WO2018057956A1

WO2018057956A1 - Electroacoustic transducer having improved esd resistance

Info

Publication number: WO2018057956A1
Application number: PCT/US2017/053054
Authority: WO
Inventors: Dr. Gholamreza Dadgar JAVID; Dr. Thomas EBNER; Dr. John JOHNSON
Original assignee: Snaptrack, Inc.
Priority date: 2016-09-26
Filing date: 2017-09-22
Publication date: 2018-03-29
Also published as: DE102016118124A1; DE102016118124B4

Abstract

An electroacoustic transducer having improved protection against ESD is proposed, wherein the ends of the electrode fingers (EF, SF) facing away from the respective bus electrode (BE) of the transducer are rounded. A dielectric (DK_S) may cover the gaps between opposing fingers.

Description

Electroacoustic transducer having improved ESD resistance

CROSS-REFERENCE TO RELATED APPLICATION

This application claims priority to German Patent Application No. 102016118124.2, filed September 26, 2016, which is expressly incorporated herein by reference in its entirety.

Description

SAW modules are in principle vulnerable to ESD discharges. Since the electrode structures of SAW transducers are very fine, damage to and total failure of structural elements are therefore to be expected in the event of ESD discharges. The probability of an ESD discharge increases with the level of the electrical potential difference at metallic structures and with decreasing spatial distance. Short distances are in particular found at the transversal gap of the electrode fingers, where the finger ends are arranged in proximity to the half-electrode of the opposite polarity. This in particular applies to resonator filters and DMS traces.

However, an enlargement of the distances at the gap is not possible since radiation losses occur for greater distance of the finger ends from the opposite half-electrode so that the transversal gap is chosen to be as small as is technologically possible. In addition to this, with increasing frequency of the SAW component, the distances become smaller and the problem is further exacerbated. In addition, the finger ends have very small curvature radii, which in turn leads to high electrical field strengths.

For a given external voltage, the field strength of a tip thus behaves proportional to the ratio 1/r, wherein r is the curvature radius of the structure. In a first approximation, a halved radius therefore leads to a field strength increased by a factor of 2.

Known measures to increase the dielectric strength exist in cascading transducers, wherein vulnerable transducers are split into sub-transducers that may then be operated in the cascade with accordingly lower voltage. However, the space requirement of such a transducer cascade may therefore also be significantly increased and outweigh the obtained advantage again.

It was additionally already proposed to arrange a dielectric in the transversal gap, that is, in the region between finger ends and counter-electrode or between finger ends and stub finger ends, which has an acoustic impedance similar to that of the material of the electrode finger. Given an enlargement of the gap, radiation losses are thereby avoided, and the probability of ESD discharges is simultaneously reduced. It is disadvantageous that more area is required due to the enlargement of the gap, which runs counter to the trend of increasing miniaturization.

It is therefore the object of the present invention to propose an electroacoustic transducer in which the susceptibility to ESD discharges is reduced, without having to accept greater surface requirement or disadvantages in the performance of the

electroacoustic transducer.

This object is achieved according to the invention by an

electroacoustic transducer according to claim 1. Advantageous embodiments of the invention are provided in additional claims.

The fundamental idea of the invention is to geometrically design electrode structures that are closely adjacent and carry

different potential such that neither point discharges may occur nor locations with strongly increased field strength in the first place.

This is achieved by rounding the ends of the electrode fingers, which were previously rectangular in design. The invention thus proposes an electroacoustic transducer that comprises two electrode combs. Each electrode comb has electrode fingers approximately parallel to one another, which are arranged with a common bus electrode. The two electrode combs are interleaved so that the electrode fingers mutually overlap in a transversal overlap region. Although the unmodified potential difference is now, as before, present between the respective end of an

electrode finger and the adjacent bus electrode lying at the opposite potential, the danger of a point discharge is avoided by the geometrically rounded shape of the finger ends. For the rounded finger ends, the field strength between the finger ends carrying different potential is proportional to the distance between these finger ends such that the higher field strength occurs at locations of greater distance, which likewise

contributes to a higher ESD resistance. The charge distribution at the finger ends, which is disadvantageous with regard to ESD, becomes significantly more uniform with rounded ends.

A rounded end is not to be understood to be a strict circular arc with constant radius. Within the scope of the invention, other rounded forms of finger ends are also suitable and

represent an improvement over the prior art. For example, a finger end may be elliptically shaped. However, it is always advantageous if no chamfer angles occur and all transitions are rounded. If chamfer angles are unavoidable, a shape is chosen with chamfer angles that are as obtuse as possible.

With modern lithography processes, such a measure is possible without additional technical cost in the production of the transducers or in the production of the metallization for the transducer. Moreover, an enlargement of the gap - i.e., the distance between the finger end and the opposite structure, i.e., opposite bus electrode or the opposite end of a stub finger - is not necessary, and therefore neither is an

additional space requirement. Unnecessarily increased radiation losses in the gap may thus also be avoided. In contrast, due to the reduced arcing danger, it is now even possible to reduce the gap width relative to the previous typical gap width, and to thereby even additionally reduce possible radiation losses.

Such radiation losses arise by a mode conversion occuring at the gap and energy of this mode being lost because it is no longer available for generation of the desired mode or wave. Moreover, it has been shown that, with finger ends designed according to the invention, the transmission response of the corresponding

transducer changes only insignificantly and certainly is not unacceptably degraded. No disadvantages at all must be accepted with the invention.

In one embodiment of the invention, stub fingers that do not overlap the finger ends are situated opposite the opposing bus electrode. It is then advantageous to also round the ends of these stub fingers. At the second electrode comb, the field strength distribution is thus also kept constant along the peripheral line of the stub fingers, in particular at their ends (see also Figure 2 in this regard, which will be discussed in detail later) .

In a further embodiment of the invention, the arc resistance or the ESD protection is further increased by applying a dielectric onto the substrate, at least in the area of the gap, so that at least the gap is completely filled therewith. "In the area of the gap" here means in the transversal area of the transducer, between the ends of the electrode fingers of an electrode comb and the opposing bus electrode of the second electrode comb, or between the ends of the electrode fingers of an electrode comb and the opposing ends of non-overlapping electrode fingers (see also in this regard Figures 3 through 7, which will be discussed in detail later) .

The dielectric is applied so that the transversal overlap area - i.e., the area in which an electrode finger is adjacent, in the longitudinal direction, to an electrode finger of different potential - remains essentially free of dielectric and is not covered therewith. However, depending on the technology, smaller and unavoidable overlaps in the finger end area may be accepted. An additional ground load across the electrode fingers in the overlap area may otherwise lead to interference effects, in particular to the excitation of different modes or frequency deviations in this area.

In one embodiment of the invention, the dielectric is structured in the form of two parallel strips that respectively extend parallel to the longitudinal direction of the transducer and are arranged at least in the gap area. The gaps of the individual electrode finger pairs then must be situated at the same

transversal level in the gap area (see also in this regard

Figures 3, 4, and 6, which will be discussed in detail later) .

The invention may also advantageously be used in transducers that have an overlap weighting in which the overlap length of adjacent fingers varies over the length of the transducer, and in

particular forms a sinusoidal overlap area. Such transducers are also better protected against ESD discharges by the invention.

A strip-shaped structured dielectric may be applied in a strip wide enough so that, in addition to the gap area, it also covers an edge area of the transducer that comprises the non-overlapping stub finger, or even extends up to the edge of the bus electrode or beyond (see also in this regard Figures 4 and 6) .

However, it is also possible that the precise transversal position of the gap fluctuates, whereby the excitation of interfering modes in the gap area may be additionally

suppressed. The fluctuation of the gap position is small relative to the overlap length and is at most 5% of said overlap length. In this instance as well, the dielectric may be

structured in strips, and a slight overlap of the exciting electrode fingers in the overlap area may be accepted.

With regard to material and layer thickness, the dielectric is preferably chosen so that an acoustic wave experiences

approximately the same acoustic impedance in the transversal overlap area and in the gap area.

In a further embodiment, S1O₂, which may alternatively or

additionally also comprise silicon nitride, is used as a

dielectric. The metallization of the electrode fingers may comprise aluminum, copper, or titanium.

The metallization of the electrode fingers may have a multi-layer structure the different sub-layers of which have different components in pure form, i.e. in the form of a pure metal, or in the form of alloys.

The height of the dielectric layer may significantly exceed the height of the metallization. The dielectric layer advantageousl has at least the same height as the metallization. However, an improved ESD protection is also already achieved with a

dielectric layer that has a lower height than the metallization

The dielectric may then comprise S1O2 or consist of S1O2. Transducers according to the invention are applied onto piezoelectric substrates. An advantageously used substrate comprises lithium tantalate. An advantageous cut of the

monocrystalline lithium tantalate is chosen so that LSAW-type waves or leaky waves are generated.

For lithium tantalate, suitable crystal cuts are cuts rotated around the yx axes, wherein the cut angle is advantageously chosen between 39° and 46°. For example, cuts that are used are at approximately 39°, 42°, or 4

A transducer according to the invention may have an overlap area that has a transversal width of less than 25 λ, wherein λ is the wavelength of the acoustic wave. The width of the overlap area (measured in the transversal direction) , which corresponds to the aperture, may advantageously be chosen between 5 λ and < 25 λ.

SAW filters having improved ESD protection may be constructed from a transducer according to the invention. A SAW filter which comprises at least one transducer with rounded finger ends thus also lies within the scope of the invention.

The invention is advantageously used for resonator filters having resonators that comprise transducers with rounded finger ends. Such resonator filters may be DMS filters or reactance filters made from SAW single-gate resonators, which are known as ladder-type filters.

The invention will be explained in greater detail below with reference to exemplary embodiments and the associated figures. The Figures are only schematic and not true-to-scale, so that neither relative nor absolute dimensions can be learned from the Figures. Rather, for the sake of better clarity, individual parts may be illustrated at an enlarged or reduced scale.

Figure 1 shows, in schematic plan view, a transducer known per se and its division into the overlap area, gap areas and border areas,

Figure 2 shows, in schematic plan view, a transducer

according to the invention,

Figure 3 shows, in the same plan view, a transducer having a strip-shaped dielectric in the gap area,

Figure 4 shows, in the same plan view, a transducer having a strip-shaped dielectric in the gap area and in the adjoining border area,

Figure 5 shows, in the same plan view, a transducer having a dielectric which, in the gap area, covers only the regions between the finger ends of overlapping and non-overlapping electrode fingers,

Figure 6 shows, in plan view, a dielectric which covers the gap area and the adjoining border area and extends longitudinally beyond the adjoining reflectors,

Figure 7 shows, in plan view, a dielectric which covers the gap area and the adjoining border area and extends further across the entirety of the adjoining reflectors,

Figure 8 shows different sections in the transversal

direction through an electrode finger and the dielectric for different layer thicknesses of the dielectric, Figure 9 shows three different sections in the transversal direction through an electrode finger, having a metallization with a terminating edge that slopes down toward the substrate in a non-vertical direction,

Figure 10 shows two different sections in the transversal direction through an electrode finger, having metallization with a terminating edge with negative chamfer angle, sloping down toward the substrate in a non-vertical direction,

Figure 11 shows, in plan view, various embodiments of a dielectric applied in dots between the finger ends,

Figure 12 shows the real part and absolute value of the admittance of a transducer according to the invention in

comparison to a conventional transducer.

Figure 1 shows a transducer known per se in schematic plan view. The transducer comprises at least two bus electrodes BE from which respective electrode fingers EF extend in the transversal

direction. The two bus electrodes with the electrode fingers attached thereon respectively form an electrode comb.

In the transducer, two electrode combs are interleaved

interdigitally so that their electrode fingers overlap in an overlap area UB that ends transversally at the ends of the electrode fingers. A gap GP in which the metallizations of the two electrode combs typically have the smallest distance from one another in the transversal direction is formed between the ends of the electrode fingers and the bus electrode or the adjacent electrode comb.

Another stub finger SF, which has no overlap with an electrode finger of the respective other electrode comb, may be arranged between the gap GP and the nearest bus electrode BE. As shown in Figure 1, the gap is then formed between the ends of the

electrode fingers and the ends of the opposite stub fingers arranged at the same longitudinal position.

The entire transducer is then subdivided into the bus electrode BE, the non-overlapping edge area RB, the gap area GB, and the overlap area UB . As shown in Figure 1, in the border area, non- overlapping stub fingers may be situated opposite the

overlapping electrode fingers. In the event of omitted stub fingers, the border area RB and gap area are identical.

The gap area GB is then a rectangular area if all gaps are located at the same height transversally and have approximately the same transversal width. The drawn coordinate system shows that the transversal direction corresponds to the y-axis and that the longitudinal direction of the propagation direction of the surface acoustic wave corresponds to the x-axis.

Figure 2 shows a simple exemplary embodiment of the invention. Except for the shape of the finger ends, the transducer

corresponds to the known transducer according to Figure 1. In contrast to the previous rectangular shape, the electrode fingers EF according to the invention are rounded at the ends, just like the stub fingers SF. That means that they at least have no corners and edges in the xy plane (see coordinate system) . Rather, the boundary line at the finger ends

corresponds to a function whose first derivative is continuous. Since such finger ends have no point with strongly increased electrical field strength if they are at a potential, the ESD risk due to point discharge is significantly minimized. For the transducer according to the invention, it is even possible to reduce the size of the gap GP, and therefore to reduce the distance between electrode structures of different potential, without the breakdown voltage reaching a critical value.

Figure 3 shows a further embodiment of the invention in which one respective strip of a dielectric DK_S is arranged per gap area GB of the transducer. The overlap area UB is not covered by the dielectric DK_S, with the exception of a minimum area at the finger ends. The edge of the strip-shaped dielectric DK_S that faces toward the overlap area UB consequently terminates with the finger end of the overlapping finger EF.

The edge of the strip-shaped dielectric DK_S that faces toward the bus electrode BE here likewise terminates with the ends of the stub fingers SF; however, it may also partially overlap. It is obvious that a flush termination of finger ends and strip-shaped dielectric is only achieved when at least the electrode fingers have steeply falling edges, in the ideal case even vertically falling edges, and the edges of the strip-shaped dielectric follow the round contour of the ends of the electrode fingers.

Figure 4 shows a transducer according to the invention in plan view, in which the strip-shaped, structured dielectric DK_S also completely covers the border area RB of the transducer in addition to the gap area GB . Therefore, the entire transducer area is covered by the dielectric, with the exception of the bus electrode BE and the overlap area UB .

Figure 5 shows an embodiment of a transducer in which the

dielectric DK_F is structured in the form of dots and is arranged exclusively within the gaps. The dots DK_F are located in the gap area between finger ends of overlapping fingers EF and stub fingers SF, but not on the electrode fingers EF in the gap area. The width of the dots DK_F may vary, but preferably corresponds approximately to the width of the electrode fingers EF or approximately to the width of the stub fingers SF. In addition to the depicted structuring of the dielectric DK_F, it may also have a different structuring and, for example, may be applied as

rectangular dots in the gap area and across the tips of the electrode fingers EF.

Figure 6 shows an embodiment similar to that in Figure 4, with the difference that the strip-shaped dielectric DK_S extends in the longitudinal x-direction beyond the reflectors REF, which are arranged on both sides of the transducer in the longitudinal direction .

Figure 7 shows an embodiment similar to that in Figure 6, in which the dielectric DK additionally extends beyond the entire

reflectors together with their longitudinal distance areas.

Figure 8 shows three cross sections a through c in the

transversal direction through an electrode finger EF, the dielectric DK, and a stub finger SF. The z-axis shown in the Figure is the normal relative to the surface of the piezoelectric substrate. The three sections differ in the height of the applied dielectric DK. Whereas the layer thickness of the dielectric DK is smaller than the metallization height of the electrode finger EF in Figure 8A, it corresponds approximately to the

metallization height in Figure 8B. In Figure 8C, the dielectric DK has a greater layer thickness than the metallization of the electrode finger EF. The impedance of the dielectric strip or dot may be adjusted via the layer thickness. Additionally or

alternatively, this may be facilitated or may already be achieved via suitable material selection of the dielectric.

Figure 9 shows three additional different cross sections through electrode finger EF, dielectric DK, and stub finger SF. In this depiction, the cross section profiles of the electrode fingers are depicted closer to reality, meaning that the cross section profile of the electrode fingers does not slope down vertically toward the substrate at the end of the electrode finger as shown, but rather is beveled or rounded. Accordingly, the dielectric applied in the gap area follows this edge profile. In this way, in plan view, an overlap or blurry area UBR of electrode finger EF and dielectric DK results, and therefore there is no sharp division between the area covered by the dielectric and the gap area.

In the cross section depiction of Figure 9A, the dielectric DK_SiF fills the gap so that the edge profile of the DK_SiF corresponds to the edge profile at the ends of the electrode fingers EF. In plan view, a blurry region UBR results in which the diagonally trailing edges of electrode fingers EF and dielectric DK_SiF overlap, such that in plan view no clear separation is to be drawn between dielectric and electrode fingers. In the instances in which a blurry region UBR exists, by definition both gap area and overlap area UB end "indistinctly" within the blurry region BR since the boundaries are effectively "blurred" across the blurry area UBR.

Figure 9B shows an electrode finger EF likewise having

diagonally falling edge profile and a dielectric DK, which is applied in the gap area and additionally covers at least a portion of the border area RB . The edge of the dielectric that faces toward the overlap area UB slopes down with the same slope as the metallization at the end of the electrode finger EF, such that a blurry region UBR at the boundary between dielectric and electrode finger is formed here as well. In this embodiment, the dielectric now extends across the upper edge of the electrode finger end, i.e., ends in a blurry region UBR.

Figure 9C likewise shows a dielectric DK applied overlapping the electrode finger or electrode fingers EF in the blurry area UBR, but here with smaller layer thickness than the metallization of the electrode finger EF.

Figure 10 shows two additional possible cross sections through electrode finger EF, dielectric DK, and stub finger SF. Here the chamfer angle of the electrode finger is greater than 90°. This is always obtained if the metallization for the electrode fingers is applied onto an existing dielectric layer, having edges sloping down diagonally. In Figure 10A, dielectric DK and

electrode finger EF have the same layer thickness. In Figure 10B, dielectric DK and electrode finger EF have different layer thicknesses .

Figure 11 shows, in plan view, various embodiments of a

dielectric DK_F applied in the form of dots only between the finger ends. In Figure 11A, the dot-form dielectric is narrower than the electrode finger. In Figure 11B, the dot-form

dielectric and electrode finger are of the same width. The dielectric is thus arranged precisely between the electrode finger and stub finger, and only there. In Figure 11C, the dot- form dielectric is wider than the electrode finger.

Figure 12 shows in the upper section the real part, and in the lower section b the corresponding absolute value, of the

admittance of a transducer designed according to the invention, for example according to Figure 2, in comparison with that of a known transducer according to Figure 1. It is evident that the curves are nearly congruent. For the invention, no additional radiation losses or other disadvantageous variations occur compared to the known transducer. In contrast, in both

depictions, the peaks for a transducer according to the

invention are sharper at resonance frequency and formed with higher amplitude, which speaks in favor of a higher admittance of the transducer according to the invention. It is thus evident that, the finger end design according to the invention, which increases the ESD protection, not only does not cause any degradation of the electrical transducer response, such that known angular finger ends may be replaced by fingers with rounded ends without losses or other disadvantages; rather, the properties of a transducer according to the invention are

improved.

The invention is particularly advantageously used in resonator filters, in particular in DMS filters and ladder-type filters having SAW resonators.

Insofar as the electrode fingers are structured by means of photolithography, given transducers that are designed for higher frequencies, a suitably rounded shaping of the finger ends is no longer exactly possible due to the smaller structures and the limited resolution of the lithography. However, even

approximately rounded finger ends are already characterized by increased ESD resistance. However, structures for frequencies up to 2 GHz may typically be safely structured in the shape

according to the invention.

rence signs bus electrode dielectric layer electrode finger gap area

edge area reflector stub finger overlap area blurry area

Claims

1. An electroacoustic transducer,

- having two electrode combs arranged on a substrate, respectively having electrode fingers (EF) connected with a bus electrode (BE) , wherein the two electrode combs are arranged interleaved with one another such that their electrode fingers (EF) mutually overlap in a transversal overlap area (UB)

- wherein the ends of the electrode fingers facing away from the respective bus electrode (BE) are rounded.

2. The transducer according to the preceding claim,

- wherein the end of the electrode finger connected with a first electrode comb and the end of the electrode finger connected with a second electrode comb are respectively oppositely situated at a given longitudinal position

- wherein the oppositely situated ends of the electrode

fingers are both rounded.

3. The transducer according to one of the preceding claims,

- wherein a gap (GP) is formed between the ends of the electrode fingers (EF) of an electrode comb and the oppositely situated bus electrode (BE) or between the ends of the electrode fingers (EF) of an electrode comb and the oppositely situated ends of non-overlapping electrode fingers (EF)

- wherein a dielectric (DK) is applied onto the substrate such that the gap (GP) is completely filled therewith, but the transversal overlap area (UB) of the electrode fingers is not covered therewith.

4. The transducer according to the preceding claim, - wherein the dielectric (DK) is structured in the form of two parallel strips (DKS) that respectively extend parallel to the longitudinal direction of the transducer,

- wherein all gaps (GP) in a transversal gap area (GB) are arranged at the same transversal height

- wherein the dielectric covers over the gap area (GB) .

5. The transducer according to claim 4,

wherein the strips of the dielectric (DKS) are wide enough that they moreover extend beyond a border area (RB) of the transducer that comprises the non-overlapping stub fingers (SF) , or beyond the bus electrode (BE) .

6. The transducer according to one of the claims 3 through 5, wherein, with regard to material and layer thickness, the dielectric (DK) is chosen such that an acoustic wave

experiences approximately the same acoustic impedance in the transversal overlap area (UB) and in the gap area (GB) .

7. The transducer according to one of the claims 3 through 6,

- wherein the dielectric (DK) comprises S1O₂ or silicon

nitride

- wherein the metallization of the electrode fingers (EF) comprises Al, Cu, or Ti,

- wherein the metallization comprises a multi-layer

structure made up of the different components in pure form or in the form of alloys formed with one another

- wherein the height of the dielectric layer corresponds to 10-500% of the height of the metallization.

8. The transducer according to one of the claims 3 through 7,

- wherein the dielectric (DK) comprises S1O₂ or is composed of Si0₂ - wherein the height of the dielectric layer corresponds to 50-150% of the height of the metallization.

9. The transducer according to one of the preceding claims,

constructed on a substrate that comprises lithium tantalate.

10. The transducer according to claim 9,

wherein the lithium tantalate has a crystal cut

LT WI rot YX, wherein WI designates the cut angle, and wherein for WI it applies that 39° < WI < 46°, in particular wherein WI is selected from 39°, 42°, and 46°.

11. The transducer according to one of the preceding claims,

wherein the overlap area (UB) has a width of less than 25λ, wherein λ is the wavelength of the acoustic wave, wherein the aperture is preferably between 5λ and less than 25λ .

12. A frequency filter comprising at least one transducer

according to

any of the preceding claims.

13. The frequency filter according to claim 12,

formed as a resonator filter, having resonators, comprising transducers according to one of the preceding claims.