CN109766571B - FBAR design and design inspection method - Google Patents
FBAR design and design inspection method Download PDFInfo
- Publication number
- CN109766571B CN109766571B CN201811435373.6A CN201811435373A CN109766571B CN 109766571 B CN109766571 B CN 109766571B CN 201811435373 A CN201811435373 A CN 201811435373A CN 109766571 B CN109766571 B CN 109766571B
- Authority
- CN
- China
- Prior art keywords
- fbar
- thickness
- piezoelectric layer
- design
- filter
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Abstract
The invention discloses an FBAR design and design inspection method. The application requirements are converted into the input of technical requirements, a simulation model is established through a theoretical basis from key indexes, optimization is carried out, and then the initial sample is prepared. Design parameters are reversely pushed to the production product from theory to design and then to production, so that the design parameters can be compared with the design before preparation, and the mutual deduction relationship between the design and the product is established, namely, the design and the design inspection are realized. The invention provides a systematic and effective method for designing an FBAR (fiber Bragg reflector) applied to a filter and a power divider, so that the FBAR can be designed, some problems in the design implementation process can be solved, and the preparation implementation can be realized by combining with the consideration of the process, thus the design inspection part is included, and the recyclable flow between the design and the preparation can be further improved.
Description
Technical Field
The invention relates to the field of radio frequency technology and MENS technology, in particular to a method for designing and manufacturing an FBAR and a method for manufacturing an FBAR filter.
Background
To meet more communication functions and better communication experience, modern communication technologies need to meet the requirements of higher efficiency and integration. The filter constructed by the FBAR cascade has the advantages of high Q value, low insertion loss, good rectangular coefficient, good direction selectivity, good zero depth, good out-of-band rejection and the like, and is small in size, compatible with the COMS process and capable of being integrated. In addition, the working frequency can be high, so that the filter constructed by the FBAR cascade can well meet the requirements of modern communication technology.
FBAR (film bulk acoustic resonator) is a structured component having a piezoelectric effect material and being capable of forming an (inverse) piezoelectric effect structure. Are manufactured by means of MEMS technology and thin-film technology using silicon backplanes. The FBAR works on the principle that, in the core part of a "sandwich" structure of electrodes-piezoelectric material-electrodes, the piezoelectric material is deformed by applying a voltage to the electrodes, and when an alternating voltage is applied, the structure has an inverse piezoelectric effect. In the process, electric energy is converted into mechanical energy, the mechanical energy is transmitted in the structure through sound waves, vibration is caused, and meanwhile, an electric signal is generated through vibration, namely, the mechanical energy is converted into the electric energy through the piezoelectric effect, and the electric signal is output. The piezoelectric effect and the inverse piezoelectric effect exist at the same time, interact with each other and can generate resonance in the interaction process, so that the signal is selected.
The emergence of FBAR technology has a wide application prospect due to its high performance, miniaturization, integratability. Communication frequency allocations have become increasingly crowded today, with the gaps between the frequency spectra becoming smaller and even in close proximity, and FBAR-constructed filters are able to better address the spectrum crowding problem. FBARs have been confirmed to be widely used in the field of radio frequency communications, and the market has a great demand for products, and meanwhile, FBARs have a wide application prospect in other fields such as sensors.
However, the research and use of FBAR technology in China is still immature, and FBARs are involved in many aspects and have many technical difficulties including design and manufacturing process. There is a need for a systematic and effective approach to FBAR design that addresses some of the design issues, combines process considerations to enable manufacturing implementation, and enables further improvements in design and manufacturing cycles. The prior art does not provide an efficient method to implement.
Disclosure of Invention
Aiming at the problems existing in the design of the FBAR in the prior art, the embodiment of the invention provides a system and effective method for designing the FBAR and applying the FBAR to a filter and a power divider, so that the FBAR can be designed, some problems occurring in the implementation process in the design can be solved, and the preparation and implementation can be realized by combining the consideration process.
The invention provides an FBAR design and design inspection method, which comprises the following steps:
(1) determining the FBAR resonance frequency f according to the requirements of the FBAR filter on the working frequency and the cascade construction principle of the FBAR filtersAnd fp；
(2) Determining the theoretical electromechanical coupling coefficient of the FBAR piezoelectric material according to the resonant frequency of the FBAR
Selecting a piezoelectric material with the electromechanical coupling coefficient not less than the theoretical value as an FBAR piezoelectric layer material;
(3) according to the resonance condition of the ideal piezoelectric layer, the ideal piezoelectric layer thickness 2H is obtaineda；
(4) Determining the actual piezoelectric layer thickness 2h of the FBAR according to the process conditions and the influence of the thickness proportion of each layer on the performancea；
(5) Obtaining equivalent piezoelectric layer thicknesses of other FBAR oscillation layers through the difference between the ideal piezoelectric layer thickness and the actually determined piezoelectric thickness, and converting the equivalent piezoelectric layer thickness into the thicknesses of other oscillation layer materials at least comprising upper and lower electrodes according to the influence of the process and the proportional relation on the performance;
(6) constructing a Mason model of the FBAR in simulation software, adjusting the thickness to meet the requirement of working frequency, further cascading and constructing a filter, performing simulation optimization, meeting the performance requirement, and obtaining the thickness and the area of each layer of the FBAR;
(7) confirming and optimizing the structure and performance through FEM multidimensional simulation, finally determining the FBAR structure and size data, drawing a layout by combining the process, and preparing a mask plate;
(8) according to the process, the FBAR is prepared by the tape-out sheet, the FBAR is applied with a filter, MBVD parameters are extracted from the FBAR test, and the index of the filter is tested;
(9) constructing an MBVD model FBAR and a filter simulation in simulation software according to the extracted MBVD parameters, and carrying out design inspection or further optimization;
(10) and deducing parameters required by the Mason model through MBVD parameter data, substituting the parameters into the Mason model FBAR and a filter for simulation, comparing design data, and performing design inspection or further optimization.
The FBAR design and design inspection method provided by the invention is implemented by converting application requirements into the input of technical requirements, establishing a simulation model through a theoretical basis from key indexes, optimizing the simulation model, and preparing a primary sample. Design parameters are reversely pushed to the production product from theory to design and then to production, so that the design parameters can be compared with the design before preparation, and the mutual deduction relationship between the design and the product is established, namely, the design and the design inspection are realized. The FBAR design and design inspection method provided by the invention can start design from the principle and theory, reduces the design threshold of the FBAR, simultaneously enables the design and manufacture to be closely related, well finds the connection for the theory and practice, and enables the analysis and preparation of the FBAR and the filter to be effective from a plurality of entry points.
Preferably, the resonant frequency of the FBAR is determined, specifically:
for FBAR filters in a cascaded configuration, the initial design is primarily concerned with operating frequency requirements. The working frequency of the FBAR filter is closely related to the resonance frequency of the FBAR, and the difference between the working frequency band of the FBAR filter and the resonance frequency of the FBAR is at most twice as large as the difference of the resonance frequency of the FBAR according to the working frequency of the FBAR filter (it needs to be noted that, considering that in the actual process, the same wafer is generally provided with uniform and equal piezoelectric layers, the electromechanical coupling coefficients are kept consistent, and the difference between the resonance frequency of the FBAR at most twice as large is the condition that the bandwidth is not exchanged by sacrificing out-of-band or other performances;
for FBAR filters, the simplest cascaded structure is one series FBAR and one parallel FBAR. For series FBAR, parallel resonant frequency fpIs the highest frequency of the FBAR filter passband, series resonance frequency fsThe intermediate frequency of the passband of the FBAR filter; for parallel FBAR, parallel resonant frequency fpFor the intermediate frequency of the passband of the FBAR filter, the series resonance frequency fsDetermining the FBAR parallel resonance frequency f for the lowest frequency of the FBAR filter passbandpAnd series resonant frequency fs。
Preferably, the theoretical electromechanical coupling coefficient of the piezoelectric material is calculated according to the following formula:
The electromechanical coupling coefficient of the piezoelectric material can be preliminarily determined by substituting the determined resonance frequency of the FBAR resonator into equation (1). However, the electromechanical coupling coefficient calculated by the above equation at this time should be considered as ideal as a theoretical electromechanical coupling coefficient for reference, because of the determination of the resonance frequency, the rectangular coefficient has been set to 1. The rectangular coefficient of the FBAR filter designed in practice is difficult to be 1 and can only be close to 1, so that a material with a slightly larger electromechanical coupling coefficient can be selected, or the electromechanical coupling coefficient of the piezoelectric layer obtained by selecting the material is larger than that obtained by calculation. In the primary simulation design, the set electromechanical coupling coefficient data should be not less than the calculated data.
Preferably, according to an ideal piezoelectric resonance condition, an ideal piezoelectric layer thickness is obtained, specifically:
the ideal piezoelectric resonance condition formula satisfies:
and calculating to obtain the thickness of the ideal piezoelectric layer according to the ideal piezoelectric resonance condition formula:
where θ is the phase shift angle, k is the wave number, HaIs half of the ideal piezoelectric layer thickness, vaThe longitudinal wave velocity of the piezoelectric material is designed.
The ideal piezoelectric layer refers to that for a basic piezoelectric core structure consisting of an electrode layer, a piezoelectric layer and an electrode layer, sound waves are totally reflected from the surface, in contact with the piezoelectric layer, of the electrode layer to the other surface through the electrode layer, the electrode layer is set to be infinitely thin, namely the electrode layer exists, analysis is convenient, the thickness of the electrode layer is supposed to be infinitely thin, and therefore the thickness calculated according to a formula is the thickness of the piezoelectric layer. This is the first step, where a concept can be built in relation to frequency and thickness, which will then be assigned.
But the thickness of the electrode layer is ignored, so that the transmission path of the sound wave is limited in the piezoelectric layer.
The thickness of each layer can be considered to be distributed from the ideal thickness of the piezoelectric material, and the piezoelectric layer of the same material can be directly subtracted from the thickness, but the thickness of different materials needs to be converted by establishing a relation between the resonant frequency and the thickness of each layer, so that the conversion of the thickness of different materials and the thickness of the piezoelectric material distributed correspondingly is equivalent.
And determining the actual thickness of the FBAR piezoelectric layer according to the process conditions and the influence of the thickness proportion of each layer on the performance. The thickness of the piezoelectric layer must not be too thin or too thick, which can be a problem for the effectiveness of the piezoelectric effect, and too thick, the difficulty of manufacturing the piezoelectric layer increases, and for multilayer structures, the thickness of other layers is limited, which can cause other problems. The same is true of the electrodes, too thin a loss, too thick a capacitive effect, or a weakening of the resonance, the resonance frequency will also decrease. The thickness of each layer is within a suitable range, and within this suitable range, the thickness is selected based on the process recipe and the equipment performance. Taken together, for an initial design, the initial piezoelectric layer thickness may take one-half the value of the ideal calculated piezoelectric layer thickness.
Preferably, the equivalent piezoelectric layer thickness is the ideal piezoelectric layer thickness minus the initial piezoelectric layer thickness; the rest thickness, namely the equivalent piezoelectric thickness, is the thickness of the equivalent piezoelectric layer except the thickness of other structural layers of the piezoelectric layer, and then the equivalent piezoelectric layer thickness is converted into the thickness of other structural layers except the piezoelectric layer.
Through the equivalent piezoelectric layer thickness, combine the material that each layer chooseed for use, obtain other each layer material thickness outside the piezoelectric layer, specifically do:
by the formula
Or
Converting the equivalent piezoelectric layer thickness into the thickness of other layers except the piezoelectric layer; wherein, 2hn(n-1, 2,3 … …) thickness of other layer materials, vn(n is 1,2,3 … …) represents the longitudinal sound velocity of other materials of each layer, and the following relationship exists:
2Ha＝2ha+2h1+2h2+......+2hn(6)
2hais the initial piezoelectric layer thickness.
After the piezoelectric layer and the thickness are determined according to the FBAR structure distribution, for other layers, it is important to determine the material according to the function of the layer, so that the remaining equivalent piezoelectric layer thickness can be considered to be converted into the thickness of the corresponding material, i.e. the thickness of the corresponding layer.
Typically, the upper and lower electrodes are each halved with respect to the calculated electrode material thickness after dispensing. The thicknesses of other material layers can be proportionally distributed according to the process conditions and are given an initial value.
Preferably, in FBAR simulation software, a Mason model of the FBAR is constructed and simulated, specifically:
in ADS software, constructing an equivalent circuit expression of a Mason model of the FBAR according to the FBAR electrical impedance model;
the FBAR electrical impedance model expression is as follows:
where ω is the angular velocity, C0Is a static capacitance, ztIs the normalized acoustic impedance, z, of the interface of the piezoelectric film of the piezoelectric layer and the upper electrode layer looking upbThe normalized acoustic impedance of the piezoelectric film of the piezoelectric layer and the lower electrode layer interface looking down is mathematically transformed into:
wherein the content of the first and second substances,
Ztinput impedance of the upper surface of the piezoelectric film being the piezoelectric layer, ZpThe input impedance of the lower surface of the piezoelectric film of the piezoelectric layer. The equivalent expression is the Mason model equivalent circuit expression.
According to the formula (8), a Mason model is constructed for FBAR in ADS, and the material parameters of each layer are input.
According to the formula (4) or (5), the thickness of each layer of material is obtained through combined calculation (6), and the calculated values are equivalent to only considering the longitudinal wave sound velocity in the Mason model, but are the product of the density and the sound velocity for the acoustic impedance, so that for different materials, only considering the thickness data obtained through equivalent calculation of the factor of the sound velocity, the thickness data is substituted into the Mason model, the resonant frequency obtained through simulation operation can deviate from the design prediction, and the thickness parameters of the FBAR are adjusted in the Mason model.
Preferably, adjusting and optimizing the initial piezoelectric layer thickness and the thicknesses of the materials of the layers except the piezoelectric layer specifically include: and inputting the obtained thickness values of the selected materials and the thicknesses of the piezoelectric layer and other layers into a Mason model, and carrying out adaptive adjustment on the thicknesses of the piezoelectric layer and other layers under the condition of considering the influence factors of acoustic impedance.
Preferably, according to the cascade structure of the FBAR filter, simulation optimization is performed to obtain the thickness and area of each layer of the FBAR meeting the performance requirement, specifically: the filter is constructed by adopting a step cascade FBAR mode and comprises FBARs with at least two frequencies, wherein the high-frequency FBARs are used for series connection, the low-frequency FBARs are used for parallel connection, the areas of the two FBARs are given initial values, the cascade series is increased by adjusting the thickness and the area of each layer of the FBARs and increasing the number of the FBARs used for series connection and/or the FBARs used for parallel connection, manual tuning or target setting automatic optimization is adopted, the simulation is carried out to obtain the data meeting the performance requirement, and the thickness and the area of each layer of the FBARs.
When the filter is constructed by cascading FBARs, at least two frequencies of FBARs are included. In the most basic structure of the FBAR cascade structure filter, high frequencies are used in series, and low frequencies are used in parallel. The two areas are given an initial value, which should not be too large or too small, and may be the same.
Generally, a step-type cascade mode is adopted, the number of stages is sequentially increased, and the observation result is simulated after each increase until the bandwidth requirement is basically met. The number of stages should not be too many, otherwise, large insertion loss is introduced, the packaging area is increased, and the subsequent typesetting becomes complicated. And after the cascade connection is preliminarily determined, optimizing, adjusting the thickness ratio of each layer of a single FBAR and the area ratio of the series connection and the parallel connection, if the adjustment and optimization find that the area of a certain FBAR is overlarge, dividing the single FBAR on the series branch into two series FBARs, and dividing the single FBAR on the parallel connection into two parallel FBARs. For the adjustment of the amplitude, many changing techniques can be obtained by some basic rules, such as connecting an FBAR in parallel to a parallel branch, artificially manufacturing a zero point, improving out-of-band rejection, and at the same time, improving in-band insertion loss, etc., but many aspects are balanced, and the final purpose is to design a filter meeting the requirements, and also to consider the process conditions.
And according to FBAR size data and a cascading mode obtained in the ADS, modeling is continuously carried out in three-dimensional electromagnetic simulation software, and the simulation design obtains an optimization result closest to the ADS. After designing a filter meeting the requirements, the FBAR is also designed, and the layout is designed by combining the process scheme. Different process schemes exist, and a plurality of sets of layouts can be drawn.
A plurality of drawing software can realize the design of the layout, but the drawing layout is the embodiment of the process scheme by combining the process scheme, so that a set of process scheme can be designed by a plurality of layouts finally and is combined into a set of layout corresponding to the set of process scheme.
For the size unit which is convenient to measure, the size of the wafer is measured in inches, the larger size of the FBAR or the filter is also measured in microns, and the number of devices converted by the unit can be theoretically contained on the wafer, so if the research is carried out, a set of layout can contain various FBARs and filters, and the layout design is reasonable. In one set of layout, different layouts can be drawn for different FBARs or filters, but the process embodiment is complete, namely, only different graphs are processed, or different processes do not influence the established process flow, and all the processes belong to the category of one set of layout. Under the condition of research or reservation detection, the bare chip can be directly tested without considering packaging, so that the design of a feed port needs to be considered during the design of a layout.
Preparing a mask plate according to the drawn layout, formulating a process flow, and manufacturing the FBAR according to the process flow. The method comprises various FBARs and filters, wherein a layout is provided, a mask plate is prepared, and the subsequent process can be realized. The filter is some embodiments of the FBAR in these aspects in consideration of application and design check, so the filter designed in the process is also considered together, and the filter is observed and compared from multiple dimensions, is intuitive and checked mutually.
According to the designed layout, a mask plate is prepared, a process flow is formulated, and then the FBAR and the FBAR filter are produced and manufactured.
Designing a layout, and preparing a mask plate, wherein the prepared mask plate comprises a part of process for manufacturing the FBAR and the filter, and the actual production and manufacturing needs to be completed into a whole process flow. The requirement of testing the device can be met, namely the design can be checked, so that the process flow can be shortened to the production of the bare chip.
The production and manufacture are finished in a dust-free room, according to a formulated process flow, a drawn mask plate is combined, materials of all layers are sequentially grown on a substrate, the thickness is given according to simulation data, the depth of a cavity is given by about 2-3 um, and after the depth is determined, all FBARs are basically consistent, so that the process is convenient to realize. The preparation is the realization of the design, the two are closely combined, and other aspects are considered in the FBAR design method provided by the invention, and the system method is described here, and other related matters are not elaborated.
Preferably, after the FBAR is manufactured and the FBAR filter constructed by using the FBAR cascade is used, MBVD parameters are extracted from an FBAR test, and a filter index test specifically includes:
and selecting qualified samples, placing the samples on a probe platform, performing point contact, and obtaining test data through a vector network analyzer.
Extracting six parameters in an MBVD model from the FBAR sample, specifically:
(1) on the S11 graph, format selects log Mag, and sets maximum and minimum two tracked mark points, where the frequency points are respectively fpAnd fs. (Note, f)s<fp)
(2) At fsNear the point, the value is 3dB greater, encompassing the bandwidth, namely Qso, at fpNear the point, the value is 3dB less inclusive of bandwidth, which is Qpo.
(3) By
Deducing
r is a capacitance ratio, defined as
The two equations are compared and the subsequent Cm calculation is used.
(4) On the S11 graph, format selects Imaginary, on leaving harmonicThe area of vibration is marked with a plurality of make points on a relatively close (flat) curve. In general, the capacitance is expressed as
The values in the diagram are frequency dependent, i.e. the values of the corresponding frequency points in the diagram are actually expressions
Therefore, to calculate more accurate Co, each corresponding frequency value (note that GHz is converted into Hz) is substituted into the formula, and mark also has one result value (i.e. the result expressed by the formula), one Co is calculated, and the average value is calculated for several frequency points to obtain the final Co result.
(5) By defining the capacitance ratio, knowing r and Co, Cm can be calculated.
(6) According to the formula
Known as fsAnd Cm, Lm can be calculated.
(7) On the S11 graph, format chooses Real, marking a number of make points on a relatively close (flat) curve away from the region of resonance. The values of these markers are averaged to obtain Rs + Ro.
(8) Following a pure mathematical calculation, let Rso be Rs + Ro, and hence Ro be Rso-Rs. By substituting the Qs expression Qe into Qso and Qpo, Rm and Ro can be calculated, and finally Rs can be calculated. Wherein the equation comprising the Q values is given as follows:
and testing the indexes of the FBAR filter, placing the probe at the port of the filter, setting different S network parameters and mark points according to the technical index requirements, and directly testing to obtain the FBAR filter.
The data obtained by testing can be directly compared with the design data, and if deviation occurs, the test data needs to be converted into the design data again for design inspection.
Preferably, the parameters required by the Mason model are derived through MBVD parameter data, and are obtained through the following relations:
substituting the acquired parameters required by the Mason model into the simulation, comparing the results and analyzing the influence of numerical value change. Further, adjustments may be made or design parameters may be modified, i.e., redesigned, due to actual manufacturing-related actual property parameters of the material.
Preferably, the MBVD model is directly established by the extracted MBVD parameters, and the MBVD parameters are converted into a Mason model to perform FBAR and filter simulation equivalently. Furthermore, the FBAR parameters are adjusted or optimized and redesigned, the MBVD model and the Mason model are equivalent in circuit simulation, and the MBVD model and the Mason model can be used in a crossed mode according to needs. The method specifically comprises the following steps:
the method comprises the steps of constructing an FBAR MBVD circuit model, substituting extracted MBVD parameters into the model for simulation, manually tuning the parameters or automatically optimizing a set target, carrying out re-simulation design, converting the obtained parameters into an FEM multidimensional simulation, and carrying out subsequent steps.
Or converting the extracted MBVD parameters into parameters required in a Mason model, reconstructing the Mason model of the FBAR in simulation software, performing simulation optimization, and performing subsequent steps.
The steps do not necessarily need to be in the order of the steps given in the method, because some steps can be skipped from different states, i.e. different cut-in points, and the flow direction will be different for different purposes, so that various combinations are possible, and the steps comprise the main points in various situations, and the main points are selected and combined as required.
Therefore, the design analysis can be carried out on the physical device, the thickness size, the area, the distribution arrangement and the like are obtained by combining experimental and observation test means such as FIB and FEM, and the like, and the analysis result is obtained by cutting into the step points given in the method.
The invention has the following advantages:
1. the design threshold of the FBAR is reduced, the design can be started from the principle and the theory, and the connection is well found for the theory and the practice.
2. The invention is a systematic and progressive design method, and multiple layers related to FBAR can be quickly grasped and positioned by constructing a frame and aiming at a series of layers or all layers.
3. And the method provides ideas and methods for the design perfection and leaves an interface for the expansion. In actual research and development and preparation processes, problems or new situations can occur, the invention provides a design inspection method, and the design and design inspection form a feedback type process, thereby continuously promoting improvement.
4. The invention considers design and preparation at the same time and fuses application in the design concept, thus, the design, preparation and application are integrated into a whole, and on one hand, the invention can quickly adapt to the changing requirements of other aspects, which can quickly respond to the market and new technology.
Drawings
FIG. 1 is a schematic diagram of an FBAR structure;
FIG. 2 is a schematic flow chart of an FBAR design and design verification method of the present invention;
FIG. 3 is a schematic diagram of a cascade structure of an FBAR filter according to an embodiment of the present invention;
FIG. 4 is a diagram illustrating the operational effect of an FBAR filter according to an embodiment;
FIG. 5a is a Mason simulation diagram of one of the FBARs;
FIG. 5b is a diagram of a FBAR circuit modular simulation;
FIG. 6 is a diagram illustrating simulation results of an FBAR operation according to an embodiment;
fig. 7 is an MBVD simulation of an FBAR of an embodiment.
In the drawings: 11. a silicon substrate; 12. a cavity; 13. a support layer; 14. a lower electrode layer; 15. a piezoelectric layer; 16. an upper electrode layer; 17. a cavity structure.
Detailed Description
The present invention will be further described with reference to the accompanying drawings and the detailed description, and it should be noted that any combination of the embodiments or technical features described below can be used to form a new embodiment without conflict.
Fig. 1 shows a structure of an FBAR according to an embodiment of the present invention, which includes a silicon substrate 11 made of a single crystal, a cavity 12 etched on the silicon substrate 11, a supporting layer 13 disposed on the silicon substrate 11, and the supporting layer 13 and the silicon substrate 11 together form a closed space for the cavity 12. On the support layer is arranged a lower electrode layer 14, a piezoelectric layer 15, and on the piezoelectric layer an upper electrode layer 16. The upper electrode layer 16, the piezoelectric layer 15 and the lower electrode layer 14 form a core layer of electrode-piezoelectric layer-electrode. In particular, a cavity structure 17 is provided on the upper electrode layer 16, i.e. a cavity 12 is provided below the core layer, a cavity structure 17 is provided above the core layer, and the cavity 12 and the cavity structure 17 constitute a total reflection surface of the oscillating acoustic wave. According to other requirements and process conditions, other layers can be added between the electrode-piezoelectric layer-electrode core layer, such as a tuning layer or a protective layer for protecting from air can be further disposed on the upper electrode layer 16 in practical preparation. The support layer 13 is also not necessarily required under the lower electrode layer 14, i.e. the lower electrode and the piezoelectric layer are directly arranged on the silicon substrate 11, or a temperature compensation layer may also be arranged at the position of the support layer.
As shown in fig. 2, an FBAR design and design verification method provided in an embodiment of the present invention includes the following steps:
21. determining the FBAR resonance frequency f according to the requirements of the FBAR filter on the working frequency and the cascade construction principle of the FBAR filtersAnd fp；
22. Determining the theoretical electromechanical coupling coefficient of the FBAR piezoelectric material according to the resonant frequency of the FBAR
Selecting a piezoelectric material with the electromechanical coupling coefficient not less than the theoretical value as an FBAR piezoelectric layer material;
23. according to the resonance condition of the ideal piezoelectric layer, the ideal piezoelectric layer thickness 2H is obtaineda；
24. Determining the actual piezoelectric layer thickness 2h of the FBAR according to the process conditions and the influence of the thickness proportion of each layer on the performancea；
25. Obtaining equivalent piezoelectric layer thicknesses of other FBAR oscillation layers through the difference between the ideal piezoelectric layer thickness and the actually determined piezoelectric thickness, and converting the equivalent piezoelectric layer thickness into the thicknesses of other oscillation layer materials at least comprising upper and lower electrodes according to the influence of the process and the proportional relation on the performance;
26. constructing a Mason model of the FBAR in simulation software, adjusting the thickness to meet the requirement of working frequency, further cascading and constructing a filter, performing simulation optimization, meeting the performance requirement, and obtaining the thickness and the area of each layer of the FBAR;
27. confirming and optimizing the structure and performance through FEM multidimensional simulation, finally determining the FBAR structure and size data, drawing a layout by combining the process, and preparing a mask plate;
28. according to the process, the FBAR is prepared by the tape-out sheet, the FBAR is applied with a filter, MBVD parameters are extracted from the FBAR test, and the index of the filter is tested;
29. and deducing parameters required by the Mason model through MBVD parameter data, substituting the parameters into the Mason model FBAR and a filter for simulation, comparing design data, and performing design inspection or further optimization.
Preferably, the resonant frequency of the FBAR is determined, specifically:
for FBAR filters in a cascaded configuration, the initial design is primarily concerned with operating frequency requirements. The working frequency of the FBAR filter is closely related to the resonance frequency of the FBAR, and the difference between the working frequency band of the FBAR filter and the resonance frequency of the FBAR is at most twice as large as the difference of the resonance frequency of the FBAR according to the working frequency of the FBAR filter (it needs to be noted that, considering that in the actual process, the same wafer is generally provided with uniform and equal piezoelectric layers, the electromechanical coupling coefficients are kept consistent, and the difference between the resonance frequency of the FBAR at most twice as large is the condition that the bandwidth is not exchanged by sacrificing out-of-band or other performances;
for FBAR filters, the simplest cascaded structure is one series FBAR and one parallel FBAR. For series FBAR, parallel resonant frequency fpIs the highest frequency of the FBAR filter passband, series resonance frequency fsThe intermediate frequency of the passband of the FBAR filter; for parallel FBAR, parallel resonant frequency fpFor the intermediate frequency of the passband of the FBAR filter, the series resonance frequency fsDetermining the FBAR parallel resonance frequency f for the lowest frequency of the FBAR filter passbandpAnd series resonant frequency fs。
Preferably, the theoretical electromechanical coupling coefficient of the piezoelectric material is calculated according to the following formula:
The electromechanical coupling coefficient of the piezoelectric material can be preliminarily determined by substituting the determined resonance frequency of the FBAR resonator into equation (1). However, the electromechanical coupling coefficient calculated by the above equation at this time should be considered as ideal as a theoretical electromechanical coupling coefficient for reference, because of the determination of the resonance frequency, the rectangular coefficient has been set to 1. The rectangular coefficient of the FBAR filter designed in practice is difficult to be 1 and can only be close to 1, so that a material with a slightly larger electromechanical coupling coefficient can be selected, or the electromechanical coupling coefficient of the piezoelectric layer obtained by selecting the material is larger than that obtained by calculation. In the primary simulation design, the set electromechanical coupling coefficient data should be not less than the calculated data.
Preferably, according to an ideal piezoelectric resonance condition, an ideal piezoelectric layer thickness is obtained, specifically:
the ideal piezoelectric resonance condition formula satisfies:
and calculating to obtain the thickness of the ideal piezoelectric layer according to the ideal piezoelectric resonance condition formula:
where θ is the phase shift angle, k is the wave number, HaIs half of the ideal piezoelectric layer thickness, vaThe longitudinal wave velocity of the piezoelectric material is designed.
The ideal piezoelectric layer refers to that for a basic piezoelectric core structure consisting of an electrode layer, a piezoelectric layer and an electrode layer, sound waves are totally reflected from the surface, in contact with the piezoelectric layer, of the electrode layer to the other surface through the electrode layer, the electrode layer is set to be infinitely thin, namely the electrode layer exists, analysis is convenient, the thickness of the electrode layer is supposed to be infinitely thin, and therefore the thickness calculated according to a formula is the thickness of the piezoelectric layer. This is the first step, where a concept can be built in relation to frequency and thickness, which will then be assigned.
But the thickness of the electrode layer is ignored, so that the transmission path of the sound wave is limited in the piezoelectric layer.
The thickness of each layer can be considered to be distributed from the ideal thickness of the piezoelectric material, and the piezoelectric layer of the same material can be directly subtracted from the thickness, but the thickness of different materials needs to be converted by establishing a relation between the resonant frequency and the thickness of each layer, so that the conversion of the thickness of different materials and the thickness of the piezoelectric material distributed correspondingly is equivalent.
And determining the actual thickness of the FBAR piezoelectric layer according to the process conditions and the influence of the thickness proportion of each layer on the performance. The thickness of the piezoelectric layer must not be too thin or too thick, which can be a problem for the effectiveness of the piezoelectric effect, and too thick, the difficulty of manufacturing the piezoelectric layer increases, and for multilayer structures, the thickness of other layers is limited, which can cause other problems. The same is true of the electrodes, too thin a loss, too thick a capacitive effect, or a weakening of the resonance, the resonance frequency will also decrease. The thickness of each layer is within a suitable range, and within this suitable range, the thickness is selected based on the process recipe and the equipment performance. Taken together, for an initial design, the initial piezoelectric layer thickness may take one-half the value of the ideal calculated piezoelectric layer thickness.
Preferably, the equivalent piezoelectric layer thickness is the ideal piezoelectric layer thickness minus the initial piezoelectric layer thickness; the rest thickness, namely the equivalent piezoelectric thickness, is the thickness of the equivalent piezoelectric layer except the thickness of other structural layers of the piezoelectric layer, and then the equivalent piezoelectric layer thickness is converted into the thickness of other structural layers except the piezoelectric layer.
Through the equivalent piezoelectric layer thickness, combine the material that each layer chooseed for use, obtain other each layer material thickness outside the piezoelectric layer, specifically do:
by the formula
Or
Converting the equivalent piezoelectric layer thickness into the thickness of other layers except the piezoelectric layer; wherein, 2hn(n-1, 2,3 … …) thickness of other layer materials, vn(n is 1,2,3 … …) represents the longitudinal sound velocity of other materials of each layer, and the following relationship exists:
2Ha＝2ha+2h1+2h2+......+2hn(6)
2hais the initial piezoelectric layer thickness.
After the piezoelectric layer and the thickness are determined according to the FBAR structure distribution, for other layers, it is important to determine the material according to the function of the layer, so that the remaining equivalent piezoelectric layer thickness can be considered to be converted into the thickness of the corresponding material, i.e. the thickness of the corresponding layer.
Typically, the upper and lower electrodes are each halved with respect to the calculated electrode material thickness after dispensing. The thicknesses of other material layers can be proportionally distributed according to the process conditions and are given an initial value.
Preferably, in FBAR simulation software, a Mason model of the FBAR is constructed and simulated, specifically:
in ADS software, constructing an equivalent circuit expression of a Mason model of the FBAR according to the FBAR electrical impedance model;
the FBAR electrical impedance model expression is as follows:
where ω is the angular velocity, C0Is a static capacitance, ztIs the normalized acoustic impedance, z, of the interface of the piezoelectric film of the piezoelectric layer and the upper electrode layer looking upbThe normalized acoustic impedance of the piezoelectric film of the piezoelectric layer and the lower electrode layer interface looking down is mathematically transformed into:
wherein the content of the first and second substances,
Ztinput impedance of the upper surface of the piezoelectric film being the piezoelectric layer, ZpThe input impedance of the lower surface of the piezoelectric film of the piezoelectric layer. The equivalent expression is the Mason model equivalent circuit expression.
According to the formula (8), a Mason model is constructed for FBAR in ADS, and the material parameters of each layer are input.
According to the formula (4) or (5), the thickness of each layer of material is obtained through combined calculation (6), and the calculated values are equivalent to only considering the longitudinal wave sound velocity in the Mason model, but are the product of the density and the sound velocity for the acoustic impedance, so that for different materials, only considering the thickness data obtained through equivalent calculation of the factor of the sound velocity, the thickness data is substituted into the Mason model, the resonant frequency obtained through simulation operation can deviate from the design prediction, and the thickness parameters of the FBAR are adjusted in the Mason model.
Preferably, adjusting and optimizing the initial piezoelectric layer thickness and the thicknesses of the materials of the layers except the piezoelectric layer specifically include: and inputting the obtained thickness values of the selected materials and the thicknesses of the piezoelectric layer and other layers into a Mason model, and carrying out adaptive adjustment on the thicknesses of the piezoelectric layer and other layers under the condition of considering the influence factors of acoustic impedance.
Preferably, according to the cascade structure of the FBAR filter, simulation optimization is performed to obtain the thickness and area of each layer of the FBAR meeting the performance requirement, specifically: the filter is constructed by adopting a step cascade FBAR mode and comprises FBARs with at least two frequencies, wherein the high-frequency FBARs are used for series connection, the low-frequency FBARs are used for parallel connection, the areas of the two FBARs are given initial values, the cascade series is increased by adjusting the thickness and the area of each layer of the FBARs and increasing the number of the FBARs used for series connection and/or the FBARs used for parallel connection, manual tuning or target setting automatic optimization is adopted, the simulation is carried out to obtain the data meeting the performance requirement, and the thickness and the area of each layer of the FBARs.
When the filter is constructed by cascading FBARs, at least two frequencies of FBARs are included. In the most basic structure of the FBAR cascade structure filter, high frequencies are used in series, and low frequencies are used in parallel. The two areas are given an initial value, which should not be too large or too small, and may be the same.
Generally, a step-type cascade mode is adopted, the number of stages is sequentially increased, and the observation result is simulated after each increase until the bandwidth requirement is basically met. The number of stages should not be too many, otherwise, large insertion loss is introduced, the packaging area is increased, and the subsequent typesetting becomes complicated. And after the cascade connection is preliminarily determined, optimizing, adjusting the thickness ratio of each layer of a single FBAR and the area ratio of the series connection and the parallel connection, if the adjustment and optimization find that the area of a certain FBAR is too large, dividing the single FBAR into two series FBARs on a series branch, and dividing the single FBAR into two parallel FBARs on a parallel connection branch. For the adjustment of the amplitude, many changing techniques can be obtained by some basic rules, such as connecting an FBAR in parallel to a parallel branch, artificially manufacturing a zero point, improving out-of-band rejection, and at the same time, improving in-band insertion loss, etc., but many aspects are balanced, and the final purpose is to design a filter meeting the requirements, and also to consider the process conditions.
And according to FBAR size data and a cascading mode obtained in the ADS, modeling is continuously carried out in three-dimensional electromagnetic simulation software, and the simulation design obtains an optimization result closest to the ADS. After designing a filter meeting the requirements, the FBAR is also designed, and the layout is designed by combining the process scheme. Different process schemes exist, and a plurality of sets of layouts can be drawn.
A plurality of drawing software can realize the design of the layout, but the drawing layout is the embodiment of the process scheme by combining the process scheme, so that a set of process scheme can be designed by a plurality of layouts finally and is combined into a set of layout corresponding to the set of process scheme.
For the size unit which is convenient to measure, the size of the wafer is measured in inches, the larger size of the FBAR or the filter is also measured in microns, and the number of devices converted by the unit can be theoretically contained on the wafer, so if the research is carried out, a set of layout can contain various FBARs and filters, and the layout design is reasonable. In one set of layout, different layouts can be drawn for different FBARs or filters, but the process embodiment is complete, namely, only different graphs are processed, or different processes do not influence the established process flow, and all the processes belong to the category of one set of layout. Under the condition of research or reservation detection, the bare chip can be directly tested without considering packaging, so that the design of a feed port needs to be considered during the design of a layout.
Preparing a mask plate according to the drawn layout, formulating a process flow, and manufacturing the FBAR according to the process flow. The method comprises various FBARs and filters, wherein a layout is provided, a mask plate is prepared, and the subsequent process can be realized. The filter is some embodiments of the FBAR in these aspects in consideration of application and design check, so the filter designed in the process is also considered together, and the filter is observed and compared from multiple dimensions, is intuitive and checked mutually.
According to the designed layout, a mask plate is prepared, a process flow is formulated, and then the FBAR and the FBAR filter are produced and manufactured.
Designing a layout, and preparing a mask plate, wherein the prepared mask plate comprises a part of process for manufacturing the FBAR and the filter, and the actual production and manufacturing needs to be completed into a whole process flow. The requirement of testing the device can be met, namely the design can be checked, so that the process flow can be shortened to the production of the bare chip.
The production and manufacture are finished in a dust-free room, according to a formulated process flow, a drawn mask plate is combined, materials of all layers are sequentially grown on a substrate, the thickness is given according to simulation data, the depth of a cavity is given by about 2-3 um, and after the depth is determined, all FBARs are basically consistent, so that the process is convenient to realize. The preparation is the realization of the design, the two are closely combined, and other aspects are considered in the FBAR design method provided by the invention, and the system method is described here, and other related matters are not elaborated.
Preferably, after the FBAR is manufactured and the FBAR filter constructed by using the FBAR cascade is used, MBVD parameters are extracted from an FBAR test, and a filter index test specifically includes:
and selecting qualified samples, placing the samples on a probe platform, performing point contact, and obtaining test data through a vector network analyzer.
Extracting six parameters in an MBVD model from the FBAR sample, specifically:
(1) on the S11 graph, format selects log Mag, and sets maximum and minimum two tracked mark points, where the frequency points are respectively fpAnd fs. (Note, f)s<fp)
(2) At fsNear the point, the value is 3dB greater, encompassing the bandwidth, namely Qso, at fpNear the point, the value is 3dB less inclusive of bandwidth, which is Qpo.
(3) By
Deducing
r is a capacitance ratio, defined as
The two equations are compared and the subsequent Cm calculation is used.
(4) On the S11 graph, format selects Imaginary, marking several make points on a relatively close (flat) curve away from the region of resonance. In general, the capacitance is expressed as
The values in the diagram are frequency dependent, i.e. the values of the corresponding frequency points in the diagram are actually expressions
Therefore, to calculate more accurate Co, each corresponding frequency value (note that GHz is converted into Hz) is substituted into the formula, and mark also has one result value (i.e. the result expressed by the formula), one Co is calculated, and the average value is calculated for several frequency points to obtain the final Co result.
(5) By defining the capacitance ratio, knowing r and Co, Cm can be calculated.
(6) According to the formula
Known as fsAnd Cm, Lm can be calculated.
(7) On the S11 graph, format chooses Real, marking a number of make points on a relatively close (flat) curve away from the region of resonance. The values of these markers are averaged to obtain Rs + Ro.
(8) Following a pure mathematical calculation, let Rso be Rs + Ro, and hence Ro be Rso-Rs. By substituting the Qs expression Qe into Qso and Qpo, Rm and Ro can be calculated, and finally Rs can be calculated. Wherein the equation comprising the Q values is given as follows:
and testing the indexes of the FBAR filter, placing the probe at the port of the filter, setting different S network parameters and mark points according to the technical index requirements, and directly testing to obtain the FBAR filter.
The data obtained by testing can be directly compared with the design data, and if deviation occurs, the test data needs to be converted into the design data again for design inspection.
Preferably, the parameters required by the Mason model are derived through MBVD parameter data, and are obtained through the following relations:
and substituting the obtained parameters required by the Mason model into the simulation, and comparing the results. Theoretical electromechanical coupling coefficient
The following describes the FBAR design and design verification method provided by the present invention with a specific FBAR design example:
in the embodiments of the present invention, for the sake of simplicity and explanation of the method of the present invention, the oscillation layer structure only includes the core structure, upper electrode layer, piezoelectric layer, and lower electrode layer, and the materials are also given in the following description, where the piezoelectric layer is AlN and the materials of the upper and lower electrode layers are Mo. In addition, a design index of the FBAR filter of the embodiment is given: the working frequency range is 2300 MHz-2400 MHz, the insertion loss is less than or equal to 3dB, and the out-of-band rejection is greater than or equal to 30 dB.
The operating frequency requirements of the manufactured FBAR filter are designed according to requirements: the working frequency range is 2300 MHz-2400 MHz, the insertion loss is less than or equal to 3dB, the out-of-band rejection is greater than or equal to 30dB, and the fundamental frequency f of the resonator is determined by the principle that the FBAR cascade structure is an FBAR filterp. Depending on the operating frequency range, the series FBAR, parallel resonance frequency f can be obtained firstps2400MHz, series resonant frequency fss2350MHz, shunt FBAR, shunt resonance frequency fpp2350MHz, series resonant frequency fsp＝2300MHz。
Calculated according to the formula (1)
Here, the first and second liquid crystal display panels are,
it should be noted that the electromechanical coupling coefficients of the serial FBARs and the parallel FBARs are not differentiated, i.e., they are the same, and this is because the AlN layer is grown in the same process, and thus, it is not necessary to distinguish them. Also, the operating frequency was also halved analytically. This is quick, convenient, efficient and feasible.
However, since the case where the coefficient of the rectangle is 1 by default in the analysis is also considered, and in the actual production, the electromechanical coupling coefficient of AlN can be made higher, and therefore,
take 0.06.
The thickness of the piezoelectric layer in the ideal case is obtained according to the conditions for resonance of the piezoelectric layer in the ideal case. From equation (3), the ideal piezoelectric layer thickness is calculated, with the series FBAR:2Hap being 2.36um and the shunt FBAR:2Haps being 2.41 um.
The thickness of the piezoelectric layer is determined based on structural considerations and thickness ratio considerations. Here, series FBAR piezoelectric layer thickness 2hapTaking 1.2um, connecting FBAR piezoelectric layer in parallel for 2has1.2um is taken.
The ideal piezoelectric layer thickness minus the determined piezoelectric layer thickness, the remaining thickness is the equivalent piezoelectric layer thickness of the electrodes and other material layers, and therefore this thickness needs to be translated into the thickness of the electrodes and other material layers. A simplified process is performed on the example, the layer structure only remains the "sandwich" structure of the core work, thus the electrode Mo layer remains except the AlN layer.
The equivalent AlN layer thicknesses of the shunt-series FBARs are 1.21um and 1.16um respectively, and the equivalent AlN layer thicknesses are converted into the total electrode thickness by the formula (4) or (5), so that the total parallel electrode thickness is 2h1s659.56nm, total thickness of series electrode 2h1p＝659.39nm。
Considering the structure and the process, the influence of the thickness ratio on the electromechanical coupling coefficient and the possible subsequent frequency deviation, the tuning is needed, so the thicknesses of the upper and lower electrodes of the parallel FBARs are respectively 259.56nm and 400nm, and the thicknesses of the upper and lower electrodes of the series FBARs are respectively 259.39nm and 400 nm.
In ADS software, a Mason model is used for establishing simulation. And opening ADS software, constructing a Mason model, having a clear hierarchical structure, and inputting parameters of materials of each layer. Firstly, the most basic FBAR filter unit is constructed, and comprises a series FBAR and a parallel FBAR, and the initial area is set to 10000um2. However, in order to make the analysis not complicated during calculation, all the consideration factors are not considered, so in the Mason model, the thickness of each layer needs to be adjusted, that is, the FBAR is adjusted and corrected to be within the required frequency range.
And (4) constructing a filter in a cascade mode, determining the area and further optimizing the thickness dimension. The cascade structure filter sequentially increases the number of the FBARs according to the serial-parallel interval, but before the FBARs are increased, the cascade structure before the increase is optimally tuned, and the cascade structure filter can be optimized by changing the area and the thickness size of the upper electrode. The parallel FBARs are not required to be consistent and are also connected in series, but the inconsistency of the series connection cannot be too large, otherwise, the passband is influenced. The subsequent need to add FBARs may be the same as the previous stage or may be an initial value since the optimization needs to be continued. With the increase of the number of the FBARs, in-band ripples, out-of-band rejection, rectangular coefficients and the like can be improved, but in-band insertion loss can also be increased, so that the number of the cascaded FBARs is not too large, and the performance requirement is met by reducing the number of the FBARs as much as possible.
The design performance can be improved by particularly optimizing FBARs with different parallel branch parallel resonance points or series inductors, but the invention mainly designs the FBARs, and designs the FBAR filter only by application and does not go deep. Also here the design of the FBAR filter can be simplified.
In the embodiment, a design meeting requirements is obtained by simulating and optimizing the thickness and the area of the FBAR and constructing a cascade mode, a cascade construction diagram of the FBAR filter is shown in fig. 3, an operation result is shown in fig. 4, a Mason simulation diagram of one of the FBARs is shown in fig. 5a, parameters can be named by self, tuning and optimization functions are turned on, modularization is performed, cascading is facilitated, and an operation simulation result is shown in fig. 6.
And obtaining the optimal FBAR filter performance, namely confirming each parameter of the FBAR. And then, simulating in three-dimensional electromagnetic software, and showing the cascading mode in the spatial distribution of the material object with the optimal performance. After the filter is designed, parameters of the FBAR of the filter with the cascade structure are also determined, including the thickness and the area of each layer, the shapes of the upper electrode layer and the lower electrode layer and the spatial distribution, and then the layout is designed. On one layout, there may be FBARs and FBAR filters, because the design of FBARs is mainly considered, so the FBARs may be more distributed.
The layout design needs to consider more processes, so the layout is embodied in the processes to a certain extent, and the layout can be drawn to prepare the mask plate only by the determined processes.
FBARs were prepared. The FBAR is prepared by a set of process flow, the mask plate prepared by combining the drawn layout is combined, each layer of material is sequentially grown on the substrate, and the thickness is given according to the simulation data. Wherein the cavity depth is additionally given, set to 3 um.
And extracting six parameters of MBVD through testing. And preparing a batch of FBARs, marking each type number, and testing the sample on a probe platform by using a vector network instrument. And (3) carrying out test setting on the vector network instrument, selecting an S11 curve as an S parameter, obtaining calculation data required by six parameters in the MBVD model by changing a data format, and deducing through a corresponding calculation formula to obtain the calculation data. Here, one kind of MBVD of FBAR is given with six parameters: rs 1.01 Ohm; r0 ═ 0.85 Ohm; rm is 0.3376 Ohm; c0 ═ 1.3132 pF; cm 56.4 fF; lm is 83.4526 nH. The MBVD model simulation is shown in fig. 7.
Through the extracted six parameters, size data of the FBAR is deduced through a formula. In the process of extracting six parameters of the MBVD, the FBAR and the filter are tested and can be compared with a simulation result. However, the deviation of the actually prepared design is corrected, and it is also known how much error exists in the prepared design, so that the error is recovered by the formulas (9), (10), (11) and (12)
A，
And comparing the design before preparation. Here, it is calculated
A＝32623um2，
And (5) reestablishing the Mason model, and substituting the deduced data. Through the previous design and preparation and corresponding data comparison, preparation errors and design errors can be found, and then adjustment is carried out, or significant change and redesign are carried out. In this way, the above process is repeated, so that the design and preparation results are continuously approached.
The above embodiments are only preferred embodiments of the present invention, and the protection scope of the present invention is not limited thereby, and any insubstantial changes and substitutions made by those skilled in the art based on the present invention are within the protection scope of the present invention.
Claims (10)
1. An FBAR design and design verification method is characterized by comprising the following steps:
(1) determining the FBAR resonance frequency according to the working frequency requirement of the FBAR filter and the cascade construction principle of the FBAR filter;
(2) determining the theoretical electromechanical coupling coefficient k of the piezoelectric material of the FBAR according to the resonant frequency of the FBARt 2Selecting the electromechanical coupling coefficient not less than the theoretical electromechanical coupling coefficient kt 2The piezoelectric material of (3) is used as a material of the FBAR piezoelectric layer;
(3) obtaining the ideal piezoelectric layer thickness according to the ideal piezoelectric layer resonance condition;
(4) determining the thickness of the FBAR actual piezoelectric layer according to the process conditions and the relation of the thickness proportion of each layer of the FBAR oscillation film to the performance;
(5) obtaining equivalent piezoelectric layer thicknesses of other oscillation layers in the FBAR oscillation film through the difference value between the ideal piezoelectric layer thickness and the actual piezoelectric thickness, and converting the equivalent piezoelectric layer thickness into the thicknesses of other oscillation layer materials according to the relation between the process condition and the proportional relation and the performance, wherein the other oscillation layers at least comprise an upper electrode layer and a lower electrode layer;
(6) constructing a Mason model of the FBAR in simulation software, adjusting the thickness to meet the requirement of working frequency, further cascading and constructing a filter, performing simulation optimization, meeting the performance requirement, and obtaining the thickness and the area of each layer of the FBAR;
(7) confirming and optimizing the structure and performance through FEM multidimensional simulation, finally determining the FBAR structure and size data, drawing a layout by combining the process, and preparing a mask plate;
(8) according to the process, the FBAR is prepared by the tape-out sheet, the FBAR is applied with a filter, MBVD parameters are extracted from the FBAR test, and the index of the filter is tested;
(9) constructing an MBVD model FBAR and a filter simulation in simulation software according to the extracted MBVD parameters, and carrying out design inspection or further optimization;
(10) and deducing parameters required by the Mason model through MBVD parameter data, bringing the parameters into the Mason model FBAR and a filter for simulation, comparing design data, and performing design inspection or further optimization.
2. The method of claim 1, wherein determining the resonant frequency of the FBAR is by:
determining the difference of the resonant frequencies of the FBARs according to the working frequency of the FBAR filter and the difference of the maximum two times of the resonant frequencies of the FBARs in the working frequency band of the FBAR filter;
for series FBAR, parallel resonant frequency fpIs the highest frequency of the FBAR filter passband, series resonance frequency fsThe intermediate frequency of the passband of the FBAR filter; for parallel FBAR, parallel resonant frequency fpFor the intermediate frequency of the passband of the FBAR filter, the series resonance frequency fsDetermining the FBAR parallel resonance frequency f for the lowest frequency of the FBAR filter passbandpAnd series resonant frequency fs。
3. The method of claim 2, wherein the theoretical electromechanical coupling coefficient of the piezoelectric material is calculated according to the formula:
4. A method according to claim 3, characterized in that, according to ideal piezoelectric resonance conditions, an ideal piezoelectric layer thickness is obtained, in particular:
the ideal piezoelectric resonance condition formula satisfies:
and calculating to obtain the thickness of the ideal piezoelectric layer according to the ideal piezoelectric resonance condition formula:
where θ is the phase shift angle, k is the wave number, HaIs half of the ideal piezoelectric layer thickness, vaThe longitudinal wave velocity of the piezoelectric material is designed.
5. The method of claim 4, wherein the equivalent piezoelectric layer thickness is an ideal piezoelectric layer thickness minus an initial piezoelectric layer thickness;
through the equivalent piezoelectric layer thickness, combine the material that each layer chose for use, according to the influence of technology and proportional relation to the performance, obtain other each layer material thickness in addition the piezoelectric layer, specifically do:
by the formula
Or
Converting the equivalent piezoelectric layer thickness into the thickness of other layers except the piezoelectric layer; wherein, 2hn，n＝1,2,3……，2hnThickness of other layers of material, vn，n＝1,2,3……，vnRepresenting the longitudinal wave sound velocity of other materials of each layer, the following relationship exists:
2Ha＝2ha+2h1+2h2+......+2hn(6)
2hathe equivalent piezoelectric layer thickness at least includes the conversion of the thickness of the upper and lower electrode materials.
6. The method of claim 5, wherein in the FBAR simulation software, a Mason model of the FBAR is constructed and simulated, specifically:
in ADS software, constructing an equivalent circuit expression of a Mason model of the FBAR according to the FBAR electrical impedance model;
the FBAR electrical impedance model expression is as follows:
where ω is the angular velocity, C0Is a static capacitance, ztIs the normalized acoustic impedance, z, of the interface of the piezoelectric film of the piezoelectric layer and the upper electrode layer looking upbIs the normalization of the interface of the piezoelectric film of the piezoelectric layer and the lower electrode layer looking downAcoustic impedance, mathematically transformed, equivalent to:
wherein the content of the first and second substances,
Ztinput impedance of the upper surface of the piezoelectric film being the piezoelectric layer, ZpAn input impedance of a lower surface of the piezoelectric film being the piezoelectric layer;
and (4) converting the model into a Mason circuit model through a formula (8), and constructing an FBAR simulation model in circuit simulation software.
7. The method of claim 6, wherein adjusting the thickness to meet the operating frequency requirement comprises: and inputting the acquired material attribute parameters and material thickness values selected by the piezoelectric layer and other layers into a Mason model, and carrying out adaptive adjustment on the thicknesses of the piezoelectric layer and other layers under the condition of considering the influence factors of acoustic impedance.
8. The method of claim 7, wherein according to the cascade structure of the FBAR filter, simulation optimization is performed to obtain the thickness and area of each layer of the FBAR that meets the performance requirement, specifically: the filter is constructed by adopting a step cascade FBAR mode and comprises FBARs with at least two frequencies, wherein the high-frequency FBARs are used for series connection, the low-frequency FBARs are used for parallel connection, the areas of the two FBARs are given initial values, the cascade series is increased by adjusting the thickness and the area of each layer of the FBARs and increasing the number of the FBARs used for series connection and/or the FBARs used for parallel connection, manual tuning or target setting automatic optimization is adopted, the simulation is carried out to obtain the data meeting the performance requirement, and the thickness and the area of each layer of the FBARs.
9. The method of claim 8, wherein after the FBAR is manufactured and the FBAR filter is constructed using FBAR cascade, MBVD parameters are extracted for FBAR testing, and the filter index is tested by:
selecting qualified samples, placing the samples on a probe platform, performing point contact, and obtaining test data through a vector network analyzer;
extracting six parameters in an MBVD model from an FBAR sample, specifically:
(1) on the S11 graph, format selects logMag, and sets the maximum and minimum two tracked mark points, where the frequency points are respectively fpAnd fsWherein f iss<fp；
(2) At fsNear the point, the value is 3dB greater, encompassing the bandwidth, namely Qso, at fpNear the point, the value is smaller by the bandwidth contained by 3dB, which is Qpo;
(3) by
Deducing
r is a capacitance ratio, defined as
Comparing the two formulas, and using the subsequent Cm calculation;
(4) on the S11 graph, format selects Imaginary, marks several mark points on a flat curve in the area away from resonance; the capacitance is expressed as
The values in the diagram are frequency dependent, i.e. the values of the corresponding frequency points in the diagram are actually expressions
So that the exact C is calculated0Substituting each corresponding frequency value into the equation, and mark also has a result value, calculating a C0Calculating the average value of a plurality of frequency points to obtain the final C0The result is;
(5) by definition of the capacitance ratioFormula, known as r and C0Cm can be calculated;
(6) according to the formula
Known as fsAnd Cm, Lm can be calculated;
(7) on the S11 graph, format selects Real, marks a plurality of mark points on a section of similar curve in the area away from resonance, and averages the values of the mark points to obtain Rs + Ro;
(8) then, pure mathematical calculation is carried out, wherein Rso is Rs + Ro, so Ro is Rso-Rs, expressions of Qs and Qe are respectively substituted into expressions Qso and Qpo, Rm and Ro can be calculated, and finally Rs is calculated;
wherein the equation comprising the Q values is given as follows:
testing indexes of the FBAR filter, placing a probe at a port of the filter, setting different S network parameters and mark points according to the technical index requirements, and directly testing to obtain the indexes;
deriving parameters required by a Mason model through MBVD parameter data, and specifically obtaining the parameters through the following relation:
when carrying out final design inspection, directly establish the MBVD model by the MBVD parameter of drawing, and the MBVD parameter turns into the Mason model and carries out FBAR and filter simulation to be equivalent, and is further, adjusts or optimizes and redesign FBAR parameter again, and MBVD model and Mason model circuit simulation are equivalent, specifically do:
the method comprises the steps of constructing an MBVD circuit model of an FBAR, substituting extracted MBVD parameters into the model for simulation, manually tuning the parameters or automatically optimizing a set target, carrying out re-simulation design, converting the obtained parameters into an FEM multidimensional simulation, and carrying out subsequent steps; or converting the extracted MBVD parameters into parameters required in a Mason model, reconstructing the Mason model of the FBAR in simulation software, performing simulation optimization, and performing subsequent steps.
10. The method of claim 1, wherein steps (1) - (10) are arranged in several combinations; the arrangement of each step is given a flow direction, and the corresponding flow direction is reversible and can have several flow directions independently.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811435373.6A CN109766571B (en) | 2018-11-28 | 2018-11-28 | FBAR design and design inspection method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811435373.6A CN109766571B (en) | 2018-11-28 | 2018-11-28 | FBAR design and design inspection method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109766571A CN109766571A (en) | 2019-05-17 |
| CN109766571B true CN109766571B (en) | 2020-10-02 |
Family
ID=66450227
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811435373.6A Active CN109766571B (en) | 2018-11-28 | 2018-11-28 | FBAR design and design inspection method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109766571B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115169272A (en) * | 2022-06-16 | 2022-10-11 | 河源市艾佛光通科技有限公司 | Simulation model optimization method of bulk acoustic wave resonator |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN205657657U (en) * | 2016-05-18 | 2016-10-19 | 华南理工大学 | Film bulk acoustic wave syntonizer and wave filter of many mode of resonances |
| CN106160693A (en) * | 2016-07-12 | 2016-11-23 | 佛山市艾佛光通科技有限公司 | A kind of FBAR filter optimization method based on Mason model |
| CN107025321A (en) * | 2016-12-12 | 2017-08-08 | 佛山市艾佛光通科技有限公司 | The design and preparation method of cavity type FBAR filter |
| CN108023567A (en) * | 2016-10-31 | 2018-05-11 | 三星电机株式会社 | Wave filter |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106446321A (en) * | 2016-07-12 | 2017-02-22 | 佛山市艾佛光通科技有限公司 | Optimization method for physical parameters of film bulk acoustic resonator |
| CN106253871B (en) * | 2016-07-26 | 2018-12-28 | 中国工程物理研究院电子工程研究所 | A kind of the BAW ladder type filter and its manufacturing method of passband broadening |
-
2018
- 2018-11-28 CN CN201811435373.6A patent/CN109766571B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN205657657U (en) * | 2016-05-18 | 2016-10-19 | 华南理工大学 | Film bulk acoustic wave syntonizer and wave filter of many mode of resonances |
| CN106160693A (en) * | 2016-07-12 | 2016-11-23 | 佛山市艾佛光通科技有限公司 | A kind of FBAR filter optimization method based on Mason model |
| CN108023567A (en) * | 2016-10-31 | 2018-05-11 | 三星电机株式会社 | Wave filter |
| CN107025321A (en) * | 2016-12-12 | 2017-08-08 | 佛山市艾佛光通科技有限公司 | The design and preparation method of cavity type FBAR filter |
Non-Patent Citations (3)
| Title |
|---|
| BAW滤波器设计方法研究;贾乐;《中国优秀硕士学位论文全文数据库 信息科技辑》;20180815(第08期);第I135-150页 * |
| 基于单晶AlN薄膜的FBAR制备研究;刘国荣;《中国优秀硕士学位论文全文数据库 信息科技辑》;20180715(第07期);第I135-224页 * |
| 薄膜体声波谐振器（FBAR）技术的若干问题研究;金浩;《中国博士学位论文全文数据库 信息科技辑》;20060915(第09期);第I135-24页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109766571A (en) | 2019-05-17 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US10755021B2 (en) | Element removal design in microwave filters | |
| Yang et al. | High $ Q $ Antisymmetric Mode Lithium Niobate MEMS Resonators With Spurious Mitigation | |
| Bi et al. | Bulk acoustic wave RF technology | |
| KR102226347B1 (en) | Hierarchical Alignment of 2D Finite Element Method Simulation for Acoustic Wave Filter Devices | |
| US9369111B1 (en) | Fabrication of surface acoustic wave filters having plate modes | |
| CN110287653B (en) | Method for extracting material parameters of bulk acoustic wave resonator in longitudinal wave mode | |
| WO2023005167A1 (en) | Thin-film bulk acoustic resonant filter assembling and using method and electronic device | |
| CN112287635A (en) | FBAR filter optimization method based on improved Mason model | |
| Neculoiu et al. | Band pass filters based on GaN/Si lumped-element SAW resonators operating at frequencies above 5 GHz | |
| CN109766571B (en) | FBAR design and design inspection method | |
| AlJoumayly et al. | 5G BAW technology: Challenges and solutions | |
| Kim et al. | Highly miniaturized RF bandpass filter based on thin-film bulk acoustic-wave resonator for 5-GHz-band application | |
| WO2023241034A1 (en) | Simulation model optimization method for bulk acoustic resonator | |
| Wu et al. | A compact low-loss high-selectivity excellent-isolation FBAR duplexer integrated chip for LTE band-3 industrial applications | |
| CN115459730A (en) | Design method of bulk acoustic wave resonator, filter, duplexer and multiplexer | |
| Ding et al. | High-performance N79 band AlScN BAW resonator and filter with the consideration of area effect | |
| CN113541644A (en) | Direct and comprehensive design method for band-pass domain of acoustic wave filter | |
| Giraud et al. | Bulk acoustic wave filters synthesis and optimization for multi-standard communication terminals | |
| Milsom et al. | Combined acoustic-electromagnetic simulation of thin-film bulk acoustic wave filters | |
| Thakar et al. | Monolithic implementation of AlN-on-silicon bandpass filters with a high-Q notch within the passband | |
| KR100451169B1 (en) | Filter design method using generic algorithm | |
| Yen | Lithium Niobate Resonators and Filters Toward Reconfigurable Filter Banks | |
| Akatimagool et al. | The development of efficient cwfd simulation tools for waveguide band-pass filter design | |
| Zou et al. | PMBVD Model Parameter Extraction and Optimization for Film Bulk Acoustic Resonator | |
| Kim et al. | New modeling of TFBAR and on-wafer inductor effects on the TFBAR ladder filter performance |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| TA01 | Transfer of patent application right | ||
| TA01 | Transfer of patent application right |
Effective date of registration: 20200409 Address after: 510000 Room 303, building 1, No. 23, Jinzhong Road, Huangpu District, Guangzhou City, Guangdong Province Applicant after: Guangzhou Everbright Technology Co.,Ltd. Address before: 517000 Guangdong high tech Development Zone, Heyuan, new high tech five road Applicant before: HEYUAN CHOICORE PHOTOELECTRIC TECHNOLOGY Co.,Ltd. |
|
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| PE01 | Entry into force of the registration of the contract for pledge of patent right | ||
| PE01 | Entry into force of the registration of the contract for pledge of patent right |
Denomination of invention: FBAR design and design inspection method Effective date of registration: 20220926 Granted publication date: 20201002 Pledgee: Agricultural Bank of China Co.,Ltd. Heyuan Yuancheng District Sub branch Pledgor: Guangzhou Everbright Technology Co.,Ltd. Registration number: Y2022980016273 |