Semiconductor device manufacturing: process – Making device or circuit emissive of nonelectrical signal – Including integrally formed optical element
Reexamination Certificate
2002-07-26
2004-05-25
Coleman, William David (Department: 2823)
Semiconductor device manufacturing: process
Making device or circuit emissive of nonelectrical signal
Including integrally formed optical element
C438S065000, C438S069000, C438S704000, C257S432000, C398S045000, C398S142000, C398S143000, C359S223100, C359S290000
Reexamination Certificate
active
06740537
ABSTRACT:
TECHNICAL FIELD
The invention relates to the field of microelectromechanical components also called MEMS standing for Microelectromechanical Systems. It relates more particularly to MEMS components used in fibre-optic communication devices. The invention relates more specifically to a process for fabricating microelectromechanical components which make it possible to optimize their performance and their manufacturing cost. This process may serve for fabricating various types of optical components which include a moving member moving under the effect of a control command. There may be optical switches, obturators or variable attenuators.
In the rest of the description, the invention will be more particularly described in respect of an optical switch, but it could easily be transposable to an optical obturator or attenuator.
PRIOR ART
In general, an optical switch receives at least one input optical fibre and at least two output optical fibres. These optical fibres are placed in optical propagation guides oriented very precisely with respect to one another, most generally at 90° with respect to one another. The optical switch comprises a mirror which can move in order to intercept the beams propagating in propagation guides. When the moving mirror is in a first position, it allows reflection of the optical beam output by an optical fibre towards a second fibre. When this mirror is in a second position, it does not modify the propagation of the beam output by the first optical fibre, which is therefore transmitted in the optical fibre located in alignment with it.
The movement of this mirror takes place by means of an actuator. Various types of actuators have already been proposed and especially electrostatic actuators, such as in particular that described in document U.S. Pat. No 6,229,640. This type of electrostatic actuator comprises a number of electrodes distributed in two interdigitated combs. These two interdigitated combs partially penetrate one another to form a capacitor thanks to their facing surfaces. Application of an electrical voltage between the two interdigitated combs causes a relative movement of one comb with respect to the other.
Since the mirror is fastened to one of the two combs of electrodes, it moves under the action of this voltage. Positional return takes place when the electrical voltage disappears, owing to the effect of return means which generally consist of one or a number of beams which connect the comb of moving electrodes to the rest of the substrate.
One of the objectives of the invention is to allow the mirror to move using an electrical voltage of a relatively limited value, while obtaining a sufficient excursion of the mirror. However, the use of a voltage of low value causes the facing surface area of the two combs of electrodes to increase.
Moreover, to obtain a movement of the greatest possible amplitude, it is important that the return means do not exert too large a force and that their stiffness be therefore relatively limited. However, this stiffness is determined inter alia by the thickness of the beams which constitute it. Therefore to increase the travel of the mirror, it is tempting to reduce the thickness of the beams of the return means of the actuator.
A problem then arises when it is desired to combine the two aforementioned effects, namely, on the one hand, an increase in the surface area of the electrodes and, on the other hand, a reduction in the thickness of the beams of the return means.
This kind of inconvenience is observed in the microcomponents produced on SOI (Silicon On Insulator)-based substrates. This is because, on SOI substrates, the definition of the electrodes and of the return means of the actuator is produced by etching down to the oxide layer. The electrodes and the return means are then freed by subsequent etching, carried out after the oxide layer has been etched. In this type of component produced from an SOI substrate, the beams of the return means and the electrodes therefore have the same height. To increase the force exerted by the actuator, it is therefore necessary to increase the number of electrodes, which results in a greater consumption of energy by the actuator and a greater occupation of the surface area of the substrate.
It has also been proposed to produce optical switches from a single-crystal silicon substrate, these also being called “bulk” switches. Various processes have been developed which depend on the crystallographic orientation of the substrate used. Thus, when the substrate used has an upper face parallel to the (100) plane of the silicon crystal structure, it is possible to carry out, in the same operation, etching of the mirror and of the propagation guides. This is because, thanks to the orientation of the crystal planes which form stop planes for the chemical etching, it is possible to obtain perfect alignment of the propagation guides lying along the same axis, and perfect perpendicularity of the orthogonal propagation guides. However, the thickness of the mirror obtained by this chemical etching depends on the etching time. The precision on the thickness of this mirror is therefore subject to the variations in the conditions under which the etching is carried out. Thus, a slight temperature drift may introduce considerable inaccuracy in the thickness of the mirror.
Wet etching operations are also carried out using substrates whose upper face is parallel to the (110) plane of the silicon crystallographic structure. In this case, the chemical etching stop planes correspond to the vertical sidewalls of the mirror, thereby making it possible to achieve very good precision on the thickness of the mirror.
However, in this situation, it is necessary to produce the propagation guides in a second phase, since the crystallographic axes do not coincide with the directions of these propagation guides. It is therefore necessary to produce them by a subsequent step, generally requiring the use of dry etching, of the reactive ion etching or RIE type.
One of the objectives of the invention is therefore to allow optical components to be produced from single-crystal silicon with a minimum number of steps.
Document U.S. Pat. No. 6,150,275 has described a process for producing microstructures from single-crystal silicon, the (111) crystallographic planes of which are parallel to the principal plane of the substrate. The process described in that document consists in linking dry etching steps for defining the contours of a microstructure on the substrate. This process is continued by a chemical etching step which makes it possible to free the structure predefined by the dry etching.
SUMMARY OF THE INVENTION
The invention therefore relates to a process for fabricating a microelectromechanical optical component which is produced from a silicon substrate. Such an optical component generally comprises:
at least two optical propagation guides, especially intended to receive optical fibres;
a wall which can move with respect to the propagation guide;
an electrostatic actuator capable of causing the moving wall to move with respect to the rest of the substrate, the said actuator comprising:
facing electrodes which can move with respect to each other, some of the electrodes being mechanically linked to the moving wall, the other electrodes being fastened to the rest of the substrate;
return means formed by at least one beam produced in the substrate and opposing the movement of the electrodes with respect to one another.
In accordance with the invention, the substrate used is made of single-crystal silicon, the (111) planes of which are parallel to the planes of the substrate. This process firstly comprises a first series of deep reactive ion etching steps during which the heights of the moving wall, of the electrodes of the actuator, and of the beams of the return means of the actuator are defined with different values. This process continues with a second wet etching step, making it possible to free the moving wall, the electrodes and the beams of the actuator from the rest of the substrate.
In oth
Berezny Neal
Coleman William David
Heslin Rothenberg Farley & & Mesiti P.C.
MEMSCAP
LandOfFree
Process for fabricating a microelectromechanical optical... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Process for fabricating a microelectromechanical optical..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Process for fabricating a microelectromechanical optical... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3190912