JPS5955615A - Elastic surface wave device and its producing method - Google Patents

Abstract

PURPOSE:To obtain an element having necessary line width easily by forming an electrode on a piezo-electric plate by a conductive film having multilayer structure. CONSTITUTION:A silicon layer 12 is formed on the piezo-electric plate 11. Aluminum is vapor deposited on the silicon layer 12 as the 1st layer 13. A photoresist film 14 having a prescribed pattern is formed on the 1st layer 13. Subsequently, an unnecessary aluminum film is removed by reactive gas sputter etching process. The sputter etching is moreover continued to remove an exposed silicon layer 12. Finally, the photoresist film 14 is removed by carbonization. The exposed time the elastic surface of the plate to ionized particles is shortened by the combination of the composition of a layer to be etched and reactive gas and the plate is prevented from coarseness. Consequently, deterioration of characteristics is reduced.

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to a surface acoustic wave device that has excellent precision in frequency characteristics and has achieved high mass production and high yield. The present invention relates to a surface wave device and its manufacturing method.

) Final Technique - As is well known, surface acoustic wave devices are made of interdigital electrodes and reflective electrodes, each of which has a different function, on the surface of a piezoelectric substrate, as well as absorbers that absorb and attenuate unnecessary surface acoustic waves. Various functional elements are selected and installed as necessary, and the input electrical signals are converted into surface acoustic waves that propagate through the surface layer of the piezoelectric substrate according to the functions and elements. The surface waves are converted back into electrical signals by other functional elements and output.

There are many types of functional elements, such as input electrodes that convert electrical signals into surface acoustic waves that are excited on the surface layer of the piezoelectric substrate, and convert the surface acoustic waves that have propagated through the surface layer of the piezoelectric substrate back into electrical signals. Output electrodes are some of the commonly used ones, and these input and output electrodes are usually comb-shaped electrodes formed from sticky metal thin films using well-known microfabrication techniques, such as photolithography. It is realized using a "blind-shaped electrode" that consists of two bodies with intersecting comb-teeth sides.

In addition, a large number of strips are formed using reflective electrodes that reflect surface acoustic waves, refractors that refract surface acoustic waves, and metal thin films that change the propagation speed in the same manner as the input and output electrodes. A grating type electrode composed of is also known.

The functional elements of such surface acoustic wave devices include aluminum (A
t)? I-films are often used because thin films of this material generally have lower electrical resistance and density than other materials, and they also have poor sound velocity (surface acoustic wave propagation velocity) dispersion and reflection properties for surface acoustic waves. This is because it has a number of industrially advantageous features, such as having small second-order surface acoustic wave effects such as modulus, being easy to process, and being able to be produced at low cost. However, the recent expansion of the field of application of surface acoustic wave devices has revealed various problems regarding this At thin film.

The first reason is that the application of surface acoustic wave devices has expanded to high frequency bands. The second reason is that it has become necessary to use a wide range of frequency bands from low frequency bands to high frequency bands. Sixthly, as with various devices in other general technical fields, the controllability of the characteristics of the function element has become a problem, not just the expansion of functions. For example, it is a technology to control the reflectance in a surface acoustic wave reflector, that is, to realize a reflector with a desired reflectance, or to control the surface acoustic wave propagation speed, that is, to control the surface acoustic wave propagation speed under a metal thin film of a functional element. When propagating, the propagation speed depends on the frequency, but - dispersion of phase velocity - technology to realize desired dispersion has become necessary.

To be more specific with regard to the above description, an efficient or well-controlled reflector of surface acoustic waves is necessary for realizing a delay line using a resonator or multiple reflections. In addition, in an IF (intermediate frequency) band filter that is often used as a conventional surface acoustic wave device, if the center frequency is 50Mt (z), the comb-like electrode fingers of the interdigital electrodes are replaced with a solid one-finger structure. When used as a type electrode finger, it usually has a width of 16 μm ± 0.2 μm (7.1 m),
When the spout IJ,y type electrode fingers had a two-finger pair structure, the width was 8 μm±0.2 μm (˜±0.3 μm).

However, in the case of 800ME (z) in the high frequency band, the electrode finger width is required to be ±1 μm ±0.1 μm.

Furthermore, for wider bands, elasticity for high frequency bands, 8.

A surface acoustic wave device (hereinafter referred to as the high frequency section), a surface acoustic wave device for the low frequency band (hereinafter referred to as the low frequency section), and a surface acoustic wave device for the intermediate frequency band (hereinafter referred to as the intermediate frequency section) are added. Therefore, it becomes necessary to coexist on one piezoelectric substrate surface.

In this case, for example, in a surface acoustic wave device having a metal strip I-IJ tube-like dagrating reflector in the low frequency region, it is necessary to make the metal film particularly thick in order to increase its reflection ability. On the other hand, the film thickness of the interdigital input/output electrodes (or wave transmitting/receiving electrodes) in the high frequency section must be made as thin as possible in order to prevent the secondary effects such as the above-mentioned sound dispersion and reflectance. . That is, due to the characteristics of the surface acoustic wave device or surface acoustic wave device, there are portions with different film thicknesses.

In the case where two different film thicknesses are required on one piezoelectric substrate, the bonding butt part that maintains the connection between the wiring part and the external circuit should be made with the former in order to avoid its own resistance loss. In the latter case, it is necessary to increase the film thickness in order to improve the reliability of the connection. Oh my,
In a surface acoustic wave device used in a wide band, functional elements having two or more different film thicknesses are generally required.

Related to the third problem, the above is a case where it is necessary to prevent secondary effects such as reflectance and sound dispersion; Techniques for controlling and actively utilizing the following effects are required.

Conventionally, with regard to reflectance, it is clear at least experimentally that the cavity blocking amount, which is the device-based display amount, and the At film thickness of a grating-type reflector satisfy a nearly linear relationship within the range of practical film thickness. and its use was also disclosed (Japanese Patent Application No. 52,965/1983).

In addition, several research results have been published regarding the dispersion of sound speed.Especially in high frequency bands, as the film thickness of the interdigital electrode increases, the effective sound speed under the metal thin film electrode changes depending on the frequency. There are concerns that dispersion will increase, so... 11.

Utilization of crystal orientation, which hardly poses a problem - Example 1: Lithium niobate (LiNbo3)-Y plate, lithium tantalate (LiTa0s) Y plate has also been experimentally clarified, and the technology for its utilization is also known. (Special application 1982)
-52965).

In this way, a surface acoustic wave device utilizing such knowledge was realized while using the complicated manufacturing technology described below.

Next, let's talk about the manufacturing method of surface acoustic wave devices.
In particular, there are many problems with the molding and manufacturing technology for functional elements mounted on piezoelectric substrates. To be more specific, the metal thin film materials used in functional components of surface acoustic wave devices, such as wave transmitting and receiving electrodes and crating-type reflectors, have relatively low electrical resistance and low density, as described above. At the same time, the second-order effects of surface acoustic waves such as reflectance and sound dispersion are relatively small and inexpensive, and the At material thin film itself is relatively easy to form and mold. Despite these advantages of this At material thin film, there are also disadvantages. Specifically, in the etching process using a chemical solution used as a molding technique, undercutting of the functional element part (excessive etching also occurs on the side walls of the At structure) due to the isotropic etching rate of the chemical solution. This makes it extremely difficult to realize the fine line width of the high-frequency functional element mentioned earlier, and the sidewall shape, that is, the cross-sectional shape, is not constant, making it completely unsuitable for industrial manufacturing technology. It's not something.

In addition, as a molding technique suitable for such a KL thin film functional element having such a fine line width, it is believed that the reactive coating technique using a chlorine-based reactive gas, which has been attracting attention in the field of semiconductor manufacturing technology in recent years, is advantageous, and the inventor of the present application have also investigated its application to the production of surface acoustic wave devices for high-frequency bands, and as a result, several problems that had not previously been considered problems became clear. To briefly describe a few experimental facts regarding this, there is 1z, especially regarding LiNbO5 material, which is the mainstream substrate material for surface acoustic wave devices today, and also LiTaO5 material, whose usage is increasing.

Although it is acceptable when used as a piezoelectric substrate, a photoresist pattern is formed on the At thin film surface on these piezoelectric substrates, and a sputter etching process is performed using a chlorine-based compound as a reactive sputtering reaction compound. In some cases, it is not possible to obtain the extremely sharp etched side surface of an At thin film that is seen when performing the same process on silicon (St), which is the mainstream of semiconductor materials, and
Extremely large side etching marks can be seen on the structure. The side etch phenomenon appears to occur as follows. In other words, since there is local non-uniformity in the etching speed distribution, some over-etching is required, but this over-etching can be completed in a short time, and when using a LiNbO5 substrate, It is presumed that during this short period of time, lateral etching occurs which is predominant in the exposed sidewall portion of the At structure formed by etching and faces the substrate. This side etching part is caused by the fact that the At structure as a whole is formed into an inverted trapezoidal shape, and if the At structure is excessively formed, its cross section becomes almost an inverted triangular shape, and the adhesion area decreases and strength is lost. Sometimes peeling also occurs.

Furthermore, a similar cause is assumed, but during the sputter etching process for forming At electrodes with fine line widths on piezoelectric substrates, wormhole-like corrosion marks appear during or after the etching process. A new group accident occurred. Specifically, when forming microelectrodes with a line width of 1 μm to 2 μm from a kA thin film with a film thickness of 0.6 μm or more, many electrode pieces are found to have already peeled off or to easily peel off. The dimensional reproducibility in the sputter etching process was also reduced. These have caused variations in the frequency characteristics of the high frequency section or high frequency elements, and have also resulted in a decrease in yield.

In addition, as a problem specific to the sputter etching process,
Damage caused by the surface of a piezoelectric substrate being struck by chlorine-based plasma gas for several tens of seconds to several minutes is also caused by surface acoustic waves propagating through the surface layer of the piezoelectric substrate.Furthermore, although this is an essential drawback of the sputter etching process, However, especially in the case of high-frequency parts or high-frequency elements, the overall process of manufacturing surface acoustic wave devices becomes very complicated, and countermeasures are required. To be more specific, this corresponds to the manufacturing process when two or more types of functional elements with different film thicknesses are required, as described above.To put it simply, first, a first thin film structure is created. A series of manufacturing steps are carried out, and then the same series of manufacturing steps as before are repeated so that a second thin film structure is formed again at a predetermined position with respect to the shape pattern of this thin film structure. I need to return it. The general difficulty in carrying out these two steps is, as is well known, when matching the shape patterns of both structures, and the shape of the photoresist of the thin film structure in the subsequent process This is known as the difficulty in accurately overlapping patterns.

・ 16. To describe these steps more specifically, when etching the second At electrode, the first At electrode is
In order to protect the electrode, it is necessary to form a film that covers only the first formed ht electrode, and to form a second At electrode.Including the first At electrode, the formation process is a total of 6 times. Film and photoresist processes are required. That is, three masks are required.

A formed first using the conventional chemical solution method and lift-off method
By omitting the film covering only the t-electrode portion and using two masks, a device structure with a certain degree of two-film thickness can be realized.
In reality, there are many limitations and problems. In other words, the chemical solution method is limited to a structure in which the thickness of the At film formed in the first step is increased and the thickness of the At film formed in the second step is decreased, and the accuracy of the At electrode formed in the first step is The 9M electrode, which is formed first during etching of the second electrode, is protected only by the natural oxide film on its surface, which prevents changes in line width and cross-sectional shape due to film thickness reduction, lateral porosity, and etching. Yes. On the other hand, with the lift-off method, the cross-sectional shape is generally not reproducible due to the shading effect of the resist, the side shape itself is tapered, and the film thickness changes depending on the line width.
It is difficult to control its size and shape.

When two or more types of thin films with different film thicknesses are required on the same substrate surface due to a large number of processes, the conventional technology is subject to extreme pressure constraints as described above, or requires complicated processes. An appropriate solution could not be found.

As mentioned above, new technology is required to improve the characteristics of surface acoustic wave devices for high-frequency bands and wide-bandwidth devices, which affects not only the constituent factors of devices but also their manufacturing methods. It is.

The object of the Japanese invention is to eliminate the drawbacks of the prior art described above, and to provide a surface acoustic wave device for high frequency bands and a wide band acoustic wave device equipped with functional elements that enable control of various surface acoustic wave characteristics. The present invention provides a surface wave device and a manufacturing method for forming the same.

- Summary of the present invention The main points of the present invention are that the electromechanical coupling coefficient 9 surface acoustic wave reflectance 1 elasticity It has a multilayer structure in which multiple types of conductor films with different surface acoustic wave properties such as sound velocity dispersion of surface waves are laminated, and the conductor film on the substrate side is used for manufacturing functional elements of surface acoustic wave devices. In the etching process for the conductor film that is not on the substrate side of the conductor film, it is also used as a protective film to prevent damage caused by the etching medium to the conductor film on the substrate side or the piezoelectric substrate. A surface acoustic wave device having an electrode made of a multilayer conductor film with etching resistance that can be used, and a surface acoustic wave device having an electrode made of a multilayer conductor film on a piezoelectric substrate. This is a method for manufacturing a surface wave device.

Hereinafter, prior to explaining the examples, the technical idea regarding the experimental results of the present inventors will be briefly explained.

A Si film is once provided on the surface of the piezoelectric substrate according to the technology of the present invention, and an At film is coated on the Si film.
Although there are -12 unconfirmed problems in the sputter etching process for forming the material function element, the process is assumed to proceed by an intermittent mechanism.

That is, the reaction compound boron trichloride (BCl
3) When a gas is introduced, it dissociates in the plasma region to form molecules or atoms such as ionic BCl2"l CL+, which are then deposited on the substrate (sample) surface to be etched under the application of accelerating voltage. This substrate is made up of a silicon (Si) layer several hundred amps thick stacked on the piezoelectric substrate surface and an A!s film several thousand amps thick on top of it. Furthermore, a photoresist film with a predetermined shape pattern (e.g. interdigital electrode shape pattern) remains at a given position on this M thin film, and the M material resists collisions of the ionized ion molecules or atomic groups during the plasma etching process. Constitutes a protective film.

, 20.

In this state, the accelerated plasma ionic molecules and atomic groups hit the surface of the M material not covered with the photoresist film, and sputter etching of the M material progresses, reaching the Si surface.

As is already well known, in the conventional sputter etching process, the vertical component (component in the film thickness direction) of the etching rate has an overwhelmingly large effect, resulting in sharp etching. It is assumed that the chemical reaction system of plasma-like ions proceeds stably, and each component is consumed while satisfying a certain relationship, producing reaction products.

Next, the reaction system after the etching process reaches the Si surface is based on experimental results regarding BCl2-ct-based plasma etching (1) M etching by reactive sputtering, Institute of Electronics and Communication Engineers, Semiconductor Transistor Research Group Material 5SD79.
28, August 22, 1979) reported that the plasma etching rate of an At film progresses at a rate of 1/4 to 1/6.

From this result, the inventor of the present application inferred the reaction system in this process and formulated the following working hypothesis. In other words, the chemical reaction system of the plasma etching process of the Si film, which progresses following the plasma etching process of the M film, does not stop instantaneously even if the material to be etched changes, but gradually increases the Si film etching process.
Since the reaction system is in a stable state unique to the film system, in the reaction system in this transition state, the plasma-like ions that were in the steady state in the reaction system during the previous etching process of the M film are Molecules and atomic groups are freed from BCl in the unreacted active state due to reduced depletion of the BCl2C6 system.
It was inferred that the 2-C1 concentration was temporarily increased. Naturally, this transition state of abnormal concentration will disappear over time.

The occurrence of such an abnormal concentration region causes an increase in the etching reaction (lateral velocity component) on the side surface of the AA film that is in contact with or in the vicinity of this reaction system region, that is, side etching progresses.

Therefore, by suppressing the reaction system in this transition state for the minimum amount of time possible, the influence of side etching can be reduced and a sharp side surface can be obtained.

On the other hand, according to the working plate theory of the prior art, in the structure where the M film is directly provided on the piezoelectric substrate surface - the M material single layer film structure, Si
Since there is no layer to replace the layer and the surface of the piezoelectric substrate is such that no etching reaction can be expected, the above transition state hardly exists.

Rather, the etching reaction stops on the piezoelectric substrate surface from the part where the etching reaction of the M film ends, so that the BCL
Because the concentration of unreacted activated ions in the m CL system increases at an accelerating rate, the steady state in the etching reaction of the M film rapidly collapses locally, and the unreacted plasma-like ion atoms and molecular groups It is easy to infer that this is directed toward the side etch on the side surface of the M film. This is considered to be a major cause of the defects in the conventional technology evaluated by the inventors of the present application.

・23. In addition, high voltage (AC,
It was thought that the effects of both direct current and direct current could be ignored.

The reason for this is that the reaction region that takes part in the etching reaction occurs within a space with a height of several hundred to several thousand R on the surface of the sample or substrate; The size of the cathode dark area provided for ionization of the reaction compound is about the mean free path depending on the partial pressure of the filled reaction gas (for example, about 10 dark areas for 1 Hg).
, is usually estimated to be several hundred, so even if a high voltage of several tens of kilovolts is applied to these hundreds of spaces, if the height of the reaction area is several hundred grams or several thousand R, it will not be possible. This is because it appears to be of such a degree that it can be regarded as a nearly uniform electric field region. However, in this case, it is known from the experimental results of the inventor of the present invention that the voltage effect cannot be ignored because the change due to the arrangement shape is obviously large in the region of the sample end, and this is a new technological development.

The technical idea of the present invention has been experimentally confirmed by the inventors as well as the etching of a Si film using a CF3-F system using CF4 (carbon tetrafluoride) as a reactive gas compound at 24°, as will be described later. did.

Therefore, based on the above-mentioned technical ideas, the inventors of the present application proposed that, as the simplest case and the most practical case of the present invention, an At film layer is added to the first metal film as the conductive film structure of the functional element. The second metal film is provided with a St film layer, and the second metal film layer controls the surface acoustic wave characteristics and at the same time controls the plasma etching reaction system. They have realized a surface acoustic wave device in which functional elements of metal film layers are provided on the same substrate.

Embodiments Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings.

FIG. 1 shows a VH of a surface acoustic wave device showing an embodiment of the present invention.
This is a top view showing the general configuration of the F-band filter.
The figure is a sectional view taken along line AA' shown in FIG. The structure according to the present invention is made clear from the cross-sectional structural diagram shown in FIG.

In Figures 1 and 2, 1a and 2a are the input side rf connection terminals of the filter, 1b e 2 b are the input side r
f wiring electrode, E is a group of input-side interdigital electrodes, 1 is a grating-type reflector, which has a structure in which a large number of linear electrodes are short-circuited with wiring electrodes at both ends, n is an output-side interdigital electrode group, 1f and 2f are output side rf wiring electrodes, 1g
, 2g are the output side rf connection terminals of the filter, 4 is a piezoelectric substrate made of lithium tantalate (LiTaOs) single crystal X-cut surface, and 5 is the surface acoustic wave propagation direction, which can be viewed from two directions of the substrate. Indicates the direction of rotation by 112”.

The operation of the VHFHF band filter according to the embodiment of the present invention will be briefly explained below. When an electric signal to be input is applied to the input side RF connection terminals 1a and 2a, the input side RF wiring electrodes 1b and 2b are connected to the input side RF wiring electrodes 1b and 2b. The voltage is applied to the input side interdigital electrode group n through the input side interdigital electrode group n. The electric signal applied to the input-side interdigital electrode group n is converted into a surface acoustic wave by the interdigital electrode group n, and propagates in the surface acoustic wave propagation direction 5 shown in FIG. 1 or 2. .

The propagating surface acoustic waves pass under the grating-type reflector 1d and reach the output-side interdigital electrode group, but during the passage, the surface acoustic waves within a part of the frequency band are attenuated, and the remaining surface acoustic waves are attenuated. Only the output reaches the interdigital electrode cluster on the output side. The reaching surface acoustic waves are converted into electrical signals again by the output side interdigital electrode group 1e, and the output side rf wiring electrode 1
f, 2f, and the output side rf connection end pads 1g, 2g, the signal is extracted as an output side electric signal having predetermined filter characteristics.

The VHF band filter given as an embodiment of the present invention is
Output side interdigital electrode u, 1! The structure has a width a1 of each electrode finger, an electrode finger interval λ01 (also referred to as an electrode wavelength), and N electrode fingers are arranged.

In addition, the grating type reflector has dimensions of an electrode finger width aG and an electrode finger interval λp (also referred to as an electrode wavelength),
It has an unacceptable number of electrode fingers.

The structure according to the present invention, as shown in the cross-section of a simple double-layer structure in FIG. 2, has a single-layer structure in the structure of the prior art, whereas the cross-sectional structure of each electrode finger and connection pad part also has a
A multilayer structure in which several thin film layers of metallic or non-metallic materials with different ultrasonic propagation characteristics are laminated is a common one. It consists of a layer 2 or a non-metallic thin film layer 2'.

FIG. 3 is a schematic explanatory diagram of characteristic measurements employed as indicators to demonstrate the effects of the technology of the present invention. This figure shows the frequency characteristics (stopping characteristics) of the surface acoustic wave amplitude passing through a grating reflector. It is a frequency value. This fa value is not necessarily measured for all samples, and may not be measured due to constraints and conditions in actual measurement. I made it possible to get the value.

fo = Upper Cf1 + f2') (1) Here, the fl value is the attenuation level on the low frequency side for a level value attenuated by 3 dB, based on the attenuation level assuming that there is no grating type reflector. In terms of frequency, the f2 value is a frequency in the high frequency range that is similarly measured. Such an fG value has a correlation with the stop frequency fa in a grating type reflector, and is a good index even when the stop frequency fa cannot be directly measured.

The grating type reflector described as the first embodiment of the present invention is a suitable example showing the effect of the technology based on the technical idea of the present invention, so if its structural dimensions are described in some detail, first the input electrode 1b , output electrode 1c1 and connection bird 1a.

As for the 1ρ structure, as shown in the schematic diagram in Fig. 2, the first upper metal thin film layer 1 is an M material thin film (the film thickness was changed in order to obtain the experimental results described below), and the second lower metal thin film layer 1 is made of M material. It has a double wind structure with a total film thickness t = 0.6 μm, in which a polycrystalline silicon (St) thin film layer with a film thickness of 500 A) is arranged as a metal or non-metal thin film. The structural dimensions of the grating reflector side as an example are as follows: Number of electrode fingers NG = 275 electrode finger pitch pc(,to + =λOG/2) = 7.2 μm 1 Ratio of electrode finger width aG to electrode finger pitch pG (hereinafter referred to as duty factor) aa/pa = 0.5. By the way, due to the structural dimensions, the blocking frequency fo = 228.4
We obtained Mf(Z (t = 0.6ttm).t =
0.6 μm is a film thickness that has often been used in the past.

FIG. 4 shows the sum of the film thicknesses t for a grating type reflector having the same structure as the above example, that is, the first layer is an M thin film layer and the second layer is a polycrystalline silicon thin film layer. The above index fG value is measured by changing the total film thickness, and the total film thickness t is plotted on the horizontal axis, and the index fG value is plotted on the vertical axis. This is an actual measurement diagram showing the film thickness dependence of the index blocking frequency fo. be. Of the total film thickness, the second polycrystalline St thin film layer is 500K.
Measurements are made by fixing it to

In the same figure, the values of fo, fl, and fo are actually measured characteristic curves obtained when changing the sum of film thicknesses t related to the present invention, and the values of fo' and f(=fi are for comparison. 2 shows an actually measured characteristic curve of a grating type reflector having a single layer structure of M material according to the conventional technology shown in FIG.

Comparing the present example group and the conventional example group, it is observed that the two exhibit different characteristic behaviors. fl+f
Regarding 1 + f2 + f2, it is shown that it should be selected appropriately according to the specifications of the target device to which it is applied, but regarding the fo value and fo' value,
The former has a gentler slope than the latter, and when it is necessary to improve frequency accuracy, it is less affected by film thickness and duty factor (acG ratio), and its characteristics are stable even against variations in the manufacturing process. This means that it is an improvement over the conventional technology.

FIG. 5 is an actual measurement characteristic diagram of the second embodiment group according to the present invention, which is the same as that of the first embodiment group. The structure of the "grating type reflector" as the second embodiment group has the number of electrode fingers NG = 180, and the electrode finger pitch PC (=λG/
2) = 111”m, ratio of electrode finger width aG to electrode finger pitch pc: ac/Pc (duty factor) =
This is a group of those with 0.6.

By the way, as a typical example, the sum of film thicknesses t=0.6
/jm, we obtained K fo = 149.0 MHz.

・ '51 ・ Stopping frequency of grating type reflector in Fig. 5 =
According to the measured characteristic diagram showing the dependence of the fo value on the film thickness, it is immediately clear that the stopping frequency fo' of the grating-type reflector according to the prior art: The characteristic curve decreases as the film thickness increases. In contrast, it has been shown that the characteristic curve of the grating-type reflector according to the present invention does not depend on the sum of film thicknesses tK for the most part.

Furthermore, although not shown, the duty factor: a
Even when the intermediate value of c/Pc between =0.4 and =0.5 was taken, the characteristic curve was exactly the same as the characteristic curve shown in FIG.

Regarding these experimental results, the stopping frequency: fo' characteristic curve of the grating-type reflector according to the prior art shows that as the duty factor: ac/Pc ratio decreases, the stopping frequency: fo' increases perpendicularly to the stopping frequency axis. There was a tendency for the overall value to shift toward higher values. That is, this means that the blocking frequency and 32° characteristic vary depending on the data factor: aa and lPa ratio.

From these points, the present invention further emphasizes the favorable effects obtained in the first embodiment group described above, and is a remarkable feature of the present invention.

A further point to note regarding this point is that in both the first example group and the second example group, the film thickness dependence of the stopping frequency index fo value is reduced compared to the conventional M material single layer structure electrode. Therefore, when compared for the same frequency, even if the thickness of the metal thin film of the grating type reflector is increased by a little more than 100 mm, it can withstand practical use. This fact suggests a technical solution to the above-mentioned problem that occurs when a plurality of pairs of interdigital electrodes having different film thicknesses are provided on the same substrate, which will be described later.

Next, a third embodiment group showing the second feature of the present invention will be described.

A second feature of the present invention discloses means for controlling the sound velocity dispersion of surface acoustic waves propagating under the functional elements.

The third group of examples is a series of samples relating to surface acoustic wave filters in which the thickness of the metal thin film layer of the interdigital electrodes is varied. For reference, the configuration and dimensions of the components of a typical surface acoustic wave filter are as follows: The piezoelectric substrate, which is the surface acoustic wave propagation medium, is made of lithium niobate (LiNbOs).
) A Y-cut single-plane two-direction propagation substrate is used, and both the input and output electrodes are regular type electrodes with an open length W = 60.
0 μm 1 electrode finger is curved, electrode finger width a = 1
．． The number of pairs of electrode fingers (N=5) per 4 μm electrode was used.

Standardized film thickness related to the present invention: t/λo = 0.5
It was 35×10. By the way, normalized film thickness: t/λ
0 is the value obtained by dividing the sum of film thicknesses t by the electrode wavelength λ0, and in the representative example of this embodiment, the thickness of the first layer M material thin film is 0.27 μm1 the second layer Si material thin film thickness is 0.05 μm, t=0.
30 μm, and the electrode wavelength λo = 5.60 μm.

FIG. 6 shows the effective sound velocity: Veff - normalized film thickness: t/ obtained from the third example group in which the input and output electrodes of the above-mentioned typical surface acoustic wave filter were manufactured by changing the film thickness sum t.
It is an actual measurement diagram (characteristic curve A) of the λ0 characteristic. The characteristic curve B of the same type is the veff t/λ0 characteristic of a surface acoustic wave filter according to the prior art, which has the same electrode configuration as the third embodiment group and whose electrode material is obtained as a single layer structure of M material. Here, the effective sound velocity Veff6 is expressed as the product of the center frequency fc in the frequency characteristics of the interdigital electrode and the electrode wavelength λ0.

The feature of the present invention is that, as is clear from comparing the characteristic curves A and B in FIG. 6, the larger the normalized film thickness t/λO, the more rapidly the effective sound velocity in the interdigital electrode according to the prior art tends to decrease. In contrast, the change in effective sound speed according to the present invention is small. This is the normalized film thickness t
This qualitatively allows us to understand that the degree of dependence on /λO is small, but this aspect can be easily evaluated quantitatively. For example, in the case of the above-mentioned typical surface acoustic wave filter to which the technology of the present invention according to the present invention is applied, normalized film thickness: t/λO (sum of film thicknesses t = 0.30 μm: M material thin film, 0.27/jm r Sl material thin film o, o3μm
), 35.

= 0.5! The effective sound speed for +5X10 is Vef
f = 3'590rr) 4 The center frequency is fc = 633
．． It becomes 6MH7. In the case of the conventional technology with the same standardized film thickness, the effective sound velocity is Ve fJ = 3377%'8,
The center frequency is fc=603.0MHz.

Therefore, when we look at the effect of variation in the film thickness of the interdigital electrode when the effective sound velocity Veff = 10 Or+1/s, we find that the film thickness is 0.
．． The variation of 5 chips of 30 μm is 0.7 with the technology of the present invention.
The conventional technology can suppress IMI to 7MHz.
(It is required that there is a change in z.

The measured characteristic diagram in FIG. 6 discloses the above-mentioned characteristics and shows that it is effective in improving the dispersion of the effective sound speed. Furthermore, if this means stabilizing the film thickness dependence of the frequency characteristics of the surface acoustic wave filter and at the same time allows for the same degree of variation in the effective sound speed, then the first embodiment group and the second embodiment group Similar to the case of the stopping frequency index fo value in the embodiment group, this suggests a technical solution to the above-mentioned problem that occurs when multiple sets of interdigital electrodes with different film thicknesses are provided on the same substrate, which will be described later. It is.

In addition to the surface acoustic wave improvements mentioned above, ■L 1
When a substrate such as NbOs is used, it has become very easy to form an M electrode by M dry etching, which had the difficulty of requiring extreme reduction of contamination inside the device. Fine M required for surface wave equipment
Electrodes can now be obtained on LiNbO3 substrates with a constant cross-sectional shape and a rectangular shape with high accuracy in line width. In this sense, frequency accuracy has also improved, and in particular, the yield of high-frequency surface acoustic wave devices has improved. Effects such as these are being produced.

In this way, the layered device structure of the present application, combined with its manufacturing method, exhibits unprecedented effects.

As mentioned above, based on the technical idea of the present invention, the embodiments have been described in the first to third embodiments regarding surface acoustic waves, but the present invention is not only applied to the above embodiments, but also briefly described below. It is also effective for multilayer structures described using a similar representation method. In addition, ・66 ・ The symbols used in the explanation of the following examples represent the contents of .

Symbol ' (A) s 10 (B) 20N (C) Output Here, (A) t: 1st layer thin film layer of material A, 2: 2nd layer thin film layer of material B, (C) sub: Using a substrate of substrate material C, the second layer Q3)2 is laminated on (C) s u b according to the above notation, and the first layer (A)+ is laminated on top of it.
This figure shows a multilayer structure in which .

Other embodiments will be shown below using the above symbols (notations).

Example 4: Prisoner 1+ (St) t ON C) sub
, A: M-based alloy eMo 1wl T1
+ Cr l C:LINbOs * LI Ta0a
5Zn021 GaAs + Crystal + Bi t2sio
A as 2o + Bi 12Ge02o.

A surface acoustic wave device constructed by appropriately selecting one constituent material of C from these groups.

Example 5: (A)1+(Ge)2ON5ub, (
Has D structure, A: AA, AA alloy +Mo +W+
T1 * CI” + C:LINb03 + LI
TFLO3+Zn02 e GaAs r crystal+Bi
A as t2siOzo +Bi tzGeo2o.

A surface acoustic wave device constructed by appropriately selecting one constituent material of C from these groups.

Example 6: (A)t + (Mo) 20NC)s
ub, A; Al, An alloy + Cr+
Au+Pt+C: A surface acoustic wave device constructed by appropriately selecting one constituent material of A and C from these groups as LiNbO3eLiTaOs, crystal+BitzSiO2o+Bi, and zGeOzo.

Example 7: (A)t+ (vV)2ON (C)s
ub, has the structure A; M, AA-based alloy t Cr+A
u, Ptt C: LiNbO3rLiTa03. Crystal r Bi t2sio2o * Bi t2Ge02o
As,A.

A surface acoustic wave device constructed by appropriately selecting one constituent material of C from these groups.

Example 8: (A)s + (Ta)z 0N(C)s
A'', A1.A1 alloy e Cr
l Au +P t + C; LiNbO3rL
I TaO3e crystal + Bi 12SiO2o e B
i 12Ge02o as A.

A surface acoustic wave device constructed by appropriately selecting one constituent material of C from these groups.

Example 9: (A)s + (Ti)2ON(C)
sub, has the structure, A; kl, M-based alloy r C
r+Au+PL C:LiNbO5゜LiTaO5+crystal+ Bi 12GeO2o+ Bi xzsio20
As,A.

A surface acoustic wave device constructed by appropriately selecting one constituent material of C from these groups.

, 39°, A: Afi, M-based alloy tAu +Pt *C;
LiNbO3 + LITa03 + ZnO2e Ga
As *Crystal + Bi 12 Ge0zo * Bi 1
25ins, the surface acoustic wave device is constructed by appropriately selecting one each of the constituent materials A and C from these groups. However, when one of the A:AL and M-based alloys is selected, the Cr layer is removed by wet etching.

Example 11: (A)s + (Pth 0N(C)
It has the structure of sub, A; All, At-based alloy + C
: LiNbO3+ LiTaO5+ZnO2+ St
A surface acoustic wave device constructed by appropriately selecting one each of constituent materials A and C from these groups as ow.

Example 12: (A)1 + (MoSi2)2ON(
C) has the structure of sub, A; u, AJ, system alloy, A
u+Ptt C: LiNb0stLiTaOs *
A surface acoustic wave device constructed by appropriately selecting constituent materials A and C from these groups as Zn0z*GaAs + crystal.

In the above embodiments, the effect of the present invention is that an electrode mainly made of M can be formed on a substrate made of GaAs or C18 that would otherwise be etched by M etching, and that damage to the substrate can be avoided. There is an advantage that less can be formed. Also, Bit*Geozo tBi12
Although it is suitable for long-time delay lines, distributed delay lines, etc. due to its slow sound velocity, such as SiO2o, it is also suitable for substrates that react with M.
It is now possible to form electrodes mainly made of M, and by taking advantage of the inherent advantages of dry etching, it is possible to form fine electrodes at higher frequencies with higher precision than in the past. In this way, the present invention can expand the range of surface acoustic wave devices that can be created. In other words, this has the effect of widening the range of application in terms of frequency and function of surface acoustic wave devices.

The following embodiments are not surface acoustic wave devices, but are devices in other fields to which the technical idea of the present invention is applied. , is shown below using the above symbols.

Example 1:5: (A)t + (Si)2ON(C
) sub, A: AX, ldt alloy + Mo+W, Ta+TLCr+C; S ion
1AL20s r P IQ resin + Si3N4*Non-magnetic spinel ferrite, magnetic spinel ferrite, with wiring and electrode parts constructed by appropriately selecting constituent materials A and C from these groups. Semiconductor integrated circuit.

Example 14: (A)l+ (Ge)z 0N(C)
sub, has the structure of ), ", A1.fiJl, system alloy # Mol we Ta l T 1 r Cr
r C; S 102 +A11I03 e PIQ-based resin, 5i3N4s non-magnetic spinel ferrite, and magnetic spinel ferrite, each of the constituent materials A and C are appropriately selected from these groups to form wiring parts and electrodes. A semiconductor integrated circuit having a part.

Example 15: (A)! + No)z 0N(C)s
ub, has the structure A; M, Al alloy + Cr + A
u+Pt+ C:5t021 pigeon 0s=PIQ resin+S
i3N4+A semiconductor integrated circuit having a wiring part and an electrode part constructed by appropriately selecting constituent materials A and C from these groups as i3N4+ non-magnetic spinel type ferrite and magnetic spinel type ferrite.

Example '6: (A)! + v) * 0N(C)su
b, has the structure A:M, M-based alloy* Cr + Au
+Pt + C:5102 + A120B +PIQ
system resin, Si3N4 *A semiconductor integrated circuit having a wiring part and an electrode part constructed by appropriately selecting constituent materials A and C from these groups as non-magnetic spinel ferrite and magnetic spinel ferrite.

Example 17: (A)t + (Ta)z 0N(C)
sub, has a structure of k'', A1. Afl alloy +
Cr+Au+Pt+ C; 5iOz, A1.20a+P
IQ resin r 5taN4. A semiconductor integrated circuit having a wiring section and an electrode section constructed by appropriately selecting constituent materials A and C from these groups as non-magnetic spinel type ferrite and magnetic spinel type ferrite.

Example 1a: (A)t+(Ti)zON(C)su
b, has the structure k; M, Al-based alloy + Cr + A
utPt+C; 5102+AQ, zOspPIQ-based resin e Si3N4+ non-magnetic spinel ferrite, magnetic spinel ferrite, select one each of the constituent materials of A and C from these groups as appropriate to form the wiring and electrode parts. A semiconductor integrated circuit with

Example 19: (A)t+(Cr)zON(C)su
b, has the structure, ・46 ・A; M, Al-based alloy + Au, Pte C; 5iOz*
AJ+zOs, PIQ resin, 813N41 non-magnetic spinel ferrite.

A semiconductor integrated circuit having a wiring part and an electrode part formed by appropriately selecting one each of constituent materials A and C from these groups as magnetic spinel type ferrite. However, A
: When one of Affi and At-based alloys is selected, the Cr layer is removed by wet etching.

Example 20: (A)t+(Pt)zON(C)su
b, has the structure A'', M, Al alloy + C: 5
t(he Altos I PIQ resin e 513N
4+ Non-magnetic spinel type ferrite.

A semiconductor integrated circuit having a wiring part and an electrode part formed by appropriately selecting one each of constituent materials A and C from these groups as magnetic spinel type ferrite.

Example 21: (A)t+(MoSi2)ON(C)
sub, structure, A; M, ld-based alloy *Au
+P t+ C:5102 +Ab on IPIQ resin+ 5isN4. As the non-magnetic spinel type ferrite and the magnetic spinel type ferrite, the constituent materials A and C are appropriately selected from each of these groups, 44.

A semiconductor integrated circuit having a wiring section and an electrode section constructed by selecting the following.

Example 22: (A)x+(St)2ON(C)su
b, has the structure, A; AA, ju-based alloy + MO + W * Ta e T 1 * Cr + C
SS iOe*AA20g+ PIQ resin* 813
A hybrid integrated circuit having a wiring section and an electrode section constructed by appropriately selecting constituent materials A and C from these groups as N4+ non-magnetic spinel type ferrite and magnetic spinel type ferrite.

Example 23: (A)1 + (Ge)zON(C)s
ub, has the structure A'', fiJl,, Al-based alloy + MO + W + Ta + Tl + Cr + C: 51021
AlzOs, PIQ resin + 813N4 + non-magnetic spinel type ferrite, magnetic spinel type ferrite, a hybrid having a wiring part and an electrode part constructed by appropriately selecting one each of constituent materials A and C from these groups. integrated circuit.

Example 24: (A)1+ (Mo)zON(C)s
It has the structure of ub, A; AA, N, alloy*
Cr + Au + P t + C; 5t02 # u=
oa + P IQ resin r 813N41 non-magnetic spinel type ferrite.

A hybrid integrated circuit having a wiring section and an electrode section constructed by appropriately selecting constituent materials A and C from these groups as magnetic spinel type ferrite.

Example 25: (A)t+(W)zON(C)sub
It has the structure of h', u, u-based alloy + Cr + Au +
Pt+ C:5i02eAl1203ePIQ resine
5ISN41 A hybrid integrated circuit having a wiring section and an electrode section constructed by appropriately selecting constituent materials A and C from these groups as non-magnetic spinel type ferrite and magnetic spinel type ferrite.

Example 26: (A)t+(Ta)zON(C)su
b, has the structure A;A1. Al-based alloy + Cr +
Au l Pt + C;s 102+ A12os
+ PIQ PIQ resin IBN4 + non-magnetic spinel ferrite, magnetic spinel ferrite, a hybrid integrated circuit having a wiring part and an electrode part constructed by appropriately selecting constituent materials A and C from these groups.

Example 27: (A)1 + (Ti)z 0N (C)
A: AM, Afi-based alloy + Cr*Au, Pt+ C: SiO, Ae, , O.

As PIQ resin + Si3N4 + non-magnetic spinel ferrite and magnetic spinel ferrite, select one each of constituent materials A and C from these groups as appropriate,
A hybrid integrated circuit that has a structured wiring section and an electrode section.

Example 28: (A)t+(Cr)2ON(C)su
b, has the structure, A: M, fiJ, system alloy + Au,
Pt+ C: 5iOz, AlzOs, PIQ resin+
5isNa + non-magnetic spinel type ferrite A wiring part constructed by appropriately selecting one each of constituent materials A and C from these groups as magnetic spinel type ferrite,
A hybrid integrated circuit having an electrode section. However, A: A11. A
When one of the II-based alloys is selected, the Cr layer is removed by wet etching.

Example 29: (A)x+(Pt)z 0N(C)s
It has the structure of ub, A; AA, Afi alloy + C:
5lo2+ A120s r PIQ resin+ 5t
A hybrid integrated circuit having a wiring part and an electrode part, in which sN4 is a non-magnetic spinel ferrite and a magnetic spinel ferrite, and constituent materials A and C are appropriately selected from each of these groups.

7 Example 30: (A)t+(MoSiz)2ON(C
) sub, A: Ag, Al-based alloy + Au + Pt + C * 5tOz + M20y + PI
As Q-based resin w Si3N4 + non-magnetic spinel ferrite and magnetic spinel ferrite, the constituent materials A and C are appropriately selected from each of these groups to form a hybrid integrated circuit having a configuration l and a wiring section and an electrode section. .

The above is an embodiment of the present invention in which a functional element having a double layer structure in which thin film layers of different materials are laminated as the first layer and second layer is used for a device using an insulating substrate or a semi-insulating substrate as the substrate material. However, the present invention, however,
In addition to these embodiments, the present invention can also be applied to device structures using semiconductor materials, semimetal materials, and good conductors for the substrate material. For example, in order to increase the effective path length of the current on a semiconductor substrate, a gold ingot whose length is aligned at right angles to the length direction of the substrate. It is used to control the electrical conductivity and specific resistance of FETs, Hall elements, etc. by using double-layer structured metal pieces in the electrode parts of FETs, Hall elements, etc. It should be clear that it is immediately applicable.

Next, examples of the method for manufacturing a surface acoustic wave device according to the present invention will be described in detail below with reference to the drawings.

FIG. 7 shows the first method of manufacturing a surface acoustic wave device of the present invention.
1 is a flowchart of process steps as an example. In the figure, 11 is a LiNbO5 substrate which is a piezoelectric substrate. 12 is a second Si thin film layer related to the present invention;
13 is a first M thin film layer, and 14 is a photoresist film for forming the first M thin film layer into a predetermined pattern shape (electrode shape). Further, ■, ■, etc. indicate the order of steps in the process flowchart. Each step will be explained below.

Step (2) is a cleaning and drying step for removing contaminants on the surface of the LiNbO5 substrate 11, and is a pretreatment step for the next step.

Step (2) is a step of forming a Si layer, which is the second layer of the present invention 8, on the surface of the LiNbO5 substrate 11 to an appropriate thickness as described below.

Step (2) is a step of forming a first M thin film layer 13 on the St layer 12 formed in step (2), and steps (2) and (2) are performed continuously. That is, the LiNbO3 substrate 11 is placed at a predetermined position in a vacuum evaporation apparatus having a well-known structure.
After heating the substrate while maintaining a predetermined degree of vacuum, the Si material is evaporated by electron beam radiation, and an 81 layer 12 is deposited on a predetermined surface of the LiNbO5 substrate 11. do. Next, the evaporation source is switched and step (3) is started, in which the M material is electron beam evaporated by a predetermined known method. Steps (1) and (2) can be performed continuously by simply changing the deposition source. L via process ■ and process ■
The iNb0a substrate 11 is taken out from the υ vacuum evaporation apparatus by a predetermined operation and transferred to the next step (2).

Step (2) is a step of forming a well-known photoresist film 14 having a predetermined pattern on the surface of the M thin film layer 13 formed on the surface of the LiNbO3 substrate 11 in the previous steps ((1) to (2)). In this example, the photoresist film 14 used in step □ has good sharpness, is suitable for high definition, has relatively high durability against dry etching atmosphere, and has easy post-processing and peeling work. From AZ
A 1350 series resist (manufactured by Brecy, UK) was used. The application of this resist film is carried out under well-known conditions by pre-baking, exposure using a pattern mask with a predetermined shape, and development, and the final protection by the photoresist film in the shape of the wave transmitting/receiving electrode is completed. . Then, move on to the next step (■).

Step (2) is a reactive gas sputtering etching step for the M thin film layer 13. In this process, the LiNbO5 substrate 11 having the photoresist film 14 in the shape of the transmitting and receiving electrodes prepared in step Boron trichloride gas added with carbon gas is used as a reaction gas,
Secure and supply the flow rate until the predetermined pressure is achieved. The reaction gas may be boron trichloride gas alone. After securing an atmosphere filled with reaction gas to a predetermined pressure, 13.56
A high frequency power of MH7 is applied to cause discharge, and a reaction is caused with the photoresist film pattern 7, thereby etching and removing the M film exposed outside the photoresist film pattern 7 portion.

If I were to discuss two or three important points in carrying out the process, I would like to explain M.
In consideration of the occurrence of non-uniformity in the etching of the layer 13, it is necessary to set an excessive etching time, and for this purpose, as shown in the cross-sectional view of step (2) in FIG. The M layer 13 is also etched to some extent. Therefore, as mentioned in step (2), the thickness of the 81 layer 16 is reduced by the LiNb
selecting a thickness that does not reach the surface of the O3 substrate 11;
At the same time, it is also important to manage the end point of M etching, taking into account the non-uniformity of etching of M layer 13. The end point of M etching is controlled by the well-known M etching in discharge plasma.
This can be done by monitoring the intensity of the emission spectrum of atoms. In addition, etching non-uniformity is particularly important for Li
When looking at the entire NbO3 substrate 11, it often occurs in the peripheral area, but this can be prevented by a technique that has already been applied for by the same applicant; a method of placing a quartz ring around the substrate 11. .

(Patent Application No. 55-529155) Next, step ① begins.

Step (2) is a step in which the exposed St layer 13 is removed by etching using the photoresist pattern used in step (2). The inventors introduced and discharged carbon tetrafluoride (CF4) gas at a predetermined flow rate to maintain a constant pressure, and accelerated the ionized particles due to the discharge of the CF4 gas to the St layer 13 in the same way as in step (2). The St layer 13 is etched by collision. The important point in this process is the technology invented by the inventor of the present application and filed by the same applicant.
(Japanese Patent Application No. 55-144371).

The "post-batter cleaning process" technology has been introduced in this technical field to block the aftereffects of the previous process, in view of the fact that the aftereffects of the previous process often do not have a negative impact on the subsequent processes, which have different characteristics. It is a process that When the M layer 13 is directly stacked on the LiNbO5 substrate 11 as in the prior art, the M layer 1 using BCl3''JID reaction gas
In the reactive sputter etching step No. 3, a surface acoustic wave device substrate, such as LiNbO5 or L
iTaO3 does not suffer from extremely large particle beam damage, but also has limitations in microfabrication of the electrode geometry of the M layer.
Furthermore, even after the device was completed, the remaining products generated in the reactive process and the quenching process caused a deterioration in reliability. The [Post Batter Cleaning Process] technology was invented by the present inventors to overcome these drawbacks, but it is important to appropriately select the reactive gas used in the [Post Batter Cleaning Process]. For example, it is important to note that it has been found that the reactive coating and etching processes of the St layer 1'5 according to the present invention can be simultaneously realized.

That is, in step (3) of the manufacturing method according to the present invention, on the one hand, the reactive sputter etching process of the second layer 81 layer 13 is realized, and at the same time, the "bost sputter cleaning" process invented by the present inventors is carried out. This means that this has also been achieved.

In addition, as for the types of reactive gases used in this step, in addition to the above-mentioned CF4, it was confirmed through experiments that the following ones are effective, although each needs to be selected depending on the conditions.

(1) Fluorocarbon gas; carbon hexafluoride (02Fg),
Carbon hexafluoride (C3Fg).

(2) Mixed gas of fluorocarbon gas and oxygen (02).

(3) Fluoride/bromide carbon gas.

(4) Silicon halide gas; silicon tetrafluoride (SiF
4).

(5) Nitrogen trifluoride gas (NF3) (6) Sulfur tetrafluoride gas (SF4) In step (3), the sputter etching rate for the LiNbO3 substrate 11 is as follows when using BCl2 and BCl3-based reactive gases. Since it is about two orders of magnitude lower than the etching speed, the etching effect is hardly recognized, and the above-mentioned inventive technique is realized. Then,
Entering the process.

The second step is a step of removing the photoresist film 14 used for forming the electrodes up to the above steps. After exhausting and removing the reactive gas, CF4 and reaction products used in step (2), 02
is introduced at an appropriate flow rate to maintain a predetermined pressure, and the photoresist film is carbonized (ashed) and removed.

Then, take out the sample and move on to the next step.

The following steps are not shown in the drawings, but can be briefly described: a step of installing an absorber for reflected waves that are reflected at end faces of surface acoustic waves and degrades the S/N ratio at a predetermined position on the LiNb0a substrate 11 on which electrodes are formed; A cutting process to make 11 into a predetermined size,
Can-package installation is a process, wire bonding wiring process, hermetic sealing process in a dry atmosphere (nitrogen gas),
Finally, the entire process is completed through an inspection process by measuring characteristics.

According to the manufacturing method of the present invention, the surface layer of the surface acoustic wave substrate is accelerated by ionized particles of a reactive gas such as chlorine-based gas or fluorocarbon-based gas, as clarified in the above embodiments. The time of exposure to particles in the etched state can be minimized in good conditions by selecting a combination of layers to be etched and a selected combination of reactive gases. Therefore, digging into the substrate, surface damage, and roughness are extremely reduced, so surface acoustic wave devices, such as M-
It is also possible to reduce propagation loss and disturbances in high-band characteristics due to spurious excitation in band-stop resonators equipped with strip grating electrodes. Furthermore, according to the manufacturing method of the present invention, a considerably complicated process assembly is required according to the prior art. A surface acoustic wave device in which a high frequency band electrode group and a low frequency band electrode group coexist on one substrate can also be realized extremely easily. A second embodiment of the method for manufacturing a surface acoustic wave device according to the present invention will be described below with reference to the drawings.

56 - FIG. 8 is a process flowchart of a manufacturing method given as a second embodiment. The surface acoustic wave device to which this embodiment is applied has an electrode for a high frequency band and a group of electrodes for a low frequency band coexisting on the surface of a single piezoelectric substrate. In the figure, 11 is a Si layer related to the LiNbO3 material piezoelectric substrate piezoelectric substrate of the present invention, 21 is a first AL thin film layer, 22 is a first photoresist film pattern, 2! I is the second Afi thin film layer, 2
4, the second photoresist film pattern 25 is a high-frequency band transmitting/receiving electrode; 26 is a wiring electrode base Si layer for the high-frequency band electrode;
7 is a wiring electrode for a high frequency electrode, 28 is a transmitting/receiving electrode for a low frequency band, and 29 is a grating electrode for a low frequency band.

To explain the process flow diagram of the manufacturing method in this example,
Briefly speaking, the first embodiment of the manufacturing method is repeated twice - however, the St thin film layer only needs to be deposited once.
It consists of the following steps.

Process ■; Cleaning and drying process of LiNbO3 substrate 11.

Step (2): Formation step of the first A1 thin film layer 21.

Step (2): Step of forming the first photoresist film pattern 22.

Step (2): Sputter etching step of the first A1 thin film layer 21.

Step (2): Step of removing the photoresist film using the ashing technique of the first photoresist film pattern 22.

Step (2): Deposition step of the Si thin film layer 12 according to the present invention.

Step: Forming the second AA thin film layer 23.

Step (2): Step of forming the second photoresist film pattern 24.

Step (2): Sputter etching step of the second A1 thin film layer 26.

Step O: Turn 24 of the photoresist film in step ■
A process of removing the exposed Si layer 12 by etching.

Step 0: Step of removing the photoresist film using the ashing technique of the second photoresist film pattern 24.

The Si thin film layer 12 in this embodiment, including the high frequency band transmitting/receiving electrode 25 formed as the first M thin film layer 21 and the high frequency band grating reflector to be installed if necessary, is formed by the sputter etching process in step (3). It is important to form the Si thin film layer 12 with an appropriate thickness so that these functional elements formed in the previous process are not damaged.

For example, in experiments conducted by the present inventors, although it depends on the conditions, when the thickness of the second M thin film layer is 0.5 μm, the thickness of the Si thin film layer 12 is required to be 0.0!1 μm or more.

The most important point in the second embodiment of the manufacturing method of the present invention is that by effectively utilizing the selectivity of the etching performance of sputter etching using the Si thin film layer 12, However, in the case of the technology of the present invention, only two photoresist masks are required and the electrode formation accuracy is improved. , it was confirmed that the accuracy of the grating reflector was significantly improved.

Note that the electrical resistance value of the Si thin film layer 12 becomes a problem60.
- It is clear that, in some cases, an optimum resistance value can be ensured by controlling in advance the electrical resistance values of the evaporation source of the Si thin film layer 12 and the sputter target plate. When the electrical resistance value of the Si thin film layer 12 is a problem, the above step 0
After the completion of the test, heat at approximately 400'C in a nitrogen or hydrogen atmosphere.
'r Si may be diffused into M by annealing for about 30 minutes.

It is also clear that the Si thin film layer can be replaced with other material layers and formed using the same process sequence.

A device with the same film structure as in the second embodiment of the manufacturing method of the present invention described above can be fabricated using conventional chemical solution etching, but generally the first layer M is electrochemically etched when the second layer is etched. The method of the present invention is clearly superior in terms of easy corrosion and accuracy. In addition, in the second embodiment of the manufacturing method of the present invention, the Si layer or the corresponding material layer is eliminated, and the conventional chemical solution method and lift-off method are used to some extent by using two masks for the first layer and the second layer. Although a two-layer device structure can be formed, the chemical solution method is limited to a structure in which the first layer M is thicker and the second layer M is thinner, and the device structure is formed from the first layer M. The precision of the structure is low, and the film thickness decreases during the second layer etching, or the line width and cross-sectional shape change due to lateral etching.In the lift-off method, the cross-sectional shape is generally poor and the film thickness depends on the line width. However, the effectiveness of the present invention without these constraints is greatly increased. As a result, only the reflector film thickness for low frequencies is increased, and the reflector
Increase the reflection efficiency of one IJ knob to create a reflector) 1
By reducing the number of 771 lines, the effect of saving substrate area and the effect of improving accuracy and yield by applying dry etching can be realized.

Furthermore, according to the present invention, the reproducibility and uniformity of the sputter etching technique for the M thin film layer have been greatly improved.

In terms of specific values, both the yield for defects and the dimensional yield for uniformity improved from 90 inches to 98 inches.

More preferably, device operating procedures in the manufacturing process,
Since maintenance was simplified, mass productivity was greatly improved, which was a major advantage for the manufacturing method.

Support from Naruto As mentioned above, according to the present invention, it has become possible to control the functional characteristics of various functional elements in the device structure, which was impossible in the prior art, and it has become possible to control the functional characteristics of various functional elements in the device structure. Since it has become possible to fabricate a complex device structure in which frequency band functional elements coexist on the same substrate, the contribution to this technical field is extremely significant.

[Brief explanation of the drawing]

FIG. 1 is a plan view showing the configuration of a first embodiment of the surface acoustic wave device of the present invention, FIG. 2 is a schematic explanatory diagram of the cross-sectional structure taken along the line ' shown in FIG. 1, and FIG. is a schematic explanatory diagram of the measured values of the characteristic index of the surface acoustic wave device of the present invention, and FIG.
Fig. 5 is an actual measurement diagram showing the film thickness dependence of the above index for the first embodiment group, Fig. 5 is an actual measurement diagram of the same kind of characteristics for the second embodiment group, and Fig. 6 is a representative surface acoustic wave 6 FIG. 7 is an actual measurement diagram of the sound velocity dispersion property with respect to the sum of film thicknesses t at the input and output electrodes of the filter, and FIG. FIG. 8 is a flowchart of process steps of a method for manufacturing a surface acoustic wave device having a structure in which one set of high-frequency band function elements and one set of low-frequency band function elements coexist, which is given as a second embodiment. 1... First metal thin film layer, 2... Second metal thin film layer, 4
．． 11... Piezoelectric substrate, 5... Surface acoustic wave propagation direction, 12... Si thin film layer, 15, 21.2
5... AII thin film layer, 14.22.24... Photoresist film pattern. 64. O 1 Bacteria O 2 Kuni O 3 Illustration 4 Evil〉 Se Wa Reyu 3rd Prisoner Not

Claims (1)

[Claims] 1. Input conversion, output conversion, interference, diffraction, reflection, resonance, attenuation or propagation, direction conversion, and electrical connection with external electronic circuits of surface acoustic waves provided on the piezoelectric substrate surface. In a surface acoustic wave device comprising a functional element of an electrically conductive thin film, at least a part of the functional element of the electrically conductive thin film realizes the functions of the first electrically conductive material thin film and the electrically conductive material thin film. A surface acoustic wave device characterized in that it has a laminated structure in which a second electrically conductive material thin film provided on the surface side of a piezoelectric substrate is laminated. 2. & The first electrically conductive material thin film forming the layered structure is an aluminum (At) material thin film, and the second electrically conductive material thin film is a semiconductor material thin film. A surface acoustic wave device according to claim 1. 3. The first electrically conductive thin film forming the stacked structure is an aluminum (At) thin film, and the second electrically conductive thin film is a transition metal thin film. A surface acoustic wave device according to claim 1. 4. The surface acoustic wave device according to claim 2, wherein the semiconductor material thin film is silicon (St). 5. The surface acoustic wave device according to claim 2, wherein the semiconductor material thin film is germanium (Ge). 6. The first electrically conductive material thin film forming the stacked structure is an aluminum (At) material thin film, and the second electrically conductive material thin film is an electrically conductive carbon material thin film having a stacked structure functional element. A surface acoustic wave device according to claim 1. l The transition metal material thin film, which is the second electrically conductive material thin film, is made of molybdenum (Mo), tungsten (W), chromium (Cr), tantalum (T&), titanium (Tt).
7. The surface acoustic wave device according to claim 6, wherein the surface acoustic wave device is one metal film selected from the group consisting of *platinum (pt). 8. Claim 6, wherein the second electrically conductive material thin film is a disilicon molybdenum material thin film.
The surface acoustic wave device described in . 9 In a method for manufacturing a surface acoustic wave device having a surface acoustic wave function element having a predetermined shape and a predetermined number of thin films of metal material at a predetermined position on the surface of a piezoelectric substrate, the thin film of metal material of the surface acoustic wave function element is The forming step is a step of forming a laminated structure of a first metal material thin film having different corrosive properties and a second metal material thin film, and a step of forming a laminated structure of a first metal material thin film having different corrosivity, and introducing a first reactive compound having a different corrosive ability into the first metal material thin film. Corrosion removal (etching) to a specified shape
and a step of cutting and introducing a second reactive compound having a corrosive ability different from the first corrosive ability, and etching the second metal material thin film into a predetermined shape. A method for manufacturing a surface acoustic wave device featuring features. , Pe. 10. As different corrosion abilities, the first metal material thin film and the second
10. The method of manufacturing a surface acoustic wave device according to claim 9, wherein the surface acoustic wave device is formed with different corrosion rates from the material thin film. 11. The method for manufacturing a surface acoustic wave device according to claim 9, wherein the selected different corrosion rates are controlled by a plurality of different corrosion reaction compounds. 12. The method for manufacturing a surface acoustic wave device according to claim 11, characterized in that gas phase compounds are switched and used as the reactive compounds having different corrosive abilities. 13. A method for manufacturing a surface acoustic wave device according to claim 11, characterized in that a gas phase compound and a liquid phase compound are switched as the reactive compounds having different corrosive abilities. 14. A surface acoustic wave according to claim 9, characterized in that a chlorine-based reactive gas is used as the first corrosion-reactive compound, and a fluorine-carbon-based reactive gas is used as the second corrosion-reactive compound. Method of manufacturing the device. 15. Boron trichloride (Bc4
) or poron trichloride (Bct3) as the main component, oxygen (02), helium (He ), carbon tetrafluoride (CF
4) A step of dry etching the first metal material thin film with a first mixed reaction compound composed of one selected as a subcomponent from the group of 4), and a step of dry etching the first metal material thin film, comprising tetrafluorocarbon (CF4) as the second corrosion reaction compound. , carbon fluoride chloride, and carbon fluoride bromide, and dry etching the second metal material thin film. A method for manufacturing a surface wave device. 16. The chlorine-based reaction gas, which is the first corrosion reaction compound, contains carbon tetrachloride (CcA4) or carbon tetrachloride (Cct4) as the main component, and helium (H
e), oxygen (02), chlorine (ct2), nitrogen (N
2) forming a first mixed reaction gas by selecting − from the group consisting of 2), and dry etching the first metal material thin film with the first mixed reaction gas;
Carbon tetrafluoride (CF4) as a second corrosion-reactive compound;
The surface acoustic wave according to claim 9, further comprising the step of dry etching the second metal material thin film by selecting - from the group consisting of carbon fluoride chloride and carbon fluoride bromide. Method of manufacturing the device.