Multi-layer polymer-metal structures for acoustic impedance matching in high-frequency broadband ultrasonic transducers design

Multi-layer polymer-metal structures for acoustic impedance matching in high-frequency broadband ultrasonic transducers design

Applied Acoustics 160 (2020) 107123 Contents lists available at ScienceDirect Applied Acoustics journal homepage: ...

2MB Sizes 0 Downloads 6 Views

Applied Acoustics 160 (2020) 107123

Contents lists available at ScienceDirect

Applied Acoustics journal homepage:

Multi-layer polymer-metal structures for acoustic impedance matching in high-frequency broadband ultrasonic transducers design Xinyu Yang, Chunlong Fei ⇑, Di Li ⇑, Xinhao Sun, Shang Hou, Jun Chen, Yintang Yang School of Microelectronics, Xidian University, Xi’an 740071, China

a r t i c l e

i n f o

Article history: Received 10 May 2019 Received in revised form 18 October 2019 Accepted 29 October 2019

Keywords: Multi-layer acoustic matching scheme Transmission line theory Mason model Smith Chart design Pulse-echo experiment

a b s t r a c t Although high-frequency (100 MHz) ultrasound has demonstrated its capability in a variety of applications, the fabrication of high frequency ultrasonic transducers with both high sensitivity and broad bandwidth remains challenging. One main reason is the mismatch of acoustic impedance between piezoelectric materials and the loading medium. Due to reliance on both specific acoustic impedance of matching materials and precise thickness control, the conventional quarter wavelength (1/4 k) matching layer design is impractical for high-frequency transducers. Based on the transmission line theory and Mason model, our work interfaced polymer-metal-polymer matching layers for transducers over 100 MHz. The modeling result comparison between transducers without matching layer and those with two conventional matching layers demonstrated that the matching performance of polymer-metalpolymer matching layers could be as good as the one of conventional matching layers. Meanwhile, unlike conventional 1/4 k matching layer design, our design is independent on materials with specific acoustic impedance, while precise thickness control of polymer and metal can be achieved by deposition. The polymer-metal-polymer matching layers scheme paves the way to high frequency ultrasonic transducers. Ó 2019 Elsevier Ltd. All rights reserved.

1. Introduction High-frequency (100 MHz) ultrasound has found its way to numerous applications, such as high resolution scanning acoustic microscopy, cellular stimulation and microparticle manipulation [1–5]. However, it remains technically challenging to fabricate high frequency transducers with both high sensitivity and broad bandwidth [6,7]. In order to achieve broad bandwidth and high transmission efficiency from high acoustic impedance piezoelectric materials such as LiNbO3, PZT, and PMN-PT, an acoustic impedance matching layer is essential [8]. To date, conventional quarter-wavelength matching layer scheme has almost been exclusively used in piezoelectric ultrasound transducers [9–12]. Both material characteristic acoustic impedance and thickness are required to have specific values. At low frequencies, both requirements were traditionally achieved by: 1) tuning the characteristic acoustic impedance of the matching layer material by mixing high impedance particles with low impedance polymer at a certain ratio, and 2) lapping it to quarter-wavelength thickness. Compared to the long wavelengths of tens to hundreds of micrometers, lapping precision of a few ⇑ Corresponding authors. E-mail addresses: [email protected] (C. Fei), [email protected] (D. Li). 0003-682X/Ó 2019 Elsevier Ltd. All rights reserved.

micrometers was tolerable. In contrast, for high-frequency ultrasound transducers, the quarter-wavelength approach becomes technically impractical by lapping because thickness discrepancies of a few micrometers cause significant performance variances. Furthermore, as the wavelength shortens and approaches the particle size for high-frequency ultrasound, the mixture properties deviate noticeably from homogenous ones, which reduce the matching performance. Homogenous materials are more preferable than the mixture of polymers and particles for a low scattering and precise thickness control through coating or deposition. Whereas in traditional matching layer design, materials with specific required impedance are rare. As a result, though high frequency (100 MHz) ultrasonic transducers have been investigated and fabricated, sensitivity and bandwidth of them are always low due to the inappropriate matching layer design, which is merely a parylene layer (characteristic acoustic impedance 2.58 MRayl) [13,14]. Recently, multi-layer polymer-metal structures for acoustic impedance matching have been investigated to avoid reliance on specific impedance of the materials. Such multi-layer structures entail polymer and metal with different impedances, achieve a specific matching effect by tuning the thickness of each layer [15–18]. Such multi-layer polymer-metal structures are also effective for low-frequency transducers fabrication [15,18].


X. Yang et al. / Applied Acoustics 160 (2020) 107123

In this work, an electromechanical equivalent circuit and a microwave transmission line method [17,19] are implemented as a guidance to study the matching effect of triple-layer polymermetal-polymer for high-frequency (100 MHz) and broad bandwidth (60%) ultrasonic transducers. The Smith chart was utilized to analyze the effective acoustic impedance as well as the reflection and transmission information. The matching performance of polymer-metal-polymer structure was compared with that of a conventional double quarter-wavelength matching layer structure with KLM modeled pulse-echo results although the quarterwavelength matching layers are hardly realizable. Moreover, a single element transducer with 100 MHz center frequency was designed and fabricated to verify the matching performance. Both the sensitivity and bandwidth were significantly improved as expected.

kp ¼ x=v p ks1 ¼ x=v s1 km ¼ x=v m Z P ¼ qP  v P Z s ¼ qs1  v s1 Z m ¼ qm  v m Z b ¼ qb  v b Z l ¼ ql  v l Z 1 ¼ jAZ p tanðkp t p =2ÞZ 2 ¼ jAZ s tanðks1 t s1 =2ÞZ 7 ¼ jAZ m tanðkm t m =2Þ Z 4 ¼ jAZ p cscðkp tp ÞZ 3 ¼ jAZ s cscðks1 ts1 ÞZ 6 ¼ jAZ m cscðkm t m Þ Z5 ¼ j


xC 0

C 0 ¼ e  A=tp

a ¼ h33  C 0

Z 2s Zl

The equivalent input characteristic acoustic impedance of each layer is defined as

Z in3 ¼

Z 2s Zl


2. Materials and methods The microwave transmission line method was applied as guidance for polymer-metal-polymer structure design and the electromechanical equivalent circuit, namely Mason model [20,21] was extended to verify the performance of ultrasonic transducers. Each ultrasonic vibration layer was equivalent to a section of circuit, and the microwave impedance matching network methodology [22] was applicable to the multi-layer transducer design. High transmission efficiency and low reflection coefficient can be achieved by tuning the matching network. As plotted in Fig. 1, the electromechanical equivalent circuit model contains two acoustic terminals with their common terminals connected to the electric transformer. The voltage transfer function derived from this model can be the criterion of the performance of matching layers and the microwave impedance matching network methodology was utilized to simplify the calculation. The equivalent input impedance and the reflection coefficient can be achieved and adjusted by tuning the matching network. The parameters of circuit are defined as follows:

ZB ¼ A  Zb ZT ¼ A 

Z in3 þ jZ m tanðkm t m =2Þ Z m þ jZ in3 tanðkm t m =2Þ


Z in2 þ jZ s tanðks1 ts1 =2Þ Z s þ jZ in2 tanðks1 ts1 =2Þ


Z in2 ¼ Z m

Z in1 ¼ Z s

where A is the area of the transducer, t is the thickness of each layer,

q is the density of each layer, v is the longitudinal velocity of each layer along the polarized axis with subscripts p, s, m and b, l indicating materials of piezoelectric element, polymer, metal, and loading medium, h33 is a piezoelectric constant, and V, F are the voltage and forces exerted by the transducer electrical and mechanical terminals, respectively. All material parameters were tested or from standard datasheet [17], and the materials were commercially available. As shown in Table 1, 36° rotated Y-cut LiNbO3 was chosen to be the piezoelectric transduction element with thickness of 29 mm, associating with a center frequency of 115 MHz. Parylene and gold were selected as the materials for polymer-metal-polymer structure due to their significant impedance difference and depositionfriendliness for accurate thickness control. Parylene and 2–3 mm silver-epoxy were selected for the conventional double quartwavelength matching layer design, which have been used widely for low frequency ultrasound transducer but unable to fabricate for high frequency (>100 MHz) transducers. By earlier work [16–18], the thickness of the outer parylene layer is set to be around 1/4 k, which is 4.7 mm. The effective impedance Zin1 dependent on the thickness of inner parylene and gold is shown in Fig. 2(a). According to the theoretical simulation, final thicknesses of the parylene-gold-parylene layers were optimized as 0.8–0.5–4.7 mm. Besides, for conventional quarter-wavelength matching layer design, the thickness of parylene and silver epoxy was 4.7 mm and 4 mm, respectively. The voltage transfer functions of two different matching schemes are shown in Fig. 2(b), the performance should be the same theoretically, however silver epoxy is not the ideal material with the specific impedance of double matching layers scheme so that the bandwidth of parylene-goldparylene matching layers is wider than that of double matching layers at the expense of sensitivity. 3. Results

Fig. 1. Schematic view of acoustic structure of the high-frequency ultrasonic transducer and its equivalent circuits.

To illustrate the feasibility of our multi-layer polymer-metal structures for acoustic matching propose, Smith chart [17,19] was utilized to illustrate the equivalent impedance and the reflection coefficient. The modeling results were compared with those of conventional double quarter-wavelength matching layers, and discussed in detail. As shown in Fig. 3, the matching loci in the Smith chart explicitly demonstrate the matching effect. The center of the


X. Yang et al. / Applied Acoustics 160 (2020) 107123 Table 1 Material properties used in the transducer designs. Material


c [m/s]


Z [MRayl]

LiNbO3 single crystal Parylene Gold Insulcast 501 with silver particles Water E-Solder 3022 EPO-TEK 301

Piezoelectric element Matching layer Matching layer Matching layer Front load Conductive backing Insulating epoxy

7360 2350 3240 1900 1540 1850 2650

4688 1100 19,700 3860 1000 3200 1150

34.5 2.58 63.8 7.3 1.54 5.92 3.05

Fig. 2. (a) Simulation result of the equivalent impedance associated with the thickness of inner parylene and gold layers for 115 MHz transducer using transmission line theory (b) Simulation result of voltage transfer functions of two different matching design schemes.

Fig. 3. (a) and (c) Loci of the transducer shown on a Smith chart normalized to the equivalent impedance of 22 MRayl, as well as (b) and (d) relative amplitude of transmission and reflection coefficient: a) and b) conventional double quarter-wavelength matching layer effects; c) and d) parylene-gold-parylene triple layer matching effects.


X. Yang et al. / Applied Acoustics 160 (2020) 107123

chart is the matched point with 100% transmission while the circle edge means 100% reflection. The phasor from the origin (center) to a position in the Smith Chart represents the reflection coefficient, length of the phasor indicating the amplitude of the reflection and the direction indicating the phase. With each layer added in sequence, the loci shift clockwise as a circle with centers on the horizontal middle line and the position dependent on the impedance of the material. Meanwhile, the impedance can be directly read from the chart. As both reflection information and the impedance information are plotted in the same chart, the reflection information can be directly converted from and to the impedance value. With matching layer added in sequence, the impedance was shifted from the edge to the center for both matching schemes (Fig. 3a and c), indicating that the impedance was matched in both cases. The two loci presented two examples of

the matching scheme, which resulted in similar bandwidth and transmission efficiency. Krimholtz–Leedom–Mattaei (KLM) model-based software PiezoCAD (Sonic Concepts, Woodinville, WA) was utilized to simulate the transducer performance with the two acoustic impedance matching schemes [23]. The modeled pulse-echo results were shown in Fig. 4. As the purpose of this work was to compare performance of the two different acoustic matching schemes, attenuation in the loading medium was excluded. Also, other parameters like the transducer diameter, properties of piezoelectric element, and backing material remained the same for all the three transducer designs (Table 1 and Table 2). A single element transducer with 100 MHz center frequency was designed and fabricated to verify the matching performance. The electrical impedance was measured by WK6500B 1J65120B

Fig. 4. Pulse-echo modeled result with KLM model-based simulation software PIEZOCAD. a) Without matching layer; b) with polymer-metal-polymer matching layers; c) with double quarter-wavelength matching layers.


X. Yang et al. / Applied Acoustics 160 (2020) 107123 Table 2 KLM modeling results of transducers without and with matching layers.

Without matching P-m-p matching Double 1/4k matching

Center frequency [MHz]

6dB Bandwidth [%]

Pk Ampl [dB re 1 V/V]




119.6 114

62 62.8

44.3 44.2

impedance analyzer (Wayne Kerr Electronics, UK), which is shown in Fig. 5. The acoustic performance was characterized by conventional pulse-echo experiment carried out in distilled water. The acoustic echo was received and analyzed by an Ultrasound Pulser/Receiver (DPR 500, USA). In the absence of the matching layer, the amplitude of the echo is 200.373 mV and the 6dB bandwidth is 36.89%. In the presence of the p-m-p structure, the magnitude of the echo increased to 791.05 mV and the 6 dB bandwidth increased to 86.6%. Both the sensitivity and bandwidth were significantly enhanced as predicted (Table 3). 4. Discussion The first part of discussion summarizes the core innovation of multi-layer polymer-metal structures compared with traditional quarter-wavelength matching layer scheme. It is demonstrated by Smith chart (Fig. 3) that p-m-p matching layer can achieve optimal value of acoustic impedance which leads to the high transmission efficient at a broad frequency range. While the loci cannot shift from edge to the center in Smith chart by conventional double matching layer design due to the lack of materials with idea characteristic acoustic impedance so that the amplitude of pressure reflection coefficient is higher than that of gold-parylene triple

Table 3 Measured results of transducers without and with matching layers.

Without matching P-m-p matching

Center frequency [MHz]

6dB Bandwidth [%]

Pk Amplitude [mV]

108.20 89.88

36.89 86.6

200.37 791.05

layer. In addition, the p-m-p matching scheme utilizes chemical deposition to control the thickness of each layer accurately which is a breakthrough in fabrication compared to what has been published. The main propose of simulation of three different transducers by KLM model is to estimate the size of piezoelement of transducers. It can be seen from Fig. 4(a–c), the electrical impedance is around 50 Ohms at operating frequency with the size of 1.5 mm * 1.5 mm so that the transducer can match the internal resistance of power source. Another aim is to obtain the acoustic performance of conventional double quarter-wavelength matching design. With the frequency increasing to 100 MHz. It is difficult to lap it to the quarter-wavelength thickness and the performance of matching layer is reduced since the thickness of matching layer shortens and approaches to the particle size. It can be found that the sensitivity and bandwidth of transducer are significantly improved by either double matching design or p-m-p scheme. The sensitivity is increased from 56.4 dB without matching layers to 44.2 dB with matching layers and the bandwidth is increased from 29.3% to 62.8% respectively. As the final point of discussion, the acoustic performance of pm-p structures is experimentally demonstrated and some drawbacks of transducers are pointed out. As predicted, the magnitude of echo and the 6dB bandwidth, which are two critical standard for diagnostic imaging, increased to 791.05 mV and 86.6% respectively. Relevant reports about the experimental achievement of

Fig. 5. a) Structure and photo of the p-m-p matching transducer, the time-domain pulse/echo response and normalized frequency spectrum of b) without matching transducer and c) p-m-p matching transducer.


X. Yang et al. / Applied Acoustics 160 (2020) 107123

high bandwidth ultrasound transducer with frequency >100 MHz have not yet been reported. However, there are still some drawbacks of transducer. The spectrum analysis of echo signal shows that there are two peaks at 75 MHz and 115 MHz respectively, which may lie in the constant error (0.5 lm) of parylene chemical deposition equipment. The thickness of inner parylene layer (0.8 lm) cannot be accurately controlled, leading to an extra phase drift and thus the other peak of acoustic energy in spectrum appears. There are some solutions to this problem such as adjusting the operating temperature of the device, using more sophisticated techniques (semiconductor process) or seeking other materials for fabrication. 5. Conclusion In conclusion, it is demonstrated that triple layer polymermetal-polymer structure enables an acoustic matching layers design for the high-frequency broad bandwidth ultrasonic transducer. Theoretically, the polymer-metal-polymer matching layers performed almost as good as the double quarter-wavelength matching layers for over 100 MHz transducer design presented in our modeling work. For practical considerations, it bypasses the fabrication limitations of the traditional matching methods in high frequency transducers. The polymer-metal-polymer matching layer design pioneers a new path to developing an effective matching strategy for high-frequency transducers. Conflicts of interest All authors declare no conflicts of interest. Acknowledgements Financial support from the National Natural Science Foundations of China (11604251, 11174230), the National Key Research and Development Program of China (2017YFC0109703), the National Key Project of Intergovernmental Cooperation in International Scientific and Technological Innovation (2016YFE0107900), the Natural Science Foundations of Shannxi Province (2017JQ1006) are greatly appreciated References [1] Xia J, Yang Y, Hu C, Meng R, Jiang Q, Liu R, et al. Evaluation of brain tumor in small animals using plane wave-based power Doppler imaging. Ultrasound Med Biol 2019;45:811–22. [2] Lee J, Lee C, Kim HH, Jakob A, Lemor R, Teh S-Y, et al. Targeted cell immobilization by ultrasound microbeam. Biotechnol Bioeng 2011;108:1643–50.

[3] Khuri-Yakub BT. Scanning acoustic microscopy. Ultrasonics 1993;31:361–72. [4] Fei C, Li Y, Zhu B, Chiu CT, Chen Z, Li D, et al. Contactless microparticle control via ultrahigh frequency needle type single beam acoustic tweezers. Appl Phys Lett. 2016;109:173509. [5] Wang XY, Seetohul V, Chen RM, Zhang ZQ, Qian M, Shi ZH, et al. Development of a mechanical scanning device with high-frequency ultrasound transducer for ultrasonic capsule endoscopy. IEEE Trans Med Imaging 2017;36:1922–9. [6] Zhou Q, Lau S, Wu D, Kirk Shung K. Piezoelectric films for high frequency ultrasonic transducers in biomedical applications. Prog Mater Sci 2011;56:139–74. [7] Lockwood GR, Turnball DH, Christopher DA, Foster FS. Beyond 30 MHz [applications of high-frequency ultrasound imaging]. IEEE Eng Med Biol Mag 1996;15:60–71. [8] Desilets CS, Fraser JD, Kino GS. The design of efficient broad-band piezoelectric transducers. IEEE Trans Sonics Ultrason 1978;25:115–25. [9] Cannata JM, Williams JA, Qifa Z, Ritter TA, Shung KK. Development of a 35-MHz piezo-composite ultrasound array for medical imaging. IEEE Trans Ultrason Ferroelectr Freq Control 2006;53:224–36. [10] Bezanson A, Adamson R, Brown JA. Fabrication and performance of a miniaturized 64-element high-frequency endoscopic phased array. IEEE Trans Ultrason Ferroelectr Freq Control 2014;61:33–43. [11] Foster FS, Mehi J, Lukacs M, Hirson D, White C, Chaggares C, et al. A new 15–50 MHz array-based micro-ultrasound scanner for preclinical imaging. Ultrasound Med Biol 2009;35:1700–8. [12] Brown JA, Foster FS, Needles A, Cherin E, Lockwood GR. Fabrication and performance of a 40-MHz linear array based on a 1–3 composite with geometric elevation focusing. IEEE Trans Ultrason Ferroelectr Freq Control 2007;54:1888–94. [13] Zhou QF, Sharp C, Cannata JM, Shung KK, Feng GH, Kim ES. Self-focused high frequency ultrasonic transducers based on ZnO piezoelectric films. Appl Phys Lett 2007;90:113502. [14] Fei C, Zhao T, Wang D, Quan Y, Lin P, Li D, et al. High frequency needle ultrasonic transducers based on lead-free Co doped Na0.5Bi4.5Ti4O15 piezoceramics. Micromachines 2018;9:291. [15] Toda M, Thompson M. Novel multi-layer polymer-metal structures for use in ultrasonic transducer impedance matching and backing absorber applications. IEEE Trans Ultrason Ferroelectr Freq Control 2010;57:2818–27. [16] Brown J, Sharma S, Leadbetter J, Cochran S, Adamson R. Mass-spring matching layers for high-frequency ultrasound transducers: a new technique using vacuum deposition. IEEE Trans Ultrason Ferroelectr Freq Control 2014;61:1911–21. [17] Fei C, Ma J, Chiu CT, Williams JA, Fong W, Chen Z, et al. Design of matching layers for high-frequency ultrasonic transducers. Appl Phys Lett 2015;107:123505. [18] Toda M, Thompson M. Detailed investigations of polymer/metal multilayer matching layer and backing absorber structures for wideband ultrasonic transducers. IEEE Trans Ultrason Ferroelectr Freq Control 2012;59:231–42. [19] Ma J, Martin KH, Li Y, Dayton PA, Shung KK, Zhou Q, et al. Design factors of intravascular dual frequency transducers for super-harmonic contrast imaging and acoustic angiography. Phys Med Biol 2015;60:3441–57. [20] Kossoff G. The effects of backing and matching on the performance of piezoelectric ceramic transducers. IEEE Trans Sonics Ultrason 1966;13:20–30. [21] Hou S, Yang X, Fei C, Sun X, Chen Q, Lin P, et al. Fabrication of PMN-PT/epoxy 2–2 composite ultrasonic transducers and analysis based on equivalent circuit model. J Electron Mater 2018;47:6842–7. [22] Qi-Jun Z, Gupta KC, Devabhaktuni VK. Artificial neural networks for RF and microwave design – from theory to practice. IEEE Trans Microw Theory Tech 2003;51:1339–50. [23] Sherar MD, Foster FS. The design and fabrication of high frequency poly (vinylidene fluoride) transducers. Ultrason Imaging 1989;11:75–94.