of a rotation about the center axis.

For example, a coating of the underside of the first spring is provided with a material having a compressive stress and / or a coating of the top of the second spring with a material having a compressive stress. As an alternative or in addition thereto, a coating of an upper side of the first spring with a material with a tensile stress and / or a coating of a lower side of the second spring with a material with a tensile stress can be provided.

Preferably, the actuating element is suspended by means of two spring devices. The two spring devices ensure high bending stiffness with low torsional rigidity. This ensures that the actuator moves on application of a voltage between the two electrode units on a desired adjustment path.

The features described in the upper paragraphs can be realized as a component by a corresponding micromechanical manufacturing method. By structuring the combs in a plane, it is not necessary to align wafers with an accuracy of less than 10 μm. The micromechanical component described here can thus be manufactured in large series.

Brief description of the drawings

Further features and advantages of the present invention will be explained below with reference to the figures. Show it:

<figref>1A</figref> to <figref>1C</figref> a plan view, a cross section and a side view illustrating a first embodiment of the micromechanical device;

<figref>2</figref> a plan view of a second embodiment of the micromechanical component;

<figref>3A</figref> and <figref>3B</figref> Top views for illustrating a third embodiment of the micromechanical component;

<figref>4A</figref> to <figref>4D</figref> three-dimensional views of the basic structure of another embodiment;

<figref>5</figref> a cross section through a complete system;

<figref>6A</figref> to <figref>6N</figref> Cross sections through a substrate to illustrate an embodiment of the manufacturing process; and

<figref>7</figref> a plan view of an embodiment of an electrode unit of the micromechanical device.

Embodiments of the invention

<figref>1A</figref> to <figref>1C</figref> show a plan view, a cross section and a side view illustrating a first embodiment of the micromechanical device.

This in <figref>1A</figref> shown in plan view micromechanical component <figref>10</figref> has a control element as opposed to a base substrate <figref>12</figref> adjustable mirror plate <figref>14</figref> on. Instead of the mirror plate<figref>14</figref> can the micromechanical component <figref>10</figref> but also have another actuator. Preferably, an upper side of the mirror plate<figref>14</figref> at least partially covered with a reflective layer. The mirror plate<figref>14</figref> is by means of two torsion springs <figref>16</figref> with an inner frame <figref>18</figref> connected. The two torsion springs<figref>16</figref> are on two opposite sides of the mirror plate <figref>14</figref> attached. The central longitudinal axes of the two torsion springs<figref>16</figref> lie on a first axis of rotation <figref>20</figref>around which the mirror plate <figref>14</figref> is adjustable by means of a mechanism not described here.

The inner frame <figref>18</figref> is stuck with an outer frame <figref>22</figref> connected. Inside the outer frame<figref>22</figref> is a variety of actuator electrode fingers <figref>24</figref> educated. The actuator electrode fingers<figref>24</figref> are aligned parallel to each other. In addition, the actuator electrode fingers run<figref>24</figref> parallel to the first axis of rotation <figref>20</figref>, Preferably, the actuator electrode fingers<figref>24</figref> arranged equidistantly to each other.

The actuator electrode fingers <figref>24</figref> are clamped on both sides. Here are the actuator electrode fingers<figref>24</figref> with a first end to a mounting bar <figref>18a</figref> of the inner frame <figref>18</figref> or on a mounting bar <figref>22a</figref> of the outer frame <figref>22</figref> attached. Also, the second end opposite the first end of the electrode fingers<figref>24</figref> is firmly attached to one of the fastening webs <figref>18a</figref> or <figref>22a</figref> arranged. Preferably, the fastening webs extend<figref>18a</figref> and or <figref>22a</figref> perpendicular to the first axis of rotation <figref>20</figref>, To increase the stability of the frame<figref>18</figref> and <figref>22</figref> are at the ends of the two fastening webs <figref>18a</figref> and <figref>22a</figref> supporting webs <figref>18b</figref> and <figref>22b</figref> appropriate. The width of the support bars<figref>18b</figref> and <figref>22b</figref> can be significantly larger than the width of the actuator electrode fingers <figref>24</figref>,

Each of the actuator electrode fingers <figref>24</figref> is a stator electrode finger <figref>26</figref> assigned. The stator electrode fingers<figref>26</figref> and the actuator electrode fingers <figref>24</figref> are designed so that a non-zero voltage between the electrode fingers <figref>24</figref> and <figref>26</figref> can be applied. Is a voltage not equal to zero between the electrode fingers<figref>24</figref> and <figref>26</figref> on, so are the actuator electrode fingers <figref>24</figref> around a second axis of rotation <figref>28</figref> in the direction of the stator electrode fingers <figref>26</figref> turned. About this adjustment movement of the actuator electrode fingers<figref>26</figref> is the position of the mirror plate <figref>14</figref> opposite the base substrate <figref>12</figref> variable. The stator electrode fingers<figref>26</figref> are stuck on the ground substrate <figref>12</figref> arranged so that an adjustment movement of the actuator electrode fingers <figref>24</figref> the position of the stator electrode fingers <figref>26</figref> to the basic substrate <figref>12</figref> not changed.

In the illustrated micromechanical component <figref>10</figref> is the second axis of rotation <figref>28</figref> perpendicular to the first axis of rotation <figref>20</figref> aligned. In a preferred embodiment, the second axis of rotation is located<figref>28</figref> centered to the actuator electrode fingers <figref>24</figref>, In this case, at least one center of a central longitudinal axis of the actuator electrode fingers<figref>24</figref> on the second axis of rotation <figref>28</figref>,

<figref>1B</figref> shows a cross section along the line AA 'the <figref>1A</figref>, The line AA 'runs parallel to the first axis of rotation<figref>20</figref> and perpendicular to the second axis of rotation <figref>28</figref>,

In the illustrated example of <figref>1B</figref> There is no voltage between the actuator electrode fingers <figref>24</figref> and the stator electrode fingers <figref>26</figref> at. The actuator electrode fingers<figref>24</figref> are therefore in their initial position to the basic substrate <figref>12</figref>, In the initial position are the actuator electrode fingers<figref>24</figref> by an initial inclination angle α inclined to the stator electrode finger <figref>26</figref> arranged. The angle between the central longitudinal axes of the actuator electrode fingers<figref>24</figref> and the stator electrode finger <figref>26</figref> is thus equal to the initial inclination angle α.

In the manufacture of the micromechanical component <figref>10</figref> become the mirror plate <figref>14</figref>, the torsion springs <figref>16</figref>, The frames <figref>18</figref> and <figref>22</figref> and the electrode fingers <figref>24</figref> and <figref>26</figref> preferably from one to the base substrate <figref>12</figref> applied electrode material layer produced. Subsequently or simultaneously, the integrally formed components<figref>14</figref>. <figref>16</figref>. <figref>18</figref>. <figref>22</figref> and <figref>24</figref> around the initial inclination angle α with respect to the longitudinal axes of the stator electrode fingers <figref>26</figref> and the basic substrate <figref>12</figref> adjusted. On an embodiment of a manufacturing method for the micromechanical component<figref>10</figref> will be discussed in more detail below.

Will a non-zero voltage between the electrode fingers <figref>24</figref> and <figref>26</figref> applied, so are the actuator electrode fingers <figref>24</figref> in one direction of rotation <figref>30</figref> turned. The direction of rotation<figref>30</figref> is due to the position of the stator electrode fingers <figref>26</figref> to the actuator electrode fingers <figref>24</figref> established. The rotational movement causes the actuator electrode fingers<figref>24</figref> adjusted from its initial position to a position in which they have a tilt angle to the stator electrode fingers <figref>26</figref> which is smaller than the initial inclination angle α. Are the actuator electrode fingers<figref>24</figref> in a position parallel to the stator electrode fingers <figref>26</figref>, so the maximum adjustment angle is reached. A further adjustment of the actuator electrode fingers<figref>24</figref> from this position is no longer possible. Thus, the initial inclination angle α defines the maximum displacement angle of the actuator electrode fingers<figref>24</figref>, or the mirror plate <figref>14</figref>,

The stator electrode fingers <figref>26</figref> are each at one end via a first fastening part <figref>32</figref> firmly with the basic substrate <figref>12</figref> connected. The other end of the stator electrode fingers<figref>26</figref> can, for example via a second fastening part <figref>34</figref>, on a cover plate <figref>36</figref> be fixed. Alternatively, both ends of the stator electrode fingers<figref>26</figref> by means of the fastening parts <figref>32</figref> and <figref>34</figref> on the basic substrate <figref>12</figref> be attached. This ensures that the stator electrode fingers<figref>26</figref> even when applying a relatively high voltage between the actuator electrode fingers <figref>24</figref> and the stator electrode fingers <figref>26</figref> hardly from their position to the basic substrate <figref>12</figref> be adjusted.

Preferably, the micromechanical component <figref>10</figref> designed so that already the application of a comparatively low voltage to the electrode fingers <figref>24</figref> and <figref>26</figref> a relatively high torque for adjusting the mirror plate <figref>14</figref> causes. In order to increase the torque while maintaining a predetermined voltage, it is advantageous, the electrode fingers<figref>24</figref> and <figref>26</figref> to train as long as possible.

However, cantilever electrode fingers become more labile with increasing length. Conventionally, the length of cantilevered electrode fingers is therefore limited by the risk of vibrations of the electrode fingers occurring. The deflection of a cantilever electrode finger in a uniform area or mass load increases with the fourth power to the length of the electrode finger. A pressure, an electrostatic force or an acceleration may therefore be sufficient to vibrate a cantilevered electrode finger.

The electrode fingers clamped on both sides <figref>24</figref> and <figref>26</figref> However, compared to a cantilever electrode fingers have the advantage that they have an advantageous bending stiffness even at a greater length. For example, clamped on both sides electrode fingers<figref>24</figref> and <figref>26</figref> Withstand four times greater force than one-sided clamped electrode fingers of the same length. This ensures that the electrode fingers clamped on both sides<figref>24</figref> and <figref>26</figref> should not be vibrated by external disturbances.

The longer the electrode fingers <figref>24</figref> and <figref>26</figref> are formed, the greater is the torque applied at the same applied voltage. The bilateral clamping of the electrode fingers<figref>24</figref> and <figref>26</figref> thus ensures the advantage that in addition to an increased torque when applying a voltage sufficient stability of the electrode fingers <figref>24</figref> and <figref>26</figref> is guaranteed. The electrode fingers clamped on both sides<figref>24</figref> and <figref>26</figref> can be formed ten times longer than the cantilever electrode fingers with the same bending stiffness. The achievable torque is thus by a factor<figref>100</figref> be increased. For example, the electrode fingers<figref>24</figref> and or <figref>26</figref> at a height of 20 microns and a width of 5 microns, a length in a range between 250 microns and 2 mm.

The micromechanical component <figref>10</figref> includes two spring devices <figref>38</figref> (please refer <figref>1A</figref>). Each of the two spring devices<figref>38</figref> is with a first end <figref>40</figref> with the basic substrate <figref>12</figref> connected. The other end<figref>42</figref> of the two spring devices <figref>38</figref> is on an outer surface of the outer frame <figref>22</figref> attached. The other end<figref>42</figref> can do this on the second axis of rotation <figref>28</figref> lie. The two spring devices<figref>38</figref> keep the outer frame <figref>22</figref> with the mirror plate <figref>14</figref> in its initial position with respect to the base substrate <figref>12</figref>, provided no voltage between the electrode fingers <figref>24</figref> and <figref>26</figref> is applied. Preferably, the two spring devices<figref>38</figref> designed so that the mirror plate <figref>14</figref> in its initial position by the initial inclination angle α with respect to the base substrate <figref>12</figref> is inclined. On the advantages of such an arrangement of the outer frame<figref>22</figref> opposite the base substrate <figref>12</figref> through the two spring devices <figref>38</figref> will be discussed in more detail below. Furthermore, the two spring devices<figref>38</figref> designed so that when applying the voltage between the electrode fingers <figref>24</figref> and <figref>26</figref> be at least partially bent. The bending stiffness of parts of thespring means <figref>38</figref> allows an adjustment of the outer frame <figref>22</figref> by applying the voltage.

Each of the two spring devices <figref>38</figref> includes two springs <figref>44</figref> and <figref>46</figref>which are arranged in two sections to each other. Preferably, the central longitudinal axes of the two springs<figref>44</figref> and <figref>46</figref> perpendicular to the second axis of rotation <figref>28</figref> aligned. In a preferred embodiment, the two springs<figref>44</figref> and <figref>46</figref> via a connecting part <figref>48</figref> be connected, both springs <figref>44</figref> and <figref>46</figref> with one end at the connecting part <figref>48</figref> are attached. The longitudinal direction of the connecting part<figref>48</figref> runs preferably parallel to the second axis of rotation <figref>28</figref>, The two springs<figref>44</figref> and <figref>46</figref> can be parallel to the adjacent side of the outer frame <figref>22</figref>, For example, the support bar <figref>22b</figref>be arranged.

Preferably, the layer has the electrode combs with the electrode fingers <figref>24</figref> and <figref>26</figref> be formed, a thickness between 60 microns and 80 microns. When the voltage is applied, the electrode fingers become<figref>24</figref> and <figref>26</figref> against the force of the spring devices <figref>38</figref> brought into a common plane. This results in an increasing overlap area between the actuator electrode fingers<figref>24</figref> and the stator electrode fingers <figref>26</figref> on. The highest moment of force is exerted when the adjustment of the mirror plate<figref>14</figref> counteracting force of the spring devices <figref>38</figref> is maximum.

<figref>1C</figref> shows a side view of a spring device <figref>38</figref>, It can be seen that the two springs<figref>44</figref> and <figref>46</figref> to the surface of the base substrate <figref>12</figref> are bent differently. The spring device<figref>38</figref> has a convex spring <figref>44</figref> and a concave spring <figref>46</figref> on. The curvature of the convex spring<figref>44</figref> points from the top of the base substrate <figref>12</figref> path. In contrast, the curvature of the concave spring<figref>46</figref> on the basic substrate <figref>12</figref> directed.

The convex spring <figref>44</figref> and the concave spring <figref>46</figref> can be made by adding for each of the two springs <figref>44</figref> and <figref>46</figref> a first subunit with a first residual stress and a second subunit with a deviating from the first residual stress second residual stress are joined together. The residual stresses of the subunits can be extrinsic and / or intrinsic stresses. For example, a difference in the thermal expansion coefficients of the various subunits causes an extrinsic stress. If, as an alternative or in addition, an epitaxial layer is used for at least one of the two subunits, an intrinsic stress can be applied to the respective spring due to the lattice mismatches<figref>44</figref> or <figref>46</figref> be applied.

In the illustrated example, each of the two springs comprises <figref>44</figref> and <figref>46</figref> a core material layer as a first subunit and at least one coating applied to the core material layer as a second subunit. The at least one coating may be on a base substrate<figref>12</figref> facing side and / or on one of the base substrate <figref>12</figref> opposite side of the springs <figref>44</figref> or <figref>46</figref> be arranged. Depending on the material of the coating and the coated side will be any spring<figref>44</figref> and <figref>46</figref> Curved in a specified direction without external forces.

The feathers <figref>44</figref> and <figref>46</figref> can have a spring length of about 1.5 mm. The width of the springs<figref>44</figref> and <figref>46</figref> may be less than 500 μm, for example 250 μm. Preferably, the thickness of the springs<figref>44</figref> and <figref>46</figref> perpendicular to the base substrate <figref>12</figref> about 15 μm. The coating of the springs<figref>44</figref> and <figref>46</figref>, for example, with an oxide, may have a layer thickness of about 1.5 microns. In this way, the coating causes an intrinsic stress of about 350 MPa on the two springs<figref>44</figref> and <figref>46</figref>,

This results in a bimorph effect of the springs <figref>44</figref> and <figref>46</figref>, For example, the base substrate<figref>12</figref> facing side of the spring <figref>44</figref> and that of the basic substrate <figref>12</figref> opposite side of the spring <figref>46</figref> provided with a layer with a tensile stress. As an alternative or as a supplement thereto may be based on that of the basic substrate<figref>12</figref> opposite side of the spring <figref>44</figref> and on the basic substrate <figref>12</figref> facing side of the spring <figref>46</figref> a layer with a compressive stress are applied.

For those of the connector <figref>48</figref> opposite ends of the springs <figref>44</figref> and <figref>46</figref> can be tangents <figref>50</figref> and <figref>52</figref> define. The convex spring<figref>44</figref> associated tangent <figref>50</figref> runs parallel to the surface of the base substrate <figref>12</figref>, In contrast, the tangent<figref>52</figref> the concave spring <figref>46</figref> inclined to the surface of the base substrate <figref>12</figref> aligned. The two tangents<figref>50</figref> and <figref>52</figref> intersect at an intersection <figref>54</figref>through which the second axis of rotation (not shown) is perpendicular to the tangents <figref>52</figref> and <figref>50</figref> runs.

The outer frame <figref>22</figref> is so on the two spring device <figref>38</figref> arranged that the components <figref>14</figref>. <figref>16</figref>. <figref>18</figref>. <figref>22</figref> and <figref>24</figref> parallel to the tangent <figref>52</figref> run. The angle between the two tangents<figref>50</figref> and <figref>52</figref> thus corresponds to the angle of inclination of the components <figref>14</figref>. <figref>16</figref>. <figref>18</figref>. <figref>22</figref> and <figref>24</figref>, There is no voltage between the electrode fingers<figref>24</figref> and <figref>26</figref> on, that's the angle between the two tangents <figref>50</figref> and <figref>52</figref> equal to the initial inclination angle α. Will a non-zero voltage between the electrode fingers<figref>24</figref> and <figref>26</figref> put on, so will the springs <figref>44</figref> and <figref>46</figref> bent and the angle between the tangents <figref>50</figref> and <figref>52</figref> gets smaller. Onthis way, the mirror plate <figref>14</figref> opposite the base substrate <figref>12</figref> around the second axis of rotation <figref>28</figref> be adjusted.

Preferably, the spring means <figref>38</figref> designed so that the angle between the tangents <figref>50</figref> and <figref>52</figref> even at a high voltage greater than 15 °. The associated advantage will be discussed below.

In the manufacture of the micromechanical component <figref>10</figref> become the related components <figref>14</figref>. <figref>16</figref>. <figref>18</figref>. <figref>22</figref> and <figref>24</figref>Separately attached stator electrode fingers <figref>26</figref> and the spring devices <figref>38</figref> preferably from one to the base substrate <figref>12</figref> etched out the applied layer. About bending the springs<figref>44</figref> and <figref>46</figref> become the related components <figref>14</figref>. <figref>16</figref>. <figref>18</figref>. <figref>22</figref> and <figref>24</figref> in a preferred starting position relative to the base substrate <figref>12</figref> arranged. In the process, the components become<figref>14</figref>. <figref>16</figref>. <figref>18</figref>. <figref>22</figref> and <figref>24</figref> without an external force effect around the output inclination angle α from the plane of the stator-electrode fingers <figref>26</figref> unscrewed. In this way, it is no longer necessary, a complex adjustment of the components<figref>14</figref>. <figref>16</figref>. <figref>18</figref>. <figref>22</figref> and <figref>24</figref> perform.

<figref>2</figref> shows a plan view of a second embodiment of the micromechanical device.

The illustrated micromechanical component <figref>60</figref> has the above-mentioned components <figref>12</figref> to <figref>18</figref>. <figref>22</figref> to <figref>26</figref> and <figref>38</figref> on. Also, the operation of the micromechanical component<figref>60</figref> is similar to the embodiment described above. However, the width of the inner frame is<figref>18</figref> parallel to the first axis of rotation <figref>20</figref> equal to the width of the outer frame <figref>22</figref> parallel to the first axis of rotation <figref>20</figref>, For example, the width is the frame<figref>18</figref> and <figref>22</figref> 1.5 mm.

The length of the inner frame <figref>18</figref> parallel to the second axis of rotation <figref>28</figref> can also be 1.5 mm. The total length of the outer frame<figref>22</figref> parallel to the second axis of rotation <figref>28</figref> is for example 3.5 mm.

The micromechanical component <figref>60</figref> also points to the outer surfaces of the outer frame <figref>22</figref> Actuator electrode fingers <figref>24</figref> on. The length of these outer actuator electrode fingers<figref>24</figref> can be between 0.5 mm to 1 mm. These actuator electrode fingers<figref>24</figref> are also stator electrode fingers <figref>26</figref> on the basic substrate <figref>12</figref> assigned. In this way it is possible, the micromechanical component shown<figref>60</figref> with relatively long electrode fingers <figref>24</figref> and <figref>26</figref> to train comparatively small.

Each of the two spring devices <figref>38</figref> has a convex spring <figref>44</figref> and a concave spring <figref>46</figref> on. The concave feathers<figref>46</figref> are adjacent to the outer frame <figref>22</figref> arranged. The convex springs are on those of the outer frame<figref>22</figref> directed sides of the spring devices <figref>38</figref> educated.

<figref>3A</figref> and <figref>3B</figref> show plan views illustrating a third embodiment of the micromechanical device.

This in <figref>3A</figref> shown micromechanical component <figref>70</figref> includes the components already described above <figref>12</figref> to <figref>18</figref>. <figref>22</figref> to <figref>26</figref> and <figref>38</figref>, The training of the frame<figref>18</figref> and <figref>22</figref> corresponds to the embodiment of the <figref>2</figref>,

In contrast to the embodiment described above, the micromechanical component <figref>70</figref> on a first side a spring device <figref>72</figref> on, at which the convex spring <figref>44</figref> adjacent to the outer frame <figref>22</figref> is trained. The concave spring<figref>46</figref> is at the outer frame <figref>22</figref> opposite side of the spring device <figref>72</figref> arranged. In addition, the micromechanical component comprises<figref>70</figref> on its second page the spring already described above <figref>38</figref> with an outer convex spring <figref>44</figref> and an inner concave spring <figref>46</figref>, The suspension of the outer frame<figref>22</figref> by means of the spring devices <figref>38</figref> and <figref>72</figref> leads to an improved balance of the weight of the related components <figref>14</figref> to <figref>18</figref> and <figref>22</figref> to <figref>24</figref> by compensation of the forces acting in different directions. In particular, should be prevented in this way a total moment of force in a particular direction.

<figref>3B</figref> shows the micromechanical component <figref>70</figref> with some tracks formed on it <figref>74</figref>, The tracks<figref>74</figref> For example, they may contain NiCrSi or aluminum. About the tracks<figref>74</figref> is a contacting of the actuator electrode fingers <figref>24</figref> possible. It can on the inner frame<figref>18</figref> also actuator-electrode fingers <figref>24</figref>which is perpendicular to the first axis of rotation <figref>20</figref> are aligned, be formed.

Furthermore, sensor elements <figref>76</figref> for determining a bending of the springs <figref>44</figref> or <figref>46</figref> over the tracks <figref>74</figref> be powered. Similarly, control or sensor signals via the tracks<figref>74</figref> between the sensor elements <figref>76</figref> and an (not outlined) evaluation device for determining a current position of the mirror plate <figref>14</figref> to get redirected. On the functioning of the sensor elements<figref>76</figref> will be discussed in more detail below.

<figref>4A</figref> to <figref>4D</figref> show three-dimensional views of the basic structure of another embodiment.

The partially shown micromechanical component <figref>80</figref> has a mirror plate <figref>14</figref>, Torsion springs <figref>16</figref>a frame <figref>82</figref> with attached actuator electrode fingers <figref>24</figref>, corresponding stator electrode fingers <figref>26</figref> and a spring device <figref>38</figref>, The number of actuator electrode fingers<figref>24</figref> and the stator electrode finger <figref>26</figref> is arbitrary. The sake of clarity are in the<figref>4A</figref> to <figref>4D</figref> but only a few electrode fingers <figref>24</figref> and <figref>26</figref> located.

The spring device <figref>38</figref> has the convex spring <figref>44</figref> and the concave spring <figref>46</figref> on. The feathers<figref>44</figref> and <figref>46</figref> are each made up of at least two subunits with different residual stresses. The residual stresses can be extrinsic and / or intrinsic residual stresses. Due to the different residual stresses have the springs<figref>44</figref> and <figref>46</figref>if no force is applied to it, a convex or a concave initial shape. In the production of springs<figref>44</figref> and <figref>46</figref> These take the convex or concave initial shape without a foreign force effect. About one on the spring device<figref>38</figref> applied force can be the springs <figref>44</figref> and <figref>46</figref> be bent from their initial shape into another shape, preferably a planar shape.

In <figref>4A</figref> There is no voltage between the actuator electrode fingers <figref>24</figref> and the stator electrode fingers <figref>26</figref> at. The mirror plate<figref>14</figref> is thus in its initial position. In the initial position is the mirror plate<figref>14</figref> due to the bulges of the convex spring <figref>44</figref> and the concave spring <figref>46</figref> inclined at an initial inclination angle to the base substrate (not shown). The advantage of a relative to the base substrate inclined starting position of the mirror plate<figref>14</figref> will be described in more detail below.

In <figref>4B</figref> is a voltage not equal to 0 between the actuator electrode fingers <figref>24</figref> and the stator electrode fingers <figref>26</figref> at. Due to the attraction caused by the applied voltage between the electrode fingers<figref>24</figref> and <figref>26</figref> becomes the mirror plate <figref>14</figref> adjusted so that their angle of inclination relative to the base substrate decreases.

The feathers <figref>44</figref> and <figref>46</figref> can be designed with regard to their bending stiffnesses so that they can be pressed into a nearly flat final shape (see <figref>4B</figref>). The mirror plate<figref>14</figref> is located at this final form of the springs <figref>44</figref> and <figref>46</figref> in their maximum adjustment position. The basic deflection of the mirror plate<figref>14</figref> The bimorph effect is almost completely eliminated. Alternatively, the spring stiffness of the springs<figref>44</figref> and <figref>46</figref> also be set so that the mirror plate <figref>14</figref> always has a minimum inclination angle relative to the base substrate. The advantages of such an embodiment will be discussed in more detail below.

As by comparing the <figref>4A</figref> and <figref>4B</figref> it can be seen, the springs are <figref>44</figref> and <figref>46</figref> when adjusting the mirror plate <figref>14</figref> especially in their areas near the connecting part <figref>48</figref> bent and / or deformed. The to the connecting part<figref>48</figref> adjacent surfaces of the springs <figref>44</figref> and or <figref>46</figref> Therefore, they are well suited for attaching a sensor element <figref>76</figref> for determining a shape of a spring <figref>44</figref> or <figref>46</figref> and / or for determining a current position of the mirror plate <figref>14</figref>,

<figref>4C</figref> shows two on the feathers <figref>44</figref> and <figref>46</figref> attached sensor elements <figref>76</figref>, The sensor elements<figref>76</figref> For example, piezoresistive resistors are located on or within the regions of the springs <figref>44</figref> and <figref>46</figref>which is adjacent to the connecting part <figref>48</figref> are formed, are arranged. Preferably, the piezoresistive resistors are NiCrSi metallizations, which are in the polycrystalline material of the springs<figref>44</figref> and <figref>46</figref> are formed. Furthermore, the sensor elements<figref>76</figref> include Wheatstone bridges. To recognize are in<figref>4D</figref> over the springs <figref>44</figref> and <figref>46</figref> guided tracks <figref>74</figref>, Each strand represents its own interconnect element. About the connecting web metal tracks can be guided, via which the sensor elements<figref>76</figref> be supplied with electricity. In addition, via the conductor tracks<figref>74</figref> Sensor signals from the sensor elements <figref>76</figref> be forwarded to a central control and evaluation system. Preferably, the control and evaluation system is designed to adjust the to the electrode fingers<figref>24</figref> and <figref>26</figref> applied voltage a current position of the mirror plate <figref>14</figref> to determine.

Of course, the micromechanical component <figref>80</figref> even more piezoresistive and / or capacitive sensor elements <figref>76</figref> to determine the current position of the mirror plate <figref>14</figref> have, which at the locations of the micromechanical component <figref>80</figref> are formed, at which when adjusting the mirror plate <figref>14</figref> a deformation takes place.

<figref>5</figref> shows a cross section through a complete system.

The micromechanical component <figref>90</figref> has the adjustable mirror plate <figref>14</figref>, which by means of (not shown) spring means in their initial position relative to the base substrate <figref>12</figref> is inclined at an initial inclination angle α. About the above-described, not shown here electrostatic drive the mirror plate<figref>14</figref> so adjustable that the angle between the mirror plate <figref>14</figref> and the basic substrate <figref>12</figref> is smaller than the initial inclination angle α.

To the shock sensitivity of the micromechanical component <figref>90</figref> can reduce the micromechanical component <figref>90</figref> in a housing <figref>92</figref> which are arranged on a vibration-damped table <figref>94</figref> is stored. For example, a closed-loop counter-regulation is executed. One is on the micromechanical component<figref>90</figref> acting acceleration detected and on the vibration-damped table <figref>94</figref> compensated (with piezo control). components<figref>96</figref> a control and Auswertesystems can also in the housing <figref>92</figref> be arranged.

In the case <figref>92</figref> is also a laser <figref>97</figref> attached, which a light beam <figref>98a</figref> emitted. The emitted light beam<figref>98a</figref> meets a light window <figref>99</figref> of the micromechanical component <figref>90</figref> and is partially reflected. The transmitted light beam<figref>98b</figref> falls on the mirror plate <figref>14</figref> and is about the current position of the mirror plate <figref>14</figref> directed in a desired direction and / or on a preferred point. The at the light incident window<figref>99</figref> reflected light beam <figref>98c</figref> also forms a point of light, often referred to as a reflection point.

To prevent the reflection point of the reflected light beam <figref>98c</figref> on the image plane of the transmitted light beam <figref>98b</figref> meets, is the micromechanical component <figref>90</figref> designed so that the mirror plate <figref>14</figref> always in an inclined position to the basic substrate <figref>12</figref>, and thus to the light window <figref>99</figref>, is located. Preferably, the mirror plate<figref>14</figref> to that parallel to the base substrate <figref>12</figref> aligned light window <figref>99</figref> always a minimum angle of inclination. In this case, the mirror plate<figref>14</figref> in the operating mode of the micromechanical component <figref>90</figref> adjustable in no position in which their angle of inclination to the light incident window <figref>99</figref> is less than the minimum inclination angle. The minimum inclination angle is for example 15 °.

The disturbing reflection on the glass surface is thus masked out by the mirror plate even when the glass plate is arranged in a plane-parallel arrangement relative to the chip surface <figref>14</figref> always has a positive angle to the chip surface.

This ensures that the mirror plate <figref>14</figref> deflected light beam <figref>98b</figref> does not fall on a plane on which the at the light incident window <figref>99</figref> reflected light beam <figref>98c</figref> occurs. The means of the micromechanical component<figref>90</figref> generated image thus has no disturbing reflection point.

<figref>6A</figref> to <figref>6N</figref> show cross-sections through a substrate to illustrate one embodiment of the manufacturing process.

In a first step of the method becomes a surface of a base substrate <figref>100</figref> at least partially with an insulating layer <figref>102</figref> covered. The basic substrate<figref>100</figref> may be a silicon substrate. Preferably, the insulating layer<figref>102</figref> a silicon oxide film formed by thermal oxidation of the silicon substrate.

On the insulating layer <figref>102</figref> becomes an electrode material layer <figref>104</figref> applied. The electrode material layer<figref>104</figref> may contain silicon, for example. The top of the electrode material layer<figref>104</figref> is covered with a nitride layer. Subsequently, for example using a suitable mask, the nitride layer is partially removed from the top of the electrode material layer<figref>104</figref> removed, so that only at least one point, which marks a later spring of the spring devices, a nitride partial layer <figref>106</figref> remains.

Not from the at least one nitride sublayer <figref>106</figref> covered surfaces of the electrode material layer <figref>104</figref> are subsequently treated with a silicon oxide layer <figref>108</figref> covered. This can be done, for example, by means of a thermal oxidation of the electrode material layer<figref>104</figref> respectively. The at least one nitride sublayer<figref>106</figref> acts as an oxidation barrier.

<figref>6A</figref> shows the basic substrate <figref>100</figref> after etching at least one trench <figref>110</figref>, for example using a structured mask. The etching time is chosen such that the maximum depth of the at least one trench<figref>110</figref> greater than the total thickness of the layers <figref>102</figref>. <figref>104</figref> and <figref>108</figref> is. On the function of the at least one trench<figref>110</figref> will be discussed below.

In a further process step, the nitride partial layer <figref>106</figref> and the silicon oxide layer <figref>108</figref> with a polysilicon layer <figref>112</figref> covered. At the same time, at least one trench will be formed<figref>110</figref> filled with polysilicon. Preferably, the polysilicon layer<figref>112</figref> an epi polysilicon layer.

As already described above, at least one further nitride sublayer is formed <figref>114</figref> on the polysilicon layer <figref>112</figref> formed, which marks a surface of a later produced spring of the spring device. Subsequently, by means of a thermal oxidation, a second silicon oxide layer<figref>116</figref> formed, which are the uncovered sub-surfaces of the polysilicon layer <figref>112</figref> covers. The result is in<figref>6B</figref> shown.

On the at least one nitride sublayer <figref>114</figref> and the second silicon oxide layer <figref>116</figref> a structured mask (not shown) is formed. Subsequently, recesses are made <figref>118</figref> etched through which the second silicon oxide layer <figref>116</figref> and the polysilicon layer <figref>112</figref> extend. The recesses<figref>118</figref> divide the previous polysilicon layer into edge regions <figref>120</figref> and in at least one feather area <figref>122</figref>, The border areas<figref>120</figref> extend over the outer edges of the first silicon oxide layer <figref>108</figref>, Every feather area<figref>122</figref> has on its surface the nitride partial layer <figref>114</figref> and a silicon oxide sub-layer formed from the previous second silicon oxide layer <figref>126</figref> on. From the at least one feather area<figref>122</figref> Later, the springs are formed for a spring device.

As in <figref>6C</figref> can be seen, the first silicon oxide layer acts <figref>108</figref> when etching the recesses <figref>118</figref> as an etch stop layer. The mask for etching the recesses<figref>118</figref> can be removed.

<figref>6D</figref> shows the basic substrate <figref>100</figref> after applying a first mask <figref>128</figref> and a second mask <figref>130</figref> on a bottom of the base substrate <figref>100</figref>, Preferably, at least one of the masks<figref>128</figref> or <figref>130</figref> a photoresist mask. The mask<figref>130</figref> is structured so that it has a recess under the spring area <figref>122</figref> having. Using the structured second mask<figref>130</figref> becomes a recess <figref>132</figref> etched, which extends through the entire height of the basic substrate <figref>100</figref> extends. The insulating layer<figref>102</figref> can act as Ätzstoppschicht.

Subsequently, the first insulating layer <figref>102</figref> from the bottom of the recess <figref>132</figref> away. The first mask<figref>128</figref> is for etching of further separation trenches <figref>134</figref> structured. At the same time the recess<figref>132</figref> deepened. The recess<figref>132</figref> extends now as in <figref>6E</figref> to see, up to the first silicon oxide layer <figref>108</figref>or to the nitride partial layer <figref>106</figref>,

The second mask <figref>130</figref> will be removed. The result is in<figref>6F</figref> shown.

By means of an HF etching, oxide layers (preferably with the first mask <figref>128</figref>) from the bottom of the base substrate <figref>100</figref> away. <figref>6G</figref> shows the basic substrate <figref>100</figref> after a subsequent thermal oxidation of the underside. The bottom of the ground substrate<figref>100</figref> is with an oxide layer <figref>136</figref> covered. The thermal oxidation after HF etching ensures that identical stress conditions on the top and bottom of the spring area<figref>122</figref> to rule.

As in <figref>6G</figref> can also be seen becomes a partial surface <figref>138</figref> the electrode material layer <figref>104</figref> exposed. The subarea<figref>138</figref> marks the mirror surface of the adjustable mirror plate produced in the following process steps. This one comes with a reflective layer<figref>140</figref>, preferably with aluminum, covered.

Likewise, other tracks <figref>142</figref> on the first silicon oxide layer <figref>108</figref>, on the silica sub-layer <figref>126</figref> and / or on the nitride sublayer <figref>114</figref> educated. At the same time materials of layers provide<figref>108</figref>. <figref>114</figref> and <figref>126</figref> an electrical insulation of the newly formed conductor tracks <figref>142</figref>, The tracks<figref>142</figref> can with the at least one filled trench <figref>110</figref> be connected. In this way, an electrical signal can be transmitted through the tracks<figref>142</figref> and the filled trench <figref>110</figref> to get redirected.

Before or after the formation of the tracks <figref>142</figref> become dividing trenches <figref>144</figref> for structuring the electrode material layer <figref>104</figref> through the first silicon oxide layer <figref>108</figref> and the electrode material layer <figref>104</figref> etched. The result is in<figref>6H</figref> to see.

In a subsequent process step becomes a mask <figref>146</figref> on the bottom of the base substrate <figref>100</figref> applied. Also the exposed areas of the nitride sublayer<figref>106</figref> and the first silicon oxide layer <figref>108</figref> be from the mask <figref>146</figref> covered. <figref>6I</figref> shows the attachment positions of the mask <figref>146</figref>,

In a subsequent etching step, the result in <figref>6J</figref> is shown, the separation trenches <figref>134</figref> deepened. The first insulating layer<figref>102</figref> serves as Ätzstoppschicht. After this etching step, the mask becomes<figref>146</figref> from the bottom of the textured base substrate <figref>100</figref>but not of the nitride sublayer <figref>106</figref> and the bottom of the first silicon oxide layer <figref>108</figref>, away.

The exposed areas of the insulating layer <figref>102</figref> are etched away. The thus separated areas of the earlier layers<figref>104</figref> and <figref>112</figref> can adjust to each other. In particular, each spring range can<figref>122</figref> due to the different residual stresses of the nitride partial layers <figref>106</figref> and <figref>114</figref> and the silica sublayers <figref>126</figref> and <figref>148</figref> bend out of its flat starting position in a curved end position. This bending of the spring area<figref>122</figref> causes an adjustment of the spring area <figref>122</figref> and the mirror plate coupled thereto without an external force effect.

The mirror plate with the reflective layer <figref>140</figref> will be there, as in <figref>6K</figref> can be seen, compared to their initial position rotated by an initial inclination angle α. The mirror plate is thus in an inclined position even without a voltage applied to the electrode fingers, not shown.

At the bottom of the textured base substrate <figref>100</figref> becomes a bottom plate <figref>150</figref>, preferably a Pyrex plate, attached (see <figref>6L</figref>). Using a sealglass bonding, a glass plate<figref>152</figref> on a sealglass layer <figref>154</figref> on the edge areas <figref>120</figref> be arranged firmly (see <figref>6M</figref>). The mirror plate, the electrode units (not shown) and the springs are thus hermetically sealed.

As in <figref>6M</figref> can be seen, is the mirror plate with the reflective layer <figref>140</figref> in an inclined position relative to the surface of the glass plate <figref>152</figref>, In this way it is ensured that a reflection beam from the glass plate<figref>152</figref> does not fall on an image plane of a light beam deflected by the mirror plate. Preferably, the mirror plate in each position by an angle of inclination of at least 15 ° relative to the glass plate<figref>152</figref> inclined. This ensures the advantage described above.

<figref>6N</figref> shows a hole through the bottom plate <figref>150</figref> for a via <figref>156</figref> the electrodes, not shown. In this way, a voltage for adjusting the mirror plate between the electrodes can be applied.

In order to guarantee an equal curvature of the springs, in the method described above, the processing of the layers responsible for the curvature is carried out simultaneously. The result is two bimorphs, with a first spring coated on the underside with the nitride and on the top with the oxide. In a second spring, the layer sequence is reversed. The nitride sublayers<figref>106</figref> and <figref>114</figref> has a tensile stress while the silica sublayers <figref>126</figref> and <figref>148</figref> exerts a compressive stress on each of the two springs. The result is two bimorphs with the same curvature but opposite curvature directions. Instead of the bimorphs, a metallization for the tensile or compressive stress can be used.

The initial inclination angle α varies due to manufacturing variations. By means of an electrical bias, however, a constant offset angle can be forced.

<figref>7</figref> shows a plan view of an embodiment of an electrode unit of the micromechanical device.

The illustrated electrode unit <figref>200</figref> may be an actuator electrode comb or a stator electrode comb. The electrode unit<figref>200</figref> includes four crossbars <figref>202</figref>, which by means of three branches <figref>202</figref> and <figref>204</figref> to a common point of contact <figref>208</figref> connected si nd. Two transverse struts<figref>202</figref> are over a first branch <figref>204</figref> connected with each other. A second branch<figref>206</figref> connects the first two branches <figref>204</figref> with the contact point <figref>208</figref>,

At each of the cross struts <figref>202</figref> are electrode fingers <figref>210</figref> formed, which in two opposite directions perpendicular to the respective transverse struts <figref>202</figref> protrude. Between the electrode fingers<figref>210</figref> the electrode unit <figref>200</figref> and electrode fingers of a counter electrode, not shown, a voltage can be applied. It is sufficient, the voltage source with the contact point<figref>208</figref> the electrode unit <figref>200</figref> and connect a contact point of the counter electrode.

The electrode unit <figref>200</figref> has opposite to a electrode comb, whose electrode fingers at least eight times the length of the electrode fingers <figref>210</figref> have the advantage that due to a subsequent interaction of the electrode fingers <figref>210</figref> the recoverable torque is increased by a factor of sixteen while maintaining the same voltage. All electrode fingers pull<figref>210</figref> preferably an actuator in the same direction. This again increases the torque by a factor of two. For example, if an applied voltage is cut in half by a factor of two, the loss of the producible force is only a factor of four.