JP2011182299A - Mems transducer and method for manufacturing the same - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To reduce manufacturing cost by enhancing the flexibility of design of a Q-value and sound pressure in an ultrasonic conversion device configured as an MEMS. <P>SOLUTION: An ultrasonic conversion device configured of an MEMS is provided with an ultrasonic element for generating ultrasonic vibration, a barrier formed integrally with the ultrasonic element, a first substrate formed with holes with the barrier as a bottom, and a closing member jointed to the face with the openings of the holes of the first substrate and formed, with recessed parts or projecting parts in a region where the openings of the holes are closed. <P>COPYRIGHT: (C)2011,JPO&INPIT

Description

  The present invention relates to a MEMS transducer (Micro Electro Mechanical Systems) and a manufacturing method thereof.

  Conventionally, an ultrasonic transducer is known as a MEMS transducer that generates sound with high directivity by vibrating a diaphragm at two frequencies in the ultrasonic range. A structure for preventing backward sound leakage is provided behind the diaphragm of the ultrasonic transducer (see Patent Document 1). The volume and shape of the cavity formed between such a structure and the diaphragm affect the magnitude and Q value of the sound pressure generated by the ultrasonic transducer.

JP 2008-20429 A

  In such a MEMS transducer, the design freedom of the cavity formed behind the diaphragm is governed by the thickness of the wafer. Specifically, for example, when a silicon wafer is anisotropically etched to form a straight hole with the diaphragm as the bottom, and the cavity is formed by closing the opening of the straight hole with a package substrate, the volume of the cavity Is substantially determined by the thickness of the wafer and the opening area of the straight hole. However, the options for wafer thickness supplied to the market are limited. In particular, when the Q value becomes excessive due to the thickness of the wafer, there arises a problem that the efficiency decreases due to nonlinear absorption of sound waves, and the sound pressures of ultrasonic waves having different frequencies cannot be made uniform.

  Further, in the MEMS transducer, the cavity behind the diaphragm is configured by closing the opening of the hole in a process after dicing. However, in the process after dicing, the process cost per product increases and the manufacturing tolerance also increases.

  An object of the present invention is to increase the design freedom of the Q value and sound pressure in a MEMS transducer and reduce the manufacturing cost.

(1) A MEMS transducer for achieving the object includes a substrate on which a non-through hole having an opening on a main surface, a piezoelectric conversion unit coupled to a bottom portion of the non-through hole, and the main surface of the substrate And a closing member having a recess or a protrusion formed in a region closing the opening.
According to the present invention, the volume and shape of the cavity behind the diaphragm can be adjusted by the concave portion or the convex portion formed in the closing member. Therefore, according to the present invention, the degree of freedom in designing the cavity behind the diaphragm can be increased, thereby increasing the degree of freedom in designing the Q value and the sound pressure.

(2) In the MEMS transducer for achieving the above object, the closing member includes an inorganic substrate bonded to the substrate and a photosensitive resin film bonded to the inorganic substrate, and is formed from the photosensitive resin film. The convex portion may be formed in a region closing the opening.
In this case, since the convex portion made of the photosensitive resin film can be formed by forming the photosensitive resin film in the wafer process, exposure and development, a cavity having a predetermined shape and volume can be accurately formed at a low cost behind the diaphragm. Can do.

(3) A method for manufacturing a MEMS transducer for achieving the object includes a step of preparing a first substrate, a step of forming a non-through hole in a main surface of the first substrate, and the main substrate of the first substrate. Forming a piezoelectric conversion portion on the non-through hole on the back surface of the surface, preparing a second substrate, forming a concave or convex portion on the main surface of the second substrate, and the non-through hole Joining the main surface of the first substrate and the main surface of the second substrate so that the bottom of the substrate and the concave portion or the convex portion face each other.
According to the present invention, the degree of freedom in designing the cavity behind the diaphragm can be increased, thereby increasing the degree of freedom in designing the Q value and the sound pressure.

(4) In the method of manufacturing a MEMS transducer for achieving the above object, the second substrate is an inorganic substrate, and the concave portion or the convex portion is formed by etching the inorganic substrate using photolithography. May be.
In this case, the volume and shape of the cavity behind the diaphragm can be accurately formed.

(5) In the method of manufacturing a MEMS transducer for achieving the above object, the second substrate is an inorganic substrate, a photosensitive resin film is formed on the inorganic substrate, and the photosensitive resin film is exposed and developed. By doing so, the convex portion made of the photosensitive resin film may be formed.
In this case, the volume and shape of the cavity behind the diaphragm can be accurately formed.
In addition, the execution order of each process described in the claims is not limited to the description order, and may be performed in reverse order as long as there is no technical obstruction factor, or a plurality of processes may be performed simultaneously.

1A is a plan view according to the first embodiment of the present invention, and FIG. 1B is a cross-sectional view taken along line 1B-1B shown in FIG. 1A. It is sectional drawing concerning 1st embodiment of this invention. It is sectional drawing concerning 1st embodiment of this invention. It is sectional drawing concerning 1st embodiment of this invention. It is sectional drawing concerning 1st embodiment of this invention. It is sectional drawing concerning 1st embodiment of this invention. It is sectional drawing concerning 1st embodiment of this invention. It is sectional drawing concerning 1st embodiment of this invention. It is sectional drawing concerning 2nd embodiment of this invention. It is sectional drawing concerning 3rd embodiment of this invention.

  Hereinafter, embodiments of the present invention will be described in the following order with reference to the accompanying drawings. In addition, the same code | symbol is attached | subjected to the corresponding component in each figure, and the overlapping description is abbreviate | omitted.

1. First embodiment (Configuration)
FIG. 1 shows a parametric speaker 1 as a first embodiment of a MEMS transducer according to the present invention. The parametric speaker 1 is a MEMS transducer manufactured using a semiconductor manufacturing process. The parametric speaker 1 is housed in a case (not shown) and is a super-directional speaker incorporated in a voice guide device, AV equipment, IT equipment, etc. in a facility such as a museum. Used. The parametric speaker 1 includes a second substrate 160, a first substrate 10, and a piezoelectric element 11 formed on the first substrate 10 as a piezoelectric conversion unit. When an ultrasonic wave having two frequency components is transmitted by driving a plurality of diaphragms 12 formed on the first substrate 10 by a piezoelectric element 11, a sound having an extremely audible range is generated by a parametric array effect. be able to.

  The plurality of diaphragms 12 are portions constituting the bottoms of the plurality of holes 10 a formed in the first substrate 10. The first substrate 10 includes a relatively thick single crystal silicon layer 101, a relatively thin single crystal silicon layer 103, and an insulating layer 102 sandwiched therebetween. The thickness of the thick single crystal silicon layer 101 is, for example, 100 to 2000 μm, the thickness of the thin single crystal silicon layer 103 is, for example, 0.5 to 10 μm, and the insulating layer 102 is, for example, silicon dioxide having a thickness of 0.05 to 10 μm. To do. The diaphragm 12 is made of a thin single crystal silicon layer 103 and has a thickness of 0.5 to 100 μm. The hole 10 a penetrates the thick single crystal silicon layer 101 and the insulating layer 102. The hole 10a is a circular hole having a cross section of 10 to 10,000 μm in diameter and a straight hole having a depth of 100 to 2000 μm. The vibration end of the diaphragm 12 is determined by the cross-sectional shape of the hole 10a.

  A piezoelectric element 11 as a piezoelectric conversion unit is bonded to the surface of the thin single crystal silicon layer 103 of the first substrate 10 on each diaphragm 12. The piezoelectric element 11 drives each of the plurality of diaphragms 12 in the same phase. The piezoelectric element 11 includes a lower electrode layer 111, an upper electrode layer 113, and a piezoelectric layer 112 sandwiched therebetween. The piezoelectric layer 112 has a thickness of 0.1 to 100 μm and is made of lead zirconate titanate (PZT) or the like. The lower electrode layer 111 and the upper electrode layer 113 each have a thickness of 0.05 to 10 μm, for example, and are made of platinum, gold, or the like.

  Lower electrode layer 111 is formed on the entire surface of thin single crystal silicon layer 103. The piezoelectric layer 112 and the upper electrode layer 113 are formed with through holes 11 b for connecting a conductive wire to the lower electrode layer 111. A through hole 11 c is formed in the piezoelectric layer 112 and the upper electrode layer 113 at a portion overlapping the central portion of the diaphragm 12. Since the first substrate 10 has a sufficient thickness at portions other than the diaphragm 12, it behaves substantially as a rigid body. Therefore, only the part of the diaphragm 12 of the first substrate 10 is bent integrally with the piezoelectric element 11. Further, only the portion of the piezoelectric element 11 that overlaps the diaphragm 12 is bent when a driving voltage is applied to the upper electrode layer 113 and the lower electrode layer 111. When the piezoelectric layer 112 contracts in the in-plane direction, each of the diaphragms 12 is deformed so that the central portion rises toward the piezoelectric element 11 side. When the piezoelectric layer 112 expands in the in-plane direction, each of the diaphragms 12 is deformed so that the center portion rises toward the hole 10a. Therefore, when ultrasonic vibration is generated in the piezoelectric element 11, the ultrasonic wave is transmitted from the diaphragm 12 that vibrates integrally with the piezoelectric element 11.

  The main surface of the second substrate 160 is joined to the surface (main surface) in which the opening of the hole 10a of the first substrate 10 is formed, and the opening of the hole 10a is airtightly closed. The second substrate 160 and the first substrate 10 may be joined via an adhesive layer 150 made of polyimide or an epoxy resin material, or may be joined directly. The second substrate 160 is made of an inorganic material such as single crystal silicon, glass, ceramics.

  Convex portions 161 are respectively formed in regions that block the openings of the plurality of holes 10 a on the main surface of the second substrate 160. The convex portion 161 protrudes inside the hole 10a and narrows the cavity C formed by the hole 10a. By adjusting the shape of the convex portion 161, the volume and shape of the cavity C can be adjusted. Specifically, by projecting the convex portion 161 to the inside of the hole 10a, depending on the cross section of the hole 10a and the thickness of the first substrate 10 (specifically, the thickness of the thick single crystal silicon layer 101 and the insulating layer 102). The volume of the cavity C can be reduced more than the determined volume. When the volume of the cavity C is reduced, the Q value decreases, and as a result, it becomes easy to drive the diaphragm 12 so that the sound pressures of the ultrasonic waves of two frequencies are equal.

  The parametric speaker 1 configured as described above operates as follows. When a driving voltage of a modulated wave obtained by amplitude-modulating an ultrasonic wave carrier wave with an audible sound wave is applied to the lower electrode layer 111 and the upper electrode layer 113 of the piezoelectric element 16 via a lead wire (not shown), the piezoelectric element 16 vibrates integrally. The modulated ultrasonic wave is transmitted from the diaphragm 12, and an audible sound is generated by the parametric array effect in a narrow range near the vibration axis of the diaphragm 12 due to the strong directivity of the ultrasonic wave. The parametric array effect is a phenomenon in which an audible aerial sound source corresponding to the original sound wave is continuously generated in a linear region near the vibration axis by self-demodulation due to nonlinearity of sound waves. Due to such a parametric array effect, even a relatively small parametric speaker 1 has a very strong directivity and a practical sound pressure of audible sound can be obtained. Note that two ultrasonic waves having a frequency difference corresponding to an audio wave in the audible range may be transmitted from the diaphragm 12, and the audio wave in the audible range may be demodulated by a beating phenomenon generated between the two ultrasonic waves. In this case, it is necessary to vibrate two or more diaphragms 12 at different frequencies by the two piezoelectric elements 16 to which drive voltages can be input independently.

  The Q value and sound pressure of the parametric speaker 1 correlate with the volume and shape of the cavity C formed behind the diaphragm 12. According to the present embodiment, the Q value and the sound pressure are adjusted independently from the thickness of the first substrate 10 and the area and shape of the diaphragm 12 by adjusting the volume and shape of the convex portion 161 of the second substrate 160. Can do.

Next, a method for manufacturing the parametric speaker 1 will be described with reference to FIGS.
First, an SOI (Silicon On Insulator) wafer to be the first substrate 10 is prepared, and a lower electrode layer 111, a piezoelectric layer 112, and an upper electrode layer 113 are sequentially formed on the surface of the thin single crystal silicon layer 103 as shown in FIG. To do. The lower electrode layer 111 and the upper electrode layer 113 are formed, for example, by stacking platinum by a sputtering method. The piezoelectric layer 112 is formed, for example, by laminating lead zirconate titanate by sputtering.

  Next, as shown in FIG. 3, a pattern of a protective film R1 made of a photoresist is formed on the surface of the upper electrode layer 113, and the upper electrode layer 113 and the piezoelectric layer 112 are etched by ion milling using the protective film R1. At this time, the etching depth of the piezoelectric layer 112 is controlled so that the lower electrode layer 111 is not exposed.

  Next, as shown in FIG. 4, the side wall of the piezoelectric layer 112 formed by ion milling in the previous step is covered, and the protective layer is made of a photoresist opened in the central portion of the diaphragm 12 and the conductive wire connection region of the lower electrode layer 111. A pattern of the film R2 is formed on the surface of the upper electrode layer 113. Subsequently, the lower electrode layer 111 is exposed by etching the piezoelectric layer 112 by wet etching using the protective film R2. Thereby, the piezoelectric element 11 is formed.

  Next, as shown in FIG. 5, a pattern of a protective film R3 made of a photoresist is formed on the surface (main surface) of the thick single crystal silicon layer 101 of the first substrate 10, and Deep-RIE (Reactive) using the protective film R3 is formed. A hole 10a as a non-through hole with the thin single crystal silicon layer 103 as a bottom is formed in the thick single crystal silicon layer 101 and the insulating layer 102 by Ion Etching. As a result, the diaphragm 12 made of the thin single crystal silicon layer 103 and the hole 10 a having the diaphragm 12 as a bottom are formed in the first substrate 10.

  Next, a wafer to be the second substrate 160 is prepared separately from the SOI wafer to be the first substrate 10, and a protective film R <b> 4 made of a photoresist is formed on the main surface of the second substrate 160. Next, as shown in FIG. 6, the protective film R4 is exposed and developed to form a pattern of the protective film R4 for leaving the convex portions 161. Subsequently, the concave surface 162 and the convex portion 161 are formed on the main surface of the second substrate 160 by etching the main surface of the second substrate 160 using the protective film R4. Etching of the second substrate 160 may be a wet process or a dry process. Thus, by forming the convex part 161 using photolithography in the wafer process before dicing, the convex part 161 with high dimensional accuracy can be formed at low cost.

  Next, an adhesive layer 150 made of photosensitive polyimide is formed on the surface on which the convex portion 161 and the concave portion 162 are formed in the previous step, and exposure and development are performed as shown in FIG. 7 to expose the adhesive layer from the surface of the convex portion 161. 150 is removed.

  Next, as shown in FIG. 8, the first substrate 10 and the second substrate 160 are bonded to each other by pressing and heating the main surface of the first substrate 10 and the main surface of the second substrate 160 through the adhesive layer 150. . As a result, the openings of the numerous holes 10a formed in the wafer of the first substrate 10 are simultaneously closed by the second substrate 160, and a cavity C is formed between each of the diaphragms 12 and the second substrate 160. The convex portion 161 of the substrate 160 and the diaphragm 12 face each other.

  After the above wafer process is completed, the parametric speaker 1 is completed when the first substrate 10 and the second substrate 160 are cut into individual pieces by a dicer and a subsequent process such as packaging is performed.

  When the cavities C are thus formed before dicing, the shapes and volumes of the cavities C are simultaneously adjusted by the protrusions 161 formed in advance on the wafer of the second substrate 160. Therefore, according to the manufacturing method described above, the cost of adjusting the volume and shape of the cavity C can be reduced independently of the thickness of the first substrate 10 and the form of the diaphragm 12.

2. Second Embodiment FIG. 9 shows a parametric speaker 2 as a second embodiment of the ultrasonic transducer according to the present invention. As shown in FIG. 9, the volume of the cavity C behind the diaphragm 12 may be enlarged by a recess 163 formed in the second substrate 160. That is, the concave portion 163 may be formed in the second substrate 160 in a region that closes the openings of the plurality of holes 10 a formed in the first substrate 10. When the volume of the cavity C behind the diaphragm 12 is enlarged, the Q value increases. When the first substrate 10 that is thick enough to obtain an appropriate Q value cannot be prepared, the volume of the cavity C can be enlarged by the recess 163 of the second substrate 160 to optimize the Q value. Such a recess 163 can be formed by etching the second substrate 160 using photolithography as in the first embodiment.

3. Third Embodiment FIG. 10 shows a parametric speaker 3 as a third embodiment of the ultrasonic transducer according to the present invention. As shown in FIG. 10, the cavity C behind the diaphragm 12 may be narrowed by coupling a convex portion 165 made of a photosensitive resin film on the second substrate 160. That is, a convex film 165 made of a photosensitive resin film may be formed by forming a photoresist film on the surface of the wafer of the second substrate 160 and exposing and developing the photoresist film. Thus, after forming the convex part 165 by a wafer process, also when joining the 2nd board | substrate 160 which the convex part 165 couple | bonded with the 1st board | substrate 10, the convex part 165 with high dimensional accuracy can be formed at low cost. it can.

4). Other Embodiments The technical scope of the present invention is not limited to the above-described embodiments, and it goes without saying that various modifications can be made without departing from the scope of the present invention. For example, the materials, dimensions, shapes, film forming methods, and pattern transfer methods shown in the above embodiment are merely examples, and descriptions of addition and deletion of processes and replacement of process order that are obvious to those skilled in the art are omitted. Has been.

The diaphragm 12 may be made of an organic material such as polyimide, PVDF (PolyVinylidene DiFluoride), rubber, silicon oxide, polycrystalline silicon, ceramic (zirconia (ZrO ₂ ), alumina (Al ₂ O ₃ ), etc.), metal You may comprise with inorganic materials, such as (Cu etc.). When the diaphragm 12 is formed of these materials, a plate material in which a layer to be the diaphragm 12 and a base layer that supports the diaphragm 12 are stacked constitutes a substrate on which non-through holes according to the present invention are formed. And the board | plate material which comprises this board | substrate may be the thing integrally formed like the SOI wafer, or the thing which joined the film | membrane material to the single crystal silicon wafer. Further, the outer peripheral form of the diaphragm 12 is not limited to a circle, but may be a rectangle, a through-hole may be formed in the diaphragm 12, or the diaphragm 12 may be a band shape. A through hole may be formed in the lower electrode layer 111 of the piezoelectric element 11 at a portion overlapping the central portion of the diaphragm 12. Further, the pattern of the upper electrode layer 113 may be changed, and the central portion of the diaphragm 12 may be driven by the piezoelectric element 11.

  The shapes of the convex portions 161 and 165 that narrow the cavity C and the concave portion 163 that widens the cavity C may be trapezoidal, columnar, dome-shaped, or increase the surface area. Such a shape including many irregularities may be used, or a porous shape may be used. Further, a material with high sound absorption may be used for the convex portion 165 joined to the second substrate 160. Further, the convex portion 161 and the concave portion 163 may be formed by a mechanical processing method such as sandblasting or laser processing. Further, the convex portion 165 may be formed by printing a resin material. Further, the pattern of the convex portions 161 and 165 and the concave portion 163 may be different for each diaphragm 12. Thus, the Q value and the sound pressure can be adjusted for each diaphragm 12, and the Q value and the sound pressure can be adjusted for each diaphragm 12 at a low cost. Further, by making the patterns of the convex portions 161 165 and the concave portion 163 different for each diaphragm 12, it is possible to transmit different ultrasonic waves from the plurality of diaphragms 12 using one piezoelectric element 11.

Further, a through hole connecting the cavity C behind the diaphragm 12 and the external space may be formed in the second substrate 160.
The MEMS transducer according to the present invention may be applied to a sensor such as a sonar.

1: parametric speaker, 2: parametric speaker, 3: parametric speaker, 10: first substrate, 10a: hole, 11: piezoelectric element, 11b: through hole, 11c: through hole, 12: diaphragm, 16: piezoelectric element, 101 : Single crystal silicon layer, 102: insulating layer, 103: single crystal silicon layer, 111: lower electrode layer, 112: piezoelectric layer, 113: upper electrode layer, 150: adhesive layer, 150: adhesive layer, 160: second substrate , 161: convex portion, 162: concave portion, 163: concave portion, 165: convex portion, C: cavity, R1: protective film, R2: protective film, R3: protective film, R4: protective film

Claims (5)

A substrate in which a non-through hole having an opening on a main surface is formed;
A piezoelectric transducer coupled to the bottom of the non-through hole;
A closing member bonded to the main surface of the substrate and having a recess or a protrusion formed in a region closing the opening;
A MEMS transducer comprising: The closing member includes an inorganic substrate bonded to the substrate and a photosensitive resin film bonded to the inorganic substrate,
The convex portion made of the photosensitive resin film is formed in a region closing the opening,
The MEMS transducer according to claim 1. Preparing a first substrate;
Forming a non-through hole in the main surface of the first substrate;
Forming a piezoelectric conversion part on the non-through hole on the back surface of the main surface of the first substrate;
Preparing a second substrate;
Forming a concave portion or a convex portion on the main surface of the second substrate;
Bonding the main surface of the first substrate and the main surface of the second substrate so that the bottom of the non-through hole and the concave portion or the convex portion face each other;
A method of manufacturing a MEMS transducer including: The second substrate is an inorganic substrate,
Forming the concave portion or the convex portion by etching the inorganic substrate using photolithography;
A method for manufacturing the MEMS transducer according to claim 3. The second substrate is an inorganic substrate,
Forming a photosensitive resin film on the inorganic substrate, and exposing and developing the photosensitive resin film to form the convex portion made of the photosensitive resin film;
A method for manufacturing the MEMS transducer according to claim 3.

Images