Rayleigh Wave Suppression in Al0.6Sc0.4N-on-SiC Resonators

Basic Information

• Paper ID: 2407.20286
• Title: Rayleigh Wave Suppression in Al0.6Sc0.4N-on-SiC Resonators
• Authors: Marco Liffredo, Silvan Stettler, Federico Peretti, Luis Guillermo Villanueva
• Classification: cond-mat.mtrl-sci (Condensed Matter Physics - Materials Science)
• Publication Date/Venue: July 2024 (arXiv preprint)
• Paper Link: https://arxiv.org/abs/2407.20286

Abstract

This paper reports the fabrication of a hybrid SAW/BAW resonator comprising a thin layer of scandium-doped aluminum nitride (AlScN) with 40 at% Sc concentration on a 4H-SiC substrate. The Sezawa mode is excited via vertical electric field, propagating longitudinal acoustic waves utilizing the d31 piezoelectric coefficient in AlScN. The resonant frequency is determined by the spacing of interdigital transducers (IDTs) defined through deep ultraviolet (DUV) lithography. Due to acoustic velocity mismatch between the piezoelectric material and substrate, the resonant modes propagate within the piezoelectric layer without leaking into the substrate. The study demonstrates the influence of piezoelectric layer and IDT layer thickness on the two observed modes, importantly showing how thin piezoelectric and electrode layers effectively suppress the Rayleigh mode.

Research Background and Motivation

Problem Context

1. 5G and Future Communication Requirements: With the deployment of 5G and future generations of communication technology, greater power density is required, increasing power handling and thermal dissipation demands in the acoustic filter design space.
2. Limitations of Existing Technologies:
   • Suspended resonators may suffer from thermal drift and nonlinearity at high power, as generated heat can only dissipate through thin anchor points
   • Standard surface acoustic wave (SAW) resonators have operational frequency limitations due to energy leakage into the substrate
   • Bulk acoustic wave (BAW) resonators require complex acoustic Bragg reflector layers
3. Research Motivation:
   • Develop non-suspended devices capable of effective heat dissipation
   • Achieve high-frequency operation while avoiding energy leakage
   • Leverage the superior piezoelectric properties of AlScN material

Technical Approach Selection

The authors selected a hybrid SAW/BAW resonator architecture, representing an evolution of the "third-class FBAR" concept proposed in 2013, combining advantages of both SAW and BAW resonators.

Core Contributions

1. Successful Fabrication of AlScN-on-SiC Hybrid Resonators: Achieved high-quality growth of 40% Sc-doped AlN thin films on 4H-SiC substrates
2. Effective Rayleigh Mode Suppression: Successfully suppressed unwanted Rayleigh modes through optimization of piezoelectric and electrode layer thickness
3. Verification of Sezawa Mode Wave Confinement: Demonstrated that Sezawa modes can be well-confined within the piezoelectric layer at appropriate geometric parameters
4. Establishment of Design Guidelines: Provided systematic analysis of layer thickness effects on mode behavior
5. Stopband Engineering Implementation: Demonstrated how to engineer stopband through IDT reflectivity adjustment to prevent unwanted mode excitation

Methodology Details

Device Design Principles

Material Selection

• Piezoelectric Material: Al0.6Sc0.4N, selected for:
  • Larger electromechanical coupling coefficient compared to standard AlN
  • Lower acoustic velocity (~9000 m/s), favorable for energy confinement
• Substrate Material: 4H-SiC, selected for:
  • Bulk shear wave velocity (~7000 m/s) faster than longitudinal wave velocity in AlScN
  • Readily available 4-inch wafers

Mode Analysis

Two primary modes exist in the device:

1. Rayleigh Mode: Low-frequency mode propagating primarily at the substrate-piezoelectric layer interface
2. Sezawa Mode: High-frequency mode propagating within the piezoelectric layer, the desired operating mode

Rayleigh Mode Suppression Mechanism

Rayleigh mode suppression is achieved by reducing:

• Piezoelectric layer thickness
• Top aluminum electrode thickness

Device Structure Design

Electrode Configuration

• Bottom Electrode: Ti adhesion layer + Pt electrode, formed as floating electrode only beneath IDT aperture
• Top Electrode: Al electrode forming IDT structure
• Electric Field Direction: Pure vertical field excitation utilizing d31 piezoelectric coefficient

Geometric Parameters

• Target Spacing: 500 nm (wavelength 1 μm)
• Piezoelectric Layer Thickness: Two designs with 250 nm and 150 nm
• Electrode Thickness: 100 nm for 250 nm device, 75 nm for 150 nm device

Fabrication Process

Process Flow

1. Bottom Electrode Preparation: Sputter Ti adhesion layer and Pt bottom electrode
2. Patterning: Lithography and etching to define bottom electrode pattern
3. Piezoelectric Layer Deposition: AlScN deposited at 300°C, Al top electrode at 100°C
4. Top Electrode Patterning: Cl2 ICP RIE etching with CF4/O2 mixture for photoresist removal

Material Quality Characterization

• 250 nm AlScN Film: X-ray diffraction rocking curve FWHM of 1.2°
• 150 nm AlScN Film: X-ray diffraction rocking curve FWHM of 1.6°
• SiC substrate demonstrates superior AlScN growth verticality compared to identical process on Si

Experimental Results

Simulation Verification

Simulation results confirmed design concepts:

• Thin piezoelectric and electrode layers effectively suppress Rayleigh modes
• Sezawa mode maintains good coupling at appropriate thickness
• Stopband position and width controllable through electrode thickness

Device Measurements

Measurement Setup

• GSG probes and RS-ZNB20 vector network analyzer
• Standard SOLT calibration procedure

Key Results

1. Rayleigh Mode Suppression: Experiments confirmed Rayleigh mode disappearance in thin piezoelectric layer devices
2. Quality Factor: Achieved loaded Q value of 390 at 6.3 GHz frequency
3. Coupling Coefficient: Measured keff² lower than simulated values, potentially due to fabrication challenges with ultra-thin AlScN layers

Performance Analysis

• Wave Confinement: IDT with 142 fingers and only 5 dummy electrodes on each side; wave confinement primarily from IDT grating itself
• Parasitic Modes: Observed parasitic modes originate from transverse wave numbers, suppressible through electrode weighting, apodization, or slanted fingers

Technical Innovations

Design Innovations

1. Simplified Structure: Compared to original hybrid SAW/BAW structures, eliminates need for defined piezoelectric pillars, simplifying to direct IDT placement on AlScN layer
2. Floating Bottom Electrode: Promotes formation of pure vertical electric field
3. Thickness Optimization Strategy: Systematic investigation of layer thickness effects on mode behavior

Material Innovations

1. High Sc Content AlScN: 40% Sc doping concentration achieving high-quality growth on SiC substrate
2. Material Matching: AlScN/SiC material combination achieving ideal acoustic velocity matching conditions

Process Innovations

1. Simplified Manufacturing Flow: Eliminates release steps compared to suspended structures
2. DUV Lithography: Enables fine patterning at 500 nm spacing

In-Depth Evaluation

Strengths

Technical Advantages

1. Successful Proof-of-Concept: Successfully demonstrated feasibility of AlScN-on-SiC hybrid resonators
2. Theory-Experiment Consistency: Simulation predictions highly consistent with experimental results in mode suppression
3. Simplified Manufacturing Process: Simpler process flow compared to traditional FBAR devices
4. Excellent Material Quality: Achieved high-quality AlScN thin film growth on SiC substrate

Academic Contributions

1. Systematic Analysis: Comprehensive analysis of layer thickness effects on mode behavior
2. Design Guidance: Established important design criteria for hybrid resonators
3. Material Combination Innovation: First systematic investigation of AlScN/SiC material combination in resonator applications

Limitations

Technical Constraints

1. Low Coupling Coefficient: Measured keff² significantly lower than simulated values, affecting device performance
2. Thin Film Quality Challenges: Deposition quality of ultra-thin AlScN layers requires improvement
3. Frequency Limitations: Must account for mode dispersion to avoid bulk radiation

Experimental Gaps

1. Limited Device Samples: Results presented for only two thickness configurations
2. Long-term Stability: No long-term stability and reliability data provided
3. Power Handling Capability: Although power handling advantages mentioned, specific test data not provided

Impact Assessment

Academic Impact

1. Field Advancement: Provides important experimental validation for hybrid SAW/BAW resonator field
2. Material Applications: Expands AlScN material applications in high-frequency resonators
3. Design Methodology: Establishes methodological foundation for such device design

Practical Value

1. 5G Application Potential: Offers new resonator solutions for 5G and future communication systems
2. Manufacturing Feasibility: Simplified process flow facilitates commercial production
3. Performance Trade-offs: Achieves good balance between frequency, quality factor, and manufacturing complexity

Applicable Scenarios

1. High-Frequency Communication Systems: Suitable for RF front-ends requiring GHz-band resonators
2. Power Handling Applications: Suitable for high-power applications requiring excellent thermal dissipation
3. Integrated Circuits: Resonator solutions compatible with existing semiconductor processes

Conclusions and Outlook

Main Conclusions

1. Successfully demonstrated Rayleigh mode suppression in AlScN-on-SiC hybrid resonators
2. Verified feasibility of mode engineering through layer thickness optimization
3. Established design and manufacturing methodology for such devices

Future Directions

1. Material Optimization: Improve deposition quality of ultra-thin AlScN layers to enhance coupling coefficient
2. Device Optimization: Explore different IDT designs to further improve performance
3. Application Extension: Extend technology to higher frequencies and higher power applications
4. Reliability Research: Conduct long-term stability and reliability assessment

This work makes important contributions to hybrid SAW/BAW resonator technology development, particularly providing valuable insights in material selection, device design, and manufacturing processes, laying the foundation for future high-performance RF device development.