JP3577566B2 - Manufacturing method of semiconductor dynamic quantity sensor - Google Patents

Description

[0001]
TECHNICAL FIELD OF THE INVENTION
The present invention relates to a semiconductor dynamic quantity sensor in which a movable electrode and a fixed electrode facing the movable electrode are formed by forming a groove in a semiconductor layer of a semiconductor substrate, and a method of manufacturing the same.
[0002]
[Prior art]
For example, in a capacitance-type semiconductor acceleration sensor, displacement of a beam structure in accordance with the action of acceleration is controlled by a movable electrode provided integrally with the beam structure and a fixed electrode provided on a substrate. The change is taken out as a change in the capacitance during the period. In the case of manufacturing such a semiconductor acceleration sensor, a semiconductor having an SOI structure in which a second semiconductor layer is stacked on a first semiconductor layer via an insulating layer, as disclosed in JP-A-6-349806, has conventionally been used. A substrate is prepared, and the first semiconductor layer is patterned into a predetermined shape according to the shape of the beam structure, the fixed electrode, and the like, and the insulating layer is subjected to a process such as etching a sacrifice layer. A method of forming a beam structure having a movable electrode and a fixed electrode on a substrate has been used.
[0003]
[Problems to be solved by the invention]
In the above-described manufacturing method, since a sacrifice layer etching step using an etchant is indispensable, during the sacrifice layer etching step, a movable electrode is stuck to a fixed electrode opposed thereto by the surface tension of the etchant. The sticking phenomenon often occurs. Even if sticking does not occur in the etching step, sticking may occur due to electrostatic force or the like in a subsequent assembling step such as dicing and die bonding or during use.
[0004]
When such a phenomenon occurs, it becomes a fatal defect that it is impossible to detect a change in capacitance between the movable electrode and the fixed electrode, and therefore, the yield deteriorates in the conventional manufacturing method. There was a problem that was inevitable.
As a method for preventing such sticking, for example, as described in Technical Digest of the 15th Sensor Symposium 1997, pp. 205-208, wet etching and NH ₄ F ₂ A method has been reported in which, after a release (beam structure release) process using sublimation, the film is introduced into a plasma polymerization apparatus, and a fluorine-based thin film for preventing sticking is deposited on the sensor surface.
[0005]
However, in such a method, since the wet etching process is performed, there is a possibility that sticking failure may occur at that stage, and the process is complicated because the release process and the process of depositing the fluorine-based thin film are performed separately. There is a problem.
The present invention has been made in view of the above circumstances, and an object of the present invention is to reliably prevent a phenomenon in which a movable electrode and a fixed electrode stick to each other at the stage of manufacture or during use, and at the same time, during the manufacture, Another object of the present invention is to provide a method of manufacturing a semiconductor dynamic quantity sensor, which can effectively prevent a situation in which a material for a fixed electrode is damaged, thereby realizing an improvement in yield.
[0006]
[Means for Solving the Problems]
In order to achieve the above object, a manufacturing method as described in claim 1 can be adopted. According to this manufacturing method, the movable electrodes (10a, 10b) and the movable electrodes (10a, 10b) are formed in the trench forming step with respect to the second semiconductor layer (14b) laminated on the first semiconductor layer (14a) via the insulating layer (14c). A trench (16) for defining the fixed electrode (4b, 5b) is formed to reach the insulating layer.
[0007]
Next, in a first etching step, a portion of the first semiconductor layer corresponding to the formation region of the movable electrode and the fixed electrode is wet-etched from a surface opposite to the insulating layer, and a portion of the etching region is etched. Etching is stopped when the thickness of the first semiconductor layer reaches a preset thickness. Then, in a subsequent second etching step, the first semiconductor layer having the set film thickness remaining as described above is removed by etching in a gas phase atmosphere, so that the back surface of the insulating layer is exposed. become.
[0008]
As described above, in the first etching step, the first semiconductor layer is left with a predetermined thickness between the first semiconductor layer and the insulating layer. As a result, the insulating layer and the second semiconductor layer are less likely to be destroyed. In addition, since the second etching step for exposing the insulating layer is performed in a gaseous-phase atmosphere, the possibility that the insulating layer and the second semiconductor layer are broken is reduced even when the step is performed. . For this reason, it is possible to generally prevent the yield from deteriorating during manufacturing.
[0009]
After the second etching step is performed, the insulating layer is removed by etching in a gas phase atmosphere in the electrode forming etching step, so that an opening continuous with the trench is formed and the movable layer is removed. An electrode and a fixed electrode are formed. In this case, since the electrode forming step, which is a step for forming the movable electrode, is performed by etching in a gas phase atmosphere, as in the case of the conventional configuration in which wet etching is performed in the step of forming the movable electrode, In addition, sticking, in which the movable electrode sticks to the fixed electrode due to the surface tension caused by the etching solution, does not occur.
[0010]
Further, by forming a hydrophobic thin film (17) on the surfaces of the movable electrode and the fixed electrode in the thin film forming step, sticking due to electrostatic force or the like occurs in an assembling step such as dicing and die bonding after this step. Is gone. This is because the thin film formed on the surfaces of the movable electrode and the fixed electrode is hydrophobic, and the surface energy can be reduced as compared with the surfaces of the movable electrode and the fixed electrode without the thin film. This is because even if the thin films on the surface of the fixed electrode come into contact with each other, they are easily separated from each other, and are unlikely to remain stuck.
[0011]
In addition, since the hydrophobic thin film remains even during the operation of the sensor, it is possible to prevent sticking during the operation of the sensor. As a result, as a result of the effect of the hydrophobic thin film, it is possible to improve the yield at the time of manufacturing and to prevent sticking at the time of use, thereby improving the reliability of the sensor.
In order to achieve the above object, the manufacturing method described in claim 2 can be adopted.
[0012]
According to this manufacturing method, the movable electrodes (10a, 10b) and the movable electrodes (10a, 10b) are formed in the trench forming step with respect to the second semiconductor layer (14b) laminated on the first semiconductor layer (14a) via the insulating layer (14c). A trench (16) for defining the fixed electrode (4b, 5b) is formed to reach the insulating layer. Next, in an etching step, a portion of the first semiconductor layer corresponding to the formation region of the movable electrode and the fixed electrode is anisotropically dry-etched from a surface opposite to the insulating layer, and a back surface of the insulating layer is formed. The side becomes exposed.
[0013]
As described above, since the etching step for exposing the insulating layer is performed in a gaseous-phase atmosphere, the insulating layer and the second semiconductor layer are destroyed by the pressure of the etchant as in the case of performing wet etching. Possibility is lost, and it is possible to prevent the yield from being deteriorated at the time of manufacturing. Further, since exposing the insulating layer is performed only by etching in a gas phase atmosphere, the process is simplified.
[0014]
Further, after the etching step for exposing the insulating layer is performed, an electrode forming etching step by etching in a gas phase atmosphere and a thin film forming step are performed in the same manner as in the manufacturing method of claim 1. Similar to the production method described in 1, the effects of these steps can be obtained.
According to a third aspect of the present invention, the thin film forming step according to the first or second aspect is performed simultaneously with the electrode forming etching step.
[0015]
Then, while controlling the etching conditions in the electrode formation etching step to form the movable electrodes (10a, 10b) and the fixed electrodes (4b, 5b), a gas used for etching is used as a material for the movable electrodes and the fixed electrodes. By depositing on the surface, a hydrophobic thin film (17) is formed. This makes it possible to remove the insulating layer (14c) by etching in a gas phase atmosphere and at the same time form a hydrophobic thin film on the electrode surface, thereby simplifying the process in addition to the effects of the first or second aspect of the present invention. Is done.
[0016]
According to the manufacturing method of the ninth aspect, when the etching is performed by a parallel plate type reactive etching (RIE) apparatus in the electrode forming etching step, the insulating layer (14c) on the semiconductor substrate is exposed. Since the surface opposite to the side, that is, the surface side of the sensor does not contact the electrode (54) of the RIE device due to the intermediate member (20) interposed, the surface of the sensor, the gap between the movable electrode and the fixed electrode, etc. It is possible to prevent dust from getting into the dust, and as a result, it is possible to realize an improvement in the yield during manufacturing.
[0017]
According to the manufacturing method of the tenth aspect, the intermediate member (20) is made of a conductive substance (20a). Therefore, in the electrode forming etching step of the manufacturing method of the ninth aspect, the movable electrode of the sensor is provided. Between the fixed electrode and the fixed electrode can be prevented from generating a potential difference due to static electricity generated in the RIE device. As a result, sticking of the movable electrode to the fixed electrode does not occur, and the yield during manufacturing is reduced. Improvements can be realized.
[0018]
According to the manufacturing method of the eleventh aspect, in the electrode formation etching step, the material of the silicon oxide film to be etched and the material of other parts in the RIE device are unified, so that the silicon oxide film in the RIE device can be formed. The in-plane uniformity of the etching rate can be improved, and as a result, the yield during manufacturing can be improved.
[0019]
Further, according to the manufacturing method of the thirteenth and fourteenth aspects, the same effect as that of the invention of the ninth aspect can be obtained in the electrode formation etching step.
Further, according to the manufacturing method of the present invention, the thin film forming step is continuously performed in a series of processes in the electrode forming etching step without leaving the chamber (50), so that the process load can be reduced. is there.
[0020]
In addition, the code | symbol in a parenthesis mentioned above is an example which shows the correspondence with the concrete means of embodiment described later.
[0021]
BEST MODE FOR CARRYING OUT THE INVENTION
Hereinafter, embodiments of the present invention shown in the drawings will be described.
(1st Embodiment)
The first embodiment will be described on the assumption that the present invention is applied to a capacitive semiconductor acceleration sensor. FIG. 1 shows a planar structure of a semiconductor acceleration sensor 1 as a semiconductor dynamic quantity sensor (however, a hatched band in FIG. 1 does not show a cross section, and is used to make it possible to easily distinguish each structural element. FIG. 2 shows a schematic cross-sectional structure along the line AA in FIG.
[0022]
1 and 2, a support substrate 2 made of, for example, single-crystal silicon is formed in a rectangular frame shape having an opening 2a, and a beam structure 3 made of single-crystal silicon is formed on the upper surface thereof. In addition, a pair of fixed electrode structures 4 and 5 are arranged via an insulating film 6 made of a silicon oxide film.
The single-crystal silicon constituting the beam structure 3 and the fixed electrode structures 4 and 5 is preliminarily diffused with impurities in order to reduce the resistivity.
[0023]
The beam structure 3 has a configuration in which both ends of a rectangular mass 7 are integrally connected to anchors 9a and 9b via rectangular frame-shaped beams 8a and 8b. 9 b is supported on the opposite side of the support substrate 2 via the insulating film 6. Thus, the mass portion 7 and the beam portions 8a and 8b are in a state facing the opening 2a of the support substrate 2.
[0024]
The beam portions 8a and 8b displace the mass portion 7 in the direction of the arrow X when receiving acceleration including a component in the direction of the arrow X in FIG. 1, and return to the original state in accordance with the disappearance of the acceleration. It has a spring function of restoring.
Furthermore, the beam structure 3 is provided with, for example, three movable electrodes 10a and 10b integrally protruding from both side surfaces of the mass portion 7 in a direction orthogonal to the mass portion 7, respectively. 10b also faces the opening 2a of the support substrate 2. The movable electrodes 10a and 10b are formed in a rod shape having a rectangular cross section.
[0025]
On the support substrate 2, a movable electrode wiring portion 11 integrally formed with one anchor portion 9 b of the beam structure 3 is formed via the insulating film 6. At a predetermined position, an electrode pad 11a for wire bonding is formed of, for example, aluminum.
The fixed electrode structure 4 is in a state in which the fixed electrode wiring portion 4a formed on the support substrate 2 via the insulating film 6 is in parallel with one side surface of the movable electrode 10a with a predetermined detection gap. , For example, three fixed electrodes 4b are integrally provided, and each fixed electrode 4b is supported in a cantilever manner by the fixed electrode wiring portion 4a. Thus, the fixed electrode 4b faces the opening 2a of the support substrate 2.
[0026]
Further, the fixed electrode structure 5 includes a fixed electrode wiring portion 5a formed on the support substrate 2 with an insulating film 6 interposed therebetween, and one side surface of the movable electrode 10b (the side of the detection gap in the movable electrode 10a). ) And, for example, three fixed electrodes 5b arranged in parallel with a predetermined detection gap, and each fixed electrode 5b is 5a is in a state of being supported in a cantilever manner. Thus, the fixed electrode 5b faces the opening 2a of the support substrate 2.
[0027]
The fixed electrodes 4b and 5b are formed in a rod shape having a rectangular cross section.
At predetermined positions on the fixed electrode wiring portions 4a and 5a, wire bonding electrode pads 4c and 5c are formed of, for example, aluminum. Further, a shielding thin film 12 made of single-crystal silicon, which is a base material of the beam structure 3 and the fixed electrode structures 4 and 5, is arranged on the periphery of the support substrate 2 in a state of being separated by the insulating separation trench 13. Have been.
[0028]
As shown in FIG. 2 and FIG. 3 (h) described later, the surface of the support substrate 2 on the side opposite to the insulating film 6 side, the beam structures 3 and both structures 3 to 5 in the fixed electrode structures 4 and 5 Are on the side surface facing, for example, an organic thin film such as a fluorocarbon film or ammonium fluoride (NH ₄ A hydrophobic thin film 17 made of an inorganic thin film such as F) is formed. The thin film 17 is a hydrophobic film, that is, a film having a small surface energy, and preferably has a contact angle with water of 70 degrees or more. As such a film, a fluorine-based thin film containing fluorine as a component can be used.
[0029]
In the semiconductor acceleration sensor 1 configured as described above, a first capacitor is formed between the movable electrode 10a and the fixed electrode 4b, and a second capacitor is formed between the movable electrode 10b and the fixed electrode 5b. A capacitor will be formed. Each capacitance of the first and second capacitors changes differentially according to the displacement of the movable electrodes 10a and 10b when an acceleration including a component in the direction of the arrow X in FIG. The acceleration can be detected by extracting such a change in capacitance through the electrode pads 4c, 5c, and 11a.
[0030]
FIG. 3 is a schematic cross-sectional view illustrating an example of a manufacturing process of the semiconductor acceleration sensor 1 as described above, which will be described below. Note that the semiconductor acceleration sensor 1 shown in FIG. 3H is a partial cross-sectional configuration model of the semiconductor acceleration sensor 1 (for convenience of description, each of the portions indicated by dashed-dotted lines Q1, Q2, and Q3 in FIG. 1). 3 (a) to 3 (g) schematically show a part corresponding to such a cross-sectional structure model in the course of manufacturing. FIG.
[0031]
First, an SOI substrate 14 (corresponding to a semiconductor substrate in the present invention) as shown in FIG. 3A is prepared. In the SOI substrate 14, a single-crystal silicon wafer 14a (corresponding to a first semiconductor layer in the present invention) serving as a base finally becomes the support substrate 2, and is formed on the single-crystal silicon wafer 14a. In which a single-crystal silicon thin film 14b (corresponding to the second semiconductor layer in the present invention) is provided via a silicon oxide film 14c (corresponding to the insulating layer in the present invention, which eventually becomes the insulating film 6). Has become.
[0032]
The single crystal silicon wafer 14a has a surface orientation set to (100) and has a low impurity concentration with a thickness of at least about 300 μm or more. The single-crystal silicon thin film 14b also has a surface orientation of (100), and is set to a thickness of, for example, about 1 μm. The single-crystal silicon thin film 14b has, for example, a high concentration of phosphorus (1 × 10 4) in order to lower its resistivity and make ohmic contact with the electrode pads 4c, 5c, 11a. ¹⁹ / Cm ³ Or more).
[0033]
Next, an electrode pad forming step as shown in FIG. In this step, aluminum is vapor-deposited on the entire surface of the single-crystal silicon thin film 14b to a thickness of, for example, about 1 μm, and the aluminum film is patterned by using a photolithography technique and an etching technique to form an electrode pad. 4c, 5c and 11a (11a is not shown in FIG. 3). In this electrode pad forming step, a known heat treatment (sintering) for obtaining ohmic contacts of the electrode pads 4c, 5c, and 11a is performed as necessary.
[0034]
From this state, the dimension adjusting step shown in FIG. In this step, the thickness of the single crystal silicon wafer 14a is adjusted to, for example, 300 μm by performing cutting and polishing on the surface (the surface opposite to the insulating film 6) of the single crystal silicon wafer 14a. Apply a mirror finish to the processing surface. The reason why the thickness dimension of the single crystal silicon wafer 14a is reduced to 300 μm in this manner is that the etching depth is reduced when the opening 2a is formed by anisotropic etching, as described later. This is to prevent an increase in chip design dimensions due to anisotropic etching.
[0035]
Next, a mask forming step as shown in FIG. In this step, a silicon nitride film is deposited on the entire surface (mirror-finished surface) of the single-crystal silicon wafer 14a to a thickness of about 0.5 μm by, for example, a plasma CVD method, and then the silicon nitride film is photo-etched. By patterning using a lithography technique and an etching technique, a mask 15 for forming the opening 2a by etching is formed.
[0036]
Thereafter, a trench forming step shown in FIG. In this step, a resist (not shown) having dry etching resistance is formed on the single-crystal silicon thin film 14b and the electrode pads 4c, 5c, and 11a in a predetermined pattern (beam structure 3, fixed electrode structures 4 and 5, shielding thin film). 12), a trench 16 reaching the silicon oxide film 14c is formed in the single-crystal silicon thin film 14b by performing anisotropic dry etching using a dry etching apparatus.
[0037]
From this state, a first etching step as shown in FIG. In the first etching step, the single crystal silicon wafer 14a is selectively etched from the front side (the side opposite to the silicon oxide film 14c) using the mask 15 and, for example, using a KOH aqueous solution. In this case, when such etching is advanced to the silicon oxide film 14c, the possibility that the silicon oxide film 14c is broken and the single crystal silicon thin film 14b is destroyed by the pressure of the etchant becomes very high. The etching time is controlled so as not to progress to the silicon oxide film 14c.
[0038]
Note that such management of the etching time is performed by calculation based on, for example, the thickness dimension of the single crystal silicon wafer 14a and the etching rate of the etchant. In the present embodiment, the management of the single crystal silicon wafer 14a is performed. Time management is performed so that the film thickness remains about 10 μm.
Next, a second etching step as shown in FIG. In the second etching step, the film left between the single crystal silicon wafer 14a and the silicon oxide film 14c in the first etching step is subjected to dry etching using, for example, a plasma etching apparatus. The single-crystal silicon wafer 14a having a thickness of about 10 μm is removed, thereby exposing the back surface (lower surface) of the silicon oxide film 14c. Note that, with such dry etching, the mask 15 is also removed at the same time.
[0039]
Although not specifically shown, the surface of the SOI substrate 14 is covered with a resist before the first and second etching steps are performed. After the process is completed, it is removed.
Then, an electrode forming etching step (third etching step) as shown in FIG. 3H is performed by reactive ion etching (RIE) in plasma.
[0040]
In this electrode formation etching step, for example, a parallel plate type dry etching apparatus as shown in FIG. 4 or FIG. 5 is used. In the etching apparatus, an upper electrode 52 disposed in an upper portion of the chamber 50 and supported by a support portion 51 and a lower electrode 54 disposed in a lower portion of the chamber 50 and supported by a base 53 face each other. Reference numeral 52 indicates that a gas from a reactive gas inlet 55 can be introduced into the chamber 50, and an object to be etched can be set on the lower electrode 54. Then, a high frequency electric field (RF) is applied between the upper and lower electrodes 52 and 54 by a power source 56 to generate plasma between the electrodes 52 and 54, and the ionized gas is incident on the body to be etched. .
[0041]
Here, the SOI substrate 14 (shown in FIG. 3 (g)) in which the silicon oxide film 14c is exposed from the back surface after performing the second etching step is used as an object to be etched, and the object to be etched is silicon oxide. The film 14c is arranged so that the surface opposite to the exposed side faces the lower electrode 54 (ie, upside down from the state shown in FIG. 3G).
[0042]
Further, a ring-shaped plate member 20a made of a conductive material is interposed and set as an intermediate member 20 at a portion where both the body to be etched and the lower electrode 54 do not come into contact with each other. The intermediate member 20 need not be ring-shaped, and may be made of a material having conductivity, such as a silicon substrate.
If the intermediate member 20 is an insulating material, it is considered that both may be electrostatically stuck due to a potential difference between the movable electrode and the fixed electrode generated at the time of dry etching. It is preferable that
[0043]
Further, as shown in FIG. 5, the intermediate member 20 is provided on the periphery of the plate member 20a where the body to be etched is not placed, for the purpose of improving the in-plane uniformity of the etching rate of the silicon oxide film 14c in the dry etching. It is desirable to have a two-layer structure with a plate 21 made of an oxide film (quartz plate).
In such an apparatus configuration, in the electrode formation etching step, the silicon oxide film 14c is removed by performing dry etching from the back surface (the surface on the side of the single crystal silicon wafer 14a) of the silicon oxide film 14c.
[0044]
According to the execution of the electrode forming etching process, the opening 2a is formed in a state continuous with the trench 16, and the mass 7, the beams 8a and 8b of the beam structure 3, the movable electrode 10a, 10b (the mass portion 7, the beam portions 8a and 8b, and the movable electrode 10b are not shown in FIG. 3H) will be released.
At this time, the fixed electrodes 4b and 5b of the fixed electrode structures 4 and 5 (the fixed electrode 5b is not shown in FIG. 3H) are also released, and the fixed electrode wiring portions 4a and 5a are cantilevered. It will be in the state that was done. In this way, the beam structure 3 and the fixed electrode structures 4 and 5 are formed according to the execution of the electrode formation etching process.
[0045]
Also, by controlling the etching conditions during the dry etching in this step, the gas used for the etching is simultaneously deposited as a material on at least the surfaces of the movable electrodes 10a and 10b and the fixed electrodes 4b and 5b, whereby the hydrophobicity is improved. A thin film 17 having a property is formed (thin film forming step). That is, in the present embodiment, the thin film forming step is performed simultaneously with the electrode forming etching step.
[0046]
For example, CF ₄ And CHF ₃ When dry etching using a gas applicable to dry etching such as Ar or Ar is performed, a deposit such as a fluorocarbon film is deposited on the electrode surface after etching by controlling the etching conditions. To leave as. As etching conditions, SiO having a high selectivity to Si ₂ Is preferably used as the etching gas.
[0047]
Since the hydrophobic thin film 17 such as a fluorocarbon film can reduce the surface energy, it is possible to suppress the sticking phenomenon (sticking), which is an intermolecular force acting on the solid.
Further, in this electrode formation etching step, the surfaces of the mass portion 7, the beam portions 8a, 8b, and the movable electrodes 10a, 10b of the beam structure 3 are not directly contacted with the surface of the lower electrode 54 of the dry etching apparatus. Since it is placed on the intermediate member 20, it is possible to prevent foreign substances from adhering to the sensor surface, and it is possible to expect an improvement in yield during manufacturing.
[0048]
After the electrode formation etching process is performed, a dicing process of cutting the SOI substrate 14 into a predetermined sensor chip shape is performed, thereby completing the basic structure of the semiconductor acceleration sensor 1.
By the way, according to the manufacturing method as described above, the electrode formation, which is the final step for releasing the beam structure 3 including the mass portion 7 as the movable portion, the beam portions 8a and 8b, and the movable electrodes 10a and 10b, is performed. The etching step (third etching step) is performed by dry etching (RIE in this example), which is an etching in a gaseous atmosphere, and the hydrophobic thin film 17 such as a fluorocarbon film is deposited. As in the case of the conventional configuration in which wet etching is performed in the final step, sticking occurs in which the beam structure 3 sticks to a fixed electrode such as the fixed electrode structures 4 and 5 due to the surface tension caused by the etching solution. As a result, the yield during manufacturing can be improved.
[0049]
In addition, the movable electrodes 10a and 10b, the beams 8a and 8b, and the fixed electrodes 4b and 5b are released to be movable, and at the same time, a hydrophobic thin film 17 (for example, a fluorocarbon film or the like) for reducing the intermolecular force acting on the side surface is provided. Since sticking can be suppressed by the formation, sticking due to electrostatic force or the like at the time of driving (at the time of using the sensor) can be suppressed, while improving the yield during manufacturing.
[0050]
In the first etching step using a KOH aqueous solution as an etching solution, a single-crystal silicon wafer 14a having a predetermined thickness is left between the silicon oxide film 14c and the second etching step. Since the remaining single crystal silicon wafer 14a is removed by dry etching in the step, in the first etching step, the pressure of the etchant increases the pressure of silicon oxide film 14c and single crystal silicon wafer 14a. As a result, the possibility of destruction of the silicon oxide film 14c and thus the single-crystal silicon thin film 14b is reduced.
[0051]
In addition, since the second etching step for exposing the silicon oxide film 14c is also performed by dry etching, there is a possibility that the silicon oxide film 14c and thus the single crystal silicon thin film 14b may be broken during the execution of the step. , So that the overall yield can be prevented from deteriorating. Further, in the completed state, the movable portions (mass portion 7, beam portions 8a and 8b, and movable electrodes 10a and 10b) of beam structure 3 and fixed electrodes 4b and 5b of fixed electrode structures 4 and 5 are opened. 2a, there is also an advantage that the visual inspection can be easily performed from both front and back sides.
[0052]
Here, the reason for performing the cutting / polishing processing as shown in FIG. 3C will be described in more detail with reference to FIG. That is, as shown in FIG. 3 (f), assuming that the opening design dimension of the opening 2a is a, in order to make the dimension a accurate, in the first etching step, the progress of etching in the lateral direction is required. It is desirable to perform anisotropic etching that can be suppressed as much as possible. In this example, an aqueous KOH solution is used to perform such anisotropic etching on the single crystal silicon wafer 14a.
[0053]
Such anisotropic etching uses an angle θ (= 54.7) from the etched surface as shown in FIG. 3F when the single crystal silicon wafer 14a having the plane orientation (100) is used as in this example. °). Therefore, the relationship between the opening design dimension a, the mask dimension b, and the etching depth d shown in FIG. 3F can be obtained by b = a + 2 × (d / tan 54.7 °). For this reason, for example, when the etching depth d is 500 μm, the mask size b must be about 700 μm larger than the opening design size a, and the chip size of the semiconductor acceleration sensor 1 increases.
[0054]
Therefore, in order to reduce the etching depth d and reduce the difference between the opening design dimension a and the mask dimension d, the present embodiment is configured to execute the dimension adjustment step as described above. However, when the thickness dimension of the single crystal silicon wafer 14a is extremely reduced, the thickness variation may increase, and the yield may decrease due to the possibility of breakage during handling. It is important to set the optimum thickness (300 μm in this example) in consideration of the polishing process capability.
[0055]
In the above-described first embodiment, if the thickness of the single-crystal silicon wafer 14a is set to about 300 μm from the beginning, only the mirror finish needs to be applied to the surface, and the thickness is reduced. Needless to say, the above-described dimensional adjustment step becomes unnecessary, so that the entire manufacturing step is simplified. However, in such a setting, it is necessary to pay attention to the handling of the single crystal silicon wafer 14a.
[0056]
Furthermore, in the first embodiment described above, if the SOI substrate 14 in which the silicon oxide film is formed in advance on the surface of the single crystal silicon wafer 14a is used, the silicon oxide film can be used as an etching mask. In FIG. 3D, the step of depositing a silicon nitride film is not required, and the manufacturing process can be further simplified.
[0057]
(2nd Embodiment)
FIG. 6 is a schematic cross-sectional view showing an example of the manufacturing process according to the second embodiment of the present invention. Hereinafter, this embodiment will be described mainly on portions different from the first embodiment.
That is, in the first embodiment, in order to form the opening 2a in the silicon wafer 14a, after performing the first etching step by wet etching (see FIG. 3F), the second and electrode by dry etching are formed. Although the formation etching steps (see FIGS. 3 (g) and (h), and FIGS. 4 and 5) are sequentially performed, the etching may be performed in a gaseous atmosphere from the beginning. In this case, a dry etching apparatus is generally used as an etching method. In order to accurately control the size of the opening 2a, it is desirable to perform anisotropic dry etching.
[0058]
Specifically, in the second embodiment, an SOI substrate 14 similar to that of the first embodiment is prepared as shown in FIG. 6A, and then an electrode pad forming step shown in FIG. The dimension adjusting step shown in FIG. 6C, the mask forming step shown in FIG. 6D, and the trench forming step shown in FIG. 6E are respectively performed in the same manner as in the first embodiment. However, in the case of the present embodiment, in the mask forming step, a resist having dry etching resistance is provided as the mask 15 '.
[0059]
Then, in the etching step shown in FIG. 6F, the single crystal silicon wafer 14a is subjected to anisotropic dry etching from the surface on the mask 15 'side, thereby removing the wafer 14a and removing the silicon oxide film 14c. The back surface (lower surface) is exposed.
Next, as shown in FIG. 6G, the mask 15 'is removed by ashing, and further, in the electrode formation etching step shown in FIG. 6H, the silicon oxide film 14c is removed in the same manner as in the first embodiment. The beam structure 3 and the fixed electrode structures 4 and 5 are formed by dry etching from the back side, and the hydrophobic thin film 17 is formed.
[0060]
According to the manufacturing method of the present second embodiment, similarly to the first embodiment, the effect of the electrode formation etching can be exerted.
Further, according to the present manufacturing method, since the opening 2a can be formed only by dry etching, the manufacturing process is simplified, and there is no possibility that sticking will occur in the movable parts such as the movable electrodes 10a and 10b. Things. When anisotropic dry etching is performed as described above, the etching proceeds in a direction almost perpendicular to the surface of the single-crystal silicon wafer 14a, so that the mask size is increased as in the case of performing wet etching. This eliminates the need to perform this operation, and eliminates the risk of increasing the chip size.
[0061]
However, dry etching as in the present embodiment has a smaller etching rate than wet etching using a KOH aqueous solution. Therefore, in order to improve the throughput, the thickness of the single crystal silicon wafer 14a is reduced. It is desirable to adjust to about 300 μm.
(Other embodiments)
In the above embodiment, the electrode formation etching step may be performed by anisotropic dry etching if possible.
[0062]
In the above embodiment, the thin film forming step is performed simultaneously with the electrode forming etching step, so that both steps can be performed using the same apparatus. For example, the steps can be simplified. After the step, a thin film forming step may be performed. For example, without removing the workpiece from the etching apparatus after the electrode formation etching step, the composition of the introduced gas, the gas pressure, and the like are changed, and the conditions under which deposition (deposition) occurs are performed so that the hydrophobic thin film 17 is formed. Alternatively, after the electrode formation etching step, the work may be taken out of the etching apparatus, installed in a plasma polymerization apparatus, and the hydrophobic thin film 17 may be formed by plasma polymerization.
[0063]
In addition, the present invention is applicable to a semiconductor dynamic quantity sensor having a movable part and a fixed part, for example, an angular velocity sensor and the like, in addition to the semiconductor acceleration sensor.
[Brief description of the drawings]
FIG. 1 is a plan view showing a configuration of a semiconductor acceleration sensor according to an embodiment of the present invention.
FIG. 2 is a schematic view showing an AA cross section in FIG. 1;
FIG. 3 is a process chart showing a manufacturing method according to the first embodiment of the present invention.
FIG. 4 is a diagram showing a configuration of an RIE apparatus of the present invention.
FIG. 5 is a view showing another example of the intermediate member of the present invention.
FIG. 6 is a process chart showing a manufacturing method according to a second embodiment of the present invention.
[Explanation of symbols]
2 ... support substrate, 4b, 5b ... fixed electrode, 10a, 10b ... movable electrode,
14 ... SOI substrate, 14a ... single crystal silicon wafer,
14b: single-crystal silicon thin film, 14c: silicon oxide film, 16: trench,
17: hydrophobic thin film, 20: intermediate member, 20a: ring-shaped plate,
21: silicon oxide film, 52: upper electrode, 54: lower electrode.

Claims (15)

A movable electrode (10a, 10b) made of a semiconductor material, which is supported on the support substrate (2) in an electrically insulated state and changes according to the action of a mechanical quantity;
A semiconductor dynamic quantity sensor that is supported in an electrically insulated state on the support substrate and includes fixed electrodes (4b, 5b) made of a semiconductor material and opposed to the movable electrode with a predetermined gap. In the manufacturing method,
Finally, a semiconductor substrate (14) in which a second semiconductor layer (14b) is laminated via an insulating layer (14c) on the first semiconductor layer (14a) serving as the support substrate is prepared, and the second semiconductor Forming a trench (16) in the layer to define the movable electrode and the fixed electrode so as to reach the insulating layer;
A portion of the first semiconductor layer corresponding to the formation region of the movable electrode and the fixed electrode was wet-etched from the surface opposite to the insulating layer, and the thickness of the first semiconductor layer in the etching region was set in advance. A first etching step of stopping etching when the film thickness is reached;
A second etching step of exposing the insulating layer by removing the first semiconductor layer having the set film thickness remaining after performing the first etching step by etching in a gas phase atmosphere;
An electrode formation etching step of forming the movable electrode and the fixed electrode by forming an opening continuous with the trench by removing the insulating layer by etching in a gas phase atmosphere;
A thin film forming step of forming a hydrophobic thin film (17) on the surfaces of the movable electrode and the fixed electrode. A movable electrode (10a, 10b) made of a semiconductor material, which is supported on the support substrate (2) in an electrically insulated state and changes according to the action of a mechanical quantity;
A semiconductor dynamic quantity sensor that is supported in an electrically insulated state on the support substrate and includes fixed electrodes (4b, 5b) made of a semiconductor material and opposed to the movable electrode with a predetermined gap. In the manufacturing method,
Finally, a semiconductor substrate (14) in which a second semiconductor layer (14b) is laminated via an insulating layer (14c) on the first semiconductor layer (14a) serving as the support substrate is prepared, and the second semiconductor Forming a trench (16) in the layer to define the movable electrode and the fixed electrode so as to reach the insulating layer;
An etching step of exposing the insulating layer by anisotropically dry-etching a portion of the first semiconductor layer corresponding to the formation region of the movable electrode and the fixed electrode from a surface opposite to the insulating layer;
An electrode formation etching step of forming the movable electrode and the fixed electrode by forming an opening continuous with the trench by removing the insulating layer by etching in a gas phase atmosphere;
A thin film forming step of forming a hydrophobic thin film (17) on the surfaces of the movable electrode and the fixed electrode. The thin film forming step is performed simultaneously with the electrode forming etching step,
The movable electrode (10a, 10b) and the fixed electrode (4b, 5b) are formed by controlling etching conditions in the electrode forming etching step, and the movable electrode and the fixed electrode are formed using a gas used for etching as a material. The method according to claim 1, wherein the hydrophobic thin film is formed by depositing the thin film on a surface of a semiconductor. The method according to any one of claims 1 to 3, wherein the electrode forming etching step is performed by reactive ion etching (RIE) in plasma. 2. An SOI substrate in which the first and second semiconductor layers (14a, 14b) are made of silicon and the insulating layer (14c) is made of a silicon oxide film as the semiconductor substrate (14). 5. The method for manufacturing a semiconductor physical quantity sensor according to any one of items 4 to 4. The method according to any one of claims 1 to 5, wherein the hydrophobic thin film (17) has a contact angle with water of 70 degrees or more. 7. The method according to claim 1, wherein the hydrophobic thin film is an organic thin film made of an organic material. 8. The method according to any one of claims 1 to 7, wherein the hydrophobic thin film (17) is a fluorine-based thin film. In the electrode forming etching step, an object to be etched is arranged on one electrode (54) between a pair of electrodes (52, 54) facing each other, and reactive ion etching (RIE) is performed in plasma.
The semiconductor substrate on which the insulating layer (14c) to be etched is exposed is disposed such that a surface of the insulating layer (14c) opposite to an exposed side faces the one electrode,
An intermediate member (20) is interposed between the one electrode and the surface of the semiconductor substrate opposite to the exposed side of the insulating layer such that the opposite surface does not directly contact the one electrode. The method for manufacturing a semiconductor dynamic quantity sensor according to any one of claims 1 to 8, wherein: The method according to claim 9, wherein the intermediate member (20) is made of a conductive substance (20a). The method according to claim 9 or 10, wherein the intermediate member (20) has a two-layer structure of a conductive substance (20a) and a silicon oxide film (21). The method according to claim 11, wherein the conductive material (20a) comprises a silicon substrate. A movable electrode (10a, 10b) made of a semiconductor material, which is supported in an electrically insulated state on the support substrate (2) and changes according to the action of a mechanical quantity, is electrically insulated on the support substrate. A method of manufacturing a semiconductor dynamic quantity sensor comprising a movable electrode and a fixed electrode (4b, 5b) made of a semiconductor material opposed to the movable electrode with a predetermined gap.
Finally, a semiconductor substrate (14) in which a second semiconductor layer (14b) is laminated via an insulating layer (14c) on the first semiconductor layer (14a) serving as the support substrate is prepared, and the second semiconductor Forming a trench (16) in the layer to define the movable electrode and the fixed electrode so as to reach the insulating layer;
A portion of the first semiconductor layer corresponding to the formation region of the movable electrode and the fixed electrode was wet-etched from the surface opposite to the insulating layer, and the thickness of the first semiconductor layer in the etching region was set in advance. A first etching step of stopping etching when the film thickness is reached;
A second etching step of exposing the insulating layer by removing the first semiconductor layer having the set film thickness remaining after performing the first etching step by etching in a gas phase atmosphere;
Removing the insulating layer by etching in a gas phase atmosphere to form an opening in a state continuous with the trench to form the movable electrode and the fixed electrode, and an electrode forming etching step.
In the electrode forming etching step, an object to be etched is arranged on one electrode (54) between a pair of electrodes (52, 54) facing each other, and reactive ion etching (RIE) is performed in plasma.
The semiconductor substrate on which the insulating layer (14c) to be etched is exposed is disposed such that a surface of the insulating layer (14c) opposite to an exposed side faces the one electrode,
An intermediate member (20) is interposed between the one electrode and the surface of the semiconductor substrate opposite to the exposed side of the insulating layer such that the opposite surface does not directly contact the one electrode. A method for manufacturing a semiconductor physical quantity sensor, characterized by comprising: A movable electrode (10a, 10b) made of a semiconductor material, which is supported in an electrically insulated state on the support substrate (2) and changes according to the action of a mechanical quantity, is electrically insulated on the support substrate. A method of manufacturing a semiconductor dynamic quantity sensor comprising a movable electrode and a fixed electrode (4b, 5b) made of a semiconductor material opposed to the movable electrode with a predetermined gap.
Finally, a semiconductor substrate (14) in which a second semiconductor layer (14b) is laminated via an insulating layer (14c) on the first semiconductor layer (14a) serving as the support substrate is prepared, and the second semiconductor Forming a trench (16) in the layer to define the movable electrode and the fixed electrode so as to reach the insulating layer;
An etching step of exposing the insulating layer by anisotropically dry-etching a portion of the first semiconductor layer corresponding to the formation region of the movable electrode and the fixed electrode from a surface opposite to the insulating layer;
Removing the insulating layer by etching in a gas phase atmosphere to form an opening in a state continuous with the trench to form the movable electrode and the fixed electrode, and an electrode forming etching step.
In the electrode forming etching step, an object to be etched is arranged on one electrode (54) between a pair of electrodes (52, 54) facing each other, and reactive ion etching (RIE) is performed in plasma.
The semiconductor substrate on which the insulating layer (14c) to be etched is exposed is disposed such that a surface of the insulating layer (14c) opposite to an exposed side faces the one electrode,
An intermediate member (20) is interposed between the one electrode and the surface of the semiconductor substrate opposite to the exposed side of the insulating layer such that the opposite surface does not directly contact the one electrode. A method for manufacturing a semiconductor physical quantity sensor, characterized by comprising: The thin film forming step is performed continuously in the electrode forming etching step without being taken out of the chamber (50) by a series of processes. After the electrode forming etching is performed, by changing etching conditions, Semiconductor dynamics according to claim 1 or 2, characterized in that the hydrophobic thin film (17) is formed by depositing on the surfaces of the movable electrode and the fixed electrode (4b, 5b, 10a, 10b). Manufacturing method of quantity sensor.

Images