METHOD FOR MANUFACTURING A LOW-NOISE ELECTROACUSTIC TRANSCURTAINER

Description

DOMAINE TECHNIQUE DE L'INVENTION

[0001] The technical field of the invention is that of microelectromechanical systems (MEMS) or nanoelectromechanical systems (NEMS). More particularly, the invention relates to a method for manufacturing an electroacoustic transducer comprising a device for transmitting motion and force between two sealed areas. Such an electroacoustic transducer can be used as a microphone or loudspeaker. ARRIERE-PLAN TECHNOLOGIQUE DE L'INVENTION

[0002] Microelectromechanical or nanoelectromechanical microphones represent a rapidly expanding market, particularly due to the development of mobile devices, such as tablets, smartphones and other connected objects, in which they are gradually replacing electret microphones.

[0003] Microphones measure a rapid change in atmospheric pressure, also called acoustic pressure. They therefore have at least one part in contact with the outside.

[0004] Most MEMS or NEMS microphones currently manufactured are capacitive sensing microphones. Patent publication FR3059659A1 describes an example of a capacitive sensing microphone comprising a moving element, capacitive sensing means, and a device for transmitting motion between the moving element and the capacitive sensing means. Publication EP2541222A1 discloses a similar device.

[0005] The moving element is capable of detecting pressure variations. It can be formed by a rigid piston comprising a diaphragm, also called a thin film, and a diaphragm stiffening structure. The diaphragm forms a separation between a cavity open to the external environment and a rear volume of the microphone, also called the reference volume because it contains a reference pressure. Thus, one face of the diaphragm is subjected to the reference pressure, and the opposite face is subjected to atmospheric pressure (whose variation we wish to detect). The moving element is connected to the motion transmission device in a first section of the microphone.

[0006] Capacitive sensing devices measure piston displacement and thus the pressure variation. They are located in a second zone, hermetically sealed from the first zone. Each device comprises a moving electrode and at least one fixed electrode positioned opposite the moving electrode. The electrodes form the plates of a capacitor whose capacitance varies with piston displacement. The second zone is a chamber under controlled atmosphere (typically vacuum) to minimize viscous friction and associated noise.

[0007] The transmission device comprises at least one first transmission arm extending into the first zone and at least one second transmission arm extending into the second zone. The piston is coupled to one end of the first transmission arm, while the moving electrode of the capacitive sensing means is coupled to one end of the second transmission arm. The first and second transmission arms are connected at their second ends by means of a pivot joint. This pivot joint allows rotation of the transmission arms relative to the microphone frame and simultaneously ensures a seal between the first and second zones.

[0008] Such a capacitive detection microphone can suffer from a damping phenomenon, called "squeeze-film damping," caused by air being pinched between the piston and the frame during microphone operation. This damping phenomenon generates mechanical noise and leads to a decrease in microphone performance. RESUME DE L'INVENTION

[0009] More generally, there is a need to manufacture an electroacoustic transducer in which the damping phenomenon is reduced, the electroacoustic transducer comprising: a frame; a movable element relative to the frame, the movable element comprising a membrane and a membrane stiffening structure; a first transmission arm, the movable element being coupled to one end of the first transmission arm.

[0010] This need is satisfied by the present invention as defined in the method for manufacturing an electroacoustic transducer according to claim 1 and the electroacoustic transducer according to claim 15.

[0011] According to a first aspect of the invention, this need is met by providing a manufacturing process comprising the following steps: form a first sacrificial layer on a substrate, the first sacrificial layer comprising a first portion and a second portion adjacent to the first portion, the first portion having a thickness greater than that of the second portion; form a first structural layer on the first sacrificial layer, thus obtaining a multilayer stack; form a second sacrificial layer on the first structural layer in a first zone of the multilayer stack; form a second structural layer on the first structural layer in a second zone of the multilayer stack and on the second sacrificial layer in the first zone of the multilayer stack, the second zone being adjacent to the first zone;etch the second structural layer so as to expose the second sacrificial layer in the first zone and to delimit the periphery of the stiffening structure of the moving element in the second zone; etch the first structural layer in the second zone up to the first sacrificial layer so as to delimit the membrane of the moving element; etch the second sacrificial layer so as to expose one face of the membrane; etch the substrate so as to delimit the first transmission arm, the first sacrificial layer serving as a stop layer for the etching of the substrate; and etch the first sacrificial layer so as to expose a second opposite face of the membrane and detach the membrane from the substrate; the first portion of the first sacrificial layer being configured so that, after the etching step of the first structural layer, a peripheral portion of the membrane is positioned opposite the first portion of the first sacrificial layer.

[0012] The first portion of the first sacrificial layer, which is thicker than the second portion, increases the distance between the substrate and the membrane of the moving element, thus reducing the damping phenomenon (called "squeeze-film damping" in English) responsible for mechanical noise.

[0013] In a preferred embodiment, the manufacturing process includes, before the step of forming the first sacrificial layer, a step of etching a cavity in the substrate, the first portion of the first sacrificial layer completely filling the cavity and the second portion of the first sacrificial layer being disposed outside the cavity.

[0014] The first structural layer can be formed by epitaxy on the first sacrificial layer.

[0015] Alternatively, the first structural layer formation step includes the following sub-steps: provide a transfer substrate including the first structural layer; glue the first structural layer of the transfer substrate to the first sacrificial layer; thin the transfer substrate until reaching the first structural layer.

[0016] The manufacturing process may further include, after the etching step of the second sacrificial layer and before the etching step of the substrate, the following steps: place a cover on the second structural layer, thus forming an assembly; and turn the assembly over.

[0017] In addition to the characteristics mentioned in the preceding paragraphs, the manufacturing process according to the invention may have one or more additional characteristics from among the following, considered individually or in all technically possible combinations: The cavity has a depth greater than or equal to 1 µm, preferably between 2 µm and 10 µm; the cavity has an annular shape and preferably extends over more than 80% of the periphery of the moving element; the cavity and the membrane have the same shape, preferably rectangular; the cavity has dimensions greater than or equal to the dimensions of the membrane; the first structural layer and the second structural layer are etched simultaneously so as to delimit the membrane and the stiffening structure of the moving element; the substrate is silicon, the first sacrificial layer is silicon oxide and the first structural layer is silicon; the second sacrificial layer is silicon oxide; the first structural layer has a thickness between 100 nm and 10 µm; the stiffening structure of the moving element rests at least partially on the membrane;and the stiffening structure of the moving element is in contact with the membrane.

[0018] A second aspect of the invention relates to an electroacoustic transducer comprising: a frame; a movable element relative to the frame, the movable element comprising a membrane and a membrane stiffening structure; a first transmission arm, the movable element being coupled to one end of the first transmission arm; the membrane being formed by a first part of a first structural layer, the stiffening structure being formed by a first part of a second structural layer disposed on the first structural layer and the frame comprising a substrate, a second part of the first structural layer and a second part of the second structural layer, and wherein the distance between the substrate and a peripheral portion of the membrane is greater than the distance between the substrate and the second part of the first structural layer, said distances being measured perpendicular to faces of the membrane.

[0019] In a preferred embodiment of the electroacoustic transducer, the substrate includes a cavity arranged opposite the peripheral portion of the membrane.

[0020] The transducer may further include a device for transmitting motion and force between a first zone and a second zone with a controlled atmosphere, the first and second zones being hermetically sealed from each other, the transmission device comprising, in addition to the first transmission arm extending into the first zone, a second transmission arm extending into the second zone.

[0021] The invention and its various applications will be better understood by reading the following description and examining the accompanying figures. BREVE DESCRIPTION DES FIGURES

[0022] Other features and advantages of the invention will become clear from the description given below, which is by way of example and not limitation, with reference to the accompanying figures, including: [ Fig. 1 ] schematically and partially represents an example of an electroacoustic transducer comprising a piston connected to two initial transmission arms; [Fig. 2A] à [Fig. 2H] represent, in cross-section, the stages of a manufacturing process for the electroacoustic transducer according to the figure 1 ; [Fig. 3A] à [Fig. 3J] represent, in cross-section, the steps of a manufacturing process for an electroacoustic transducer according to a preferred embodiment of the invention, including the steps of figures 2B à 2H ; And [ Fig. 4 ] represents an implementation variant of the step of the figure 3C .

[0023] For clarity, identical or similar elements are identified by identical reference symbols across all figures. DESCRIPTION DETAILLEE

[0024] There figure 1 shows an example of an electroacoustic transducer 1, of the capacitive sensing microphone type.

[0025] The electroacoustic transducer 1 comprises a frame (not shown) delimiting at least partially a first zone 11 and a second zone 12, a movable element 13 relative to the frame, and a transmission device 14 for movement between the first zone 11 and the second zone 12. The first and second zones 11-12 of the electroacoustic transducer 1 are subjected to different pressures. They are hermetically sealed from each other.

[0026] The moving element 13, hereinafter referred to as the piston, is in contact with the first zone 11. It comprises a diaphragm 131 and a diaphragm stiffening structure 132, also called a skeleton or frame. The diaphragm 131 of the piston 13 serves to collect a pressure difference between its two faces across its entire surface, thereby deducing a variation in atmospheric pressure.

[0027] The diaphragm 131 of the piston 13 partially delimits a closed volume, called the reference volume, where a reference pressure prevails. It separates this reference volume from a cavity open to the external environment, in this case, air. One face of the diaphragm 131 is therefore subjected to the reference pressure, and the opposite face of the diaphragm 131 is subjected to atmospheric pressure (the variation of which we wish to detect in the case of a microphone).

[0028] The first zone 11 may correspond to the cavity open to the external environment, and therefore be subject to atmospheric pressure. Alternatively, the first zone 11 may be the reference volume subjected to the reference pressure.

[0029] Furthermore, in this particular example, the electroacoustic transducer 1 includes capacitive sensing means 15 arranged in the second zone 12. These capacitive sensing means 15 allow the displacement of the piston 13 to be measured, and therefore the pressure difference between its two faces. They preferably include a movable electrode 151 and at least one fixed electrode positioned opposite the movable electrode 151. The electrodes form the plates of a capacitor whose capacitance varies according to the displacement of the piston 13.

[0030] The second zone 12 is a chamber under a controlled atmosphere to reduce viscous friction and associated acoustic noise. By "chamber under a controlled atmosphere," we mean a chamber under reduced pressure, typically less than 10 mbar, and preferably under vacuum.

[0031] The transmission device 14 is mounted to rotate flexibly relative to the frame by means of one or more pivot joints 16. The transmission device 14 comprises at least one first transmission arm 141 extending into the first zone 11, at least one second transmission arm 142 extending into the second zone 12, and at least one transmission shaft 143 extending partly into the first zone 11 and partly into the second zone 12. In the example of the figure 1 , the transmission device 14 comprises two first transmission arms 141, two second transmission arms 142 and two transmission shafts 143, each transmission shaft 143 connecting a first transmission arm 141 to a second transmission arm 142.

[0032] Each first transmission arm 141 comprises a first end coupled to the piston 13 and a second opposite end coupled to the associated transmission shaft 143. Each second transmission arm 142 comprises a first end coupled to the moving electrode 151 of the capacitive sensing means 15 and a second opposite end coupled to the associated transmission shaft 143.

[0033] The transmission shaft 143, for example, is in the shape of a straight cylinder. The transmission arms 141-142 preferably have the shape of a rectangular beam, with a dimension (length) significantly greater than the others (width and thickness). The piston 13, for example, has a rectangular shape. The first transmission arms 141 preferably extend perpendicularly to one side of the piston 13, advantageously a longer side. They can be anchored inside the periphery of the piston 13, as shown by the figure 1 , for example by means of a first end of cylindrical shape.

[0034] Each pivot joint 16 preferably comprises a watertight insulation element 161, through which a transmission shaft 143 passes, and two aligned blades 162 extending between the transmission shaft 143 and the frame. The watertight insulation element 161 is, for example, in the form of a sealing membrane. It provides a seal between the first zone 11 and the second zone 12 at the pivot joint 16. The blades 162 are dimensioned to allow torsional deformation and permit rotation of the transmission device 14. They are preferably arranged diametrically opposite each other with respect to the transmission shaft 143. Preferably, the watertight insulation element 161 is also designed to deform under the rotational movement of the transmission device 14.

[0035] The frame may include a support (formed by a first substrate), a structural layer (for example made of silicon) placed on the support and a cover attached to the structural layer (for example formed by a second substrate).

[0036] The stiffening structure 132 of the diaphragm of the piston 13 advantageously includes at its periphery a rim which extends in a direction perpendicular to the diaphragm 131. This rim increases the path of the air around the piston and reduces leakage between the external environment and the closed reference volume.

[0037] THE figures 2A à 2H These figures represent steps S1 to S8 of a manufacturing process for the electroacoustic transducer 1. They show how the piston 13 can be formed and separated from the frame. Therefore, only a portion of the electroacoustic transducer, near the periphery of the piston 13, is shown. For the sake of simplicity, reference will be made only to a single first transmission arm 141, a single pivot joint 16, a single sealing membrane 161, etc., it being understood that all elements of the same type can be formed simultaneously.

[0038] The first step S1 represented by the figure 2A consists of providing a stack of layers 20 serving as the starting material for the realization of the transducer. The stack 20 successively comprises a substrate 21, a first sacrificial layer 22 and a first structural layer 23, also called a "thin layer".

[0039] The substrate 21 is used in particular to fabricate the first transmission arm 141 and part of the frame (the support). It initially has a thickness that can range from 500 µm to 700 µm. The substrate 21 can be made of a semiconductor material, for example silicon.

[0040] The first structural layer 23 is used to form the diaphragm 131 of the piston 13. It can also be used to form the sealing diaphragm 161 of the pivot joint 16 and / or the moving electrode 151 of the capacitive sensing means 15. It has a thickness less than that of the substrate 21, preferably between 100 nm and 10 µm, for example equal to 1 µm. It is preferably made of the same material as the substrate, for example silicon.

[0041] The first sacrificial layer 22 is intended to be partially removed during the transducer fabrication process. This layer is particularly useful for defining the first transmission arm 141. It can also serve as the lower air gap in the capacitive sensing region of the transducer. It can also be used to mechanically bond the substrate 21 and the first structural layer 23. The first sacrificial layer 22 can be made of a dielectric material, preferably a silicon nitride or a silicon oxide, for example silicon dioxide (SiO₂). Its thickness is, for example, between 100 nm and 10 µm.

[0042] The 20 stack can notably be a multilayer structure of the silicon on insulator or SOI type ("Silicon On Insulator" in English), commonly called SOI substrate.

[0043] Although not shown in the figures, the manufacturing process may then include an etching step of the first structural layer 23. This etching step of the first structural layer 23 may in particular be used to form release holes for the moving electrode 151 (to allow the etching solution of the first sacrificial layer 22 to subsequently infiltrate).

[0044] At stage S2 of the figure 2B A second sacrificial layer 24 is formed on the first structural layer 23 in a first zone 20A of the stack 20. To achieve this, the second sacrificial layer 24 can first be deposited so as to completely cover the first structural layer 23 and then partially etched into a second zone 20B of the stack 20, for example, through a resin mask formed by photolithography. The first and second zones 20A-20B of the stack 20 are adjacent. The etching of the second sacrificial layer 24 is preferably selective with respect to the first structural layer 23. The second sacrificial layer 24 is advantageously formed from the same dielectric material as the first sacrificial layer 22, for example, a silicon oxide. Its thickness can be between 100 nm and 10 µm.

[0045] The second sacrificial layer 24 can serve as the upper air gap for capacitive sensing. Etching the second sacrificial layer 24 can lead to the etching of the first sacrificial layer 22, where the first structural layer 23 has previously been etched (not shown in the figures).

[0046] At stage S3 of the figure 2C A second structural layer 25 is deposited on the first structural layer 23 (in the second zone 20B of the stack 20) ​​and on the second sacrificial layer 24 (in the first zone 20A of the stack 20), for example by epitaxy. The second structural layer 25 is intended to form one or more (structural) elements of the transducer, in particular the stiffening structure 132 of the piston 13. It is advantageously made of the same material as the first structural layer 23, for example silicon. The thickness of the second structural layer 25 is preferably between 5 µm and 50 µm, for example 20 µm.

[0047] Then, during an S4 step represented by the figure 2D The second structural layer 25 is etched to define the contours of the stiffening structure 132 (piston contouring) and to lighten the piston 13. In the first zone 20A of the stack 20, the second sacrificial layer 24 (e.g., silicon oxide) acts as a stop layer for the etching of the second structural layer 25 (e.g., silicon), thus preserving the underlying first structural layer 23 (e.g., silicon). The etching of the second structural layer 25 is therefore selective with respect to the second sacrificial layer 24. Conversely, in the second zone 20B of the stack 20, the etching of the second structural layer 25 to define the periphery (or outer contour) of the stiffening structure 132 leads to the first structural layer 23.Since the etching of the second structural layer 25 is not selective with respect to the first structural layer 23 (but only with respect to the first sacrificial layer 22), the first structural layer 23 is etched at the same time as the second structural layer 25 up to the first sacrificial layer 22.

[0048] Thus, at the bottom of the trench corresponding to the periphery of the stiffening structure 132, the first structural layer 23 was engraved and the first sacrificial layer 22 is exposed.

[0049] At the end of step S4, the first structural layer 23 comprises a first part 23a and a second part 23b, separated from each other. The first part 23a of the first structural layer 23 (on the left of the figure 2D ) is intended to form the membrane 131 of the piston 13. It is covered by the second sacrificial layer 24 and by a portion decoupled from the second structural layer 25 forming the stiffening structure 132 of the piston 13.

[0050] The engraving technique used in step S4 of the figure 2D is advantageously deep reactive ion etching (or DRIE, for "Deep Reactive Ion Etching" in English).

[0051] With reference to the figure 2E The manufacturing process then includes a step S5 of etching the second sacrificial layer 24 so as to (partially) expose the first part 23a of the first structural layer 23 (in other words, so as to expose a first face of the membrane 131). This step S5 can be described as the first step in releasing the piston 13.

[0052] The etching of the second sacrificial layer 24 is preferably an isotropic etching selective with respect to the substrate 21, the first structural layer 23 and the second structural layer 25. The second sacrificial layer 24 is preferably etched chemically, for example by immersing the stack in a bath of hydrofluoric acid (HF) in liquid or vapor phase (in the case of a silicon oxide layer) for a controlled time.

[0053] On the other hand, a portion of the first sacrificial layer 22 located in line with the periphery of the stiffening structure 132 is etched at the same time as the second sacrificial layer 24, which forms a cavity 22' in the first sacrificial layer 22. The etching can be controlled in time so that this cavity 22' is small.

[0054] The etching of the sacrificial layers 22 and 24 can also be used to release the mobile electrode 151 from the capacitive sensing means 15 (before it is enclosed in the chamber under controlled atmosphere).

[0055] Although not shown in the figures, the manufacturing process may subsequently include a step S6 of transferring a cover onto the second structural layer 25, thus forming the chamber under controlled atmosphere, i.e., the second zone 12. The cover can be made by machining a silicon substrate. It can, in particular, be attached to the second structural layer 25 by direct bonding (e.g., Si-Si) or by eutectic sealing (e.g., Au-Si or Al-Ge).

[0056] Then, at stage S6' of the figure 2F , the assembly formed by the stack of layers 20 and the hood (not shown) is turned over, to facilitate the subsequent etching of the substrate 21. After this turning over, the substrate 21 is advantageously thinned, for example by DRIE etching, grinding and / or chemical mechanical polishing (CMP), preferably until a thickness of between 30 µm and 300 µm is reached, i.e. the desired thickness for the first transmission arm 141.

[0057] Stage S7 of the figure 2G consists of etching the substrate 21 (possibly thinned) down to the first sacrificial layer 22 in order to create access to the piston 13 and to delimit, in an area not shown on the figure 2G , the first transmission arm. The substrate etching is preferably selective with respect to the first sacrificial layer 22. The substrate 21 can be etched by DRIE.

[0058] As illustrated on the figure 2G , the engraving of the substrate 21 to create access to the rear face of the piston 13 can be inscribed inside the periphery of the piston 13, so as not to open onto the cavity 22' formed in step S5 (cf. Fig. 2E ) by the partial (and unintentional) etching of the first sacrificial layer 22. Thus, the etching in step S7 does not extend to the piston 13, which includes the first part 23a of the first structural layer 23 (membrane 131) and the detached portion of the second structural layer 25 (stiffening structure 132). Inside the periphery of the piston 13, the first sacrificial layer 22 (e.g., silicon oxide) acts as a barrier to the etching of the substrate 21 (e.g., silicon), thereby preserving the first part 23a of the underlying first structural layer 23 (e.g., silicon). This prevents significant air leaks between the cavity open to the external environment (and therefore subject to atmospheric pressure) and the reference volume (subject to the reference pressure), also called the back volume, on either side of the membrane 131.

[0059] Finally, at step S8 (cf. Fig. 2H The first sacrificial layer 22 is etched to expose the first part 23a of the first structural layer 23 (in other words, to expose a second opposite face of the membrane 131) and detach it from the substrate 21. Following step S8, the piston 13 is free to move. Step S8 can therefore be described as the second step in the release of the piston 13.

[0060] The etching of the first sacrificial layer 22 is preferably an isotropic etching selective with respect to the substrate 21, the first structural layer 23 and the second structural layer 25. The first sacrificial layer 22 is preferably etched chemically, for example by immersing the assembly in a bath of hydrofluoric acid (HF) in liquid or vapor phase (in the case of a silicon oxide layer) for a controlled time.

[0061] During the movement of the piston 13, air is pinched between the membrane 131 and the remaining portion of the substrate 21, and this over almost the entire periphery of the piston 13 (cf. Fig.2H This air pinch is due to the fact that the etching of substrate 21 is located inside the periphery of piston 13. It causes a damping phenomenon, called "squeeze-film damping," which generates mechanical noise and leads to a decrease in transducer performance. The strength of this damping phenomenon is inversely proportional to the cube of the distance (called the "gap") between piston 13 and the frame (here, the remaining portion of substrate 21).

[0062] THE figures 3A à 3J illustrate modifications to the manufacturing process, allowing for a considerable reduction in the damping phenomenon.

[0063] Steps S11 to S13, illustrated by the figures 3A à 3C aim to recreate the multilayer stack 20 comprising the substrate 21, the first sacrificial layer 22 and the first structural layer 23, but further equipping it with a cavity 210 arranged on the surface of the substrate 21 and filled by the first sacrificial layer 22.

[0064] Thus, with reference to the figure 3A The manufacturing process begins with a step S11 of etching the cavity 210 into the substrate 21. The substrate 21 is preferably a bulk semiconductor substrate, for example, silicon. The cavity 210 is located in a region 21A of the substrate 21 dedicated to the formation of the piston 13. It preferably has a depth P greater than or equal to 1 µm, preferably between 2 µm and 10 µm. The depth P of the cavity 210 is measured from the initial surface of the substrate 21, hereinafter referred to as the reference surface R.

[0065] Then, at stage S12 of the figure 3B , the first sacrificial layer 22 is formed on the substrate 21. The first sacrificial layer 22 here comprises a first portion 22a completely filling the cavity 210 and a second portion 22b adjacent to the first portion 22a and disposed outside the cavity 210. The first portion 22a has a greater thickness than the second portion 22b.

[0066] The first sacrificial layer 22 advantageously has a flat surface. The thickness of the first portion 22a is then equal to the sum of the depth P of the cavity 210 and the thickness of the second portion 22b. The thickness of the second portion 22b of the first sacrificial layer 22 is, for example, between 100 nm and 10 µm.

[0067] The first sacrificial layer 22 is advantageously deposited so as to completely fill the cavity 210 and to form an excess thickness on the reference surface R of the substrate 21. For example, the first sacrificial layer 22 is formed by plasma-enhanced chemical vapor deposition (PECVD) followed by annealing. This deposition operation is advantageously followed by a planarization operation, for example by chemical mechanical polishing (CMP), to obtain the flat surface.

[0068] At stage S13 of the figure 3C The first structural layer 23 is formed on the first sacrificial layer 22. The first structural layer 23 (which, as a reminder, is intended to form the diaphragm 131 of the piston 13) covers at least the first portion 22a of the first sacrificial layer 22. It is preferably deposited so as to cover the first and second portions 22a-22b of the first sacrificial layer 22, as illustrated in the figure 3C .

[0069] The first structural layer 23 can be formed at step S13 by silicon epitaxy on the first sacrificial layer 22, typically in SiO2. The first structural layer 23 is then in polycrystalline silicon.

[0070] In one implementation variant of step S13 represented by the figure 4 , the first structural layer 23 initially belongs to a transfer substrate 40. It is transferred onto the first sacrificial layer 22 using a direct bonding technique.

[0071] The transfer substrate 40 is bonded to the substrate 21 covered with the first sacrificial layer 22 by bringing the first structural layer 23 into contact with the first sacrificial layer 22. Then, the transfer substrate 40 is thinned down to the first structural layer 23. The thinning of the transfer substrate 40 can be accomplished by etching or CMP.

[0072] The carrier substrate 40 is preferably a multilayer SOI-type structure comprising successively a support layer 41 (typically silicon), a buried oxide layer 42 (typically SiO2) and a thin film of monocrystalline silicon forming the first structural layer 23. The carrier substrate 40 is thinned by successively removing the support layer 41 and the buried oxide layer 42.

[0073] This implementation variant allows the formation of a first structural layer 23 in monocrystalline silicon, rather than in polycrystalline silicon.

[0074] With reference to figures 3D à 3J The manufacturing process then includes steps S2 to S8 described previously, in relation to the figures 2B à 2H .

[0075] There figure 3D represents step S2 of the formation of the second sacrificial layer 24 on the first structural layer 23. The second sacrificial layer 24 is formed only in the region 21A dedicated to the formation of the piston 13, preferably in the manner described in relation to the figure 2B (full plate deposit + engraving).

[0076] There figure 3E represents step S3 of the formation of the second structural layer 25 (intended to form the stiffening structure 132 of the piston 13), for example by epitaxy. As described previously in relation to the figure 2C , the second structural layer 25 covers both the second sacrificial layer 24 (in region 21A) and the first structural layer 23.

[0077] Before step S3 of formation of the second structural layer 25, the manufacturing process may include a so-called opening step of the first structural layer 23 and the second sacrificial layer 24. This opening step consists of forming a well 30 which extends through the first structural layer 23 and the first sacrificial layer 22 to the substrate 21. This well 30 is formed by successively etching a portion of the first structural layer 23 and a portion of the first sacrificial layer 22, preferably through the second sacrificial layer 24. The etched portions are located in region 21A (cf. Fig.3D The well 30 allows the material of the second structural layer 25 (for example silicon) to grow from the substrate 21 to the first structural layer 23, thus forming a pillar which passes through the first sacrificial layer 22. This pillar will ensure the connection between the diaphragm 131 of the piston 13 (formed in the first structural layer 23) and the first transmission arm 141 (formed in the substrate 21).

[0078] There figure 3F represents the S4 step of etching the second structural layer 25, after its deposition on the first structural layer 23 and the second sacrificial layer 24. The second structural layer 25 is etched so as to expose the second sacrificial layer 24 (piston lightening) and to delimit the stiffening structure 132.

[0079] As described previously with reference to the figure 2D The first structural layer 23 can be etched during the same step S4 to delimit the membrane 131 (this is generally the case when the first structural layer 23 and the second structural layer 25 are formed of the same material). The first structural layer 23 is thus separated into two portions: a first portion 23a forming the diaphragm 131 of the piston and a second portion 23b belonging to the transducer frame.

[0080] An alternative is to etch the second structural layer 25 and the first structural layer 23 separately, using different etching chemistries, when the two layers are formed of different materials.

[0081] The engraving of the first structural layer 23 leads advantageously to the first portion 22a of the first sacrificial layer 22.

[0082] Between the S3 step of forming the second structural layer 25 and the S4 step of etching the second structural layer 25, the manufacturing process may also include steps of deposition and structuring of a germanium layer 26 on the second sacrificial layer 25. The germanium layer 26 is preferably formed at the periphery of the first region 21A.

[0083] At stage S5 of the figure 3G , the second sacrificial layer 24 is engraved in the manner described with reference to the figure 2E (selective etching with respect to substrate 21, the first structural layer 23, and the second structural layer 25), until it is completely removed. Simultaneously, a portion of the first sacrificial layer 22 is etched, starting from the bottom of the trench corresponding to the periphery of membrane 131, thus forming cavity 22'. Cavity 22' extends at least partially beneath membrane 131.

[0084] More specifically, the engraving of the second sacrificial layer 24 can extend to the first portion 22a of the first sacrificial layer 22, as illustrated by the figure 3G .

[0085] There figure 3H represents the S6 step of transferring the hood 27 onto the second structural layer 25, to form the chamber under controlled atmosphere and, in this example, the rear volume BV of the microphone.

[0086] The hood 27 can be covered with an aluminum layer 28 which is brought into contact with the germanium layer 26 in order to achieve a eutectic aluminum-germanium bond.

[0087] There figure 3I represents step S7 of etching substrate 21 (after step S6' of flipping substrate 21 followed by possible thinning of substrate 21), allowing access to the opposite face of piston 13 and defining the first transmission arm 141. As on the figure 2G The substrate 21 is etched using the first sacrificial layer 22 as a stop layer. The etching of the substrate 21 is located inside the periphery of the piston 13, so as not to open into the cavity 22'.

[0088] With reference to the figure 3J The manufacturing process concludes with step S8, which involves etching the first sacrificial layer 22 (and more specifically its first portion 22a), as described in relation to the figure 2H (second stage of piston release). The cavity 210 of the substrate 21 then becomes completely free again.

[0089] Air pinching, a source of damping and therefore noise, occurs between a peripheral portion of the membrane 131 and the substrate 21. To reduce this damping, the cavity 210 formed in step S11 within the substrate 21 is configured so that the first portion 22a of the first sacrificial layer 22 is located opposite the peripheral portion of the membrane 131 (the portion overlapping the substrate 21). "Appropriately positioned" means that the peripheral portion of the membrane 131 and the first portion 22a extend at least partially opposite each other (in other words, there is partial or total overlap). The peripheral portion of the membrane 131 is preferably covered by the stiffening structure 132. The width of the peripheral portion is, for example, between 2 µm and 15 µm.

[0090] Such a configuration makes it possible to obtain (after the S8 step of etching the second sacrificial layer 22; cf. Fig.3J ) a larger gap G between the diaphragm 131 of the piston 13 and the remaining part of the substrate 21, compared to the situation where the first sacrificial layer 22 is of constant thickness ( Fig.2H ). The damping phenomenon, due to air pinching, is therefore considerably reduced.

[0091] In other words, the substrate 21 and the peripheral portion of the membrane 131 are separated by a distance (gap G) greater than the distance between the substrate 21 and the second part 23b of the first structural layer 23. These distances are measured perpendicular to the faces of the membrane 131.

[0092] The formation of the cavity 210 results in a local thickening of the first sacrificial layer 22. Using a first sacrificial layer 22 of greater thickness in the region 21a of the substrate dedicated to the formation of the piston 13 (or at least opposite the peripheral portion of the membrane 131) is a more advantageous solution than increasing the thickness of the entire first sacrificial layer 22, because it does not impact the manufacture of the other parts of the microphone (which use this first sacrificial layer 22), such as the capacitive sensing means.

[0093] The cavity 210 of the substrate 21 (and therefore the first portion 22a of the first sacrificial layer 22) may have an annular shape. It is then dimensioned so as to extend relative to the peripheral portion of the membrane 131. The annular cavity 210 preferably extends over more than 80% of the periphery of the piston 13, advantageously over the entire periphery of the piston 13.

[0094] Alternatively, cavity 210 can have the same shape as membrane 131, for example a rectangular shape (cf. Fig.1 ).

[0095] To minimize damping, the dimensions of cavity 210 (external dimensions in the case of an annular cavity 210) are advantageously greater than or equal to those of membrane 131.

[0096] The first sacrificial layer 22, comprising the first and second portions 22a-22b (the first portion 22a being thicker than the second portion 22b), can be formed without first etching a cavity in the substrate, in other words, starting from a substrate 21 with a flat surface. For example, the (sacrificial) material that constitutes the first sacrificial layer 22 is first deposited as a layer of constant thickness, then a portion of this layer is thinned, preferably by etching, to form the second portion 22b (the unthinned portion of the layer forming the first portion 22a).

[0097] The manufacturing process for the electroacoustic transducer according to the invention has been described using a capacitive sensing microphone as an example ( Fig.1 ), one face of which is subjected to atmospheric pressure and the other face to a reference pressure. The manufacturing process described in relation to the figures 3A-3J , however, is applicable to other types of microphones and other types of electroacoustic transducers, including loudspeakers (sound emitters) or ultrasonic emitters.

[0098] More generally, a microphone in the second zone 12 (controlled atmosphere chamber) includes means for measuring the movement of the transmission device and / or a force applied to the transmission device. These measuring means include, for example, a vibrating beam (resonant detection microphone).

[0099] In the case of a loudspeaker or ultrasonic transmitter, an actuator (for example capacitive) replaces the measuring means in the second zone 22. The actuator sets in motion the first end of the second transmission arm 142. This movement is transmitted by the transmission device 14 to the piston 13 attached to the first end of the first transmission arm 141. The movement of the diaphragm 131 of the piston 13 allows to emit a sound (or ultrasound).

Claims

1. A method for manufacturing an electroacoustic transducer (1) comprising: - a frame; - an element (13) movable relative to the frame, the movable element (13) comprising a membrane (131) and a membrane rigidifying structure (132); - a first transmission arm (141), the movable element (13) being coupled to one end of the first transmission arm (141); method comprising the following steps of: - forming (S12) a first sacrificial layer (22) on a substrate (21), the first sacrificial layer comprising a first portion (22a) and a second portion (22b) adjacent to the first portion (22a), the first portion (22a) having a greater thickness than the second portion (22b); - forming (S13) a first structural layer (23) on the first sacrificial layer (22), thus obtaining a stack (20) of layers; - forming (S2) a second sacrificial layer (24) on the first structural layer (23) in a first region (20A) of the stack (20); - forming (S3) a second structural layer (25) on the first structural layer (23) in a second region (20B) of the stack (20) and on the second sacrificial layer (24) in the first region (20A) of the stack (20), the second region (20B) being adjacent to the first region (20A); - etching (S4) the second structural layer (25) so as to expose the second sacrificial layer (24) in the first region (20A) and to delimit the periphery of the rigidifying structure (132) of the movable element (13) in the second region (20B); - etching (S4) in the second region (20B) the first structural layer (23) up to the first sacrificial layer (22) so as to delimit the membrane (131) of the movable element (13); - etching (S5) the second sacrificial layer (24) so as to expose a first face of the membrane (131); - etching (S7) the substrate (21) so as to delimit the first transmission arm (141), the first sacrificial layer (22) acting as a stop layer to etching of the substrate; and - etching (S8) the first sacrificial layer (22) so as to expose a second, opposite face of the membrane (131) and to detach the membrane (131) from the substrate (21) Method wherein the first portion (22a) of the first sacrificial layer (22) is configured to, after etching (S4) of the first structural layer (23), a peripheral portion of the membrane (131) is arranged facing the first portion (22a) of the first sacrificial layer (22).

2. The method according to claim 1, comprising, prior to the step (S12) of forming the first sacrificial layer (22), a step (S11) of etching a cavity (210) in the substrate (21), the first portion (22a) of the first sacrificial layer (22) completely filling the cavity (210), and the second portion (22b) of the first sacrificial layer (22) being arranged outside the cavity (210).

3. The method according to claim 2, wherein the cavity (210) has a depth (P) greater than or equal to 1 µm, preferably between 2 µm and 10 µm.

4. The method according to any of claims 2 and 3, wherein the cavity (210) has an annular shape.

5. The method according to claim 4, wherein the cavity (210) extends over more than 80% of the periphery of the movable element (13).

6. The method according to any of claims 2 and 3, wherein the cavity (210) and the membrane (131) are of the same, preferably rectangular, shape.

7. The method according to any of claims 2 to 6, wherein the cavity (210) has dimensions greater than or equal to the dimensions of the membrane (131).

8. The method according to any of claims 1 to 7, wherein the first structural layer (23) and the second structural layer (25) are simultaneously etched so as to delimit the membrane (131) and the rigidifying structure (132) of the movable element (13).

9. The method according to any of claims 1 to 8, further comprising, after the step (S5) of etching the second sacrificial layer (24) and before the step (S7) of etching the substrate (21), the following steps of: - arranging (S6) a cap (27) on the second structural layer (25), thereby forming an assembly; and - turning over (S6') the assembly.

10. The method according to any of claims 1 to 9, wherein the substrate (21) is of silicon, the first sacrificial layer (22) is of silicon oxide, and the first structural layer (23) is of silicon.

11. The method according to any of claims 1 to 10, wherein the second sacrificial layer (24) is of silicon oxide.

12. The method according to any of claims 1 to 11, wherein the first structural layer (23) is formed by epitaxy on the first sacrificial layer (22).

13. The method according to any of claims 1 to 11, wherein the step of forming the first structural layer (23) comprises the following substeps of: - providing a transfer substrate (40) comprising the first structural layer (23); - bonding the first structural layer (23) of the transfer substrate (40) to the first sacrificial layer (22); - thinning the transfer substrate (40) until the first structural layer (23) is reached.

14. The method according to claim 13, wherein the transfer substrate is a multilayer silicon-on-insulator type structure.

15. An electroacoustic transducer (1) comprising: - a frame; - an element (13) movable relative to the frame, the movable element (13) comprising a membrane (131) and a membrane rigidifying structure (132); - a first transmission arm (141), the movable element (13) being coupled to one end of the first transmission arm (141); the membrane (131) being formed by a first part (23a) of a first structural layer (23), the rigidifying structure (132) being formed by a first part of a second structural layer (25) arranged on the first structural layer (23), and the frame comprising a substrate (21), a second part (23b) of the first structural layer (23) and a second part of the second structural layer (25), characterized in that the distance (G) between the substrate (21) and a peripheral portion of the membrane (131) is greater than the distance between the substrate (21) and the second part (23b) of the first structural layer (23), said distances being measures perpendicularly to faces of the membrane (131).

16. The electroacoustic transducer (1) according to claim 15, wherein the substrate (21) comprises a cavity (210) arranged facing the peripheral portion of the membrane (131).