Search

In this study, we focus on the development of key processes towards wafer-scale 3-dimentional/vertical (3D) integration of Indium-Phosphide (InP) photonic membranes on InP electronics via adhesive bonding. First, we identified the most critical steps and optimized them to achieve high thermal and mechanical compatibility of components for the co-integration process. Next, we developed a strategy for InP-to-InP wafer bonding with high topology tolerance, and introduced hard benzocyclobutene (BCB) anchors to preserve the alignment and BCB thickness uniformity after bonding. The resulting bond layer is homogeneous in terms of physical and mechanical properties. Finally, we developed a novel method to selectively remove the InP substrate from the photonics side via wet etching while protecting the electronics carrier wafer with hermetic multi-layer coatings. The investigation of these key steps is essential for scalable 3D integration of photonics and electronics at ultra short distances (< $15 ~\mu \text{m}$). [ABSTRACT FROM AUTHOR]

MEMS resonators with interdigital transducer (IDT) based on aluminum nitride (AlN) thin-film (a few micrometers thickness) materials gradually arouse researchers’ interest due to their ability to excite multiple modes of acoustic waves. However, the low electromechanical coupling coefficient is still one of the main reasons that limit the AlN-based resonators. In this article, we demonstrate a design journey for 10% scandium-doped AlN-based dual-mode acoustic wave resonators. Also, we investigated the influence of stack architectures with and without introduced AlN interlays, as well as the influence of stacked film thicknesses, on the performance of resonators. Simulation results show that changes in the SiO2 film thickness have different effects on the TCF of the two acoustic wave modes, providing a way for sensors to decouple two parameters that jointly affect the frequency drift. Moreover, measured results show that the introduced AlN interlayer (AlN-IL) induces a better crystalline orientation of the scandium-doped AlN (AlScN) thin film and a higher coupling coefficient of the leaky longitudinal (LL) mode (${K}^{{2}}$ : from 4.19% to 4.96%), which make the resonators have great potential in surface acoustic wave (SAW)-based applications such as sensors and high-frequency applications. [ABSTRACT FROM AUTHOR]

Network slicing is the key to enable virtualized resource sharing among vertical industries in the era of 5G communication. Efficient resource allocation is of vital importance to realize network slicing in real-world business scenarios. To deal with the high algorithm complexity, privacy leakage, and unrealistic offline setting of current network slicing algorithms, in this paper we propose a fully decentralized and low-complexity online algorithm, DPoS, for multi-resource slicing. We first formulate the problem as a global social welfare maximization problem. Next, we design the online algorithm DPoS based on the primal-dual approach and posted price mechanism. In DPoS, each tenant is incentivized to make its own decision based on its true preferences without disclosing any private information to the mobile virtual network operator and other tenants. We provide a rigorous theoretical analysis to show that DPoS has the optimal competitive ratio when the cost function of each resource is linear. Extensive simulation experiments are conducted to evaluate the performance of DPoS. The results show that DPoS can not only achieve close-to-offline-optimal performance, but also have low algorithmic overheads. [ABSTRACT FROM AUTHOR]

In this letter, HfO2/Al2O3/AlN hybrid stack capacitor with ferroelectric switching is experimentally demonstrated for the first time. The metal-insulator-semiconductor (MIS) capacitors and InGaZnO4 (IGZO) channel thin-film transistors (TFTs) with the HfO2/Al2O3/AlN hybrid gate stack are demonstrated in this letter. With the hybrid gate stack, the TFT exhibits the clear counter-clockwise memory windows (MWs) of 6.5 V, 6.2 V and 28.7 V respectively measured at 368 K, RT, and 80 K under the sweep voltage of ±40 V. Moreover, the retention over $10^{{4}}$ s and the endurance over $10^{{4}}$ cycles are demonstrated under the program (P)/erase (E) pulsed height of ±45 V. The results exhibit the applicable possibility of the HfO2/Al2O3/AlN stack for the high voltage memory device applications. [ABSTRACT FROM AUTHOR]

This letter reports on the demonstration of recessed-gate ultra-wide bandgap semiconductor Al0.6Ga0.4N MOSFET with Hf0.5Zr0.5O2 (HZO) ferroelectric charge storage dielectric structure. A positive threshold voltage ($\text{V}_{\text {TH}}$) of 3.550 V and maximum gate swing of 15 V is achieved due to the trapped electrons induced by remnant polarization of HZO in gate stack. The Al0.6Ga0.4N MOSFET exhibits a specific on-resistance ($\text{R}_{\text {on,sp}}$) of ${17.8} \text {m}~\Omega \cdot \text {cm}^{{2}}$. Moreover, an 840 V breakdown voltage (BV) is obtained with a gate-to-drain ($\text{L}_{\text {gd}}$) length of $2 ~\mu \text{m}$ , indicating an average breakdown electric field ($\text{E}_{\text {av}}$) beyond 4 MV/cm. In addition to the high $\text{E}_{\text {av}}$ , a high BV > 3 kV is also derived on a transistor with $\text {L}_{\text {gd}} = {8}\,\,\mu \text {m}$ , translating to a power figure-of-merit (P-FOM=BV2/ $\text{R}_{\text {on,sp}}$) of 208 MW/cm2. These high device performances indicate a great potential of Al-rich E-mode AlGaN MOSFET for future high-power applications. [ABSTRACT FROM AUTHOR]

This article presents the lamb wave contour mode resonators (CMRs) based on 22% aluminum scandium nitride (AlScN) thin film with $\lambda $ of 12–24 $\mu \text{m}$ , and operating in $\text{S}_{{0}}$ mode. We report the design, fabrication, and characterization of 500 nm-thick AlScN CMRs, which take advantage of optimized stress control of co-sputtered AlScN thin films and vertical inductively coupled plasma (ICP) etching profile. The experimental results are compared to theoretical predictions by finite element analysis (FEA). All Al $_{{0.78}}$ Sc $_{{0.22}}\text{N}$ devices show excellent agreement with simulations in piezoelectric coupling using modified AlScN film parameters. The best Al $_{{0.78}}$ Sc $_{{0.22}}\text{N}$ CMR has achieved an electromechanical coupling coefficient (${k}_{t}^{{2}}{)}$ of 5.24% and loaded quality factor (${Q}$) of 1219 with an operating frequency at approximately 300 MHz, which exhibits a high Figure-of-Merit (FoM) of 63.88 in piezoelectric microelectromechanical system (MEMS) lamb wave CMR. This article also presents the co-sputtering characteristics of the AlScN thin films under $\text{N}_{{2}}$ gas to achieve low-stress and high-quality piezoelectric materials, and the etching optimization of high concentration Sc doping aluminum nitride (AlN) thin films under Cl2/BCl3/Ar chemistry to obtain record profile angle of 77°, high selectivity of 1:1 with SiO2 hard mask. [ABSTRACT FROM AUTHOR]

Polycrystalline aluminum nitride has extensive application prospects in substrate materials, insulating layer materials, and packaging materials. Polycrystalline aluminum nitride is sintered from powder. The traditional mechanical polishing method is easy to cause aluminum nitride grains to fall off to form pits. In this paper, to improve the polishing effect, based on the hydration of polycrystalline aluminum nitride, a sol-gel (SG) polishing tool was proposed to polish polycrystalline aluminum nitride. The polishing process was divided into rough polishing and fine polishing. The diamond particle size of the SG tool used for rough polishing was $40\mathbf {\mathrm {\mu }}\text{m}$. The diamond particle size of the SG tool used for fine polishing was $5\mathbf {\mathrm {\mu }}\text{m}$. Moreover, the polycrystalline AlN was processed with or without deionized water. The results showed that in the rough polishing stage, the material removal rate (MRR) of the polishing method without deionized water was four times higher than that of the polishing method with deionized water. However, in the fine polishing stage, the MRR of the polishing method with deionized water was slightly higher than that of the polishing method without deionized water. The results showed that mechanical removal is the primary material removal method in the rough polishing stage. In the fine polishing stage, hydration can improve the polishing effect of polycrystalline aluminum nitride. [ABSTRACT FROM AUTHOR]

This letter presents a $D$ -band low-noise amplifier (LNA) in 250-nm InP HBT technology for the next-generation wireless applications. The LNA has a measured peak gain of 13 dB, a 3-dB bandwidth greater than 20 GHz (120–140 GHz), and a measured noise figure (NF) of less than 6 dB in the band. A reduction in the 3-dB bandwidth from simulation was observed during the measurements which was attributed to the substrate waves using full chip electromagnetic (EM) simulation. EM simulations show that a partial or complete removal of the back side metallization of the InP substrate, holes in metal-1 ground plane, or a strategic placement of through-substrate vias suppress these substrate waves. To the authors’ knowledge, this is the first 120–140-GHz LNA in the InP 250-nm HBT technology. [ABSTRACT FROM AUTHOR]

In this brief, a high-speed wideband track-and-hold amplifier (THA) is implemented in a $0.8~{\mu }\text{m}$ InP double-heterojunction bipolar transistor (DHBT) process. This THA is based on a base-collector (b-c) switching diode architecture and achieves a −3 dB bandwidth of 35 GHz at a sampling rate of 30 GSa/s. For the first time, an enhanced unity-gain frequency (${f_{T}}$) doubler and an active peaking technique are used together in the design of the b-c switching diode sampler. The dominant pole of the THA is canceled by one zero produced by the active peaking structure. This is one step to improve the bandwidth. Secondly, the enhanced ${f_{T}}$ -doubler structure improves the driving capability of the input buffer at high frequencies. Therefore, the bandwidth of the THA is further extended. The measurement results show a total harmonic distortion (THD) of <−30.0 dBc and a maximum spurious-free dynamic range (SFDR) of 47.07 dB at a 30-GSa/s sampling rate. S-parameter tests show a better than −43 dB isolation within 45 GHz in hold-mode, and a −3 dB bandwidth of 35 GHz in track-mode under a 165 GHz device ${f_{T}}$. The proposed THA has the potential for high-speed broadband applications due to its competitive performance. [ABSTRACT FROM AUTHOR]

In this article, a unique structure of InP/InGaAs uni-traveling-carrier photodiode (UTC-PD) is proposed. Compared with the one-sided junction photodiode, the UTC-PD has the advantages of a simpler epitaxial layer structure while maintaining the characteristics of high bandwidth and high output power loss profile. Simulated results show that a large built-in electric field can be generated under illumination, which aids UTC carrier velocity. The merits of the new structure are compared to a standard UTC-PD photodiode in terms of improved electric field and carrier concentration response. It is demonstrated that the photogeneration and light absorption of UTC-PDs are improved by incorporating a step-like internal texture of cones and dots profile in the photo absorption layer. The simulated device shows a peak electrical 3-dB bandwidth of 65 GHz at a low light intensity and reverse bias voltage. The performance characteristics of 1-D and 2-D UTC-PD simulations, including internal electric field distribution, energy band diagram, carrier concentration, and power loss distribution, are carefully studied. A theoretical discussion of the working principle and key performance characteristics of a UTC-PD enhanced by heterojunction design is reported. The entire physical environment is modeled and simulated through the finite element method (FEM) using commercial software. The proposed photodiode structure configuration is designed and optimized for photodetectors for high RF frequencies at different light intensities. To the best of our knowledge, the obtained bandwidth and electric field response is the fastest among those reported for other various higher wavelength photodiodes. [ABSTRACT FROM AUTHOR]

This work demonstrates the high-frequency characteristics of In0.53Ga0.47As nanowire with scaled wire width by implementing TCAD simulations. The physical models and correlated parameters have been calibrated to the experiments (Ko et al., 2022). As the width of the nanowire is scaled to 10 nm, the electron density peaks are no longer located close to the oxide/semiconductor interface. Instead, the peaks merge and volume inversion effects appear due to the strong quantum confinement. The volume inversion effects lead to higher cut-off frequency due to the reduced total transport delay time. To have a better understanding of this phenomena, the high-frequency properties of nanowire were quantified with the assistance of small-signal analysis and delay time analysis using TCAD. It is found that the channel charging delay increases with narrower wire width due to the raise of source/drain resistance. Regarding the extrinsic and intrinsic delay, they increase with smaller wire width and drop at width of 10 nm due to the volume inversion effects. Electron distribution which aims to clarify the above-mentioned observation is also plotted.

Power electronic systems employing wide-bandgap GaN transistors promise high efficiency operation and power density but require minimized parasitic circuit elements and an effective cooling concept. This article presents a half-bridge module integrating two 600 V/170 mΩ gallium nitride (GaN) high-electron mobility transistors with their gate drive stages and a fraction of the dc-link capacitance on a patterned multilayer polycrystalline AlN-substrate. The high-voltage isolation at a layer distance of 10 μm and a dense chip-by-chip integration on the GaN half-bridge module enable a compact lateral commutation loop design combined with improved cooling capability of the power transistors. Consequently, the GaN half-bridge module allows for higher load currents at lower device temperature while most parasitic circuit elements are reduced compared to a conventional printed circuit board (PCB) design. The parasitic circuit elements of the GaN half-bridge module and a reference four-layer PCB half-bridge are evaluated using 3D-FEM field simulation and in-circuit measurements. Selected finite element method (FEM) simulation results are validated by S-parameter measurements and further used to parametrize a lumped commutation loop model. The thermal characterization of the GaN half-bridge module validates the improved cooling capability of the GaN half-bridge power module. Transient switching characteristics are studied in hard-switching mode. The device temperature and converter efficiency are evaluated in dc/dc buck-converter operation. [ABSTRACT FROM AUTHOR]

This letter presents a piezoelectric tuning fork resonator vibrating in in-plane flexural mode with a high quality factor (Q) and low motional impedance. A novel double-sided actuating design is implemented based on unimorph structure with a 2.5- $\mu \text{m}$ -thick AlN film as passive layer. The design significantly improves the Q and motional impedance compared with the conventional single-sided actuating design thanks to an increase of static capacitance without change of material interfaces. The proposed double-sided actuating device shows an excellent Q measured as 11272 with an extracted motional impedance of 7.2 $\text{k}\Omega $ in vacuum. The measured Q-pressure relationship shows that a vacuum level <100 Pa would ensure a sufficiently high Q value, which can be easily achieved through wafer-to-wafer bonding with getters in the future. Furthermore, the impact of AlN passive layer thickness on the resonator performance is analyzed. The proposed device would find applications in sensing fields as miniaturized and high-performance resonator-based solutions, such as gyroscopes, accelerometers and pressure sensors. [ABSTRACT FROM AUTHOR]

Low-damage plasma oxidation treatment for AlN/GaN high-electron-mobility transistors (HEMTs) was developed in this letter, with which high-performance enhancement-mode (E-mode) AlON/AlN/GaN HEMTs were demonstrated. After removal of in situ SiN cap layer within gate area, remote plasma oxidation (RPO) treatment of 4.9nm AlN barrier layer at 300 °C leads to formation of AlON insulator layer and effective depletion of 2-D electron gas channel, without etch-induced damage or serious interfacial issues compared with the commonly used gate- recess metal-insulator-semiconductor HEMTs. Thanks to the low-damage RPO treatment, the achieved E-mode HEMTs with threshold voltage of 0.4 V exhibit quite a small loss of output current (7.8%) compared with the depletion-mode devices. In addition, the AlON gate insulator layer results in a remarkable increase in breakdown voltage of devices from 20 V to 74 V, which benefits the high-voltage operation of AlN/GaN technology. Further investigation shows that the oxidation process of AlN barrier tends to be saturated with treatment time increasing, providing a promising solution to the high-performance E-mode operation of scaled GaN-based HEMTs in RF applications. [ABSTRACT FROM AUTHOR]

An In0.53Ga0.47As/indium phosphide (InP) avalanche photodiode (APD) with a separate absorption, grading, charge, and multiplication (SAGCM) structure is epitaxially grown by molecular beam epitaxy (MBE). The resulting material is studied using X-ray diffraction (XRD), photoluminescence (PL), and scanning transmission electron microscopy. The activation energy extracted from the dark current of the APD indicates that it is dominated by the generation–recombination (G–R) process. Deep low-temperature PL peaks reveal the existence of an ${E}_{v} + 0.42$ -eV deep energy level defect in the indium–gallium–arsenide (InGaAs) absorber layer, which is considered to be a result of point defects caused by Ga-poor or In-poor MBE growth conditions. The effect of the defect on the dark current is confirmed using numerical calculation methods. [ABSTRACT FROM AUTHOR]

This letter presents a wideband high-speed static 1/2 frequency divider using a 0.8- ${\mu }\text{m}$ indium phosphide (InP) process. In this design, some performance-enhanced unity-gain frequency (${f_{T}}$) doubler structures are proposed and compared. Moreover, the application of an enhanced ${f_{T}}$ -doubler technique, for the first time, improves the radio-frequency (RF) performance of the emitter-coupled logic (ECL) divider. The divider has better than 0.4– ${f_{T}}$ frequency ranges from 0.2 to >66 GHz (single-ended sine wave input) within 165-GHz transistor ${f_{T}}$. The input-referred self-oscillation frequency (SOF) of the divider is 53.9 GHz, and the single-ended signal output power is from −5 to 2 dBm. The measured results confirm a great improvement in the performance of the ECL divider by using the enhanced ${f_{T}}$ -doubler structure. [ABSTRACT FROM AUTHOR]

Network slicing has emerged as a key concept in 5G systems, allowing Mobile Network Operators (MNOs) to build isolated logical networks (slices) on top of shared infrastructure networks managed by Infrastructure Providers (InP). Network slicing requires the assignment of infrastructure network resources to virtual network components at slice activation time and the adjustment of resources for slices under operation. Performing these operations just-in-time, on a best-effort basis, comes with no guarantee on the availability of enough infrastructure resources to meet slice requirements. This paper proposes a prioritized admission control mechanism for concurrent slices based on an infrastructure resource reservation approach. The reservation accounts for the dynamic nature of slice requests while being robust to uncertainties in slice resource demands. Adopting the perspective of an InP, reservation schemes are proposed that maximize the number of slices for which infrastructure resources can be granted while minimizing the costs charged to the MNOs. This requires the solution of a max-min optimization problem with a non-linear cost function and non-linear constraints induced by the robustness to uncertainties of demands and the limitation of the impact of reservation on background services. The cost and the constraints are linearized and several reduced-complexity strategies are proposed to solve the slice admission control and resource reservation problem. Simulations show that the proportion of admitted slices of different priority levels can be adjusted by a differentiated selection of the delay between the reception and the processing instants of a slice resource request. [ABSTRACT FROM AUTHOR]

Network virtualization (NV) has emerged as a momentous facilitator for a notable triumph of future networks by allowing a flexibility, cost-efficiency and on-demand services through the deployment of heterogeneous network service requests on a shared physical infrastructure. The most major challenge of NV is to efficiently and effectively map diversified virtual network requests (VNRs), comprising a set of virtual nodes connected by virtual links, onto a shared substrate network meeting various stringent resource constraints. Most of the research papers in this field have merely focused on separate virtual node mapping (VNoM) or virtual link mapping (VLiM) with scalable heuristic algorithms for simple implementations. Unfortunately, the lack of a coordination between node and link mapping stages might cause low embedding results. In this paper, we present a new approach relied upon Genetic Algorithm (GA), that jointly coordinates virtual node and link mappings where the link mapping is based on three different path searching methods. Moreover, a novel heuristic conciliation mechanism is proposed to deal with a set of possibly infeasible link mappings while exploring embedding solutions within the operations of GA algorithm. Extensive performance results indicate that our proposed GA-based algorithms outperform state-of-the-art virtual mapping algorithms in all evaluation metrics we adopt. [ABSTRACT FROM AUTHOR]

To achieve on-chip lasers for Si-photonics, monolithic integration using selective epitaxy is a favorable option due to the unique defect engineering and resultant bufferless structure. Among the intensively investigated selective epitaxy methods, lateral aspect ratio trapping generates III-V devices in the same plane as the Si layer enabling efficient coupling with Si waveguides. Here, we demonstrate telecom InGaAs/InP quantum well lasers selectively grown on commercial silicon-on-insulator substrates with an in-plane structure using a simple structural design and growth scheme without any regrowth steps. Leveraging on the lateral aspect ratio trapping technique, uniform quantum wells with high crystalline quality were obtained and shown by comprehensive material characterizations. Room temperature pulsed lasing was achieved on the fabricated micro-ring lasers with a low threshold of 20 μJ/cm2. The results present a crucial step towards electrically pumped lasers at telecom band as well as fully integrated Si-photonics. [ABSTRACT FROM AUTHOR]

Deposition of Aluminum Scandium Nitride (AlScN) films directly on Silicon at high Sc alloying levels, with controlled stress, and free of anomalous grains (AOGs) is demonstrated using reactive co-sputtering. The average stress and AOG formation consider the Sc alloying range between 20.3 and 36.6 atomic % and for film thicknesses from 400 nm to 1000 nm for 27.3% and 31.7% Sc alloying. The combination of a gradient seed layer and proper process gas mixture is shown to inhibit formation of AOGs in AlScN with up to 36% Sc alloying even when grown directly on Si. It is demonstrated that the total flow can be utilized to control the average film stress across the 20-36% doping range while the process gas mixture primarily impacts the density of AOG formation in the AlScN films. Bulk acoustic wave resonators fabricated from 380 nm and 485 nm thick low stress and AOG free Al0.68Sc0.32N films grown directly on Si demonstrate high frequency operation of 3.6 and 4.8 GHz, high electromechanical coupling >20%, and quality factors in excess of 500. [2021-0252] [ABSTRACT FROM AUTHOR]

Sensitivity is a key parameter for microspeakers, especially in low-power applications such as consumer electronics, medical devices and Internet of Things. However, the sensitivities of most reported piezoelectric MEMS speakers are inferior to those of commercially available electrodynamic speakers. Design of high-sensitivity piezoelectric MEMS speakers requires an analytical expression of sensitivity, which is not yet available. In this paper, the analytical expression is derived, and an ultrahigh-sensitivity piezoelectric MEMS speaker is designed, fabricated and characterized. The sensitivities of the speaker in an ear simulator are 107 dB/mW, 104 dB/mW and 105 dB/mW at 250 Hz, 1 kHz and 10 kHz, respectively, which are the highest among piezoelectric MEMS speakers. The analytical calculation results and measurement data are perfectly matched. Furthermore, the speaker is based on thin-film AlN instead of commonly used thin-film PZT. This work demonstrates that the ultrahigh-sensitivity piezoelectric AlN MEMS speaker consumes 10 times less power than a typical 6 mm electrodynamic speaker; and the developed analytical expressions provide valuable design insights into piezoelectric MEMS speakers for high sensitivity and low power. [2022-0063] [ABSTRACT FROM AUTHOR]

This work reports on the measured performance of an Aluminum Scandium Nitride (AlScN) Two-Dimensional Resonant Rods resonator (2DRR), fabricated by using a Sc-doping concentration of 24%, characterized by a low off-resonance impedance ($\sim $ 25 $\Omega $) and exhibiting a record electromechanical coupling coefficient (${k}_{t}^{\textit {2}}$) of 23.9% for AlScN resonators. In order to achieve such performance, we identified and relied on optimized deposition and etching processes for highly-doped AlScN films, aiming at achieving high crystalline quality, low density of abnormally oriented grains in the 2DRR’s active region and sharp lateral sidewalls. Also, the 2DRR’s unit-cell has been acoustically engineered to maximize the piezo-generated mechanical energy within each rod and to ensure a low transduction of spurious modes around resonance. Due to its unprecedented ${k}_{t}^{\textit {2}}$ , the reported 2DRR opens exciting scenarios towards the development of next generation monolithic integrated radio-frequency (RF) filtering components. In fact, we show that $5^{th}$ -order 2DRR-based ladder filters with fractional bandwidths (BW) of $\sim $ 11%, insertion-loss (I.L) values of $\sim $ 2.5 dB and with $>$ 30 dB out-of-band rejections can now be envisioned, paving an unprecedented path towards the development of ultra-wide band (UWB) filters for next-generation Super-High-Frequency (SHF) radio front-ends. [2022-0024] [ABSTRACT FROM AUTHOR]

Aluminum nitride (AlN) has gained wide interest owing to its high values of elastic modulus, band gap, dielectric strength, resistivity, thermal conductivity and acoustic velocities, especially because it retains most of its properties and versatility in the thin-film form. This review focuses on applications where the CMOS integration of AlN MEMS has been effectively demonstrated. First, the fundamental concepts of piezoelectricity on polycrystalline $c$ -axis oriented thin-films are introduced and AlN is compared to other common piezoelectric materials, namely LiNbO3, LiTaO3, quartz, lead zirconate titanate (PZT), ZnO and GaN by thoroughly discussing the material properties, processing and technological implications. After presenting the possible MEMS-CMOS integration strategies, recent demonstrations of AlN-based devices are reviewed, namely energy harvesters, film bulk acoustic resonators (FBAR), contour mode resonators (CMR), gas sensors, imagers, microphones, transducers for chip-scale communication and calorimetric sensors. Finally, other recent applications/integration opportunities are outlined for AlN-based micro-mirrors, flexible sensors and transducers for liquid media and harsh environments. [2022-0006] [ABSTRACT FROM AUTHOR]

The success of artificial neural networks (ANNs) in machine vision techniques has driven hardware researchers to explore more efficient computing elements for energy-expensive operations such as vector-matrix multiplication (VMM). In this work, InP-based floating-gate photo-field-effective transistors (FG-PFETs) are demonstrated as computing elements that integrate both photodetection and initial signal processing at the sensor level. These devices are fabricated from semiconductor channels grown via a back-end CMOS compatible templated-liquid phase (TLP) approach. Individual devices are shown to exhibit programmable responsivity, mimicking the effect of a synapse connecting the photodetector to a neuron. Using these devices, a simulated optical neural network (ONN) where the experimentally measured performance of FG-PFETs is used as an input shows excellent image recognition accuracy for color-mixed handwritten digits. [ABSTRACT FROM AUTHOR]

We provide an overview of our progress on the development of linear mode avalanche photodiodes (LmAPDs) on InP substrates using antimony (Sb)-based multipliers for short-wavelength infrared (SWIR) spectral region. We identify the key figures of merit of LmAPDs to provide higher sensitivity and speed for applications like light detection and ranging (LiDAR) and remote sensing. We discuss the design of separate absorption, charge, and multiplication (SACM) APDs that are used for narrow gap absorption. We summarize our results on the impact ionization, multiplication gain, dark current, and excess noise of AlGaAsSb and AlInAsSb multipliers lattice-matched to InP substrates. Finally, we identify the key technical challenges associated with the development of SACM APDs on InP substrates. [ABSTRACT FROM AUTHOR]

The mainstream production of optoelectronic and electronic devices based on epitaxial III-V compound semiconductor materials largely employs metalorganic vapor phase epitaxy, a process first demonstrated for devices in the 1970’s. For this special issue in honor of Prof. P. Daniel Dapkus, this paper first briefly reviews his early contributions that spearheaded the interest of MOVPE as a III-V device epitaxy process. Next, the paper discusses MOVPE reactor development in the 1980’s, which at the time was one of the key challenges in establishing MOVPE technology as a reliable epitaxy manufacturing process. The approach combined physical modeling and numerical simulations of reactor processes followed by materials and device demonstrations. Highly uniform performance metrics of AlGaAs/AlGaAs quantum-well diode lasers and two-dimensional monolithic surface-emitting lasers were demonstrated with excellent run-to-run reproducibility. This foundational approach advanced efforts to standardize reactors and subsequent scaling for high-throughput production. [ABSTRACT FROM AUTHOR]

AlGaN germicidal ultraviolet (GUV) light emitting diodes (LEDs) are one of the most promising disinfection technologies in fighting the COVID-19 pandemic; however, GUV LEDs are still lacking in efficiency due to low p-type doping efficiency in p-AlGaN. The most successful approach for producing conductive p-type AlGaN is the implementation of a polarization-enhanced short period Al $_{\mathbf {x}}$ Ga $_{\mathbf {1-}\mathbf {x}}$ N/Al $_{\mathbf {y}}$ Ga $_{\mathbf {1-}\mathbf {y}}$ N superlattice (SL) structure, which enhances hole injection and reduces device operating voltage. In this report, we investigated different aspects of the superlattice including the Al $_{\mathbf {x}}$ Ga $_{\mathbf {1-}\mathbf {x}}$ N and Al $_{\mathbf {y}}$ Ga $_{\mathbf {1-}\mathbf {y}}$ N alloy constituent compositions, ${x}$ and ${y}$ , period thickness, total thickness, and Mg dopant concentration in terms of LED performance as well as electrical, optical, and morphological characteristics. The polarization-enhanced p-type doping in the AlGaN superlattice was also investigated computationally, giving excellent agreement with experimental results. Highly efficient UVC LEDs (279 nm) with EQE of 2% at 5 A/cm2 were demonstrated. A maximum output power of 5.5 mW (56 mW/mm2) was achieved at 100 mA. [ABSTRACT FROM AUTHOR]

AlGaN-based UV LEDs have proven to have broad applications in many fields including sterilization, disinfection, purification, and phototherapy, but the performance still needs to be further improved. Benefiting from advantages such as better current spreading, lower self-heating effect and higher light extraction efficiency achieved by the reduced LED size, UV micro-LEDs are expected to improve quantum efficiency and thus can expand more potential applications. In this article, performance enhancement techniques and applications of UV micro-LEDs will be reviewed, providing an outlook for further development of AlGaN-based UV-LEDs. The micro-LED format offers possibilities to address fundamental issues that UV LEDs confront in general, but also demonstrates specific characteristics and performance advantages opening up new areas of application including high-speed optical communication, time-resolved fluorescence lifetime measurement, optical pumping, direct writing and charge management. [ABSTRACT FROM AUTHOR]

Digital alloy and random alloy Al0.85Ga0.15As0.56 Sb0.44 avalanche photodiodes (APDs) exhibit low excess noise, comparable to Si APDs. Consequently, this material is a promising multiplication layer candidate for separate absorption, charge, and multiplication structure APDs with high gain-bandwidth product. Characterization of the impact ionization coefficients of electrons (α) and holes (β) plays an important role in the simulation of avalanche photodiodes. The multiplication gain curves of eight p+-i-n+ and n+-i-p+ APDs covering a wide range of avalanche widths have been used to determine the electric field dependence of the impact ionization coefficients of Al0.85Ga0.15As0.56Sb0.44. A large impact ionization coefficient ratio between that of electrons to holes was seen across a wide electric field range. Simulations of the avalanche multiplication in these structures using a random path length (RPL) model gave good agreement with experimental results over almost three orders of magnitude, and a mixed injection method was employed to verify the extracted impact ionization coefficients. Interestingly, no difference in the impact ionization coefficients was seen between digital alloy and random alloy Al0.85Ga0.15As0.56Sb0.44. This knowledge of impact ionization coefficients is beneficial for the future utilization of the AlxGa1-xAsySb1-y material system. [ABSTRACT FROM AUTHOR]

Avalanche photodiodes (APDs) made with the material AlGaAsSb (lattice-matched to InP) exhibit very low excess noise characteristics. We demonstrate a Separate Absorption and Multiplication APD (SAM-APD) incorporating a GaAs0.52Sb0.48 (GaAsSb) absorption region and an Al0.85Ga0.15As0.56Sb0.44 (AlGaAsSb) avalanche region. Our GaAsSb/AlGaAsSb SAM-APD exhibits a cut-off wavelength of 1.70 μm at room temperature and a responsivity of 0.39 A/W at 1.55 μm wavelength (with no antireflection coating). Temperature dependence of the breakdown voltage was obtained from avalanche gain data from multiple devices operated at 77 to 295 K. This produced a temperature coefficient of breakdown voltage of 4.31±0.33 mV/K, a factor of 10 and 5 smaller than values for comparable InP and InAlAs SAM-APDs. The very small temperature coefficient of this work is consistent with the extremely weak temperature dependence of avalanche breakdown previously observed in AlGaAsSb diodes. [ABSTRACT FROM AUTHOR]

A new packaging technology allowing the development of high-performance power modules for harsh environments and the optimization of the manufacturing process is presented. The assembly is achieved using an aluminum nitride ceramic heat sink and multilayer silver sintering to make both current track and die attach layers. In order to control the properties of silver sintered layers, a focus on the pressureless sintering process parameters at conventional and unusually high temperatures was performed. Depending on the sintering temperature and the dwell time, the relative density of the sintered layers was evaluated in the range of 56%–88%. The electrical resistivity was measured in a large temperature range for various densities, and the latter can only be 30% higher than the one of bulk silver once the paste is sintered at 500 °C for 1 h. Finally, the mechanical and microstructure properties of assemblies using several metal adhesion layers on the ceramic heat sink and silver sintered multilayers were investigated. Assemblies using Ti/Ni/Ag sputtered layers and silver ink sintered layer as adhesion layers on the AlN ceramic show acceptable shear stress values of about 20 MPa, while the values less than 10 MPa are obtained for W/Ni/Au and W/Ni/Ag adhesion layers. For W/Ni/Au adhesion layers, Ag–Au solid solution was formed at the interface between the sintered silver and the Au finish, leading to a depletion zone at the sintered silver interface and, hence, low shear stress. For W/Ni/Ag adhesion layers, the low shear stress value is mainly due to the oxidation of the Ni layer. [ABSTRACT FROM AUTHOR]

We report a packaging approach for building mm-wave massive multiple-input–multiple-output (MIMO) transmitter arrays in tiles. A single-channel 135-GHz transmitter, having a 22-nm fully-depleted (FD)-silicon on insulator (SOI) CMOS IC for baseband-RF conversion, an InP HBT power amplifier, and a linear microstrip patch antenna array, all packaged on a low-permittivity ceramic interposer (Kyocera GL771, $\varepsilon _{r}=5.2\,\,\text {and}\,\,\delta =0.003$), had 27.5-dBm measured effective isotropic radiated power (EIRP) at saturation and showed 8.5% error vector magnitude (EVM) during 30-Gb/s, 64 quadratic-amplitude modulation (QAM) transmission. An eight-channel transmitter MIMO array tile module constructed on an low-temperature-cofired ceramic (LTCC) interposer, with the same antennas and ICs, after calibration and under the control of a field-programmable gate array (FPGA), had 38.5-dBm measured EIRP at saturation and showed −13-dB EVM in 1.92-Gb/s 16 QAM transmission. Steerable two-simultaneous beam transmission was demonstrated at 34-dBm peak EIRP per beam. [ABSTRACT FROM AUTHOR]

This letter presents a nanowatt-level detector-first wakeup receiver using microelectromechanical system (MEMS)-based matching network (MN). The aluminum nitride (AlN) piezoelectric MEMS resonator is utilized to implement the impedance MN for high passive gain to improve the receiver sensitivity. The high- $Q$ MEMS-based MN also greatly improves the receiver’s out-of-band interference rejection. Gate-biased envelope detector (ED) with small input capacitance is used for RF-to-baseband direct conversion. The wakeup receiver is implemented in a 180-nm CMOS process. It obtains −57-dBm sensitivity with 1-kbps ON-OFF keying (OOK) data rate at 789 MHz while consuming only 16-nW power consumption. [ABSTRACT FROM AUTHOR]

InGaN-based orange-red resonant cavity light-emitting diodes (RCLEDs), were fabricated by using an AlN current-confinement aperture and double dielectric distributed Bragg reflectors (DBRs). For realizing the structure of device, a substrate transfer technique was employed in process of fabrication. The device exhibited optical resonant effect, a high Q factor (∼3010), and a narrow emission linewidth (FMHW ∼0.2 nm), indicating low optical loss in the resonant cavity. Additionally, due to the three-dimensional (3D) optical confinement effect, the discrete modes of red emission were clearly observed in the far field via an angular resolved measurement system. And then, the energies of the photon states of a red RCLED consists generally with the simulation results based on a circular waveguide model. The saturated emission intensity of the orange-red RCLED was proportional to the area of current injection. This work demonstrated the feasibility of InGaN-based electrically injected orange-red RCLEDs that are useful for the development of displays and communication systems. [ABSTRACT FROM AUTHOR]

InP high electron mobility transistors (InP HEMTs) with different spacer thickness 1 to 7 nm in the InAlAs-InGaAs heterostructure have been fabricated and characterized at 5 K with respect to electrical dc and rf properties. The InP HEMT noise performance was extracted from gain and noise measurements of a hybrid low-noise amplifier (LNA) at 5 K equipped with discrete transistors. When biased for optimal noise operation, the LNA using 5 nm spacer thickness InP HEMTs achieved the lowest average noise temperature of 1.4 K at 4–8 GHz. The InP HEMT channel noise was estimated from the drain noise temperature which confirmed the minimum in noise temperature for the 5 nm spacer thickness InP HEMT. It is suggested that the spacer thickness acts to control the degree of real-space transfer of electrons from the channel to the barrier responsible for the observed noise variation in the cryogenic InP HEMTs. [ABSTRACT FROM AUTHOR]

An extremely wideband hybrid bandpass filter (BPF) with high selectivity and low insertion loss is proposed in this brief. The hybrid BPF is composed of three coupled lines (CLs), six matching transmission lines (TLs) and two film bulk acoustic resonator (FBAR) networks with ladder-type structure. These two FBAR networks can help greatly enhance the upper and lower sideband roll-off of the hybrid BPF. At the same time, the passband is compensated by CLs and TLs. The odd-even mode theory is used to analyze the hybrid BPF to obtain the analytical solutions. Through electromagnetic-acoustic co-simulation and optimization, a hybrid BPF with extremely wideband is fabricated and measured. A discussion on the deviation between the simulation and the measured results is presented. The measured hybrid BPF has a 3-dB fractional bandwidth of 56 %, a minimum insertion loss of 1.733 dB and a shape factor of 0.83 simultaneously. [ABSTRACT FROM AUTHOR]

The design and fabrication of InP Mie resonators as a broadband anti-reflection coating for InGaAs/InP FPA sensors have been reported in this work. Using the finite-difference time-domain (FDTD) simulator, we obtained a reflectivity lower than 5% over the whole spectrum of 900 – 1650 nm, with the optimized structure of 200 nm height, 360 nm side length and 600 nm period InP resonators. The characterization was contrasted between the InGaAs focal plane array (FPA) with and without InP Mie resonators with measurements of the blackbody response signal, the noise and the quantum efficiency. A 16.7% improvement of the peak responsivity in the structured area has further demonstrated the promising prospect of such design in the application of InGaAs sensors as well as other spectral domain detectors. [ABSTRACT FROM AUTHOR]

Doped AlN thin films, especially high Sc-ratio AlScN film, have been reported to significantly improve the piezoelectric properties and draw attention for high performance resonators, transducers and integrated ferroelectric applications. However, many demonstrated devices were limited by poor film stress control, abnormal oriented grains and lack of a good etching profile. Compared to costly single alloy target, co-sputtered AlScN films can benefit customized doping concentrations and provide a unique solution for high Sc-ratio AlScN film quality and device studies. In this work, the optimized co-sputtering deposition and ICP etching processes of 500 nm AlScN thin film were developed and released AlScN Lamb-wave resonators were demonstrated. The influence of stress control by changing N2 process gas on the crystalline orientation, abnormal orientation grains, film roughness and piezoelectric property of AlScN thin films were discussed in detail. The AlScN film with a high Sc content requires a lower deposition pressure to obtain good crystalline quality. Al0.85Sc0.15N thin films with FWHM of 1.75°, an average stress of −14.5 MPa and a stress range of 156 MPa were obtained. 130 nm/min etching rate and 77° sidewall profile were achieved by optimized ICP etching of Al0.78Sc0.22N film. Lamb-wave resonators were fabricated based on both Al0.78Sc0.22N and Al0.85Sc0.15N thin films, achieving a quality factor of over 1000 at resonant frequency of approximately 300 MHz. The electromechanical coupling coefficients were improved by 152% and 80% compared to pure AlN devices.[2021-0210] [ABSTRACT FROM AUTHOR]

Cardiac auscultation is an orthodox method to prevent and diagnose cardiovascular diseases. Traditional stethoscopes cannot offer a highly reliable diagnosis, and electronic stethoscopes suffer from low signal-to-noise ratio (SNR) and high cost. This paper reports an electronic stethoscope based on piezoelectric bimorph MEMS transducers for cardiac sound monitoring. The stethoscope prototype, measuring 37 $\times 28\times12$ mm3, contains a preamplifier, holders and polydimethylsiloxane (PDMS) packaging. The transducer is designed using a lumped element model, fabricated by MEMS process, packaged, and assembled for final use. A triangular cantilever is chosen as a basic element for the transducer, with analytical analysis showing high sensitivity per unit area as a figure of merit (FoM). Performance parameters such as sensitivity and noise are experimentally characterized and consistent with simulation results. Owing to the outstanding receiving performance of the aluminum nitride (AlN) material and bimorph triangular cantilever structure of the transducer, compared with the commercial electronic stethoscopes, the proposed stethoscope demonstrates real-time phonocardiogram (PCG) with higher SNR. With its small size, high performance and low cost, the stethoscope is developed for portable and wearable cardiac sound monitoring. [2022-0007] [ABSTRACT FROM AUTHOR]

In this paper, the design, modeling and characterization of width-extensional (WE) mode piezoelectric microelectromechanical systems (MEMS) resonators are presented. The resonators consist of piezoelectric stacked layers for transductions, highly doped single crystal silicon (HDS) device layer as vibrating structures, and oxide layers on top and bottom surfaces for passive temperature compensation. Dependencies of quality factors ($Q\text{s}$) of the WE mode resonators on structure designs, such as device length, thickness as well as crystal orientation of the HDS device layer are investigated. Effects of crystal orientation of the HDS device layer and oxide thickness on the temperature stability of resonant frequency of the WE mode resonators are also analyzed. Both aluminum nitride (AlN) and scandium-doped aluminum nitride (ScxAl1−xN, $x=9.5\%$) are adopted as the piezoelectric transduction layer in this work, in order to investigate the performance improvements by Sc doping. The designed WE mode MEMS resonators are fabricated based on a piezoelectric on cavity-silicon-on-insulator (CSOI) platform. A fabricated ScAlN-based MEMS resonator exhibits a $Q$ of 11086 and motional impedance as low as 28 $\Omega $ at its series resonant frequency of 25.044 MHz. [2021-0222] [ABSTRACT FROM AUTHOR]

This work investigates the influence of aluminum nitride (AlN) fillers on the charge dynamics of epoxy resin (EP) and its hybrid composites through space charge measurements and quantum chemical calculations (QCCs). QCCs analyze the energy-level orbitals and trap parameters of the molecules under the electric field. The pure EP exhibits a homo-charge accumulation, whereas fewer charges are accumulated in the EP hybrid composites. The AlN filler addition significantly suppresses the space charge accumulation during polarization and facilitates the charge dissipation during the depolarization due to the lower trap density than the pure EP. It is found that the orbital energy level of the combined reaction of all monomers together is not a simple superposition of individual orbital energy levels of different monomers. The increased charge injection barrier height suppresses space charge injection and accumulation. The electric field significantly influences the hole traps more than the electron traps. The positive charges with low mobility are easily trapped inside hole traps compared with the negative charges with high mobility inside the electron traps. This results in positive charge accumulation during the polarization and slow decay during the depolarization. The increase of the electrostatic potential is more uniform in hybrid composites than in the pure EP, and no visible electron migration is observed. It indicates that hybrid composites remain stable under electric fields. Moreover, it is not easy for the hole to transport due to the deep trap after the charge injection, leading to limited visible space charge accumulation near the anode. [ABSTRACT FROM AUTHOR]

In this work, we present a new class of thickness extensional microelectromechanical resonator, the overmoded bulk acoustic resonator (OBAR) for filtering applications in the 5G millimeter wave (mm-wave) spectrum. This resonator operates in a second overtone thickness mode with approximately equal thickness electrodes and piezoelectric layer so that acoustic energy is distributed evenly between the different layers. Compared to a fundamental mode at a fixed frequency, the OBAR possess a $3\times $ thicker piezo layer and 5– $10\times $ thicker metal electrodes, enabling manufacturable 30–60 GHz devices. We demonstrate the OBAR experimentally through fabrication of a Pt-AlN-Al device with electromechanical-coupling coefficient ($k_{t}^{2}$) of 1.7%, and series resonance quality factor ($Q_{s}$) of 110 at 33 GHz. [ABSTRACT FROM AUTHOR]

This letter reports on a 108-GHz bandwidth 0.5- $\mu \text{m}$ InP DHBT analog-multiplexer-driver (AMUX-driver). To the best of the authors’ knowledge, this 2:1 AMUX-driver shows unprecedented 1.9-Vppd 160-GSa/s 160-GBd non-return-to-zero (NRZ) and 2.4-Vppd 100-GSa/s 100-GBd PAM-4 output swings, with very high-quality eye diagrams, without any digital signal processing (DSP) or postprocessing. Up to 3.2 Vppd is obtained in NRZ at 100 GBd. The lumped AMUX-driver also shows record 25.7-dB gain and 2.08-THz gain-bandwidth product with 11.1-dB equalizing capabilities at 86.6 GHz. [ABSTRACT FROM AUTHOR]

We report a five-gain-stage 150–175-GHz solid-state power amplifier (PA, SSPA) monolithic microwave integrated circuit (MMIC) having modest 20–21-dBm output power $P_{\mathrm {out}}$ , high gain, and high power-added efficiency (PAE). The transistor technology is 250-nm InP HBT. The $S_{21}$ gain is 31.4 ± 1.2 dB from 100 to 178 GHz. The 3-dB $S_{21}$ bandwidth is between 96 and 181 GHz. Typical return loss values for $\vert S_{11}\vert $ and $\vert S_{22}\vert $ exceed 8.5 and 10 dB, respectively. The dc power is 0.72 W. Across 150–175 GHz, the saturated output power $P_{\mathrm {sat}}$ is 106–126 mW with 16–18-dB gain and 12.5%–16.2% PAE—peak values for $P_{\mathrm {out}}$ , gain, and PAE are at 170-GHz operation. The 1-dB gain compression $P_{\mathrm {out}}$ (OP $_{\mathrm {1\,dB}}$) is 30–60 mW. Across the $D$ -band (110–170 GHz), $P_{\mathrm {sat}}$ is 82–126 mW with 9.1%–16.2% PAE. This work establishes a 1.62– $1.89\times $ increase to PAE over 150–175 GHz operation at the output power and gain levels cited. It is also the highest RF output power demonstrated across the full 110–170-GHz WR6.5 waveguide band. [ABSTRACT FROM AUTHOR]

An aluminum nitride hybrid waveguide is proposed to reduce the challenges brought on by high temperature sensitivity of microresonator-based devices. The waveguide is temperature–insensitive (athermal) and has an effective thermo-optic coefficient value of ±2 × 10−6/K in the wavelength range of 1412 to 1800 nm. It exhibits anomalous dispersion of less than 96 ps/nm/km over octave in the range of 1540 to 3234 nm. The nonlinear coefficient of the aluminum nitride waveguide is enhanced by the strong nonlinearity of titanium oxide. Application of the athermal waveguide in the optical frequency comb is simulated to prove its superior thermo-optic performance. [ABSTRACT FROM AUTHOR]

In monolithic photonic integrated circuits (PICs), an optimized active/passive integration is needed to provide efficient coupling and low amount of interface reflections between amplifiers and passive components. A 1300 nm semiconductor optical amplifier (SOA) on InP substrate and optimized for butt-joint reflections is investigated. Material performance were assessed from measurements of broad area lasers. Room temperature operation reveals 1.2 W single facet output power with threshold current around 100 A/cm2 per well. Characteristic temperatures of T0 = 75 K and T1 = 294 K were obtained. A compact model description of the SOA, suitable for the design of PICs and rate equation analysis, was applied to parametrize the unsaturated gain measurements. Current injection efficiency of 0.65, transparency carrier density of 0.57 1018 cm−3, and free carrier absorption losses up to 15 cm−1 were extracted from fitting the data. The model is verified with measurements of optical gain saturation. A modal gain of 15 dB for a 600 μm long narrow ridge SOA leads to output saturation power higher than 30 mW at 7 kA/cm2. This building block contributes to the development of an InP monolithic integration technology in the 1300 nm range, which can enable the use of photonic integrated circuits in many kind of applications. [ABSTRACT FROM AUTHOR]

In this work, flexible In0.7Ga0.3As metal-oxide-semiconductor field-effect-transistors (MOSFETs) are demonstrated by utilizing liquid polyimide (LPI) transfer. The flexible In0.7Ga0.3As MOSFET on polyimide showed excellent logic performances including threshold voltage (Vth) = 0.16 V, subthreshold swing (SS) = 95 mV/dec and drain induced barrier lowering (DIBL) = 64 mV/V with $\text{L}_{\mathbf {g}} \text {= 5}\,\, \boldsymbol {\mu }\text{m}$ at VDS = 0.5 V. In addition, ex-situ bending experiment was employed to demonstrate electro-mechanical performance, which include stable Ion/Ioff ratio and Ron from 85 mm to 9 mm of bending radius, and deviation of subthreshold swing is less than 5.3 % under bending radius from 85 mm to 4 mm. Our findings indicate that the LPI transfer method on fabricated devices without additional complex bonding methods would be one of effective solutions to enable flexible III-V compound semiconductor electronics and opto-electronic devices. [ABSTRACT FROM AUTHOR]

The quality enhancement of sputtering AlN has hampered a bottleneck that limits the further development and deployment of film bulk acoustic resonator (FBAR) filters. To address this bottleneck, a new two-step growth process integrating metal–organic chemical vapor deposition (MOCVD) and physical vapor deposition (PVD) technologies was carefully constructed to grow AlN films on Si substrates. It was found that both the crystallinity and the surface roughness were increased by more than double that of the one-step produced PVD AlN because of the MOCVD AlN buffer layer with smaller dislocations and grain boundaries. Afterward, using our 4-inch MEMS wafer process, high-quality AlN-based FBARs wafers were obtained accordingly with small full-width at half-maximums for two-step grown AlN(0002) X-ray rocking curves of 0.68°, showing that the Q-factor of FBARs with two-step grown AlN precedes FBARs with one-step grown PVD AlN by 57.6%. [ABSTRACT FROM AUTHOR]