FR3031096A1 - MICROELECTROMECHANICAL OR NANOELECTROMECHANICAL DEVICE COMPRISING A MEMBRANE THAT IS MOBILE IN TRANSLATION AND IS PROFILED TO REDUCE SHORT CIRCUITS AND THE FORMATION OF ELECTRIC ARCS - Google Patents

Abstract

The microelectromechanical or nanoelectromechanical device (1A) comprises a membrane (3), which is supported above a substrate (2), spaced from said substrate (2), and at least one actuating electrode (5 or 5 '), which is formed on said substrate (2), which is positioned at least partly below the membrane (3), and which electrostatically actuates the membrane (3). The membrane (3) is movable in translation along at least a first translation axis (XX ') substantially parallel to the substrate (2), with a displacement stroke along this axis (XX') which is limited between two extreme positions. The lower face (3a) of the membrane comprises an electrostatic actuation zone (31a) which, when the membrane (3) is at rest, is spaced from the upper face (5a) of the actuating electrode 5 or 5 ') of a minimum distance (DR) measured perpendicular to the substrate (2). The lower face (3a) of said membrane (3) is not entirely flat and is profiled outside the electrostatic actuation zone (31a), so that when the membrane (3) is at rest, and for all the positions in translation of the membrane (3) at rest between said extreme positions along this first translation axis (XX '), the lower face (3a) of the membrane (3), at any point on its surface facing the upper face (5a) of the actuating electrode (5 or 5 '), is spaced from the upper face (5a) of the actuation electrode (5 or 5') by a distance ( D) measured perpendicular to the substrate (2), which is greater than or equal to said minimum idle distance (DR).

Description

TECHNICAL FIELD The present invention relates to the field of microelectromechanical systems, commonly known MEMS SPRING ROLLS. In this field 10 it relates to a new microelectromechanical or nanoelectromechanical device, which is of the type comprising an electrically conductive membrane and movable in translation parallel to a substrate, and at least one actuating electrode, which is formed on said substrate, and which allows to electrostatically actuate said membrane. The invention also relates to a method of manufacturing this new microelectromechanical or nanoelectromechanical device. This microelectromechanical or nanoelectromechanical device can be used in different applications, and in particular can be used to make an ohmic or capacitive switch, and in particular an RF (Radio Frequency) switch, or can be used to make a micro-switch or nanocom. optical system. PRIOR ART Among the MEMS or NEMS devices employing an electrically conductive membrane, which can be electrostatically actuated by means of an actuation electrode, there is a category of MEMS or NEMS devices in which the membrane, which is supported at above an insulating substrate, is not anchored to the insulating substrate, but is mobile in translation relative to the substrate, along at least one axis substantially parallel to the substrate. The membrane operating electrode (s) 30 are deposited on the substrate below the membrane, with an electrically insulating space, and in particular an air gap, between the membrane and each actuation electrode. Preferably, the membrane is more particularly flexible and electrostatically deformable by using the actuation electrode or electrodes. Examples of flexible and translatable membrane device are described in particular in the international patent application WO2006 / 099945 and in the international patent application WO2010 / 105827. In order to manufacture this type of MEMS or NEMS devices, standard manufacturing technologies are used to form and etch thin layers of materials superimposed on the surface of an insulating substrate, including at least one sacrificial layer, which is interposed between the actuating electrode (s) and the membrane. This sacrificial layer makes it possible, after suppression, to produce the insulating space, and in particular an air gap between the membrane and the underlying actuation electrode or electrodes 15 formed on the surface of the substrate. When operating this type of electrostatically actuatable membrane MEMS or NEMS devices, it is important that the membrane, which is electrically conductive, does not come into direct contact with an underlying actuating electrode, in order to avoid a short circuit. - circuit. It is also important to avoid the damaging formation of arcing between the membrane and each actuating electrode. The repetition over time of these electric arcs between the membrane and the actuation electrode may further cause material transfers between the membrane and the actuating electrode, with the risk over time of forming a short circuit. circuit between the membrane and the actuation electrode. To reduce the risk of occurrence of these malfunctions (short-circuit between the membrane and an actuation electrode, formation of arcs between the membrane and an actuating electrode), a first technical solution consists of dimensioning the MEMS or NEMS device, such that the distance 3031096 3 between each actuation electrode and the membrane is sufficiently large, so that when the membrane is actuated in its extreme position closest to the actuation electrode there remains always, between the electrode and the membrane, an insulating space of sufficient size to avoid a short circuit by contact between the membrane and the actuating electrode, and to reduce the formation of electric arcs. In this solution, a compromise must nevertheless be found on the choice of this distance between each actuating electrode and the membrane, because it is preferable that this distance is as small as possible.

A second known technical solution consists in providing, under the membrane or on the surface of the actuating electrode, dielectric pads making it possible to avoid direct contact between the membrane and the actuation electrode, and mechanically maintaining a sufficient minimum distance between the membrane and the underlying actuating electrode. A third known technical solution, and disclosed in particular in the aforementioned international patent application WO2006 / 099945, consists in covering the whole of the upper face (face facing the membrane) of each actuating electrode, with a layer of dielectric material, which makes it possible to electrically isolate the membrane from the whole of the upper face of the actuation electrode. This third solution may seem more effective than the other two solutions. However, by using this third solution, it is found in use that the dielectric layer on the surface of the actuation electrode is electrically charged over time, which modifies in time, and especially attenuates, the force of the electrostatic actuation applied to the membrane. This modification is detrimental to the repeatability of the operation of the device, and may even in time lead to a malfunction of the device.

For MEMS or NEMS devices having a membrane, which is anchored to the substrate, and which is not translatable in parallel with the substrate, the above solutions are in practice relatively effective in reducing the risks of short-circuiting. In contrast, the Applicant has demonstrated that in the particular case of MEMS or NEMS devices, comprising a membrane which is mobile in translation parallel to the substrate, these solutions were not completely satisfactory, and that the risks of occurrence of the aforementioned operating defects (short circuit between the membrane and an actuation electrode and arcing between the membrane and the electrode of actuation) could remain important. OBJECT OF THE INVENTION An object of the invention is to propose a new technical solution, which makes it possible to reduce the appearance of the above-mentioned functional defects (short circuit and arcing) in a microelectromechanical or nanoelectromechanical device. membrane movable in translation parallel to a substrate, and operable electrostatically by at least one underlying actuation electrode formed on the substrate. A more particular object of the invention is to propose a new technical solution, which makes it possible to preserve the possibility of manufacturing the microelectromechanical or nanoelectromechanical device by using thin layer deposition fabrication technologies, in particular by epitaxial growth, which are standards in this area. SUMMARY OF THE INVENTION Where standard technologies are used to fabricate a MEMS or NEMS mobile membrane device in translation parallel to a substrate, and electrostatically actuatable by at least one underlying actuation electrode formed on the surface of the substrate. substrate, and in particular an air gap, between the membrane and the actuating electrode, it is practically carried out, during manufacture, to form a sacrificial layer, which covers the substrate and 3031096 the actuating electrode, and whose thickness is substantially constant. As a result, in practice, the lower face of the membrane (facing towards the substrate), on the one hand, comprises an active part, generally flat, which is positioned in line with the actuating electrode 5, and whose zone the closest to the actuation electrode is located in a first reference plane when the membrane is not actuated, and secondly also has a recess, which is oriented toward the substrate, which is normally not not positioned at the right of the actuating electrode, but extending towards the substrate 10 to a plane closer to the actuating electrode than the abovementioned reference plane of the active part of the lower face of membrane. Now the Applicant has demonstrated that this recess in combination with the translational mobility of the membrane parallel to the substrate is an important source of the aforementioned operating defects (short circuit or arcing). Indeed, it has been demonstrated that in some cases the membrane may have undergone a small displacement or displacement in translation, unintentionally, parallel to the substrate and towards the actuating electrode. This translational positioning defect of the membrane, combined with the presence of this recess in the lower face of the membrane, can thus result in a detrimental reduction of the distance between this recess and the actuation electrode, which can cause the aforementioned malfunctions (short circuit or arcing). To overcome these malfunctions, the invention thus relates to a microelectromechanical or nanoelectromechanical device having the following characteristics. It comprises a membrane, which is supported above a substrate, spaced apart from said substrate, and which has a bottom face oriented towards the substrate, and at least one actuation electrode, which is formed on said substrate, which is positioned at least partially below the membrane, with an electrically insulating space, and in particular an air gap, between the membrane and the actuation electrode, and which electrostatically actuates the membrane . The membrane is movable in translation along at least one first translation axis (XX ') substantially parallel to the substrate, with a displacement stroke along this axis (XX') which is limited between two extreme positions. The lower face of the membrane comprises an electrostatic actuation zone which, for at least one non-shifted position of the membrane, is entirely positioned facing the actuation electrode, and which, when the membrane is at rest, is spaced apart the upper face of the actuating electrode a minimum distance (DR) measured perpendicularly to the substrate. The lower face of said membrane is not entirely flat and is profiled outside the electrostatic actuation zone, so that, when the membrane is at rest, and for all the positions in translation of the membrane at rest. between said extreme positions along this first translation axis (XX '), the lower face of the membrane, at any point on its surface facing the upper face of the actuating electrode, is spaced from the upper face of the actuating electrode of a distance (D) measured perpendicular to the substrate, which is greater than or equal to said minimum rest distance (DR). Optionally, the device of the invention may comprise the following optional technical features, taken alone or in combination with each other: when the membrane (3) is actuated, said electrostatic actuation zone (31a) is spaced apart the upper face (5a) of the actuating electrode (5 or 5 ') having a minimum distance (DA) measured perpendicular to the substrate (2); the lower face (3) of the membrane is profiled such that, when the membrane (3) is actuated and for all the positions in translation of the membrane (3) in the actuated state, between the two extreme positions the along said first axis (XX '), the distance between each point of the lower face (3a) of the membrane (3) and the point closest to the operating electrode (5 or 5') is greater or equal to the minimum distance (DA) in the actuated state. - the membrane (3) being movable in translation along said first translation axis (XX ') in a first direction (X1), the electrostatic actuation zone (31a) of the lower face (3a) of the membrane ( 3) is extended laterally, in a second direction (X2) opposite the first direction (X1), by a first recess (320a) oriented in the opposite direction to the substrate 10 (2) - the electrostatic actuation zone (31a) the lower face (3a) of the membrane (3) is extended laterally in the second direction (X2) by a first groove portion (32a), a portion of which forms said first recess (320a), which extends longitudinally substantially perpendicular to the first translation axis (XX '), and whose longitudinal edge (321a) opposite the electrostatic actuation zone (31a) is spaced from the upper face (5a) of the actuating electrode (5). or 5 '), when the membrane is at rest, of a tance (D1) measured perpendicular to the substrate (2), which is greater than or equal to, and preferably greater than, said minimum rest distance (DR). - The first groove portion (32a) extends laterally in the second direction (X2) by a first planar wall (33a) substantially parallel to the substrate (2). When the membrane (3) is at rest, the distance (D '), measured perpendicularly to the substrate (2), between said first plane wall (33a) and the plane of the upper face (5a) of the electrode Actuation (5 or 5 ') is greater than the minimum distance at rest (DR). The first groove portion (32a) extends laterally in the second direction (X2) by a recess (34a) facing the substrate (2). in the non-shifted position of the membrane (3), a first portion (50a) of the wafer (50) of the actuation electrode (5 or 5 ') is positioned below the first recess (320a), and if necessary below the first groove portion (32a). - the membrane (3) being movable in translation along said first translation axis (XX ') in the second direction (X2), the electrostatic actuation zone (31a) of the lower face (3a) of the membrane (3 ) is extended laterally in the first direction (X1) by a second recess (320b) oriented in the opposite direction to the substrate (2). the electrostatic actuation zone (31a) of the lower face (3a) of the membrane (3) is extended laterally in the first direction (X1) by a second groove portion (32b), a portion of which forms second recess (320b), which extends longitudinally substantially perpendicularly to the first translation axis (XX '), and whose longitudinal edge (321b) opposite the electrostatic actuation zone (31a) is spaced from the upper face (5a ) of the actuation electrode (5 or 5 '), when the membrane is at rest, a distance (D2) measured perpendicular to the substrate (2), which is greater than or equal to, and preferably greater than, said minimum distance at rest (DR). - The second groove portion (32b) extends laterally in the first direction (X1) by a second plane wall (33b) substantially parallel to the substrate (2). when the membrane (3) is at rest, the distance (D "), measured perpendicular to the substrate (2), between said second plane wall (33a) and the plane of the upper face (5a) of the electrode of Actuation (5 or 5 ') is greater than the minimum distance at rest (DR) - the second groove portion (32b) extends laterally in the first direction (X1) by a recess (34b) facing the substrate (2): in the non-shifted position of the membrane (3), a second portion (50b) of the wafer (50) of the actuation electrode (5 or 5 ') is positioned below the second recess (320b), and where appropriate below the second groove portion (32b) .- the membrane (3) is movable in translation along a second translation axis (YY ') substantially parallel to the substrate (2) and perpendicular to the first translation axis (XX '), with a displacement stroke 10 along this second axis (YY') which is limited between two extreme positions, in which the lower face (3a) of the membrane (3) is profiled outside the electrostatic actuation zone (31a), so that when the membrane (3) is at rest, and for all the positions in translation of the membrane 15 (3) at rest between said extreme positions along this second axis of translation (YY '), the lower face (3a) of the membrane (3), at any point of its surface facing the upper face (5a) of the actuating electrode (5 or 5 '), is spaced from the upper face (5a) of the actuating electrode 5 or 5') by a distance 20 (D) measured perpendicular to the substrate (2), which is greater than or equal to said minimum distance at rest (DR). when the membrane (3) is actuated, said electrostatic actuation zone (31a) is spaced from the upper face (5a) of the actuation electrode (5 or 5 ') by a minimum distance (DA) 25 measured perpendicular to the substrate (2), and wherein the lower face (3) of the membrane is profiled such that, when the membrane (3) is actuated and for all the positions in translation of the membrane (3) to the actuated state, between the two extreme positions along said second axis (YY '), the distance 30 between each point of the lower face (3a) of the membrane (3) and the point closest to the electrode Actuation (5 or 5 ') is greater than or equal to the minimum distance (DA) in the actuated state. the diaphragm (3) being movable in translation along said second translation axis (YY ') in a third direction (Y1), the electrostatic actuation zone (31a) of the lower face (3a) of the diaphragm ( 3) is extended laterally, in a fourth direction (Y2) opposite the third direction (Y1), by a third recess (320c) oriented in the opposite direction to the substrate (2). the electrostatic actuation zone (31a) of the lower face (3a) of the membrane (3) is extended laterally in the fourth direction (Y2) by a third groove portion (32c), a portion of which forms third recess (320c), which extends longitudinally substantially perpendicularly to the second translation axis (YY '), and whose longitudinal edge (321c) opposite the electrostatic actuation zone (31a) is spaced from the upper face ( 5a) of the actuation electrode (5 or 5 '), when the membrane is at rest, a distance (D3) measured perpendicular to the substrate (2), which is greater than or equal to, and preferably greater than, said minimum distance at rest (DR). The third groove portion (32c) extends laterally in the fourth direction (Y2) by a third plane wall (33c) substantially parallel to the substrate (2). when the membrane (3) is at rest, the distance (D '), measured perpendicular to the substrate (2), between said third plane wall (33c) and the plane of the upper face (5a) of the electrode Actuation (5 or 5 ') is greater than the minimum distance at rest (DR). - The third groove portion (32c) extends laterally in the fourth direction (Y2) by a recess (34c) facing the substrate (2). in the non-shifted position of the membrane (3), a third portion (50c) of the wafer (50) of the actuation electrode (5 or 5 ') is positioned below the third recess (320c ), and if necessary below the third groove portion (32c). - the membrane (3) being movable in translation along said second translation axis (YY ') in the fourth direction (Y2), the electrostatic actuation zone (31a) of the lower face (3a) of the membrane ( 3) is extended laterally in the third direction (Y1) by a fourth recess (320d) oriented in the opposite direction to the substrate (2). The electrostatic actuation zone (31a) of the lower face (3a) of the membrane (3) is extended laterally in the third direction (Y1) by a fourth groove portion (32d), a portion of which forms fourth recess (320d), which extends longitudinally substantially perpendicularly to the second translation axis (YY '), and whose longitudinal edge (321b) opposite the electrostatic actuation zone (31a) is spaced from the upper face ( 5a) of the actuation electrode (5 or 5 '), when the membrane is at rest, a distance (D4) measured perpendicular to the substrate (2), which is greater than or equal to, and preferably greater than, at said minimum distance at rest (DR). - The fourth groove portion (32d) extends laterally in the third direction (Y1) by a fourth plane wall (33d) substantially parallel to the substrate (2). when the membrane (3) is at rest, the distance (D "), measured perpendicular to the substrate (2), between said fourth plane wall (33d) and the plane of the upper face (5a) of the electrode d The actuation (5 or 5 ') is greater than the minimum rest distance (DR) - the fourth groove portion (32d) extends laterally in the third direction (Y1) by a recess (34d) facing the substrate (2) 3031096 12 - in the non-shifted position of the membrane (3), a fourth portion (50d) of the wafer (50) of the actuating electrode (5 or 5 ') is positioned below of the fourth recess (320d), and if necessary below the fourth groove portion (32d) .- the four groove portions (32a, 32b, 32c and 32d) form a groove (32), which surrounds the the operating electrode (5 or 5 ') over its entire periphery - in the non-shifted position of the diaphragm (3), the groove (32) is positioned relative to the electrode 5 or 5 'underlying, so that the wafer (50) of the electrode 5 (or 5') is positioned below the groove (32) over the entire periphery of the the actuation electrode (5 or 5 '). - At least the lower face (3a) of the membrane (3) is electrically conductive. The diaphragm (3) is flexible and in which the actuation electrode or electrodes (5; 5 ') make it possible to deform the membrane (3) electrostatically. the actuation electrode (5 or 5 ') is positioned entirely below the membrane (3). The electrically conductive upper face (3a) of the actuating electrode (5 or 5 ') is bare on at least one electrically conductive central part (51) or on its entire surface. the membrane (3), when actuated by means of an underlying actuating electrode (5 or 5 '), does not touch the electrically conductive surface of the actuating electrode (5 or 5) '). the electrostatic actuation zone (31a) is substantially flat. - The electrostatic actuation zone (31a) is centered relative to the actuating electrode (5 or 5 '), when the membrane is in its non-shifted position. For at least a portion of the translationally offset positions of the membrane (3), the electrostatic actuation zone (31a) is only partially positioned facing the underlying actuation electrode (5 or 5); 5 '). the dimension (Lx) of the electrostatic actuation zone (31a), measured parallel to said first translation axis (XX '), is smaller than the dimension (LEx) of the underlying actuation electrode (5); or 5 ') measured parallel to said first translation axis (XX'). the dimension (Ly) of the electrostatic actuation zone (31a), measured parallel to said second translation axis (YY '), is smaller than the dimension (LEy) of the underlying actuation electrode (5); or 5 ') measured parallel to said second translation axis (YY'). the membrane (3) makes it possible to perform an ohmic, capacitive or optical switch function. The device comprises two electrically conductive tracks (S1 and S2) formed on the substrate (2) and spaced apart from one another, and the membrane (3) performs an ohmic switch function, so that when it is at rest, it is not in contact with the two tracks (S1 and S2), and when it is electrostatically actuated by means of at least one actuation electrode (5), it is in contact with the two tracks (S1 and S2) and makes it possible to electrically connect the two tracks (S1 and S2) to one another. Another object of the invention is a microelectromechanical or nanoelectromechanical system comprising at least one aforementioned microelectromechanical or nanoelectromechanical device. Another subject of the invention is a first method of manufacturing a microelectromechanical or nanoelectromechanical device referred to above and comprising the following manufacturing steps: (1) a multilayer structure is produced comprising a substrate on which at least one electrode is formed; actuation, (2) depositing a dielectric layer (CL3) covering the substrate and the electrode, (3) depositing a layer of photoresist resin (CL4), above the dielectric layer (CL3), 5 ( 4) the multilayer structure is etched so as to keep on the substrate, the actuation electrode and a part of the dielectric layer (CL3) straddling at least on the edge of the actuation electrode and on the upper substrate face, (5) a sacrificial layer (CL5) is deposited, which covers the substrate, the electrode and the remaining portion of the dielectric layer (CL3), the upper face of this sacrificial layer (CL5), not being flat and having a non-rectilinear profile corresponding to the profile of the lower face of the membrane of the device, (6) depositing at least one layer of material (CL6) above the sacrificial layer (CL5), and more particularly a layer (CL6) of electrically conductive material, (7) etching of the sacrificial layer (CL5) so as to remove it, and forming the membrane by means of at least the layer of material (CL6) deposited in the previous step. Another object of the invention is a second method of manufacturing a microelectromechanical or nanoelectromechanical device referred to above and comprising the following manufacturing steps: (1) a multilayer structure is produced comprising a substrate on which at least one electrode is formed; actuation, (2) depositing a first sacrificial layer (CL3 or CL3 '), covering the substrate and the electrode, (3) depositing a layer of photoresist resin (CL4 or CL4'), above the layer dielectric (CL3 or CL3 '), (4) an etching of the multilayer structure is carried out so as to keep on the substrate, the actuation electrode and a part of the first sacrificial layer (CL3 or CL3') to at least on the edge of the actuation electrode and on the substrate, (5) a second sacrificial layer (CL5 or CL5 ') is deposited, which covers the substrate, the electrode and the remaining portion of the first sacrificial layer ficielle (CL3 or CL3 '), the upper face of this second sacrificial layer (CL5 or CL5'), not being flat and having a non-rectilinear profile corresponding to the profile of the lower face of the membrane of the device, (6 at least one layer of material (CL6 or CL6 ') is deposited above the second sacrificial layer (CL5), and more particularly a layer (CL6 or CL6') of electrically conductive material, (7) a etching the remaining portion of the first sacrificial layer (CL3 or CL3 ') and the second sacrificial layer (CL5 or CL5') so as to remove them, and form the membrane by means of at least the layer of material (CL6 or CL6) ') deposited in the previous step.

Another object of the invention is to provide a third method of manufacturing a microelectromechanical or nanoelectromechanical device referred to above and comprising the following manufacturing steps: (1) a multilayer structure is produced comprising a substrate on which at least one electrode is formed; actuation, (2) depositing a first sacrificial layer (CL2 -) covering the electrode, (3) depositing on the first sacrificial layer (CL2 -) a second sacrificial layer (CL3 -), (4) depositing, on the second sacrificial layer (CL3 -), a layer (CL4 -) of a photoresist resin, (5) an etching of the multilayer structure from the previous step is carried out so as to preserve on the substrate (2) the actuation electrode, the first sacrificial layer (CL2 ") and a part of the second sacrificial layer (CL3"), so that the upper face of the multilayer structure 5 is not flat and prese In a non-rectilinear profile corresponding to the profile of the underside of the device membrane, (6) a layer of material (CL6 ") is deposited on top of the first sacrificial layer (CL2") and the remaining portion of 10 the second sacrificial layer (CL3 "), and more particularly a layer (CL6") of electrically conductive material, (7) the first sacrificial layer (CL2 ") and the remaining portion of the second sacrificial layer (CL3") are etched, so as to completely remove them, and form the membrane.

BRIEF DESCRIPTION OF THE DRAWINGS The features and advantages of the invention will appear more clearly on reading the following detailed description of several particular embodiments of the invention, which particular variants of embodiment are described as non-exemplary embodiments. and non-exhaustive of the invention, and with reference to the accompanying drawings in which: - Figure 1 is a top view of a first embodiment of a microelectromechanical or nanoelectromechanical device of the invention, the membrane being represented in transparency; FIG. 2 is a cross-sectional view of the device of FIG. 1 in the sectional plane 11-11 of FIG. 1 or in the sectional plane 11'-11 'of FIG. 1, the membrane being at rest and in non-shifted position, the thicknesses and dimensions in this figure not being scaled; Figure 3 is a cross-sectional view of the device of Figure 1 in section plane 111-111; FIG. 4 represents the device of FIG. 2, when the membrane 3031096 17 is in the non-shifted position and is electrostatically actuated by the underlying actuation electrode; - Figure 5 shows the device of Figure 2, when the membrane is at rest, and is slightly offset in translation in the direction X1 (or Y1); - Figure 6 shows the device of Figure 2, when the membrane is at rest and is offset in translation in the direction (X1 or Y1) to a position in extreme translation; FIG. 7 represents the device of FIG. 2, when the membrane 10 is electrostatically actuated by the underlying actuating electrode, and is offset in translation in the direction (X1 or Y1) to an extreme translation position ; FIG. 8 is a cross-sectional view of a second variant embodiment of a microelectromechanical or nanoelectromechanical device of the invention, the membrane being at rest and in the non-shifted position, the thicknesses and dimensions in this FIG. being not to scale; FIG. 9 shows the device of FIG. 8, when the membrane is in the non-shifted position, and is electrostatically actuated by the underlying actuation electrode; - Figure 10 shows the device of Figure 8, when the membrane is at rest, and is slightly offset in translation in the direction X1 (or Y1); FIG. 11 represents the device of FIG. 8, when the diaphragm 25 is at rest and is offset in translation in the direction (X1 or Y1) to a position in extreme translation; FIG. 12 represents the device of FIG. 8, when the membrane is electrostatically actuated by the underlying actuating electrode, and is shifted in translation in the direction (X1 or Y1) to a position in extreme translation; FIG. 13 is a cross-sectional view of a third embodiment of a microelectromechanical or nanoelectromechanical device of the invention, the membrane being at rest and in an unshifted position, the thicknesses and dimensions in this FIG. being not to scale; FIG. 14 shows the device of FIG. 13, when the membrane is in the non-shifted position, and is electrostatically actuated by the underlying actuation electrode; FIG. 15 represents the device of FIG. 13, when the membrane is at rest and is slightly offset in translation in the direction X 1 (or Y 1); - Figure 16 shows the device of Figure 13, when the membrane is at rest and is shifted in translation in the direction (X1 or Y1) to a position in extreme translation; FIG. 17 shows the device of FIG. 13, when the membrane is electrostatically actuated by the underlying actuating electrode, and is shifted in translation in the direction (X1 or Y1) to a position in extreme translation. ; FIG. 18 is a cross-sectional view of a fourth embodiment of a microelectromechanical or nanoelectromechanical device of the invention, the diaphragm being at rest and in the non-shifted position, the thicknesses and dimensions in this FIG. being not to scale; FIG. 19 shows the device of FIG. 18 when the membrane is in the non-shifted position and is electrostatically actuated by the underlying actuation electrode; - Figures 20 to 28 show the main manufacturing steps of the device of Figure 2 (1st variant); - Figures 29 to 36 show the main manufacturing steps of the device of Figure 8 (2nd variant); FIGS. 37 to 44 represent the main manufacturing steps of the device of FIG. 13 (third variant).

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENT FIG. 1 to 7 With reference to the particular variant embodiment of FIGS. 1 and 2, the microelectromechanical or nanoelectromechanical device 1A 5 comprises: an insulating substrate 2, comprising a substantially flat upper face 2a a flexible membrane 3, which is supported above the substrate 2, being placed on two pillars 4 spaced along the axis) (X ', 10 - actuating electrodes 5, 5' which are formed in the form of thin layers, on the upper face 2a of the substrate 2, and which are entirely positioned below the membrane 3, with an air space 6 formed between the lower face 3a of the membrane 3 and the upper face 5a substantially flat electrodes 5, 5 '.

Usually, the insulating substrate 2 is a monolayer structure in an electrically insulating material or a multilayer structure, at least the top layer of which is in an electrically insulating material. For example and without limitation to the invention, the substrate 2 is a plate 20, made of semiconductor material, such as silicon, covered with an insulating upper layer 21, such as, for example, silicon nitride. . In the particular example illustrated, the device 1 comprises: - two inner electrodes 5, which are positioned under the membrane 3 between the two pillars 4, each electrode 5 being close to one of the two pillars 4, 25 - two outer electrodes 5 ', which are respectively positioned under the ends of the membrane 3 outside the two pillars 4, each electrode 5' being close to one of the two pillars 4. In this example, the space 6 between the membrane 3 and the electrodes 5, 5 'contain air, which forms a dielectric constituting a good electrical insulator. In another variant, this air could be replaced by any other gas, especially an inert gas, making it possible to produce an electrically insulating space 6. With reference to FIG. 1, the membrane 3 is not fixed or anchored on the pillars 4, but is free by being mobile in translation on the one hand in the two opposite directions X 1 and X 2, along an axis longitudinal translation) (X 'parallel to the upper face 2a of the substrate 2, and secondly in the two opposite directions Y1 and Y2, along a transverse axis YY', parallel to the upper face 2a of the substrate 2 and perpendicular to the longitudinal axis) (X '.

The lower face 3a of the membrane 3 is electrically conductive. The membrane 3 is for example a monolayer or multilayer membrane made in one or more electrically conductive materials, and more particularly in a metal, such as for example gold or aluminum, or in a metal alloy. In this case, the membrane 3 acts as a moving electrode. The membrane 3 may also be a multilayer membrane, comprising a lower face 3a, which is made of an electrically conductive material, and more particularly in a metal, such as for example gold or aluminum, or in an alloy of metals, the upper layer or layers of the membrane not necessarily being electrically conductive. In this case, only the layer forming the lower face of the membrane 3 acts as a moving electrode. The pillars 4 are preferably made of an electrically conductive material, and for example a metal, such as for example gold or aluminum, or in a metal alloy, and are connected to the mass of the device. . The actuation electrodes 5, 5 'are made of an electrically conductive material, and more particularly a metal, such as for example gold or aluminum, or in a metal alloy. Each pillar 4 further comprises, on its upper face, two L-shaped abutments 40 which act as abutments in translation to the membrane 3 in the plane parallel to the substrate 2, defined by the longitudinal axis) (X 'and by When the membrane 3 is centered between the abutments 40, as shown in FIG. 1, there is a space E between the edge of the membrane 3 and each abutment 40, such that the membrane 3 is centered between the stops 40, as shown in FIG. so that the diaphragm 3 at rest is supported on the two pillars 4 while being free in translation in the two opposite directions X1 and X2 along the longitudinal axis) (X 'on a limited translation stroke, and in both opposite directions Y1 and Y2 along the transverse axis YY 'on a limited translation stroke The stops 40 thus make it possible to limit the travel of the translation of the membrane 3, in order to maintain the membrane above the pillars 4 Pillars 4 also feature on their college e upper stops 41 (Figures 1 and 3) which are used to lock the membrane 3 vertically, that is to say to block the membrane 3 in translation above the pillars 4 along an axis ZZ '(Figure 3) perpendicular to the upper face 2a of the substrate 2. With reference to FIG. 1, the device 1 is more particularly an ohmic switch, and comprises a signal line S comprising two electrically conductive coplanar tracks Si, S2, which are formed on the surface of the substrate 2, and which are spaced apart and in the extension of one another. The space G between the two tracks Si and S2 is positioned under the membrane 3, and between the two pillars 4, preferably halfway between each pillar 4. These tracks 51, S2 consist, for example, of a layer of metal and for example a layer of gold or aluminum deposited on the upper face 2a of the substrate 2.

In a known manner, in this particular embodiment, when all the electrodes 5 and 5 'are not electrically powered, the switching membrane 3 is at rest being substantially flat and spaced from the electrode 5, 5' (configuration at rest illustrated in Figure 2). In this idle configuration, the signal line S being interrupted between the two tracks S, no electrical signal can pass through this signal line S. When the inner electrodes 5 are electrically powered, they exert on the membrane 3 electrostatic forces, which are oriented towards the substrate 2, and which pull the central portion of the membrane 3, between the two pillars 4, in the direction of the substrate 2 and bring it closer to the actuating electrodes 5 (FIG. 4), without The electrostatic forces are strong enough that the flexible switching membrane 3 deforms by bending downwards between the two pillars 4, and touches the two conductive tracks S1, S2, so that the membrane 3 contacts the electrodes 5. to electrically connect the two conductive tracks S1, S2. During its deformation, the membrane 3 is not anchored on the pillars 4 at its two ends, it also slides slightly on the two pillars 4. In this switching configuration of the membrane, called ON, the signal line S is then closed and can be used to pass an electrical signal. When the flexible membrane 3 is in its ON switching configuration, and the internal electrodes 5 are no longer electrically powered, the aforementioned electrostatic forces are eliminated, and the membrane 2, by its flexibility, resumes its position at The external electrodes 5 'are electrically energized, they exert on the ends of the membrane 3 electrostatic forces which are oriented in the direction of the substrate 2, and which pull them ends of the membrane 3 towards the substrate 2, and bring them closer to the external actuating electrodes 5 ', without the membrane 3 does not touch the electrodes 5'. These electrostatic forces are sufficiently strong for the flexible switching membrane 3 to bend upwards while being supported on the two pillars 4, which makes it possible to move the central part of the membrane 3 away from the two conductive tracks S1 and S2 a distance greater than that of the membrane at rest. In this so-called OFF configuration, the signal line S is interrupted between the two tracks S, and no electrical signal can pass through this signal line S. The invention is not limited to a device 1A having a function of 3031096 23 ohmic switch. For example and in a nonlimiting and non-exhaustive manner of the invention, the device 1A can be modified so that its membrane 3 can act as a capacitive switch or act as a variable capacitance with an electrode replacing the signal line S, or so that its membrane 3 can act as an optical switch. The invention is not limited either to a device 1A employing both internal operating electrodes 5 and external actuating electrodes 5 '. For example and in a nonlimiting and non-exhaustive manner of the invention, in another variant, the device could comprise only internal actuating electrodes 5 or a single internal actuating electrode 5, positioned between the two pillars 4 and preferably centered under the membrane 3; the device could comprise only external operating electrodes 5 '. In addition, the actuating electrodes 5, 5 'do not necessarily have the same geometry or the same dimensions, and in the context of the invention an actuating electrode 5 or 5' may have a geometry and / or dimensions different from those of the appended figures. With reference to FIG. 2, the lower face 3a of the membrane 3 (face 3a facing towards the substrate 2) comprises, for each actuating electrode 20, 5 ', a part 31, called the active part, whose lower face 31a forms a main electrostatic actuation zone, which is provided to cooperate with the underlying actuating electrode 5 (or 5 ') to allow the electrostatic actuation of the membrane 3. In the non-shifted translation position of 3, this electrostatic actuation zone 31a is fully positioned facing the underlying actuating electrode 5 (or 5 '). In the particular example illustrated, but non-imperatively, in the non-shifted translational position of the membrane 3, this electrostatic actuating zone 31a is centered above the underlying actuating electrode 5 (FIG. or 5 '). More particularly, with reference to FIG. 4, the dimension Lx 3031096 24 (respectively Ly) of the electrostatic actuation zone 31a, measured parallel to said first translation axis) (X '(respectively parallel to said second translation axis YY') is smaller than the dimension LEx respectively LEy) of the underlying actuation electrode 5 (or 5 ') measured parallel to said first translation axis) (X' (respectively parallel to said second translation axis YY). membrane 3 is at rest, that is to say when the membrane is not actuated, this electrostatic actuating zone 31a is spaced from the upper face 5a of the actuating electrode 5 (or 5 ') a minimum distance DR at rest, measured perpendicular to the substrate 2. In this example, this electrostatic actuation zone 31a being substantially flat and parallel to the substrate 2, it defines a first reference plane P1 parallel to the substrate 2. In FIG. 2, the vertical distance measured perpendicularly to the substrate 2 between this reference plane P1 and the upper face 5a of the electrode 5 or 5 'at rest corresponds to this minimum distance DR at rest. With reference to FIG. 4, the device 1A is designed in a manner known per se so that, when the membrane 3 is not shifted (position of FIG. 1), and it is electrostatically actuated by means of FIG. an electrode 5 (or 5 ') and approaches the electrode 5 (or 5'), the electrostatic actuation zone 31a of the lower face 3a of the membrane 3 does not touch the upper face 5a of the electrode 5, in order to avoid a short circuit, and there remains between this electrostatic actuation zone 31a of the membrane 3 and the upper face 5a of the actuation electrode 5 (or 5 '), a minimum distance DA in the actuated state of the membrane 3, which has been calculated so as to avoid arcing between the membrane 3 and the electrode 5 (or 5 '). It should be noted that in the particular example of FIGS. 1, 2 and 4, the "non-shifted" translation position of the membrane 3 corresponds to a centered position of the membrane 3 with respect to the abutments 40 of the abutments 4. because in this particular example the residual space E (FIG. 1) between the edge of the membrane and the translational stops 40 of the pillars 4 is substantially identical on the two pillars 4. More generally, this "non-shifted" position corresponds to practice at the initial position in translation of the membrane 3 at the end of the manufacturing process of the device, the membrane 3 in the "non-shifted" position, not being necessarily centered with respect to the abutments 40 of the two pillars 4. The "non-shifted" position of the membrane 3 is in practice the optimal positioning for the operation of the membrane 3.

The membrane 3 being movable in translation in a plane parallel to the substrate 2, it is common that the membrane 3 does not remain perfectly on the pillars 4, but shifts in translation along the longitudinal axis) (X 'and / or along the transverse axis YY 'This shift can occur when the membrane 3 is at rest, or during actuation.

FIG. 5 shows the device 1A when the membrane 3 is at rest, and is shifted in translation in the direction X1 (or Y1) in an intermediate position between the two extreme positions in translation along the axis ) (X '(or YY'). FIG. 6 shows the device 1A when the membrane 3 is at rest, and is displaced in its extreme (abutment) position in translation in the X1 (or Y1) direction. along the axis) (X '(or YY') For a part of the translational offset positions of the diaphragm 3 (Fig. 6), the electrostatic actuation zone 31a is offset from the electrode 5 (or 5 ') so that the electrostatic actuation zone 31a is no longer fully positioned opposite the underlying actuating electrode 5 (or 5'), but only part of the electrostatic actuation zone 31a is positioned facing the actuating electrode 30 Underlying 5 (or 5 ') The extreme position in translation of the membrane 3 in the opposite direction X2 (or Y2) along the axis (X' (or YY ') is not shown in the appended figures, but corresponds to the position symmetrical to that of FIG. 6. In FIGS. 5, 6 and 7, the area of the lower face 3a of the membrane 3, which is located opposite the upper face 5a of the actuating electrode 5 (or 5 '). Referring to Figures 5, 6 and 7, according to the invention, the lower face 3a of the membrane is not flat over its entire surface, but is profiled outside the electrostatic actuation zone 31a, so when the membrane 3 is at rest, and for all the positions in translation of the diaphragm 3 at rest between the extreme positions along the translation axis) (X 'and along the axis of translation YY ', the lower face 3a of the membrane 3, at any point on its surface facing the upper face 5a of the actuating electrode 5 (or 5') [that is to say, with reference in the example of FIGS. 5, 6 and 7, at any point of the zone Z] is spaced from the upper face 5a of the actuating electrode 5 or 5 'by a distance D, measured perpendicular to the substrate 2, which is greater than or equal to the minimum distance at rest DR The minimum distance at rest Dr referred to above thus corresponds to the minimum value This characteristic makes it possible to reduce the risks of short-circuiting by contact between the membrane 3 and the upper face 5a of an underlying actuating electrode 5 (or 5 '), as well as the risks. forming electric arcs between the membrane 3 and the upper face 5a of an underlying actuating electrode 5 (or 5 ').

More particularly, in the particular example of FIGS. 2, 4, 5 and 6, the electrostatic actuation zone 31a of the lower face 3a of the membrane 3 is extended laterally in the direction X2 (respectively Y2) by a portion groove 32a (respectively 32c), which extends longitudinally substantially perpendicular to the axis of translation) (X '(respectively YY') .This groove portion 32a (respectively 32c) comprises a recess 320a (respectively 3031096 27 320c ), which extends the electrostatic actuation zone 31a in the direction X2, and which is oriented in the opposite direction to the substrate 2. In another variant, this recess 320a (respectively 320c) could be extended in the direction X2 by a wall plane, the membrane 3 having in this case no groove 32a (respectively 32c), but a single recess 320a (respectively 320c) oriented in the opposite direction to the substrate 2. When the membrane rane 3 is at rest, the longitudinal edge 321a (respectively 321c) of this groove portion 32a (respectively 32c), opposite to the electrostatic actuation zone 31a, is spaced from the upper face 5a of the electrode 5 (or 5 '), with a distance D1 (respectively D3) (cf. Figures 5 and 6) measured perpendicular to the substrate 2, which is greater than or equal to, and preferably greater than said minimum distance at rest DR.

More particularly, in the non-shifted translational position of the membrane 3 illustrated in FIG. 2, the portion 50a (50c) of the wafer 50 of the electrode 5 (or 5 ') is positioned below this portion of throat 32a (32c). In the variant of FIGS. 2, 4, 5 and 6, this groove portion 32a (respectively 32c) is extended laterally in the direction X2 (respectively Y2) by a plane wall 33a (respectively 33c), which is situated in a plane P2, parallel to the substrate 2, and farther from the substrate 2 than the abovementioned reference plane P1. Thus, when the membrane 3 is at rest, the distance D '(FIG. 2), measured perpendicular to the substrate 2, between said plane wall 33a (respectively 33c) and the plane of the upper face 5a of the actuation electrode 5 or (5 ') is greater than the minimum distance at rest DR. More particularly, in the particular example of FIGS. 2, 4 and 5, 6, the electrostatic actuation zone 31a of the lower face 3a of the membrane is also extended laterally, in the direction X1 (respectively Y1) by a portion groove 32b (respectively 32d) 3031096 28 extends longitudinally substantially perpendicular to the axis of translation) (X '(respectively YY') .This groove portion 32b (respectively 32d) comprises a recess 320b (respectively 320d), which extends the electrostatic actuation zone 31a in the direction X1, and which is oriented in the opposite direction to the substrate 2. In another variant, this recess 320b (respectively 320d) could be extended in the direction X1 by a flat wall , the membrane 3 having in this case no groove 32b (respectively 32d), but a single recess 320b (respectively 320d) oriented 10 in the opposite direction to the substrate 2. When the membrane 3 is at rest, the longitudinal edge 321b (respectively 321d) of this groove portion 32b (respectively 32d), opposite the electrostatic actuation zone 31a, is spaced from the upper face 5a of the electrode 5 (or 5 '), a distance D2 (respectively D4) measured perpendicular to the substrate 2, which is greater than or equal to, and preferably greater than said minimum distance at rest DR. The distance D2 (respectively D4) is not necessarily equal to the distance D1 (respectively D3). More particularly, in the non-shifted translation position 20 of the membrane 3 illustrated in FIG. 2, the portion 50b (50d) of the wafer 50 of the electrode 5 (or 5 ') is positioned below this portion of 32b throat (32d). In the variant of FIGS. 2, 4, 5, 6 this groove portion 32b (respectively 32d) is extended laterally, in the direction X1 (respectively Y1) by a plane wall 33b (respectively 33d), which is situated in a plane P3, parallel to the substrate 2, and farther away from the substrate 2 than the abovementioned reference plane P1. Thus, when the membrane 3 is at rest, the distance D "(FIG. 2), measured perpendicular to the substrate 2, between said plane wall 33b (respectively 33d) and the plane of the upper face 5a of the operating electrode 5 or (5 ') is greater than the minimum distance at rest DR.

The plane P3 is not necessarily coplanar with the plane P2, the above-mentioned distances D 'and D "being different.In the variants of FIGS. 2, 4, 5 and 6, the four groove portions 32a, 32b, 32c and 32d form a groove 32, which surrounds the electrode 5 over its entire periphery, and which in the non-shifted translation position of the diaphragm 3 of FIG. 2, is positioned relative to the electrode 5 (or ') underlying such that the wafer 50 of the electrode 5 (or 5') is positioned below the groove 32 over the entire periphery of the electrode 5 (or 5 ').

In the particular embodiment of FIGS. 2, 4, 5 and 6, the particular profile of the lower face 3a of the membrane 3 has been obtained by depositing on the substrate 2, during manufacture of the device 1, a dielectric layer 7. of uniform thickness and greater than the thickness of the electrode 5 (or 5 '). At the end of the manufacturing process, this dielectric layer 7 is formed on the upper face 2a of the substrate 2, so that: - it covers at least the wafer 50 of the electrode 50 in a continuous manner over all its periphery, that is to say the four abovementioned portions 50a, 50b, 50c, 50d slice 50, 20 - and secondly that at least a central zone 51 of the upper face 5a of the electrode of actuation 5 or 5 ', which is located opposite the electrostatic actuation zone 31a of the lower face 3a of the membrane 3, is not covered by this dielectric layer 7 and is bare.

This dielectric layer 7 may consist of any electrically insulating material, and in particular of any polymer having a very low electrical conductivity. Preferably, this dielectric material is selected to have a higher dielectric permittivity than the gas, and in this case air, in the space 6 between the membrane 3 and the electrodes 5 or 5 '. By way of non-limiting example of the invention, the dielectric layer 7 is for example made of silicon nitride.

This dielectric layer 7 contributes to further reducing the risks of short-circuiting by contact between the membrane 3 and an underlying actuating electrode 5 (or 5 '), as well as the risks of arcing between the membrane 3 and an underlying actuating electrode 5 (or 5 '). With reference to FIG. 4, preferably in this variant, the dielectric layer 7 also covers for each portion 50a, 50b, 50c, 50d of wafer 50 of the electrode 5 or 5, an edge zone respectively referenced Z1, Z2, Z3, Z4 of the upper surface 5a of the 5 or 5 'actuating electrode, which edge region Z1, Z2, Z3, Z4 extends from said corresponding wafer portion 50a, 50b, 50c, 50d, respectively on a width L1, L2, L3, L4 limited. In the appended figures, the widths L1, L2, L3 and L4 are identical. In another variant, they may not be identical, the important thing being that there remains preferably at least one central zone 51 of the upper face 5a of the actuating electrode 5 or 5 ', which is positioned in view of the electrostatic actuation zone 31a of the lower face 3a of the membrane, which is bare and is not covered by this dielectric layer 7.

In the variant of FIGS. 1, 2, 4, 5 and 6, since the central zone 51 of the upper face 5a of each actuation electrode 5 or 5 'is not covered by the dielectric layer 7, it follows that the electrostatic actuation of the membrane 3 is not altered in time, which is favorable to the repeatability of the operation of the device 1. This advantageously overcomes the operating problems inherent in the technical solution disclosed in the patent application. International application W02006 / 099945 referred to above, and resulting from the presence of a dielectric layer covering the entire upper face of each actuating electrode.

In the embodiment of FIGS. 1, 2, 4, 5 and 6, the dielectric layer 7 also covers the substrate 2 outside the electrodes 5 and 5 ', which advantageously allows passivation of this substrate 2. In the context of the invention, the region of the upper face 2a of the substrate 2 covered by the dielectric layer 7 may be more or less extended. 2nd variant embodiment - FIGS. 8 to 12 FIGS. 8 to 12 show a second alternative embodiment of a microelectromechanical or nanoelectromechanical device 1B, which is in accordance with the invention, and which differs from the first variant of Figures 1 to 7, mainly by the absence of dielectric layer 7.

Furthermore, in this variant, and with reference to FIG. 8, the groove portion 32a (respectively 32c) of the lower face 3a of the membrane is extended laterally in the direction X2 (respectively Y2), away from the electrostatic actuation zone 31a, by a recess 34a (respectively 34c), which is oriented towards the substrate 2, 15 and which extends to a plane P4, parallel to the substrate 2 and closer to the substrate 2 than the plan P1. This recess 34a (respectively 34c) is extended laterally in the direction X2 (respectively Y2), away from the electrostatic actuation zone 31a, by a flat wall 35a (respectively 35c) in the plane P4.

The groove portion 32b (respectively 32d) of the lower face 3a of the membrane 3 is extended laterally, in the direction X1 (respectively Y1), away from the electrostatic actuation zone 31a, by a recess 34b ( 34d), which is oriented towards the substrate 2, and extends to a plane P5, parallel to the substrate 2 and closer to the substrate 2 than the plane P1. This recess 34b (respectively 34d) is extended laterally, in the direction X1 (respectively Y1), away from the electrostatic actuation zone 31a, by a flat wall 35b (respectively 35d) in the plane P5. The plane P5 is not necessarily coplanar with the plane P4. 3rd variant embodiment - FIGS. 13 to 17 FIGS. 13 to 17 show a third variant of a microelectromechanical or nanoelectromechanical device 1C, which is in accordance with the invention, and which differs from the second variant of Figures 8 to 12, by the profile of the groove 32 and the recesses 34a, 34b, 34c, 34d) of the lower face 3a of the membrane 3. 4th variant embodiment - Figures 18 and 19 It is shown on the FIGS. 18 and 19, a fourth variant embodiment of a microelectromechanical or nanoelectromechanical device 1D, which is in accordance with the invention, and which differs from the first variant of FIGS. 1 to 7 solely by the absence of a layer 10 dielectric 7. Preferably, in the 1st 2nd 3rd and 4th variants aforementioned embodiment, to further reduce the risk of short circuit by contact between the membrane 3 and an actuation electrode 5 (or 5 '), as well as the risks of arcing between the membrane 3 and an underlying actuating electrode 5 (or 5'), the lower face 3a of the membrane is profiled such that, for all the possible positions in translation of the membrane 3 in the actuated state, the distance between each point of the lower face 3a of the membrane 3 and the closest point of the actuation electrode 5 (or 5 ') underlying is greater than or equal to the minimum distance (DA) in the actuated state. Thus, the smallest distance between the lower face 3a of the membrane 3 and the actuation electrode 5 (or 5 '), that is to say not only with respect to the upper face 5a of the electrode 5 (or 5 ') of actuation, but also with respect to the slice 50 of the actuating electrode 5 (or 5'), is never less than said minimum distance DA in the actuated state of the diaphragm . For example, referring to FIGS. 12 and 17, outside the electrostatic actuation zone 31a which defines the minimum distance DA in the actuated state, the smallest distance between the lower face 3a of the membrane 3 and the electrode 5 (or 5 ') is referenced D5 and corresponds, in this position offset in particular translation of the membrane 3, at most 3031096 33 small gap between the recess 34a (respectively 34c) and the wafer 50 of the electrode 5 (or 5 '). This smallest spacing D5 outside the electrostatic actuation zone 31a is less than the minimum distance DA in the actuated state of the membrane. 1st variant - manufacturing method - FIGS. 20 to 28 An example of a method for manufacturing the device 1A of FIG. 2 will now be described with reference to the successive steps of FIGS. 20 to 28. These manufacturing steps are focused on the manufacture of the part of the device 1 at an electrode 5 (or 5 '). The steps 10 corresponding to the manufacture of the pillars 4 and conductive tracks S1 and S1 of the device 1A are also already known, and therefore will not be described. FIG. 20 - Step 1 Starting from a multilayer structure comprising an electrode 5 (or 15 ') deposited in the form of a thin layer on the insulating upper face 2a of the substrate 2. The steps for manufacturing such a multilayer starting structure are otherwise known and are not detailed. Figure 21 - Step 2 A thin layer CL3 of dielectric material is deposited, covering substrate 2 and electrode 5 (or 5 '). This deposition is carried out for example by epitaxial growth. The thickness of this layer CL3 is greater than the thickness of the electrode 5 (or 5 '). More particularly, the thickness of this CL3 layer is substantially uniform. This layer CL3 is intended in part to form the dielectric layer 7 of the final device 1A. Figure 22 - Step 3 A layer of photoresist resin CL4 is deposited on top of the layer CL3 of dielectric material. More particularly, the thickness of this photoresist layer CL4 is substantially uniform. FIGS. 23 and 24 - Step 4 The photoresist resin layer CL4 is illuminated by means of a radiation of wavelength adapted to the resin of the layer CL4, for example ultraviolet radiation, using an M2 mask comprising an annular through opening 02, which is positioned relative to the electrode 5 (or 5 '), so that the wafer 50 of the electrode 5 (or 5') is positioned at the right of this opening O 2. thus forming in the layer CL4 (FIG. 22), in line with this opening 02 of the mask M2, a cured region C2 which surrounds the electrode 5 (or 5 '), and in which, in a manner known per se, For example, the structure of the resin has been modified by radiation. FIG. 25 - Step 5 The upper face of the multilayer structure of FIG. 24 is etched with a solution that is inert with respect to the metal actuation electrode 5 and the cured region C2, but adapted to Chemically etching the unmodified portion of the CL4 layer that was not illuminated in the previous step, as well as the exposed portion of the dielectric layer CL3 that is not protected under the cured region ("cured") C2. The portion of the dielectric layer CL3 which is protected under the cured region C2 is ultimately intended to form the dielectric layer 7. FIG. 26 - Step 6 The upper face of the multilayer structure of the 25, with a suitable solution, so as to chemically attack the cured region C2 of the CL4 layer.

At the end of this step, the cured region C2 of the layer CL4 is removed, and there remains on the substrate 2, the actuation electrode 5 (or 5 ') and the dielectric layer 7. formed of the remaining dielectric material of the layer CL3 and straddling at least on the wafer 50 of the actuation electrode 5 (or 5 ') and on the upper face 2a of the substrate 2.

Figure 27 - Step 7 A CL5 sacrificial resin layer is deposited covering the substrate 2, the electrode 5 (or 5 ') and the remaining portion of the dielectric layer CL3. More particularly, the thickness of this sacrificial resin layer CL5 is substantially uniform. The upper face SUP of this layer of sacrificial resin CL5 5 is not flat and has a non-rectilinear profile corresponding to the profile of the lower face 3a of the membrane 3 of the device 1A. This non-planar upper face SUP of the sacrificial resin layer CL5 thus makes it possible subsequently to form the profile of the lower face 3a of the membrane 3.

More particularly, in this example, the upper face SUP of this layer of sacrificial resin CL5 comprises at the right of the upper face 5a of the electrode 5 (or 5 ') a recess R surrounded on its entire periphery by a beaded B forming a protruding boss, the wafer 50 of the electrode 5 (or 5 ') being positioned in line with this bead B. The recess R subsequently makes it possible to form the electrostatic actuation zone 31a of the lower face 3a of the membrane 3 and the protruding bead B subsequently makes it possible to form the groove 32 in the lower face 3a of the membrane 3. FIG. 28 - Step 8 At least one CL6 layer of electrically conductive material is deposited above the sacrificial layer CL5 Subsequently, to form the membrane 3 Step 9 - Figure 2 The sacrificial layer CL5 is etched by means of a suitable solution, so as to completely remove this sacrificial layer. the CL5, and form the membrane 3 by means of at least the layer of CL6 material, which provides the device 1A of Figure 2, the sacrificial layer CL5 being replaced by the air space 6. 2nd variant A manufacturing method of the device 1B of FIG. 8 will now be described with reference to the successive steps of FIGS. 3031096, 36 to 36. These fabrication steps are focused on the fabrication. of the part of the device 1B at an electrode 5 (or 5 '). Figure 29 - Step 1: We start from a multilayer structure identical to that of Figure 20, and is deposited on the insulating upper face 2a of the substrate 2, a sacrificial layer CL3 'covering the electrode 5 (or 5' ). The thickness of this sacrificial layer CL3 'is greater than the thickness of the electrode 5 (or 5'). More particularly, the thickness of this sacrificial layer CL3 'is substantially uniform.

Figure 30 - Step 2 A layer CL4 'of a photoresist resin is deposited on the sacrificial layer CL3'. More particularly, the thickness of this layer CL4 'is substantially uniform. FIGS. 31 and 32 - Step 3 The photoresist resin layer CL4 'is illuminated by means of wavelength radiation adapted to the resin of the layer CL4', for example ultraviolet radiation, using an M2 mask having an annular through opening 02, which is positioned relative to the electrode 5, so that the wafer 50 of the electrode 5 (or 5 ') is positioned at the right of this opening O2. the layer CL4 ', in line with this opening 02 of the mask M2 (FIG. 30), a cured region C2 of limited width L which surrounds the electrode 5 (or 5') and in which, in a known manner in itself, the structure of the resin has been modified by radiation.

FIG. 33 - Step 4 The upper face of the multilayer structure of FIG. 32 is etched so as to etch the unmodified portion of the layer CL4 'which has not been illuminated in the previous step, as well as the exposed portion of the sacrificial layer CL3 'which is not protected under the cured region C2. Figure 34 - Step 5 The upper face of the multilayer structure of Figure 33 is etched with a solution adapted to etch the remaining portion of the resin layer CL4 '. The remaining portion of the sacrificial layer CL3 'of limited width L is not removed, and is positioned astride the upper face of 5a of the electrode 5 (or 5'), and on the upper face 2a of the substrate covering the wafer 50 of the electrode 5 (or 5 '). The central portion 51 of the electrode 5 (or 5 ') is bare and is not covered by this remaining portion of the sacrificial layer CL3'. FIG. 35 - Step 6 A sacrificial layer CL5 'is deposited above the multilayer structure of FIG. 32. The upper face SUP of this sacrificial resin layer CL5' is not flat and has a non-rectilinear profile corresponding to the profile of the lower face 3a of the membrane 3 of the device 1B. This non-planar upper face SUP of the sacrificial resin layer CL5 'thus makes it possible subsequently to form the profile of the lower face 3a of the membrane 3. More particularly, in this example, the upper face SUP of this layer of sacrificial resin CL5 comprises, in particular at the right of the upper face 5a of the electrode 5 (or 5 '), a recess R surrounded on its entire periphery by a bead B forming a bump in projection, the wafer 50 of the electrode 5 (or 5 ') being positioned in line with this bead B. The recess R subsequently makes it possible to form the electrostatic actuation zone 31a of the lower face 3a of the membrane 3, and the protruding bead 25 B subsequently makes it possible to form the groove 32 in the lower face 3a of the membrane 3. FIG. 36 - Step 7 At least one layer CL 6 'of electrically conductive material is deposited at above the sacrificial layer CL5 ', so as to subsequently form the membrane. Step 8 - FIG. 8 The sacrificial layers CL3' and CL5 'are etched with a suitable solution so as to remove them completely. , and forming the membrane 3 to obtain the device 1B of Figure 8, the sacrificial layers CL3 'and CL5' being replaced by the air space 6. 5th variant - manufacturing method - Figures 37 to 44 An example of manufacturing method of the device 1C of Figure 13 will now be described with reference to the successive steps of Figures 37 to 44. These manufacturing steps are focused on the manufacture of the part of the device 1C at a level of electrode 5 (or 5 ').

Figure 37 - Step 1 starts from a multilayer structure identical to that of Figure 20, and is deposited on the insulating upper face 2a of the substrate 2, and the electrode 5 (or 5 '), a first layer sacrificial CL2 "covering the electrode 5 (or 5 ').

The thickness of this sacrificial layer CL 2 "is greater than the thickness of the electrode 5 (or 5 '). More particularly, the thickness of this sacrificial layer CL 2" is substantially uniform. Figure 38 - Step 2 A second sacrificial layer CL3 "is deposited on the first sacrificial layer CL2". More particularly, the thickness of this second sacrificial layer CL3 "is substantially uniform Figure 39 - Step 3 A layer CL4" of a photoresist resin is deposited on the second sacrificial layer CL3 ". of this layer CL4 "is substantially uniform. FIGS. 40 and 41- Step 4 The photoresist resin layer CL4 "is illuminated by means of wavelength radiation adapted to the resin of the CL4" layer, for example ultraviolet radiation, using an M2 mask. 3031096 39 having a through annular opening 02, which is positioned relative to the electrode 5, so that the wafer 50 of the electrode 5 (or 5 ') is positioned at the right of this opening 02. It thus forms in the CL4 "layer, at the right of this opening 02 of the mask M2 (FIG. 30), a cured region C2 of limited width L which surrounds the electrode 5 (or 5 ') and in which, In a manner known per se, the structure of the resin has been modified by radiation Figure 42 - Step 5 The upper face of the multilayer structure of FIG. 41 is etched so as to etch the unmodified portion of the CL4 layer. "who was not enlightened in the previous step, as well that the exposed portion of the second sacrificial layer CL3 "which is not protected under the cured region C2. Figure 43 - Step 6 The upper face of the multilayer structure of Figure 42 is etched with a solution adapted to remove the remaining portion of the resin layer CL4 "The remaining portion of the sacrificial layer CL3" of width limited L is not removed. The upper face SUP of the multilayer structure of FIG. 43 is not flat and has a non-rectilinear profile corresponding to the profile of the lower face 3a of the membrane 3 of the device 1C. This non-planar upper face SUP of the sacrificial resin layer CL5 thus makes it possible subsequently to form the profile of the lower face 3a of the membrane 3.

More particularly, in this example the upper face SUP of the multilayer structure of FIG. 43 comprises, at the right of the upper face 5a of the electrode 5 (or 5 '), a recess R surrounded on its entire periphery by a bead B forming a protruding boss, the wafer 50 of the electrode 5 (or 5 ') being positioned in line with this bead B. The recess R subsequently makes it possible to form the electrostatic actuation zone 31a of the lower face 3a of the diaphragm 3, and the protruding bead 3031096 B subsequently makes it possible to form the groove 32 in the lower face 3a of the membrane 3. FIG. 44 - Step 7 A layer of material CL6 "is deposited on top of the sacrificial layers CL2" and CL3 in a subsequent manner to form the membrane Step 8 - Figure 13 The sacrificial layers CL2 "and CL3" are etched by means of a suitable solution, so as to remove them completely, and form the membrane 3 to obtain the device 1C of FIG. 13, the sacrificial layers CL2 "and CL3" being replaced by the air space 6. 4th variant - Production method To manufacture the device 1D of FIG. 18, the same steps are carried out 1 8 to 20, which have been previously described to manufacture the device 1A of the first embodiment, and replacing in its steps the dielectric layer CL3 by a sacrificial layer CL3. Then, in a last manufacturing step 9, the two superposed sacrificial layers CL3 and CL5 are etched by means of a suitable solution, so as to completely remove them, and obtain the device 1D of FIG. 18, the two layers sacrificial being replaced by the air space 6. In the particular embodiments which have been described with reference to the appended figures, the switching membrane 3 is flexible and electrostatically deformable under the action of the actuating electrodes 5 or 5 '. In another variant embodiment of the invention, it is possible to produce a microelectromechanical or nanoelectromechanical device whose actuation electrode or electrodes make it possible to move the switching membrane 3 or to modify the orientation of the switching membrane 3 without distorting it. In this case, the membrane 3 may be thicker.

In another variant, the membrane 3 could be anchored or mechanically connected to the substrate 2 or to the pillars 4, by means of one or

Claims (34)

REVENDICATIONS1. Microelectromechanical or nanoelectromechanical device (1A; 1B; 1C; 1D) having a membrane (3), which is supported above a substrate (2), being spaced from said substrate (2) and having a face bottom (3a) facing the substrate (2), and at least one operating electrode (5 or 5 '), which is formed on said substrate (2), which is positioned at least partly below the membrane (3), with an electrically insulating space (6), and in particular an air gap, between the membrane (3) and the actuation electrode (5 or 5 '), which electrostatically actuates the membrane (3), wherein the membrane (3) is translatable along at least a first translation axis (XX ') substantially parallel to the substrate (2), with a displacement stroke along this axis (XX ') which is limited between two extreme positions, in which the lower face (3a) of the membrane comprises an electrostatic actuation zone (31a) which, for at least one non-shifted position of the membrane (3) is fully positioned facing the actuating electrode (5 or 5 '), and which, when the membrane (3) is at rest, is spaced from the upper face (5a) of the actuating electrode 5 or 5 ') by a minimum distance (DR) measured perpendicularly to the substrate (2), wherein the lower face (3a) of said membrane (3) ) is not entirely flat and is profiled outside the electrostatic actuation zone (31a), so that when the membrane (3) is at rest, and for all the translational positions of the membrane (3), ) at rest between said extreme positions along this first translation axis (XX '), the lower face (3a) of the membrane (3), at any point on its surface facing the upper face (5a) of the the actuating electrode (5 or 5 ') is spaced from the upper face (5a) of the actuating electrode (5 or 5') by a distance (D) perpendicular to the substrate (2), which is greater than or equal to said minimum rest distance (DR). 5 2. Device according to claim 1, wherein when the membrane (3) is actuated, said electrostatic actuation zone (31a) is spaced from the upper face (5a) of the actuating electrode (5 or 5 '). a minimum distance (DA) measured perpendicular to the substrate (2), and in which the lower face (3) of the membrane is profiled so that when the membrane (3) is actuated and for all positions in translation of the membrane (3) in the actuated state, between the two extreme positions along said first axis (XX '), the distance between each point of the lower face (3a) of the membrane (3) and the 15 closest point to the actuating electrode (5 or 5 ') is greater than or equal to the minimum distance (DA) in the actuated state. 3. Device according to any one of claims 1 or 2, wherein the membrane (3) being movable in translation along said first axis of translation (XX ') in a first direction (X1), the actuating zone electrostatic (31a) of the lower face (3a) of the membrane (3) is extended laterally, in a second direction (X2) opposite the first direction (X1), by a first recess (320a) oriented in the opposite direction 25 to the substrate (2). 4. Device according to claim 3, wherein the electrostatic actuation zone (31a) of the lower face (3a) of the membrane (3) is extended laterally in the second direction (X2) by a first portion of throat (32a), a part of which forms said first recess (320a), which extends longitudinally substantially perpendicular to the first translation axis (XX '), and whose longitudinal edge (321a) opposite the electrostatic actuation zone (31a) is spaced from the upper face (5a) of the actuating electrode (5 or 5 '), when the membrane is at rest, by a distance (D1) measured perpendicular to the substrate (2), which is greater than or equal to, and preferably greater than, said minimum rest distance (DR). The device (1A, 1D) according to claim 4, wherein the first groove portion (32a) extends laterally in the second direction (X2) by a first planar wall (33a) substantially parallel to the substrate (2). The device (1A; 1D) according to claim 5, wherein when the membrane (3) is at rest, the distance (D '), measured perpendicular to the substrate (2), between said first planar wall (33a) and the plane of the upper face (5a) of the actuating electrode (5 or 5 ') is greater than the minimum distance at rest (DR). 20 7. Device (1B, 1C) according to claim 4, wherein the first groove portion (32a) extends laterally in the second direction (X2) by a recess (34a) oriented towards the substrate (2). 25 8. Device according to any one of claims 3 to 7, wherein, in the unshifted position of the membrane (3), a first portion (50a) of the wafer (50) of the actuating electrode (5). or 5 ') is positioned below the first recess (320a), and optionally below the first groove portion (32a). 3031096 45 9. Device according to any one of claims 3 to 8, wherein the membrane (3) being movable in translation along said first translation axis (XX ') in the second direction (X2), the electrostatic actuation zone. (31a) of the lower face (3a) 5 of the membrane (3) is extended laterally in the first direction (X1) by a second recess (320b) oriented in the opposite direction to the substrate (2). 10. Device according to claim 9, wherein the electrostatic actuating zone (31a) of the lower face (3a) of the membrane (3) is extended laterally in the first direction (X1) by a second portion of throat (32b), a part of which forms said second recess (320b), which extends longitudinally substantially perpendicular to the first translation axis (XX '), and whose longitudinal edge (321b) is opposite to the electrostatic actuation zone ( 31a) is spaced from the upper face (5a) of the actuation electrode (5 or 5 '), when the membrane is at rest, by a distance (D2) measured perpendicular to the substrate (2), which is greater or equal to, and preferably greater than, said minimum idle distance (DR). 11. Device (1A, 1D) according to claim 10, wherein the second groove portion (32b) extends laterally in the first direction (X1) by a second plane wall (33b) substantially parallel to the substrate (2). 12. Device (1A, 1D) according to claim 11, wherein when the membrane (3) is at rest, the distance (D "), measured perpendicular to the substrate (2), between said second plane wall (33a) and the plane of the upper face (5a) of the actuating electrode (5 or 5 ') is greater than the minimum distance at rest (DR) 3031096 46 13. Device (1B; 1C) according to claim 10, wherein the second groove portion (32b) extends laterally in the first direction (X1) by a recess (34b) oriented towards the substrate (2). 14. Device according to any one of claims 9 to 13, wherein, in the unshifted position of the membrane (3), a second portion (50b) of the wafer (50) of the actuating electrode 10 ( 5 or 5 ') is positioned below the second recess (320b), and optionally below the second groove portion (32b). 15 15. Device (1A, 1B, 1C, 1D) according to any preceding claim, wherein the membrane (3) is movable in translation along a second translation axis (YY ') substantially parallel to the substrate ( 2) and perpendicular to the first translation axis (XX '), with a displacement stroke along said second axis (YY') which is limited between two extreme positions, in which the lower face (3a) of the membrane ( 3) is profiled outside the electrostatic actuation zone (31a), so that, when the membrane (3) is at rest, and for all the positions in translation of the membrane (3) at rest 25 between said extreme positions along this second axis of translation (YY '), the lower face (3a) of the membrane (3), at any point on its surface facing the upper face (5a) of the actuation electrode (5 or 5 '), is spaced from the upper face (5a) of the actuating electrode 5 or 5') of a 30 istance (D) measured perpendicular to the substrate (2), which is greater than or equal to said minimum distance at rest (DR). 3031096 47 16. Device according to claim 15, wherein when the membrane (3) is actuated, said electrostatic actuation zone (31a) is spaced from the upper face (5a) of the actuating electrode (5 or 5 '). a minimum distance (DA) measured perpendicular to the substrate (2), and in which the lower face (3) of the membrane is profiled such that when the membrane (3) is actuated and for all positions in translation of the membrane (3) in the actuated state, between the two extreme positions along said second axis (YY '), the distance 10 between each point of the lower face (3a) of the membrane (3) and the closest point to the actuating electrode (5 or 5 ') is greater than or equal to the minimum distance (DA) in the actuated state. 17. Device according to any one of claims 15 or 16, wherein the membrane (3) being movable in translation along said second translation axis (YY ') in a third direction (Y1), the operating zone electrostatic (31a) of the lower face (3a) of the membrane (3) is extended laterally, in a fourth direction (Y2) opposite the third direction (Y1), 20 by a third recess (320c) oriented in the opposite direction to the substrate (2). 18. Device according to claim 17, wherein the electrostatic actuation zone (31a) of the lower face (3a) of the membrane (3) is extended laterally in the fourth direction (Y2) by a third portion of the throat. 32c), a part of which forms said third recess (320c), which extends longitudinally substantially perpendicularly to the second translation axis (YY '), and whose longitudinal edge (321c) opposite the electrostatic actuation zone ( 31a) is spaced from the upper face (5a) of the actuation electrode (5 or 5 '), when the membrane is at rest, by a distance (D3) measured perpendicular to the substrate (2), which is greater than or equal to, and preferably greater than, said minimum rest distance (DR). 5 19. Device (1A, 1D) according to claim 18, wherein the third groove portion (32c) extends laterally in the fourth direction (Y2) by a third plane wall (33c) substantially parallel to the substrate (2). 10 20. Device (1A, 1D) according to claim 19, wherein when the membrane (3) is at rest, the distance (D '), measured perpendicular to the substrate (2), between said third plane wall (33c) and the plane of the upper face (5a) of the actuating electrode (5 or 5 ') is greater than the minimum distance at rest (DR). 21. The device (1B; 1C) according to claim 18, wherein the third groove portion (32c) extends laterally in the fourth direction (Y2) by a recess (34c) facing the substrate (2). 22. Device according to any one of claims 15 to 21, wherein, in the unshifted position of the membrane (3), a third portion (50c) of the wafer (50) of the actuation electrode 25 ( 5 or 5 ') is positioned below the third recess (320c), and optionally below the third groove portion (32c). 23. Device according to any one of claims 15 to 22, wherein the membrane (3) being movable in translation along said second translation axis (YY ') in the fourth direction (Y2), the operating zone electrostatic (31a) of the lower face 3031096 49 (3a) of the membrane (3) is extended laterally in the third direction (Y1) by a fourth recess (320d) oriented in the opposite direction to the substrate (2). 5 24. Device according to claim 23, wherein the electrostatic actuation zone (31a) of the lower face (3a) of the membrane (3) is extended laterally in the third direction (Y1) by a fourth groove portion ( 32d), a part of which forms said fourth recess (320d), which extends longitudinally substantially perpendicularly to the second translation axis (YY '), and whose longitudinal edge (321b) opposite the electrostatic actuation zone (31a) ) is spaced from the upper face (5a) of the actuating electrode (5 or 5 '), when the membrane is at rest, by a distance (D4) measured perpendicular to the substrate (2), which is greater or equal to, and preferably greater than, said minimum distance at rest (DR). 25. Device (1A; 1D) according to claim 24, wherein the fourth groove portion (32d) extends laterally in the third direction (Y1) by a fourth plane wall (33d) substantially parallel to the substrate (2). 26. Device (1A, 1D) according to claim 25, wherein when the membrane (3) is at rest, the distance (D "), measured perpendicular to the substrate (2), between said fourth plane wall (33d) and the plane of the upper face (5a) of the actuating electrode (5 or 5 ') is greater than the minimum distance at rest (DR). 27. Device (1B, 1C) according to claim 24, wherein the fourth groove portion (32d) extends laterally in the third direction (Y1) by a recess (34d) facing the substrate (2). 28. Device according to any one of claims 23 to 27, wherein, in the non-shifted position of the membrane (3), a fourth portion (50d) of the wafer (50) of the actuation electrode ( 5 or 5 ') is positioned below the fourth recess (320d), and optionally below the fourth groove portion (32d). 10 29. Device according to any one of claims 24 to 28, wherein the four groove portions (32a, 32b, 32c and 32d) form a groove (32), which surrounds the actuating electrode (5 or 5 ' ) on its entire periphery. 15 30. Device according to claim 29, wherein in the non-shifted position of the membrane (3), the groove (32) is positioned relative to the underlying actuating electrode (5 or 5 '), such as so that the wafer (50) of the electrode 5 (or 5 ') is positioned below the groove (32) over the entire periphery of the actuating electrode (5 or 5'). 31. Device according to any one of the preceding claims, wherein at least the lower face (3a) of the membrane (3) is electrically conductive. 32. Device according to any one of the preceding claims, wherein the membrane (3) is flexible, and wherein the one or more actuating electrodes (5; 5 ') can electrostatically deform the membrane (3). 25 30 33. Apparatus according to any one of the preceding claims, wherein the actuating electrode (5 or 5 ') is positioned 3031096 completely below the membrane (3). 34. Device according to any one of the preceding claims, in which the electrically conductive upper face (3a) of the actuating electrode (5 or 5 ') is exposed on at least one electrically conductive central part (51). or on its entire surface. 1035. 36. 37. The device according to claim 34, wherein the membrane (3), when actuated by means of an actuating electrode (5 or 5 ') subdivided into it does not touch the electrically conductive surface of the actuating electrode (5 or 5 '). Apparatus according to any one of the preceding claims, wherein the electrostatic actuation zone (31a) is substantially planar. Apparatus according to any one of the preceding claims, wherein the electrostatic actuation zone (31a) is centered with respect to the actuating electrode (5 or 5 ') when the membrane is in its non-shifted position. Device according to any one of the preceding claims, in which for at least part of the translational offset positions of the diaphragm (3), the electrostatic actuation zone (31a) is only partly positioned facing the electrode of the diaphragm (3). underlying actuation (5 or 5 '). Device according to any one of the preceding claims, in which the dimension (Lx) of the electrostatic actuation zone (31a), measured parallel to said first translation axis (XX '), is smaller than the dimension (LEx) of the underlying actuating electrode (5 or 5 ') measured parallel to said first translation axis (XX'). 40. Device according to claim 39, wherein the dimension (Ly) 5 of the electrostatic actuation zone (31a), measured parallel to said second translation axis (YY '), is smaller than the dimension (LEy) of the underlying actuating electrode (5 or 5 ') measured parallel to said second translation axis (YY'). 41. Device according to any one of the preceding claims, wherein the membrane (3) makes it possible to perform an ohmic, capacitive or optical switch function. 42. A device according to any one of the preceding claims, comprising two electrically conductive tracks (S1 and S2) formed on the substrate (2) and spaced from one another, and wherein the membrane (3) fills a ohmic switch function, so that when it is at rest, it is not in contact with the two tracks (S1 and S2), and that when it is electrostatically actuated by means of at least one electrode actuator (5), it is in contact with the two tracks (S1 and S2) and electrically connects the two tracks (S1 and S2) between them. 43. Microelectromechanical or nanoelectromechanical system comprising at least one microelectromechanical or nanoelectromechanical device according to any one of the preceding claims. 44. A method of manufacturing a microelectromechanical or nanoelectromechanical device (1A) as claimed in any one of claims 1 to 42, and comprising at least the following steps: (1) a multilayer structure having a substrate is produced; (2) on which is formed at least one actuating electrode (5 or 5 '), (2) depositing a dielectric layer (CL3) covering the substrate (2) and the electrode (5 or 5'), ( 3) depositing a layer of photoresist resin (CL4), above the dielectric layer (CL3), (4) etching of the multilayer structure so as to preserve on the substrate (2), the electrode (5 or 5 ') and a portion of the dielectric layer (CL3) straddling at least on the wafer (50) of the actuating electrode (5 or 5') and on the upper face (2a). ) substrate (2), (5) is deposited a sacrificial layer (CL5), which covers the substrate (2), the electrode (5 or 5 ') and the res of the dielectric layer (CL3), the upper face of this sacrificial layer (CL5), not being flat and having a non-rectilinear profile corresponding to the profile of the lower face (3a) of the membrane (3) of the device (1A), (6) depositing at least one layer of material (CL6) above the sacrificial layer (CL5), and more particularly a layer (CL6) of electrically conductive material, (7) etching is carried out of the sacrificial layer (CL5) so as to remove it, and form the membrane (3) by means of at least the layer of material (CL6) deposited in the preceding step. 45. The method according to claim 44, wherein the upper face (SUP) of the sacrificial layer (CL5) comprises a recess 3031096 54 (R) which is positioned in line with the upper face (5a) of the electrode (5). or 5 '), and which is surrounded over its entire periphery by a bead (B) forming a protruding boss, the wafer (50) of the electrode (5 or 5') being positioned at the right of this bead (B) , Said recess (R) subsequently making it possible to form the electrostatic actuation zone (31a) of the lower face (3a) of the membrane (3) of the device (1A), and the said protruding bead (B) subsequently making it possible to forming the groove (32) in the lower face (3a) of the membrane (3) of the device (1A). 46. A method of manufacturing a microelectromechanical or nanoelectromechanical device (1D or 1B) according to any one of claims 1 to 42, and comprising at least the following steps: (1) a multilayer structure comprising a substrate is manufactured (2) on which is formed at least one actuating electrode (5 or 5 '), (2) depositing a first sacrificial layer (CL3 or CL3'), covering the substrate (2) and the electrode (5 or 5 '), (3) depositing a layer of photoresist resin (CL4 or CL4'), above the dielectric layer (CL3 or CL3 '), (4) etching the multilayer structure so as to storing on the substrate (2), the actuation electrode (5 or 5 ') and a portion of the first sacrificial layer (CL3 or CL3') straddling at least on the wafer (50) of the electrode actuation (5 or 5 ') and on the substrate (2), (5) a second sacrificial layer (CL5 or CL5') is deposited, which covers the substrate (2), the electrode (5 or 5 ') and the remaining portion of the first sacrificial layer (CL3 or CL3'), the upper face (SUP) of this second sacrificial layer (CL5 or CL5 '), n' being not flat and having a non-rectilinear profile 3031096 55 corresponding to the profile of the lower face (3a) of the membrane (3) of the device (1D or 1B), (6) at least one layer of material (CL6 or CL6) is deposited ') above the second sacrificial layer (CL5), and more particularly a layer (CL6 or CL6') of electrically conductive material, (7) etching the remaining part of the first sacrificial layer (CL3 or CL3) ') and the second sacrificial layer (CL5 or CL5') so as to remove them, and form the membrane (3) by means of at least the layer of material (CL6 or CL6 ') deposited in the previous step. 47. The method of claim 46, wherein the upper face (SUP) of the second sacrificial layer (CL5 or CL5 ') comprises a recess (R) which is positioned in line with the upper face (5a) of the electrode (5 or 5 '), and which is surrounded on its periphery by a bead (B) forming a protruding boss, the edge (50) of the electrode (5 or 5') being positioned at the right of this bead ( B), said recess (R) subsequently making it possible to form the electrostatic actuation zone (31a) of the lower face (3a) of the membrane (3) of the device (1D or 1B), and said bead (B) in projection subsequently making it possible to form the groove (32) in the lower face (3a) of the membrane (3) of the device (1D or 1B). 48. A method of manufacturing a microelectromechanical or nanoelectromechanical device (1C) according to any one of claims 1 to 42, and comprising at least the following steps: (1) a multilayer structure having a substrate (2 ) on which at least one actuating electrode 3031096 56 (5 or 5 ') is formed, (2) a first sacrificial layer (CL2 ") covering the electrode (5 or 5') is deposited, (3) on the first sacrificial layer (CL2 "), a second sacrificial layer (CL3"), (4) a layer (CL4 ") of a photoresist resin is deposited on the second sacrificial layer (CL3"), ( 5) an etching of the multilayer structure resulting from the preceding step is carried out so as to keep on the substrate (2) the actuation electrode (5 or 5 '), the first sacrificial layer (CL2 ") and a part of the second sacrificial layer (CL3 "), so that the upper face (SUP) of the stru layer is not flat and has a non-rectilinear profile corresponding to the profile of the lower face (3a) 15 of the membrane (3) of the device (1C), (6) a layer of material (CL6 ") is deposited at above the first sacrificial layer (CL2 ") and the remaining portion of the second sacrificial layer (CL3"), and more particularly a layer (CL6 ") of electrically conductive material, (7) the first sacrificial layer is etched (7) (CL2 ") and the remaining portion of the second sacrificial layer (CL3"), so as to remove them completely, and form the membrane (3). 49. The method of claim 48, wherein the upper face (SUP) of the multilayer structure obtained in step (5) comprises a recess (R) which is positioned at the right of the upper face (5a) of the electrode (5 or 5 '), and which is surrounded over its entire periphery by a bead (B) forming a protruding boss, the wafer (50) of the electrode (5 or 5') being positioned at the right of this 30 bead (B), said recess (R) subsequently making it possible to form the electrostatic actuation zone (31a) of the lower face (3a) of the membrane (3) of the device (1D or 1B), and said bead ( B) projecting subsequently to form the groove (32) in the lower face (3a) of the membrane (3) of the device (1C). 5