Search

Micro-electromechanical system (MEMS) technologies were used to develop a thickness-mode piezoelectric micromachined ultrasonic transducer (Tm-pMUT) annular array utilizing a lead magnesium niobate–lead zirconate titanate (PMN–PZT) single crystal prepared by the solid-state single-crystal-growth method. Dicing is a conventional processing method for PMN–PZT single crystals, but MEMS technology can be adopted for the development of Tm-pMUT annular arrays and has various advantages, including fabrication reliability, repeatability, and a curved element shape. An inductively coupled plasma–reactive ion etching process was used to etch a brittle PMN–PZT single crystal selectively. Using this process, eight ring-shaped elements were realized in an area of 1 × 1 cm2. The resonance frequency and effective electromechanical coupling coefficient of the Tm-pMUT annular array were 2.66 (±0.04) MHz, 3.18 (±0.03) MHz, and 30.05%, respectively, in the air. The maximum positive acoustic pressure in water, measured at a distance of 7.27 mm, was 40 kPa from the Tm-pMUT annular array driven by a 10 Vpp sine wave at 2.66 MHz without beamforming. The proposed Tm-pMUT annular array using a PMN–PZT single crystal has the potential for various applications, such as a fingerprint sensor, and for ultrasonic cell stimulation and low-intensity tissue stimulation.

In this paper, we present a flexible artificial basilar membrane (FABM) that mimics the passive mechanical frequency selectivity of the basilar membrane. The FABM is composed of a cantilever array made of piezoelectric aluminum nitride (AlN) on an SU-8 substrate. We analyzed the orientations of the AlN crystals using scanning electron microscopy and x-ray diffraction. The AIN crystals are oriented in the c-axis (0 0 2) plane and effective piezoelectric coefficient was measured as 3.52 pm V−1. To characterize the frequency selectivity of the FABM, mechanical displacements were measured using a scanning laser Doppler vibrometer. When electrical and acoustic stimuli were applied, the measured resonance frequencies were in the ranges of 663.0–2369 Hz and 659.4–2375 Hz, respectively. These results demonstrate that the mechanical frequency selectivity of this piezoelectric FABM is close to the human communication frequency range (300–3000 Hz), which is a vital feature of potential auditory prostheses.

We design and fabricate segmented annular arrays (SAAs) using piezoelectric micromachined ultrasonic transducers (pMUTs) to demonstrate the feasibility of acoustic focusing of ultrasound. The fabricated SAAs have 25 concentric top-electrode signal lines and eight bottom-electrodes for grounding to enable electronic steering of selectively grouped ultrasonic transducers from 2393 pMUT elements. Each element in the array is connected by top-crossover-to-bottom metal bridges, which reduce the parasitic capacitance. Circular-shaped pMUT elements, 120 μm in diameter, are fabricated using 1 μm-thick sol–gel lead zirconate titanate on a silicon wafer. To utilize the high-density pMUT array, a deep reactive ion etching process is used for anisotropic silicon etching to realize the transducer membranes. The resonant frequency and effective coupling coefficient of the elements, measured with an impedance analyzer, yields 1.517 MHz and 1.29%, respectively, in air. The SAAs using pMUTs are packaged on a printed circuit board and coated with parylene C for acoustic intensity measurements in water. The ultrasound generated by each segmented array is focused on a selected point in space. When a 5 Vpp, 1.5 MHz square wave is applied, the maximum spatial peak temporal average intensity () is found to be 79 mW cm−2 5 mm from the SAAs' surface without beamforming. The beam widths (−3 dB) of ultrasonic radiation patterns in the elevation and azimuth directions are recorded as 3 and 3.4 mm, respectively. The results successfully show the feasibility of focusing ultrasound on a small area with SAAs using pMUTs.

A new design methodology and fabrication process for two-dimensional (2D) piezoelectric micromachined ultrasonic transducer (pMUT) arrays using a top-crossover-to-bottom (TCTB) structure was developed. Individual sensing and actuation of pMUT elements from a small number of connection lines was enabled by the TCTB structure, and the parasitic coupling capacitance of the array was significantly reduced as a result. A 32 × 32 pMUT array with a TCTB structure was fabricated, resulting in 64 connection lines over an area of 4.8 × 4.8 mm2. The top electrodes for each pMUT element were re-connected by metal bridging after bottom-electrode etching caused them to become disconnected. A deep reactive ion etching process was used to compactify the array. Each pMUT element was a circular-shaped K31-type ultrasonic transducer using a 1 µm thick sol–gel lead zirconate titanate (PZT: Pb1.10 Zr0.52 Ti0.48) thin film. To characterize a single element in the 2D pMUT array, the resonant frequency and coupling coefficient of 20 pMUT elements were averaged to 3.85 MHz and 0.0112, respectively. The maximum measured ultrasound intensity in water, measured at a distance of 4 mm, was 4.6 µW cm−2 from a single pMUT element driven by a 5 Vpp sine wave at 2.22 MHz. Potential applications for development of a TCTB-arranged 2D pMUT array include ultrasonic medical imaging, ultrasonic communication, ultrasonic range-finding and handwriting input systems.

The study presented in this paper assessed the frequency selectivity of an artificial basilar membrane (ABM) constructed using a piezoelectric beam array with narrow supports. Three ABM samples were constructed. Each ABM contained 16 beams with various lengths in a one-dimensional array. To experimentally assess the frequency selectivity of the ABM, mechanical vibration induced either by an electrical or an acoustic stimulus was measured with a scanning laser-Doppler vibrometer. The electro-mechanical and acousto-mechanical transfer functions were defined for the same purpose. The tonotopy of each beam array was visualized by post-processing the experimental results. Finite element analyses were conducted to numerically compute the resonance frequencies, identify the associated vibrational modes, and evaluate the harmonic responses of the beams. The influence of the residual stresses existing in the beams was reflected in the geometric models by introducing three different levels of arc-shaped lateral deformations in the beams. The harmonic analyses revealed that each beam of the ABM samples presented independent band-pass characteristics. The experiments and simulations commonly showed a frequency selectivity of the fabricated ABMs in the range of 2?20?kHz. Therefore, the device is suitable for development of a totally implantable artificial cochlea, implementing a mechanical frequency analyzer. This work is part of research to develop a prototype of a totally implantable artificial cochlea.

A two-dimensional composite plate model was developed as part of the design methodology for micro-scale thin membrane structures in general and pMUTs in particular. The model was compared with a one-dimensional beam model developed earlier and experimental measurements. The two-dimensional model was shown to converge to the one-dimensional model for the structures with a large aspect ratio. Compared to the experimental data, the qualitative trends regarding the dependence of transducer performance on the aspect ratio predicted by the model were validated by the experimental measurements in all cases except that of the electromechanical coupling factor k2eff. The quantitative agreement between model and experimental data was quite good for all parameters at a transducer width of 180 µm, and became worse as the transducer width became smaller. The resonance frequency was predicted very well by the model, and did not depend on the aspect ratio. With the enhanced flexibility provided by the two-dimensional model, some optimization study was also performed. It was found that among different geometries, a square membrane, i.e. a membrane with the aspect ratio equal to 1, appears to have the highest effective coupling coefficient. On the other hand, for a given square membrane, the electrode which covers about 25% of the membrane, i.e. 50% of the width and length measured from the center, possesses the optimized coupling coefficient. In practice, the model can be used to first optimize the dimension of the membrane for pMUTs to obtain the targeted frequencies and then the electrode coverage area for optimized coupling coefficient.

The objective of the current study was to characterize and model the performance of piezoelectric micromachined ultrasonic transducers (pMUTs) with large length/width aspect ratios. Single-element pMUTs with 20 different dimensions corresponding to aspect ratios ranging from 5:1 to 23:1 were designed. Multiple samples were fabricated for each design so that statistically meaningful data could be obtained. The pMUTs were characterized by the impedance measurement combined with an equivalent circuit analysis. A one-dimensional composite beam model was also used to correlate the equivalent circuit components with the structural parameters, and gain insight into the performance characteristics of pMUTs. The resonant frequencies were observed to decrease with the width of the membrane, but have no appreciable length dependence. With the correction of parasitic capacitance, the effective coupling coefficients were observed to increase with the width up to around 150 µm and then decrease. However, they did not show clear and consistent length dependence. The variation of the coupling coefficient as a function of width of the membrane was shown to be mainly due to the relative ratios between the electrode and membrane widths rather than membrane width itself. Although the model presented in this study was a simple one-dimensional electro-mechanical model, it did seem to offer both good qualitative and quantitative insights into the performance of pMUTs and provide a convenient tool for designing thin membrane transducers with a large aspect ratio. The model can also take into consideration the residual stress effect and offer an even more realistic prediction.