University of Southern California Dissertations and Theses

Design and fabrication of ultrasound transducers: from single element to high frequency 2D array

(USC Thesis Other)

Design and fabrication of ultrasound transducers: from single element to high frequency 2D array

PDF

Download

Open document

Flip pages

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content Design and Fabrication of Ultrasound Transducers: from Single
Element to High Frequency 2D Array
By
Yizhe Sun
A Dissertation Presented to the
FACULTY OF THE USC GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
[BIOMEDICAL ENGINEERING]
May 2024
Copyright © 2024 Yizhe Sun

ii
Dedication
To my beloved
Parents
Tao Sun & Chunling Zhu
Wife
Ting Jiang
Copyright 2024 © Yizhe Sun

iii
Acknowledgements
I would like to express my heartfelt gratitude to several individuals who have played pivotal
roles in the successful completion of my doctoral journey.
First and foremost, I extend my deepest appreciation to my advisor, Professor Qifa Zhou, for
his unwavering guidance, invaluable insights, and continuous support throughout the entire
research process. His expertise and mentorship have been instrumental in shaping the direction
of my work and fostering my intellectual growth.
I would also express my sincere appreciation to my esteemed dissertation committee
members, Dr. Qifa Zhou, Dr. Yong Chen and Dr. Keyue Shen for their valuable guidance and
insights have been pivotal to the culmination of this academic endeavor. I am honored to have
had the opportunity to work under your guidance and benefit from your collective wisdom.
I extend my heartfelt thanks to my parents, Tao Sun and Chunling Zhu, for their boundless
love, encouragement, and sacrifices throughout my educational journey. Their unwavering
support has been a source of inspiration, and I am deeply grateful for the values they instilled in
me.
I am profoundly grateful to my wife, Ting Jiang, whose unwavering love, encouragement, and
understanding have been my constant pillars of strength. Her belief in my abilities and sacrifices
made this academic pursuit possible, and I am truly fortunate to have her by my side.
Special thanks to my USC lab colleague, Robert Wodnicki, Ruimin Chen, Xuejun Qian, Zeyu
Chen, Laiming Jiang, Hsiao-Chuan Liu, Runze Li, Haochen Kang, Gengxi Lu, Yushun Zeng,
Junhang Zhang, Adnan Rayes, Chi-Feng Chang, Xin Sun, Chen Gong, Jie Ji and Fan Wei, for
their collaborative spirit, camaraderie, and shared passion for research. Our collective efforts
have created a stimulating academic environment, and I am thankful for the camaraderie and

iv
shared experiences that enriched my PhD experience. Thank you for being more than just
colleagues – thank you for being friends.
Five years is long enough for a new building to stand in the parking lot in front of the
building; for a global pandemic to stop in its tracks; and for me to finish my education alone in a
foreign country, as well as my life's work.
To everyone who has been part of my academic and personal journey, thank you for your
support, encouragement, and belief in my potential. Without your help, I cannot get through the
tough time during pandemic. This achievement is as much yours as it is mine.

v
Table of Contents
Dedication.................................................................................................................................. ii
Acknowledgements................................................................................................................... iii
List of Tables.............................................................................................................................ix
List of Figures.............................................................................................................................x
Abstract .................................................................................................................................. xiii
Chapter 1 Introduction ................................................................................................................1
1.1 Background .......................................................................................................................1
1.1.1 Ultrasound Single Element Transducer with Textured Ceramic ................................1
1.1.2 Ultrasound Array Transducers..................................................................................2
1.1.3 Stereolithography 3D Printing Technology...............................................................9
1.2 Motivations and Objectives..............................................................................................12
1.3 Outline.............................................................................................................................14
Chapter 2 Design and Fabrication of 15-MHz Ultrasonic Transducers Based on A Textured
Pb (Mg1/3Nb2/3) O3-Pb (Zr,Ti)O3 Ceramic .................................................................................17
2.1 Introduction .....................................................................................................................17
2.2 Material and Methods......................................................................................................18
2.2.1 Properties of the New Material ..................................................................................18
2.2.2 Design and Fabrication of Transducers......................................................................20
2.3 Results and Discussion ....................................................................................................22

vi
2.3.1 Characterization of Transducer Performance .............................................................22
2.3.2 Transducer Imaging Performance Evaluation ............................................................25
2.4 Discussion and Conclusion ..............................................................................................28
Chapter 3 The General Design of Fabrication Process of 2D Arrays with Interposers...............30
3.1 The Linear Array Fabrication Process..............................................................................30
3.2 The 2D Array Development .............................................................................................33
3.2.1 The Advantages and Applications of 2D Arrays ........................................................33
3.2.2 The Interposer Function and Design ..........................................................................35
3.2.3 The Fabrication of Interposers...................................................................................37
Chapter 4 Design and Fabrication 3-MHz 2D array...................................................................39
4.2 Design and Fabrication of the 3-MHz 64-element 2D Array.............................................39
4.2.1 Introduction...............................................................................................................39
4.2.2 3-MHz Ultrasound 2D Array Design .........................................................................39
4.2.3 KLM Model-based Simulation ..................................................................................41
4.2.4 1-3 Composite and 3D Printed Interposer Fabrication................................................41
4.2.5 Array Assembly Process............................................................................................45
4.2.6 Array Performance Test.............................................................................................49
4.3 2D Array Based Ultrasound Retina Prosthesis Experiment...............................................52
4.4 Discussion and Conclusion ..............................................................................................53
Chapter 5 Design and Fabrication of 4-MHz 256-elements stimulation 2D array.......................55

vii
5.1 4-MHz Ultrasound 2D Array Design................................................................................55
5.2 Circuit Design..................................................................................................................56
5.3 Interposer Design and Fabrication....................................................................................57
5.4 Array Fabrication.............................................................................................................58
5.5 Array Performance Test ...................................................................................................60
5.6 2D-array Based Ultrasonic Retinal Prosthesis Experiment ...............................................61
5.7 Discussion and Conclusion ..............................................................................................63
Chapter 6 Design and Fabrication of 15-MHz 256-elements Imaging 2D Array .......................64
6.1 2D Array and Interposer Design.......................................................................................64
6.2 Interposer and Array Fabrication......................................................................................67
6.3 Performance Testing and Imaging Results .......................................................................72
6.4 Discussion and Conclusion ..............................................................................................74
Chapter 7 Design and Fabrication of 17.5 MHz 256-elements Imaging 2D Array.....................77
7.1 2D Array and PCB Design...............................................................................................77
7.2 2D Array and PCB Fabrication and Preparation ...............................................................79
7.3 2D Array Assembly Process ............................................................................................83
7.4 Performance Testing and Imaging Results .......................................................................86
7.5 Discussion and Conclusion ..............................................................................................88
Chapter 8 Summary and Future Work .......................................................................................90
8.1 Summary and Discussion.................................................................................................90

viii
8.2 Future Work ....................................................................................................................91
BIBLIOGRAPHY.....................................................................................................................95

ix
List of Tables
Table 2-1. Properties of the textured PMN-PZT ceramic and other major piezoelectric
materials ...................................................................................................................................19
Table 2-2. The acoustic design parameters for textured PMN-PZT ceramic single element
transducer .................................................................................................................................22
Table 2-3. Performance comparison of transducers using the textured PMN-PZT ceramic
and other piezoelectric materials. ..............................................................................................25
Table 4-1. Parameters of the 3MHz stimulation array ...............................................................40
Table 5-1. Parameters of the 2 & 4 MHz stimulation array .......................................................56
Table 6-1. Parameters of the 15MHz image array .....................................................................64
Table 7-1. Parameters of the 15MHz image array .....................................................................78

x
List of Figures
Figure 1-1. Illustrations of the different ultrasound 1D array probe types....................................3
Figure 1-2. Some of the 3-D ultrasound imaging and reconstruction systems..............................5
Figure 1-3. Diagram showing the basic configuration of the row-column array transducer..........7
Figure 1-4. Interposer based 2D array fabrication. ......................................................................8
Figure 1-5. Illustration of 405 nm stereolithography apparatus 3D printer................................12
Figure 2-1. (A) SEM image of the textured PMN-PZT ceramic. (B) XRD pattern of the
textured PMN-PZT ceramic. .....................................................................................................19
Figure 2-2. (Left) Schematic illustration of inner structure, and (Right) photo of the 15
MHz textured PMN-PZT ceramic single element transducer. ....................................................20
Figure 2-3. Simulated and measured results of the transducer. .................................................24
Figure 2-4. Two-way IL of the fabricated texture ceramic-based transducer. ...........................24
Figure 2-5. The UBM system diagram.....................................................................................25
Figure 2-6. Wire phantom image and line spread function of the transducer. ...........................27
Figure 2-7. UBM system image of an excised porcine eyeball using the textured PMNPZT ceramic transducers...........................................................................................................27
Figure 3-1. Ultrasound probes may be single element (left), or linear arrays (right) .................30
Figure 3-2. The cross-section view of the linear array...............................................................31
Figure 3-3. (A)Linear array after sub-dicing. (B) The linear array assembly stage ....................32
Figure 3-4. The perspective view and photo of the pitch shifted interposer designed for
the 15MHz 2D array..................................................................................................................36
Figure 4-1. Pulse-Echo impulse response simulation for 3-MHz 2D array fabricated with
(A)1-3 composite. (B) bulk PZT DL-48 material.......................................................................41

xi
Figure 4-2. (A)Composite dicing in 1st direction. (B) composite dicing in both directions.
(C)Composite after both sides sputtered. ...................................................................................42
Figure 4-3. (A)3D printed interposer grids before filling; (B) 3D printed dams. ........................43
Figure 4-4. Interposer filling and alignment marks process.......................................................45
Figure 4-5. The lab-made array assembly stage. .......................................................................46
Figure 4-6. The piezo material assembly process:.....................................................................47
Figure 4-7. The acoustic stack assembly process: .....................................................................48
Figure 4-8. The Finished 3-MHz 64-element 2D array..............................................................49
Figure 4-9. The P/E results comparison of 3-MHz 2D array: ....................................................50
Figure 4-10. The P/E magnitude mapping for (A) 3-MHz composite array and (B) 3-MHz
bulk material array. ...................................................................................................................51
Figure 4-11. The hydrophone results in XY plane of 3-MHz 2D array made with
(A)composite. (B)bulk material. ................................................................................................52
Figure 4-12. (A) Schematic diagram of the experiment setup and (B) Photo of the ongoing
experiment. ...............................................................................................................................53
Figure 5-1. (a) The layered structure of an element in the array. (b) The photo of the
fabricated array together with interposer and PCB board. ..........................................................55
Figure 5-2. High-density interconnect printed circuit board for the 2D array.............................56
Figure 5-3. The interposer filling process. ................................................................................58
Figure 5-4. The array fabrication process..................................................................................59
Figure 5-5. Array performance maps: .......................................................................................60
Figure 5-6. Acoustic field measured by hydrophone. ................................................................61
Figure 5-8. The distribution of different letter patterns..............................................................62

xii
Figure 5-7. Schematic diagram of ultrasound retina stimulation system. ...................................62
Figure 6-1. Top: Conceptual drawings for image 2D array matrix and elements’ structure........65
Bottom: Perspective View of the pitch-shifted interposer. .........................................................65
Figure 6-2. Pulse-Echo impulse response simulation for 15-MHz 2D array element. ................66
Figure 6-3. Interposer fabrication for 15-MHz 2D array ...........................................................69
Figure 6-4. Schematic of Acoustic Stack Fabrication................................................................70
Figure 6-5. The array fabrication process of the 15-MHz 2D array ...........................................71
Figure 6-6. 15 MHz array performance maps............................................................................72
Figure 6-7. Typical pulse-echo response and spectrum of the 15-MHz image array. .................73
Figure 6-8. Experiment set up and wire target with 5 micro-scale stainless steel strings............73
Figure 6-9. Images of the wires target.......................................................................................74
Figure 7-1. Schematic diagram of the 16*16 stacked PCB[112]................................................78
Figure 7-2. Pulse-Echo impulse response simulation for 17.5-MHz 2D array element. ............79
Figure 7-3. The 3D printed interposer fabrication and testing process (.....................................80
Figure 7-4. Schematic view of the stacked PCB board to drive the 2D array.............................81
Figure 7-5. The stacked PCB preparation work.........................................................................82
Figure 7-6. The reroute of the electrodes. .................................................................................83
Figure 7-7. Schematic draw of the assembly platform for the high frequency 2D array.............84
Figure 7-8. The assembly process of the array ..........................................................................85
Figure 7-9. Array performance maps: .......................................................................................86
Figure 7-10. Typical pulse-echo response and spectrum of the 17.5-MHz image array. ............87
Figure 7-11. Images of the wires target.....................................................................................88
Figure 8-1. Interposer for a 20MHz 1.75D array. a) Before filling; b) After filling. ..................92

xiii
Abstract
Ultrasound medical imaging is an entrenched and powerful tool for medical diagnosis, it
demonstrates high efficiency, reduced radiation exposure, and non-invasive characteristics. The
ultrasound transducers can be classified according to their morphology into single-element
transducers, linear arrays and 2D arrays.
The process of fabricating ultrasound single element transducers has matured over decades; the
research efforts have focused on ceramic materials with better performance. Ultrasound linear
arrays are arrays of ultrasound transducers with multiple array elements that fire the individual
elements of the array in a specific sequence to adjust the phase to direct the ultrasound waves in a
specific direction, thus achieving a larger and controllable imaging area than that of a single
element transducer. Ultrasound 2D arrays are widely used for elastography as they can generate
highly focused acoustic beams that can be tuned in 3D space. At the same time, ultrasonic 2D
arrays enable real-time volumetric imaging, and thus have been designed, fabricated, and applied
to transesophageal echocardiography and intracardiac echocardiography.
Despite all these advantages, the technical obstacle of ultrasound 2D array mainly lies in its
manufacturing process. Due to its strict limitations on element density and element size, the
fabrication process of linear arrays is difficult to be applied to 2D arrays. High-frequency
ultrasound 2D arrays require the realization of the highest possible number of array elements and
array element spacing up to 50 m level to achieve high-resolution and high-frequency operation.
Existing assembly methods to date have generally relied on vertical flexure integration, but the
vias width of electronic substrates at this stage is at least 50 m, and circuit structures that do not
use vias have difficulty in achieving the required number of array elements, making the process of
standard 2D arrays difficult to realize. For fully sampled 2D arrays, the connection and fabrication

xiv
of circuits as fine as a few hundred microns is extremely difficult and expensive. In recent years,
with the flourishing development of 3D printing technology, it is finally possible to print complex
and fine 3D structures in a relatively short period of time, which also provides a solution for
fabricating ultrasound 2D arrays.
The work in this dissertation proposal investigates the feasibility of an ultrasound transducer
based on a new textured ceramic material, the fabrication process of ultrasound 2D arrays using
3D printing technology, and its applications. In our work on ultrasound 2D arrays based on 3D
printing technology, we have continued to explore and optimize the fabrication process
successfully upgraded the 64 array elements to 256 array elements and the center frequency from
3-MHz to 15-MHz. the fabricated ultrasound 2D arrays have successfully acquired 3D
reconstructed images of wire targets and have been applied in experiments on ultrasound
stimulated rat visual prostheses. These studies demonstrate the feasibility and broad application
prospects of ultrasound 2D arrays fabricated using 3D printing technology. Moreover, our future
research will continue to focus on improving the performance and fabrication success of ultrasound
2D arrays and applying them to multiple areas of ultrasound stimulation and imaging.

1
Chapter 1 Introduction
1.1 Background
1.1.1 Ultrasound Single Element Transducer with Textured Ceramic
Ultrasound imaging is a well-established medical diagnostic tool for determining anatomical
information. Recently, ultrasound technology has become progressively more widespread.
Compared with general medical imaging methods, including magnetic resonance imaging (MRI),
X-ray, and computed tomography, ultrasonic imaging demonstrates high efficiency, reduced
radiation exposure, and non-invasive characteristics[1-3]. Unlike standard clinical ultrasound with
a central frequency of 2-10 MHz, high-frequency ultrasound imaging (with frequency higher than
15 MHz) can achieve more acceptable image resolution with tens of microns due to the reduced
imaging wavelength and pulse duration. Application areas of high-frequency ultrasound include
cardiology, obstetrics and gynecology (OB-GYN), ophthalmology, intravascular ultrasound
(IVUS) imaging, and small animal imaging [4, 5].
Ultrasonic transducers work by converting electrical and acoustic energy to each other through
the piezoelectric effect[6, 7]. The quality of ultrasonic imaging is mainly affected by the
piezoelectric material, which is the core component in the ultrasound transducer. Essential
characteristics of the piezoelectric layer used in ultrasonic transducers include high kt, the
appropriate dielectric constant that meets electrical impedance matching for the required
transducer diameter, and easy processing. Therefore, selection of the piezoelectric material
dominates the transducer performance. In biomedical imaging applications, the so-called pulseecho mode is that the piezoelectric layer of a transducer emits ultrasound waves into the acoustic
medium and receives the reflected echoes[8]. Thus, it is important to develop piezo-materials with
high piezoelectricity and large electromechanical coupling factors[9, 10]. At present, due to their

2
excellent electromechanical coupling characteristics, most popular piezoelectric materials for
medical ultrasound transducers fabrication are Pb (Zr,Ti)O3 (PZT) based piezo-ceramics, LiNbO3
piezo-crystals and Pb(Mg1/3Nb2/3)O3-PbTiO3 (PMN-PT) piezo-crystals[4, 5]. It is commonly
accepted that the piezo-crystal and composite materials have higher piezoelectricity and
electromechanical coupling coefficient than other materials[11-14]. For instance, the
micromachined PMN-PT crystal-based 1-3 piezo-composites with a resonance frequency in the
range of 15-75 MHz was prepared, and the coupling coefficient was 0.67-0.79[15]. However, the
single crystal growth process and the dice-and-fill fabrication method for composites are
complicated, high cost, and time-consuming, which are not favorable for industrial production[16,
17]. Therefore, it is necessary to explore new piezoelectric materials that can be used in ultrasonic
transducers and obtain better performance.
In the following chapters, an ultrasound transducer utilizing a new textured PMN-PZT ceramic
material was designed and fabricated. Ultrasound bio microscopy (UBM) imaging of a porcine
eyeball was performed using the fabricated device and demonstrated the potential for the use of
PMN-PZT based transducers for ultrasound imaging applications.
1.1.2 Ultrasound Array Transducers
Ultrasound transducer arrays are mainstream product in the modern commercial medical
imaging market. It is a combination of ultrasound probes, which are arranged by multiple
piezoelectric elements in a certain regular distribution. By controlling the different delay times of
the pulses emitted (or received) by each array element in the transducer array, the phase
relationship between the sound waves arriving at (or coming from) a point within the object is
changed to achieve a change in the focal point and the direction of the sound beam, thus realizing
the beam scanning, deflection and focusing of ultrasound waves.

3
Then, a combination of mechanical and electronic scanning is used to achieve imaging. It
provides greater capability than single or multiple single elements transducer systems for
determining the shape, size, and direction of discontinuities.
Figure 1-1. Illustrations of the different ultrasound 1D array probe types [18].
The main ultrasound 1D array probes currently on the market can be classified by size and shape
into the following categories(Fig. 1-1): linear arrays, phased arrays and curved arrays[19, 20].
Linear array probes usually have a higher center frequency and one wavelength-sized element
pitch at the center frequency (>7MHz with small wavelengths), with good resolution and a
rectangular field of view. However, due to the high frequency and high absorption of ultrasound,
the ultrasound depth is small (usually less than 9 cm). It is used to visualize superficial anatomical
structures (i.e., nerves, muscles, arteries, mammary glands, testes, etc[19, 21].) and to provide
needle guidance for biopsies and vascular access. The curved array is also a kind of low-frequency
probe, with applications mostly in thoracic and abdominal ultrasound [19]. Its array elements are
a series of curved rows with one wavelength pitch elements at the central frequency. The field of

4
view and imaging depth of curved arrays is relatively large. Phased array probes typically have a
lower center frequency compared to a linear array. Its element pitch is less than one wavelength at
its center frequency and has a small footprint. The shallow resolution is extremely poor. It is ideal
for echocardiography [22], has a large depth (up to 30 cm), and because of the small element
spacing, phased arrays have a wider field of view. Volumetric images of biological tissues can be
obtained by free or mechanical scanning of the 1D arrays described above (Fig. 1-2) [23-30]. After
decades of development, the fabrication process of ultrasonic 1D arrays has become nearly mature.
High frequency 1D arrays can even reach 20 MHz or even 30 MHz[31, 32]. However, the
limitation of 1D array volumetric imaging is that its frame rate is significantly limited [19, 33],
and affects its image quality. The next generation 2D ultrasound arrays that can achieve real-time
volumetric imaging have emerged in this context[34].
In modern times, clinical applications of ultrasound 2D arrays have focused on
echocardiography and fetal imaging [19, 35]. Ultrasound 2D arrays have hundreds or thousands
of array elements arranged in a plane to provide doctors with a real-time 3D image. In addition, in
cutting-edge academic fields, 3D super-resolution imaging technologies (SRI) that locate
microbubbles in blood vessels to map microvascular structures are in urgent need of the
development of ultrasound 2D arrays to help. At this stage of research, SRI techniques have been
proposed to be performed with mechanically controlled 1D ultrasound array scans[36, 37].
However, since the scans of 1D arrays can only record information within the field of view, they
cannot capture blood flow information orthogonal to the imaging plane. At the same time, onedimensional mechanical scanning of different planes and integrating them into volumetric imaging
requires very long data acquisition and processing time[38], which is very unfavorable for
microbubbles of very short duration.

5
Figure 1-2. Some of the 3-D ultrasound imaging and reconstruction systems [23]:
(a) Freehand 3D imaging with an optical tracking system[24].
(b) Freehand 3D imaging with an ultrasound transducer [25].
(c) Freehand 3D imaging using a mechanical external fixture[26] .
(d, e) Freehand 3D imaging with an electromagnetic tracking system[27-30]
Therefore, ultrasound 2D arrays that can acquire 3D image in real time can overcome the
limitations of 1D arrays in SRI studies and will become an important part of SRI in the future[38,
39]. 1.75D ultrasound arrays [40, 41], whose elements can be accessed individually in elevation

6
are specified for providing electronic focusing in elevation within the whole depth range of the
field of view and are capable of providing a highly focused image slice, thereby significantly
reducing the influence of aliasing from nearby slices, which increases contrast.
Along with the unparalleled advantages of ultrasound 2D arrays comes a very large processing
difficulty. As the center frequency rises, the array element density of 2D arrays rises exponentially,
making it impossible to apply the traditional 1D array processing method to 2D arrays. The
traditional 1D array fabrication method is to directly connect the ultrasound array elements to the
circuit board, and then add the matching layer and backing layer on both sides of the circuit board
and the piezoelectric ceramic.
Since the 1D array only requires the size of the array element in one direction relative to the
wavelength, the circuit board is not very difficult to process. However, for 2D arrays, the size of
the array element must be twice or even equal the wavelength in both directions. For example, a
2D array with a center frequency of 5 MHz has a wavelength of 300 µm, so the elements of the
2D array will be closely spaced in a space of 300 µm and connected to the driving system. This
involves an extremely high level of calibration and processing, and the cost of the board will be
very high. Therefore, the main technical barrier for 2D arrays lies in the connection between the
acoustic element and the driving system. Fiering et al [42] proposed the use of flexible printed
circuit (FPC) for 2D arrays back in 2000. However, to this day, most manufacturers of FPCs still
have difficulty in making circuits with elements and accuracy below 300 microns. In the past
decades, both academia and industry have proposed various solutions for the processing of
ultrasonic 2D arrays.
Recent work in the design and implementation of row-column arrays demonstrates a promising
solution to 2D array design issues. In the row-column array, the electrodes on both sides of the

7
piezoelectric material are aligned in a direction perpendicular to each other, thus enabling control
of individual elements to scan in two direction and acquire the 3D image (Fig. 1-3)[43]. Recently,
a forward-looking miniature endoscope designed for 3D imaging was presented by Katherine
Latham et al [44]. This device was based on a 30-MHz row-column array with 64 × 64 orthogonal
elements. Utilizing a previously reported novel imaging technique [45], this array presented a clear
3D image of two crossed, depth offset wires with good resolution. However, the novel imaging
technique requires different pulse polarity on transmit and receiving which presents further
challenges in circuit and system design[46].
Figure 1-3. Diagram showing the basic configuration of the row-column array transducer.
The probe has N columns and N rows. The rows are aligned
along the X direction and the columns are aligned along the Y direction[43].
Application-specific integrated circuits (ASICs) with pre-amplifier, multiplexer or microbeamformer for sub-array processing were designed for fully-sampled matrix arrays with the
purpose to reduce the channel number to a more reasonable value for most ultrasound systems [40,
47-52]. Such ASIC based matrix arrays have been commonly fabricated directly on the surface of
the IC substrate itself, which benefits the signal-to-noise ratio in addition to relieving the 2D
routing bottleneck. Thus, in such pitch-matched designs, the ASICs’ pads and related circuitry
need to fit into the area of an element. With the help of semiconductor technology, great research

8
has been done in academia and industry and presents different ASICs with promising performance.
Even though these designs represent significant work for trading off design parameters to fit with
array requirements and improve performance, the limited area still constrains the performance and
becomes one of the main challenges for the ASIC design. The complexity of these tradeoffs is
further compounded by the fact that in ultrasound it is necessary to utilize both low voltage and
high voltage ASIC fabrication processes.
In our previous work, a mature 2D array fabrication process-based a on 3D printed interposers
[40, 53, 54] was developed. The interposers with multiple straight channels were filled with
conductive epoxy which acts as a backing layer as well as provides electrical interconnection
between elements and locally integrated interface circuits. This design is applicable for both PCB
and ASIC substrates (Fig.1-4).
Figure 1-4. Interposer based 2D array fabrication. (a) 2D array fabricated on interposer;
(b) cross-section of the composite and the interposer channels;
(c) 2D array with PCB after assembly. [40]
As I mentioned in the above article, the traditional 1D array fabrication process is not applicable
to 2D arrays. If ultrasonic array elements with dimensions in the hundred-micron range are
connected directly to the control circuit, the impedance difference between the material of the
circuit and the piezoelectric ceramic will bring about a strong ringdown and cause poor image
quality. With interposers, the filled channels with E-solder will work as the backing layer to absorb
the acoustic wave on the back side and connect the elements to the circuit at the same time. In our

9
initial work, the channels of the interposers are straight from top to bottom (Fig.1-4), thus the size
of the acoustic area and the electrodes remain the same. With the initial designed interposers, the
scale problem we mentioned above is still exist. Therefore, a new design of the interposer ‘Pitchshifted interposer’ was developed.
In the following chapters, I will present in detail the fabrication process of the 2D array we have
explored using 3D printing technology combined with traditional transducer fabrication processes.
With the Pitch-shifted interposer structure, the ultrasound 2D array fabrication process could be
more affordable and reliable.
1.1.3 Stereolithography 3D Printing Technology
3D printing technology has been one of the hottest research directions in industry and academia
in the last decade. The traditional manufacturing process can be divided into the isomaterial
process represented by casting and forging, and the subtractive process represented by turning,
milling, and grinding.[55]. 3D printing technology, called additive manufacturing process,
appeared in the late 1980s and is one of the rapid prototyping technologies. Based on the
development of computers, it is a technology that uses 3D digital models as input data [56, 57] and
constructs objects by printing layer by layer using bondable materials such as powder or
thermoplastic.
Compared with the traditional manufacturing process, 3D printing technology has the following
main advantages: 1. Large design space. For pieces with internal cavities or topologies, 3D printing
has a huge advantage over traditional manufacturing processes. Complex structures that require
traditional manufacturing to break down and process the item before assembling it, 3D printing
can also be integrated and shaped. 2. Zero-skill manufacturing. The traditional manufacturing
process is expensive and bulky and requires skills training to operate. 3D printing is entirely

10
machine-based, with only the design of the model required by humans. Some 3D printing
technologies (e.g., FDM) have been realized to commercialize very inexpensive and small models,
which provides many non-mechanical researchers with the opportunity to conduct structural and
geometric research[56, 58]. 3. Wide range of material combinations. Multi-jet 3D printing
technologies can print on a wide range of materials, ranging from plastics, metal powders, resins
to ceramics. By stacking and combining materials, it is easy to print items with different physical
and mechanical properties of a single material[59]. To date, progress in 3D printing for highquality and low-cost micro-scaler resolution has already directly benefited a range of fields
including biotechnology [60], optical devices [61], and electronic device [62, 63] to name a few.
There are many different 3D printing devices available, and the devices are designed to work
with materials[57, 64, 65]. Some of the common 3D printing devices include:
1. Laminated Object Manufacturing (LOM) is considered as the oldest 3D printing technology.
It is an inexpensive and highly accurate method that uses paper, PVC film, and other materials.
The process involves a laser cutter that will cut the film along the cross-sectional contour of the
workpiece. The liftable table will support the workpiece, and after each layer is formed, the
material thickness will be reduced. This is done to feed the new layer of material to be bonded and
cut. Finally, the thermal bonding part will bond the film in the formed area layer by layer.
2. Stereolithography (SLA) is a 3D printing technology that employs a photosensitive resin as
the material. This resin is housed within a liquid tank and undergoes solidification, layer by layer,
under the influence of a system-controlled UV laser. A computer directs the process, utilizing
layered cross-section data of the workpiece to guide the laser beam. The laser scans the surface of
the liquid resin, line by line and point by point, with the option of a helium-cadmium or argon ion
laser.[66, 67].

11
3. Selective Laser Sintering (SLS) is a process that uses a powdered material to create a 3D
object. The material is scanned and irradiated by a laser controlled by a computer. The laser raises
the temperature of the powder to the melting point, bonding it to the molded part below. The
process repeats layer by layer until the final shape is achieved. To start, a layer of powder is laid
flat on the molded workpiece using a pressure roller. The laser then scans and irradiates the powder
layer according to the shape of the layer.
4. The Fused Deposition Molding process (FDM) involves heating and melting a filamentary
hot melt material, typically made of ABS or PLA. This molten filament is then extruded through
a micro-fine nozzle and bonded to the previous layer of material. Once one layer is done, the table
drops one thickness in predetermined increments, and the above steps are repeated until the final
part is formed. FDM is the most used 3D printer and is now available for just a few thousand
dollars. It's also known as a desktop 3D printer and has become popular amongst creators for
personal use.[68-70].
In our project, the main 3d printing technology used is Stereolithography (SLA)(Fig.1-5).
Compared to other 3D printing technologies, SLA offers excellent print resolution and has
demonstrated versatility in research and industrial work. Large 3D structures comprised of nearly
10 million 150 μm pitch periodic cells, were presented by Angkur Jyoti Dipanka Shaikeea et al
[71]. To fabricate this design, a moving projection SLA system with the micrometer-scale
resolution was realized. 3D printing technology improvements are not limited to better
resolution. A novel STA technique which supported multi-material projection was reported by
Zhenpeng Xu et al [72]. This approach enabled dissolvable supports in complex architectures,
which makes the printing of overhanging, fly-like micro-architectures possible. In addition, SLA
is widely used in patient-specific models, surgical aids and biocompatible tissue engineering

12
scaffolds [66, 73-75].
In this project, we will take advantage of the above-mentioned benefits of SLA printing to
design and print the above-mentioned pitch-shifted interposer to realize the high frequency
ultrasound 2D array fabrication.
Figure 1-5. Illustration of 405 nm stereolithography apparatus 3D printer. [76]
1.2 Motivations and Objectives
Compared with the growth process of single crystals, the textured ceramic process is relatively
simple, fast, and easy to produce. Additionally, due to the disordered orientation of its crystal
grains, the performance of ceramics prepared by the traditional solid-phase method is inferior to
that of single crystals. There is an increasing need to develop new, relative cheaper piezoelectric

13
materials with better piezoelectric performance, high-energy conversion efficiency and highdensity to optimize device performance. Therefore, textured ceramics with superior properties than
traditional ceramics because of their grain orientation is a good choice. With improved device
performance, imaging can be performed using lower acoustic energy, which saves power, reduces
the requirements and cost of the transmit and receive electronics, and minimizes ultrasonic
biological effects on imaged tissues. Recently, Yan et al. reported a highly <001> textured
Pb(Mg1/3Nb2/3)O3-Pb(Zr,Ti)O3 piezoelectric ceramic that is inexpensive, and high density, and
exhibits high piezoelectric constant and sizeable electromechanical coupling coefficient with
relatively large dielectric constant, making it beneficial as the basis for large-size transducers with
high performance[77].
Benefiting from the development of 3D printing technology, SLA 3D printing technology with
high resolution can achieve the interposer requirement for high frequency 2D arrays. In our
previous work, it has been demonstrated that filling the grids printed by 3D printing technology
with E-solder 3022 can serve as a connection between the piezoelectric material and the driving
system. Here, the role of the 3D printed interposer is twofold: the backing layer for absorbing the
back acoustic waves and the electric pathway for the connection[40, 54, 78]. However, as
mentioned above, high-frequency ultrasonic 2D arrays also face the problem of difficult processing
of circuits due to the very small size of the acoustic part. When we revisited the 3D printed
interposer, we realized that the flexibility of 3D printing technology in handling complex structures
can still be exploited. We can redesign the structure of the interposer so that it can scale up the size
of the circuit while playing the role of backing and circuitry. Therefore, we came up with the idea
of changing the structure of the interposer during 3D printing, thus realizing the process of
controlling the 2D array by using a circuit with a different size than the aperture size of the array.

14
In the next chapters, we will describe several generations of 2D arrays that we have iterated in
the lab. We continue to learn from our experience and improve our fabrication process to increase
the center frequency, number of components, and yield of 2D arrays. The main goal of our work
is to explore the feasibility of pitch-shifted interposers in the fabrication of high-frequency
ultrasound 2D arrays and their application to real-time volume imaging and corneal stimulation.
1.3 Outline
The thesis is outlined as follows:
Chapter 1 introduces the principles of ultrasound medical imaging and the potential of
ultrasound transducers made from novel textured ceramic materials for imaging. Then, the
classification, characteristics, and shortcomings of ultrasound 1D arrays, which are more widely
used in commercial medical imaging, are presented, leading to the necessity of ultrasound 2D
arrays for real-time volume imaging. Finally, the motivation and possibilities of developing
ultrasound high-frequency 2D arrays in combination with 3D printing technology are shown,
and the prospects of high-frequency ultrasound 2D arrays for imaging and stimulation
applications are discussed.
Chapter 2 utilizes the process of designing and fabricating ultrasonic single element
transducers using a new textured PMN-PZT ceramic material, including comparison of material
properties, simulations, and imaging results. In this chapter, the structure and fabrication
process of ultrasound single probes will be detailed. The vibrating unit structure of ultrasound
transducers, including single element transducers, one- and two-dimensional arrays, is basically
similar. Performance tests on the transducer include pulse-echo tests, ultrasound biomicroscopy
(UBM) imaging results on wire phantom and porcine eyeball and demonstrate the potential of
using PMN-PZT-based transducers for ultrasound imaging applications.

15
Chapter 3 shows our exploration of the fabrication process in ultrasound 2D arrays
incorporating 3D printing technology. This chapter will brief describe our group's process for
fabricating ultrasound 1D arrays and the design and fabrication process of interposers, and
proposes a scheme for realizing high-frequency ultrasound 2D arrays using this process as well
as the difficulties and solutions that may be encountered in practice.
Chapter 4 introduces the design and fabrication process of the initial generation 3MHz 2D
array with 64 elements and the performance test results. Each ultrasound array requires pulseecho and hydrophone tests to measure important information such as its yield, beam intensity,
and array emitting field. This is an important attempt in our exploration of the ultrasound 2D
array fabrication process, which has gained valuable experience for the subsequent work.
Chapter 5 describes the application of the 2D array designed and fabricated using a pitchshifted interposer for stimulation use. In this work, two 2D arrays with a center frequency of
4MHz with 256 array elements were fabricated and successfully applied to experiments on
fundus stimulation in rats with convincing results. This chapter will include detailed
information on the design, fabrication, performance testing and fundus stimulation experiments
of the 2D array.
Chapter 6 presents the details of the design, fabrication, and performance testing of the
ultrasound high-frequency 2D array for imaging using a pitch-shifted interposer. The ultrasound
2D array in this chapter has a center frequency of 15 MHz but uses the same common PCB as
the 4 MHz 2D array in Chapter 4 for the driving circuit. This strongly demonstrates the
feasibility of the variable-tuning interpolator.
Chapter 77 proposes and validates a new trend of high frequency ultrasound 2D array
fabrication method with straight channels interposer at 200 um pitch. The 3D printed interposer

16
was designed with all straight channels at 200 µm and then connected to a stacked PCB board
to realize the pitch size. The ultrasound 2D array in this chapter has a central frequency at 17.5
MHz and was also tested both pulse-echo and wire phantom imaging results.
Chapter 8 summaries the current work on high frequency ultrasound 2D array. It includes a
summary of the existing work and a discussion of the existing problems. It also looks at the
future direction of the techniques presented in this paper.

17
Chapter 2 Design and Fabrication of 15-MHz Ultrasonic
Transducers Based on A Textured
Pb (Mg1/3Nb2/3) O3-Pb (Zr,Ti)O3 Ceramic
2.1 Introduction
Ultrasound medical imaging is an entrenched and powerful tool for medical diagnosis. Image
quality in ultrasound is mainly dependent on performance of piezoelectric transducer elements,
which is further related to the electromechanical performance of the constituent piezoelectric
materials. With rising need for piezoelectric materials with better performance and low-cost, a
highly <001> textured piezo-ceramic, Pb (Mg1/3Nb2/3) O3-Pb (Zr,Ti)O3, have been developed.
Recently, textured ceramic materials can be produced at low cost and exhibit high piezoelectric
strain constants and large electromechanical coupling coefficients. In this work, 15-MHz
ultrasonic transducers with an effective aperture of 2.5 mm in diameter based on these highly <001>
textured ceramics has been successfully fabricated. The textured PMN-PZT ceramic was first
synthesized via a template grain growth method and exhibits a high degree of orientation of 90%
<001> texture, with a piezoelectric coefficient (d33) of 1100 pC/N, a relative permittivity (εs/ε0),
of 2310, a piezoelectric voltage coefficient (g33) of 53.8 × 10-3 V·m·N-1
, an electromechanical
coefficient(kt) of 0.69, and a Curie temperature (TC) of 204 ℃. The piezoelectric performance (d33
= 1100 pC/N) in this textured ceramic is superior to most polycrystalline piezoelectric ceramics
currently reported, such as PZT-5 (d33 in the range of 440-600 pC/N) and is like commercial PMNPT single crystal (d33 vary from 1200-2000 pC/N). The 15-MHz ultrasonic transducer was
designed, simulated, and fabricated utilizing the synthesized PMN-PZT ceramic. The device has a
15-MHz center frequency and a fractional bandwidth of 67% at -6 dB. Ultrasound biomicroscopy
(UBM) imaging of a porcine eyeball was performed using the fabricated device and demonstrated

18
the potential for the use of PMN-PZT based transducers for ultrasound imaging applications.
2.2 Material and Methods
2.2.1 Properties of the New Material
A textured Pb(Mg1/3Nb2/3)O3-Pb(Zr,Ti)O3 (abbreviated as PMN-PZT) piezo-ceramic was
prepared by a template grain growth (TGG) approach, with BaTiO3 seeds as the template[77]. First,
the PMN-PT matrix powder was synthesized by conventional solid state reaction method. Mixture
of PbO (99.9%, Sigma Aldrich, MgNb2O6 (99.9%, Alfa Aesar), and TiO2 (Ishihara Sangyo) was
ball-milled in ethanol for 24 hours using ZrO2 milling media. After drying process, the ball-milled
mixture was dried and calcined at 700 ℃ for 2 h. Calcined powder was ball-milled again with 1.5
wt.% excess PbO for 24 h. A pure BaTiO3 template crystal with the required morphology was
prepared using molten salt synthesis method combined with topochemical microcrystal conversion
technique. For tape casting, the slurries were prepared by ball milling the PMN-PT matrix powder
with organic binder and toluene/ethanol solvents. Next 5 vol. % BT template crystals were mixed
with the slurries by magnetic stirring and the slurries. The dried green tapes were cut, stacked, and
laminated at 80 ℃ under 20 MPa pressure for 15 min. The green samples were heated to 400 ℃
and fired for 2h to remove organic solvent and binder, and then isostatically pressed under 200
MPa for 1min. Samples were subsequently sintered at 1150 ℃ for 10 h in flowing O2. Poling
process was carried out for 30 min at 3 kV·cm−1 (direct current) in silicone oil.
Compared to the randomly oriented samples, a highly textured sample has shown a brick walllike construction, as shown in Fig. 2-1(A). Fig. 2-1(B) displays its XRD pattern of the prepared
textured ceramic.

19
Figure 2-1. (A) SEM image of the textured PMN-PZT ceramic.
(B) XRD pattern of the textured PMN-PZT ceramic.
The ceramic sample exhibits a high degree of orientation (with 90% <001> texture) and
improved piezoelectric characteristics. i.e., a higher d33 (1100 pC/N) and a larger kt (0.69), as
compared with other well-known piezoelectric single-crystal materials. Table 2-1 shows a
summary of the momentous piezoelectric and acoustic properties of this texture PMN-PZT and
other major piezoelectric systems.
Table 2-1. Properties of the textured PMN-PZT ceramic
and other major piezoelectric materials
Piezoelectric materials d33(pC/N) kt (%) ε33
S
(1KHz)
Textured PMN-PZT ceramic (this work) 1100 69 2310
PZT-5H Ceramic[81] 440 48 1802
NBT-BT Ceramic[82] 125 60 650
KNN Ceramic[83, 84] 225 58 553
KNN crystal[85] 145 69 580
PMN-PT crystal[86] 2000 62 950
PMN-PIN-PT crystal[87] 2742 59 980

20
2.2.2 Design and Fabrication of Transducers
High-frequency ultrasound transducers are important in medical imaging for diagnostic
purposes in clinical settings because they have a high resolution and play a crucial role in
determining the ultrasound image quality. Based on the prepared textured PMN-PZT piezoelectric
ceramic, ultrasonic transducers operating at a 15 MHz central frequency with active aperture size
of 2.5 mm in diameter were designed and manufactured, as shown in Fig. 2-2.
To optimize the transducer design, PiezoCAD package (Sonic Concepts Inc.) was employed to
conduct simulation based on the Krimholtz, Leedom and Mattaei (KLM) model. Fig. 2-2 illustrates
the internal configuration of the ultrasonic probe fabricated in this paper. Two quarter-wavelength
matching layers are employed to match the difference in acoustic impedance between the piezomaterial (29.44 MRayl) and the bio-tissue (1.5 MRayl) to achieve better imaging sensitivity and
fractional bandwidth.
The desirable acoustic impedance of the two matching layers (Zm1 & Zm2) were computed using
following equations:
1 = (
4
3
)
1/7
, (1)
2 = (
6
)
1/7
(2)
The first matching layer with acoustic impedance of ~ 7.3 MRayl was fabricated with silver
epoxy, which is a mixture of 2-3 μm silver powders (Sigma-Aldrich Inc.) with Insulcure 9 and
Figure 2-2. (Left) Schematic illustration of inner structure,
and (Right) photo of the 15 MHz textured PMN-PZT ceramic single element transducer.

21
Insulcast 501 (American Safety Technologies) in a ratio of 8: 0.325: 2.5. The second matching
layer was applied with a parylene C powder with acoustic impedance of 2.5 MRayl. The backing
layer (Z = 5.9 MRayl) material was comprised of a commercial conductive silver epoxy (E-Solder
3022). The process of production is illustrated experimentally as follows:
First, the textured PMN-PZT piezoelectric ceramic was laminated to a design thickness of 80
μm. The thickness of the piezo electric ceramic was determined using the designed central
frequency of the probe and sound velocity of the material. Next, two thin Cr/Au (50/100 nm)
electrodes were sputtered on the top and bottom surface of the ceramic, respectively. Then, the
materials of the backing and the first matching layer were placed on the respective sides of the
ceramic by centrifuging uncured backing layer and matching layer materials in succession with
curing steps in between material application. The backing and matching layers were laminated to
3 mm and 20 μm, respectively after solidifying. The piezoelectric stack (with electrode, backing
and matching layers), was diced to small square posts using a dicing saw (Thermocarbon,
Casselberry, FL) and then machine-turned to circular disks of 2.5 mm in diameter using a lathe.
The device parameters were optimized to satisfy the impedance of a standard 50Ω system with the
following equation:
=

2

(4)
where is the thickness of the ceramic, A is its area and

is its permittivity. The machineturned circular disk-shaped acoustic stacks were inserted and fixed into brass housings one by one.
Afterwards, the conductive backing layer was connected to a SMA connector with a lead wire.
Insulating epoxy (EPO-TEK 301) was added to strengthen the mechanical structure. After curing,
another Cr/Au (50/100nm) electrode was sputtered onto the brass housing and the first matching
layer, forming a mutual connection to ground. To optimize imaging capability, the transducers

22
were press-focused to reduce the f-number, which is described as the ratio of focal depth to
aperture size, to 1. Finally, a 10μm thick layer of parylene C, acting as the second matching layer
and the waterproof coating, was added onto the transducer via vapor-deposition.
Table 2-2. The acoustic design parameters for
textured PMN-PZT ceramic single element transducer
2.3 Results and Discussion
2.3.1 Characterization of Transducer Performance
Fig. 2-3(A) and (B) show the simulated and measured impedance spectra and pulse-echo
response of the textured PMN-PZT ceramic single element transducer, respectively. Phase angle
and electrical impedance of a representative textured PMN-PZT probe was measured using a HP
4294A impedance analyzer (Agilent Technologies) and found in Fig. 2-3(C). The resonance, fr and
antiresonance, fa frequencies were found to be located at 15 MHz and 21 MHz. The effective
electromechanical coupling coefficient eff can be estimated by the following equation based on
the IEEE standard[88],
eff = √1 −

2

2
(5)
For the measured frequencies, was evaluated to be 0.69. The electrical impedance near the
resonance is from 90 to 120 Ω. Finally, a LCR inductance-capacitance-resistance digital bridge
meter (QuadTech) was applied to evaluate the dielectric loss and capacitance.
Pulse-echo response of the fabricated probe was assessed utilizing a pulse-echo setup[89] with
the device immersed in a tank filled with deionized water and a fixture that allowed the device
Layer Material Acoustic impedance (MRayl) Thickness (mm)
Piezoelectric layer Textured PMN-PZT ceramic 29.44 0.08
Matching layer I 2-3μm silver epoxy 7.3 0.02
Matching layer II Parylene-C 2.23 0.01
Backing layer E-Solder 3022E 5.92 3

23
position to be adjusted freely. A quartz plate was placed 2.5 mm below the transducer and
functioned as the target reflector. The probe was then driven by an ultrasound pulser-receiver
(Panametrics 5900PR). The testing parameters are 200 Hz pulse repetition rate, 50 Ω damping
factor, and 1 μJ energy per pulse.
The echo response signals were digitised and shown on a laptop using a GaGe EON 12-bit
digitizer (DynamicSignals LLC). The sampling rate of the signal was 1GSPS. The received
impulse-echo response RF data were processed by the fast Fourier transform (FFT) to obtain the
calculated spectrum [Fig. 2-3(D)]. The center frequency, fc, and fractional bandwidth (FBW) at -6
dB were calculated from the computed FFT spectrum by
=
+
2
(6)
=
−

× 100% (7)
where and are the cut-off frequencies at -6 dB, respectively. The fc of the probe was tested
to be 15 MHz, and the -6 dB FBW was 67%. Numerical simulations and experimental
measurements show a good qualitative agreement. Although attenuation and machining errors
result in slightly lower sensitivity than simulation, non-amplified peak-to-peak output voltage (Vpp) was 0.8 Vp-p, which proved the good performance of this fabricated ultrasonic probe, as
demonstrated in Fig. 2-3.
The two-way insertion loss (IL), which is described as the ratio of the output voltage (VO) to the
input voltage (Vi), is an important parameter to measure the sensitivity of the transducer. The quartz
plate was used as the reflector. The IL was calculated by
= |20log

|
(8)

24
The final IL was compensated to account for losses caused by diffraction and attenuation in
water (22 × 10-4
dB/mm·MHz2
) and quartz (1.8 dB). The calculated two-way IL versus frequency
is shown in Fig. 2-4. The minimum insertion loss of 21 dB occurred at 17 MHz. From Table 2-3,
the insertion loss is shown to be lower than ordinary ceramic materials.
0 5 10 15 20 25 30
0
15
20
25
30
35
40 Insertion loss (dB)
Frequency (MHz)
21 dB @ 17 MHz
Figure 2-4. Two-way IL of the fabricated texture ceramic-based transducer.
Figure 2-3. Simulated and measured results of the transducer.
(A). Electrical impedance and phase simulated using PiezoCAD,
(B) Pulse-echo signal and frequency band simulated using PiezoCAD,
(C) Measured electrical impedance and phase.
(D) Measured pulse-echo signal and frequency band.

25
Table 2-3. Performance comparison of transducers using the textured PMN-PZT ceramic
and other piezoelectric materials.
2.3.2 Transducer Imaging Performance Evaluation
An ultrasound image test was preformed using the fabricated transducer with a UBM system
(Fig. 2-5), which was explained in previous work for imaging experiments including ex vivo
porcine eyeball imaging[16].
The UBM system includes a pulser/receiver (JSR DPR500) and a high voltage remote pulser
(RP-H4, Imaginant Inc.). The circuit consists of an amplifier with a gain of up to 50 dB, an analog
high-pass filter with frequencies from 5 to 30 MHz and an analog low-pass filter with frequencies
up to 300 MHz. The device was also fixed on a 3D electric linear stepper motor to obtain the 2D
image, and the plane perpendicular to the surface of the probe was mechanically linearly scanned
in 9 μm increments. When a trigger event is received from a function generator (AFG 3252 C), the
pulse generator/receiver drives the probe with excitation pulses of 1 µJ per pulse. After filtering
Piezoelectric Materials Insertion Loss (dB) fc (MHz) -6 dB BW (%)
PMN-28%PT single crystal[90] 24.6 3.09 89.9
PZT-5H ceramic[90] 29.8 2.95 62.2
Undoped KNN ceramic[91] 27~51 5.5 49
LiNbO3 single crystal[89] 13.4~19.5 22~23 60~72
BZT-50BCT ceramic[92] 18.7 30.5 53
Textured PMN-PZT ceramic (this work) 21 15 67
Figure 2-5. The UBM system diagram.

26
through the two filters, the RF signal is digitalized by a 12-bit digitizer (ATS9360). The raw RF
data will be processed offline in MATLAB to form the ultrasound image.
An important evaluation of ultrasound imaging is the lateral and axial spatial resolution. The
axial resolution () can be estimated using the following formula:
=

2 ×
(9)
For the 15 MHz device fabricated in this paper, the theoretical value of the axial resolution is
calculated to be 74 μm. The lateral resolution () can be valued by
= × number (10)
Therefore, the theoretical lateral resolution of this transducer is 103 μm.
The actual resolution was measured by imaging a wire phantom which consists of 4 tungsten
wires with 20-µm in diameter. The four wires were spaced at the same distance in both axial (1
mm) and lateral directions (1.5 mm) (Fig. 2-6(A)). The line spread functions in two directions are
shown in Fig. 2-6(B) and Fig. 2-6(C) demonstrate. The line spread function was obtained from the
closest wire which was located 2.5 mm away from the probe. The actual axial and lateral
resolutions at -6dB of the transducer were demonstrated to be 83 μm and 224 μm, respectively.
The axial resolution shows good agreement with the theoretical expected value. The lateral
resolution has some differences which may cause from pressed focus.
Finally, imaging performance was evaluated using an excised pig eyeball. To scan the entire
front of the porcine eyeball, the scan width and depth were set to 16 mm and 12 mm, respectively.
Compared to the frequency range of clinical applications (2-10 MHz), the fabricated 15 MHz
transducer has better resolution, which is important for to imaging the fine tissue structures of the
eyeball. Fig. 2-7 shows an ultrasound biomicroscope (UBM) image of a pig eyeball. The UBM
image clearly shows the anterior anatomical details including the iris (I), lens surface (L1),

27
conjunctiva (C2), limbus (L2), and cornea (C1) .
Figure 2-7. UBM system image of an excised porcine eyeball using the textured
PMN-PZT ceramic transducers (dynamic range: 50dB, C1: cornea, C2: conjunctiva,
C3: ciliary body, I: Iris, L1: lens, L2: limbus)
Figure 2-6. Wire phantom image and line spread function of the transducer.
(A) Wire phantom UBM images obtained by the 15 MHz textured PMN-PZT ceramic single element
transducer. (B) Lateral line spread function of the closest wire in the wire phantom obtained by the 15 MHz
textured PMN-PZT transducer. (C) Axial line spread function of the closest wire obtained by the 15MHz
textured PMN-PZT transducer.

28
2.4 Discussion and Conclusion
In this study, a recently developed <001> textured PMN-PZT ceramic was employed to
manufacture a 15-MHz center frequency ultrasound transducer for high-performance biomedical
ultrasonic imaging applications. The manufactured probe demonstrated an effective
electromechanical coupling factor of 0.69, a center frequency of 15 MHz, a -6 dB fractional
bandwidth of 67% and an insertion loss of 21 dB.
For the developed transducers, the high sensitivity is attributed to the internal enhanced
electromechanical coupling coefficients and piezoelectric constants. Ceramic texturing is an
effective strategy to realize high-performance piezo ceramics with low cost. In this work, the
piezoelectric layer is a textured PMN-PZT ceramic that was prepared by a template grain growth
(TGG) technique using BaTiO3 seeds as the template, with a high degree of orientation (with 90%
<001> texture).
Additionally, the ceramic possesses a relatively large dielectric constant, which is beneficial for
fabrication of small-size (e.g., 2.5-mm diameter in this work) transducers with high performance
50Ω matching. Good electrical impedance matching is important for high sensitivity in small
devices where large transmit voltages cannot be tolerated.
The bandwidth of the fabricated is 67%. The measured center frequency was close to 15MHz
instead of the simulated frequency of 20MHz. This range in performance is caused by the variation
in machining tolerances and material parameters during the fabrication process as well as damping
of materials. A thorough description of the ceramic preparation process and its measurement
characteristics can be found in the previous work [77, 93]. As the core component, the ceramic
material in a transducer must be lapped to a thickness of between 10 to 100 microns depending on
the required center frequency. The grain structure of the piezo material will be compromised when

29
the material is too thin. Given its relatively large grain size the performance of the material may
be changed if the material thickness is reduced below 100 μm. Lastly, in terms of material
optimization, a lower impedance can be achieved by targeting a lower dielectric constant and
higher kt through fabrication of 1-3 composite.
In summary, we have present the newly developed textured PMN-PZT ceramic with high
piezoelectric constant and large electromechanical coupling coefficient that has potential to for
building high-frequency and highly sensitive transducers for biomedical ultrasound imaging. The
fabricated needle transducer had a high center frequency of 15 MHz, an effective
electromechanical coupling factor, , of 0.69, fractional bandwidth of 67% at -6 dB, and
insertion loss of 21 dB. These initial results are promising and point to potential future clinical
use of textured PMN-PZT ceramic for the next generation high-frequency ultrasonic devices
requiring high-sensitivity.

30
Chapter 3 The General Design of Fabrication Process of
2D Arrays with Interposers
3.1 The Linear Array Fabrication Process
In the previous chapters we mentioned that different ultrasound arrays are fabricated in
completely different processes, this chapter will briefly describe the fabrication process for
ultrasound one-dimensional array. I will use a linear array with a center frequency of 5MHz made
by our group as an example to introduce the process of one-dimensional arrays here. In contrast to
an ultrasound single element transducer with a fixed focal length, a linear array essentially consists
of many individual transducer elements separated from each other by a wavelength or less. By
controlling the orderly superposition of the acoustic beams emitted by each array element, the
deflected and focused beams can be generated flexibly, and high-resolution inspection of the area
of interest can be accomplished without changing the probe, and its unique linear sweep, sector
sweep, and dynamic focus can be used to inspect the part with high efficiency without moving or
less moving the probe[94].
Therefore, compared with the traditional ultrasonic inspection of single wafer, the ultrasound
phased array has a more flexible sound beam, faster inspection speed, higher resolution, and is
more suitable for inspection of parts with complex shapes.
Figure 3-1. Ultrasound probes may be single element (left), or linear arrays (right)[94].

31
Therefore, compared with the traditional ultrasonic inspection of single element, the ultrasound
phased array has a more flexible sound beam, faster inspection speed, higher resolution, and is
more suitable for inspection of parts with complex shapes.
We fabricated several 5MHz linear arrays with Prof. Jung-Yeol Yeom from Korea University
in 2022. I will skip the design process of these linear arrays and use this as an example to introduce
our group's current process of fabricating linear arrays. Here, the materials and properties of the
piezoelectric ceramics and matching layer material will not be discussed.
As with the single element transducers, the linear array structure also contains a matching layer
and a backing layer. In our linear array fabrication process, the driving circuit is connected directly
to the electrode coated piezoelectric piezo-ceramic, and a pre-made backing layer is epoxied
behind the circuit. The cross-section of the linear array is shown in Fig. 3-2.
The first step is also to create a matching layer of piezoelectric ceramics with electrodes on
both sides and polish it to a given thickness. The next step is to sub-dice the electrodes of the
piezoelectric ceramic into the corresponding components with a precise dicing saw according to
the designed circuit spacing with the matching layer facing down as shown in Fig.3-3(A). Next,
the diced linear array is fixed on a stage with two directional degrees of freedom and the circuit
board is placed on top. Since the thickness of the circuit board is just under one hundred microns,
Figure 3-2. The cross-section view of the linear array.

32
the array elements below the board can be seen through a USB microscope (Dino-Lite, Taiwan)
to monitor the alignment process, as shown in Fig. 3-3(B).
After determining the alignment of the circuit and the array elements match, we will use a
minimal amount of epoxy (EPO-TEK 301) drops on the component surface and use a heavy weight
to laminate the circuit and the piezoelectric ceramic together. It is proven that during this process,
the epoxy will be pressed into the kerfs formed by the sub-dicing and no breakage problem will
occur.
Finally, the linear array differs from the previously mentioned single probe in the choice of
material for the backing layer. Since the circuitry is already connected to the array element, we
do not need to use expensive E-solder for the backing, where the backing layer serves only to
absorb acoustic waves. Since the backing layer is the largest part of the linear array, we chose
aluminum oxide powder mixed with epoxy resin, and used 3D printing technology and silicone
to shape the mixture to facilitate the use and operation of the linear array afterwards.
A B
Figure 3-3. (A)Linear array after sub-dicing. (B) The linear array assembly stage

33
3.2 The 2D Array Development
3.2.1 The Advantages and Applications of 2D Arrays
Ultrasound imaging has become an indispensable tool in various fields of medicine and
beyond, enabling non-invasive visualization of internal structures with remarkable detail. As
technology advances, the demand for higher resolution, faster imaging, and improved spatial
coverage continues to grow. In response to these demands, ultrasound 2D arrays have emerged
as a promising solution, offering unprecedented capabilities in imaging and diagnostics.
Nowadays, Ultrasound 2D arrays are widely used for elastography as they can generate highly
focused acoustic beams that can be tuned in 3D space. With a matrix of hundreds of elements,
these 2D arrays can provide real-time 3D image and become very helpful tools for physicians.
Besides, many advanced research fields for ultrasound applications could benefit from
developing 2D arrays, such as super-resolution imaging techniques.
Traditional ultrasound imaging systems employ single-element transducers that produce a
focused beam, requiring mechanical scanning to generate a complete image. While effective, this
approach is limited by slow acquisition times and reduced spatial resolution, particularly at
depth. In contrast, 2D array transducers consist of multiple elements arranged in a grid pattern,
allowing for electronic beam steering and focusing. This enables rapid, real-time imaging with
superior spatial resolution and improved penetration depth, revolutionizing the way we acquire
ultrasound images.
Ultrasound 2D arrays have a number of unique advantages, including the following aspects:
1. Higher spatial resolution: By electronically controlling the beam direction and focus, 2D
arrays can provide higher spatial resolution than single element transducers. This allows for
clearer visualization of fine anatomical structures and subtle abnormalities.

34
2. Real-time imaging: The ability to electronically steer and focus the ultrasound beam
facilitates real-time imaging without the need for mechanical scanning. This not only reduces
examination time, but also allows for dynamic studies such as cardiac imaging and fetal
monitoring.
3. Increased depth of penetration: 2D arrays are capable of generating high-frequency
ultrasound beams that penetrate deeper into tissue while maintaining excellent resolution. As a
result, 2D arrays are particularly well suited for imaging structures located deep within the body,
such as abdominal organs and vascular structures.
4. Versatility and Flexibility: The 2D arrays provide electronic control that allows for a variety
of imaging modes, including B-mode, Doppler, and elastography. This versatility makes it ideal
for a wide range of clinical applications, from diagnostic imaging to interventional procedures.
At this stage, the clinical applications of ultrasound 2D arrays mainly include diagnostic
imaging and intervention guidance during surgery. In addition to clinical applications, ultrasound
2D arrays play an important role in research and development, enabling scientists to study tissue
biomechanics, evaluate drug delivery and investigate physiological processes in vivo. The
versatility and high-resolution imaging capabilities of 2D ultrasound arrays make them
indispensable tools for advancing our understanding of biological systems.
In conclusion, ultrasound 2D arrays represent a significant advancement in medical imaging
technology with higher spatial resolution and real-time imaging capabilities for a wide variety of
medical specialties and research applications. The continued development of 2D arrays holds the
promise of further improving diagnostic accuracy, guiding interventions, and enhancing our
understanding of the human body.

35
Recently, some studies have indicated that ultrasound can evoke localized neural activities by
stimulating the retina [79] and VC [80]. Ultrasound, as a non-invasive neuromodulation
technique, has a very high scientific potential in the clinical response to amblyopia and
blindness.
The current challenges in fabricating ultrasound 2D arrays lies in achieving higher resolution
while maintaining cost-effectiveness and scalability. The fabrication process need to ensure
precise alignment and integration of hundreds of elements with the controlling system.
Additionally, optimizing the array’s performance across various frequencies and aperture sizes is
another challenge in the fabrication process.
Based on the background mentioned above, we would like to develop a complete ultrasound 2D
array fabrication process that is both economically practical and well maneuverable. In the
preliminary design, without changing the basic ultrasound array element, we hope to realize the
connection between the control circuit and the array elements by changing the form of the structure
of the backing layer of the array. The specific design and process will be detailed in subsequent
sections.
3.2.2 The Interposer Function and Design
Conventional ultrasound 2D array fabrication processes use high-precision FPCs or ASICs,
which are certainly good but costly solutions for high-frequency ultrasound 2D arrays with
hundreds of elements.
In a 2D array structure, since ultrasound signals need to be generated and received in the
direction of the matching layer, and having a circuit board in the way would greatly reduce the
strength of the signal, most ultrasound transducer designs will have the control circuit in the
direction of the backing layer. In our 2D array structure, the backing layer not only has the
traditional function of absorbing ultrasound signals from the back side, but also needs to be

36
connected to the control circuits and piezoelectric ceramics, and good electrical conductivity is
necessary. In addition to this, we also expect the backing layer to provide other additional functions,
such as lowering the cost of the circuit board and other potential solutions.
Therefore, we combined 3D printing technology to create an interposer structure. The interposer
has the following characteristics: its upper and lower surfaces are connected to piezoelectric
ceramics and control circuits, respectively, and in the middle is connected by a matrix of channels.
We will try to fill the channels with E-solder, which will be both conductive and ultrasonic energyabsorbing after curing.
Since the backing layer required for high-frequency ultrasound is only a few millimeters, we
can subtly make changes to the circuit at the structure beyond the backing layer. The lower design
and manufacturing costs and shorter lead times for 3D printed parts compared to FPCs and ASICs
provide a huge advantage for the design and fabrication of 2D arrays. Based on the above reasons,
we designed and proposed a novel interposer for ultrasound 2D arrays: the pitch-shifted interposer
(Fig. 3-4).
Figure 3-4. The perspective view and photo of the pitch shifted interposer
designed for the 15MHz 2D array.
As clearly seen in the figure, the upper and lower surfaces of the pitch-shifted interposer have
different isometric matrices connected by a structural channel with size amplification in the middle.

37
With such a structure, it is possible to drive a 2D array with a circuit larger than the size of the
transducer array elements, while also combining the functions of a backing layer and a circuit path.
At the same time, this device has the advantage of saving design cost. For ultrasound 2D arrays
with different center frequencies, shapes, and sizes, it is necessary to design the corresponding
PCB or ASIC, which is time-consuming and costly. As a connector between the transducer
elements and the driving system, the interposer can be adapted to use the same board to drive
different 2D arrays. It is faster, more economical, and easier to design and print new inserts than
it is to design and order new control circuits.
3.2.3 The Fabrication of Interposers
In the design, we want the interposer to be able to connect up to 256 2D array elements
simultaneously. First, we created the interposer file through the 3D modeling software
SolidWorks and searched for a suitable 3D printer to fabricate it. At the same time, we need also
prepare a dam with a size slightly larger than the interposer as an auxiliary structure when filling
it with E-solder. Next we will use 15min epoxy to fix the dam to a flat piece of glass and fill the
dam with E-solder. Finally, the interposer will be pressed into the dam all the way down to the
glass and the E-solder will spill out through the channel of the interposer, filling the channels
through this process.
After removing the interposer and leaving it to cure, the interposer was subsequently tested.
The test is mainly divided into conductivity test and short circuit test, we need to make sure that
the channels of the interposer are independent and conductive, otherwise each non-functioning
channel will lead to the loss of array elements.
In the next sections, I will detail our iterative 2D array process from low frequency, with a
smaller number of array elements, to high frequency, 256 array elements for 2D arrays for
imaging. Since the size of the arrays decreases dramatically as we move from low to high

38
frequency arrays, the fabrication process is quite different and encounters different difficulties,
which will be described in detail in the respective chapters that follow.

39
Chapter 4 Design and Fabrication 3-MHz 2D array
4.2 Design and Fabrication of the 3-MHz 64-element 2D Array
4.2.1 Introduction
Retinal degeneration (RD) disease is characterized by progressive degeneration of
photoreceptors in the retina and is one of the leading causes of vision loss and blindness worldwide.
Age-related macular degeneration (AMD) is a major cause of retinal degeneration affecting 196
million people worldwide [95].Another cause is retinitis pigmentosa, with a worldwide prevalence
of 1:3000 to 1:7000 [96]. Despite the loss of sensitivity to light, the rest of the visual pathway is
mostly intact and functional in patients with retinal degeneration, making prostheses a useful tool
to restore lost vision[97]. The current treatment strategy for blindness due to retinal degeneration
is the use of invasive retinal prostheses based on electrical stimulation. Non-invasive ultrasound
stimulation of the retina is considered a more promising technique for vision restoration and has
been studied in vitro and in vivo by several different groups [79, 98-100]. Despite the mechanism
of US neuromodulation being still under investigation [101, 102], efforts have been made to
exploit the potential of US neuromodulation to treat various nerve-related diseases [103, 104]. Our
group work reports the first observation of ultrasound-evoked visual signals in vivo in rats blinded
by retinal degeneration[105].
Based on the above motivation, we designed and fabricated the first generation of ultrasound
2D array as an attempt.
4.2.2 3-MHz Ultrasound 2D Array Design
As an initial attempt for 2D array-based retina stimulation, a 2D ultrasound array transducer
with 3 MHz central frequency, 64 active elements, and 480µm (~1λ) pitch was designed for this
study, details of the array parameter are presented in Table 3-1. The aperture geometry can be
described as an 8 by 8 matrix with a 6 by 6 opening area. Since the array was designed for retina

40
stimulation, the requirements of our array are different from the arrays used for imaging.
Table 4-1. Parameters of the 3MHz stimulation array
Unlike the transducer for imaging, the transducer for stimulation requires slightly different
performance. In the field of Acoustic Radiation Force pulse stimulation, low-frequency
transducers with narrow bandwidths that can produce long pulse durations have been widely used
[106, 107]. Therefore, narrow bandwidths will not be a problem for stimulation transducers. Since
the stimulated transducer only needs to operate at a single frequency and does not need to receive
a signal, the array is not fabricated with a matching layer to improve the energy transfer efficiency
of the transducer at the center frequency and to reduce attenuation.
Similarly, the backing of the array used for imaging has a high acoustic attenuation coefficient,
which will be traded for lower amplitude and wider bandwidth at the cost of transmitted power
and sensitivity. However, for arrays, transmitted power is the first consideration. Therefore, in this
array, our interposer pillars were designed to be reduced in size and filled with uncentrifuged Esolder. The main function of the interposer is thus an acrylic frame filled with a small amount of
E-solder with a low attenuation coefficient and thereby increase the resonance and output power
of the elements. Also, we have tried to use composite materials as a comparison with bulk
piezoelectric ceramic materials DL-48(Del-Piezo Specialties, FL, USA) in the first generation of
arrays used for stimulation.
Parameters Values
Piezo material PZT-DL48 bulk ceramic & 1-3 composite
Design Center Frequency 3 MHz
Number of Elements 64
Pitch 480 µm
Element Size 415 µm * 415 µm
Kerf 65 µm
Volume Fraction 53 %
Array Size 5 * 5 mm

41
4.2.3 KLM Model-based Simulation
To achieve the long pulse duration of the array, we used an online tool BioSono KLM 2.0
(Fremont, CA, USA) based on the KLM model to simulate the performance of the array
elements. The simulation results are shown in Fig. 4-1:
It is clear from the figure that even in the absence of a matching layer, the 1-3 composite still
has a bandwidth of 40% due to a lower acoustic impedance and a larger electromechanical
coupling coefficient(kt). In contrast, the bulk piezoelectric ceramic material has a bandwidth of
only 20% without the matching layer, but with a significantly higher pulse duration. For
stimulation 2D arrays, bulk materials are more suitable. However, as a first attempt, we made our
own 1-3 composites to compare the actual experimental results.
4.2.4 1-3 Composite and 3D Printed Interposer Fabrication
To begin with, we made our own piece of 1-3 composite material in the laboratory with a
center frequency of 3 MHz. Starting from an ordered piece of DL-48 (Del-Piezo Specialties, FL,
USA) with a thickness of 1 mm, we fixed the piezoelectric on the glass and dice it along two
directions orthogonal to each other. Due to the fragility of the piezoelectric ceramic, the two cuts
need to be made separately, and the kerfs need to be filled with epoxy resin (EPO-TEK 301) in
A B
Figure 4-1. Pulse-Echo impulse response simulation for 3-MHz 2D array fabricated with
(A)1-3 composite. (B) bulk PZT DL-48 material.

42
the middle of the cut to stabilize the structure. For the composite material with a center
frequency of 3 MHz, we cut with the parameters of a kerf width of 65 μm and a depth of cut of
700 μm. Finally, the diced and epoxy-filled composite is lapped to a thickness of 600 µm and
electrodes are plated on both sides. The resulting 1-3 composite is shown in the Fig. 4-2. In the
following fabrication process, 1-3 composites and bulk materials are the same, which will not be
described in detail here.
In the 2D array fabrication, the interposers act as the interconnection layer between the
material and the circuit. The most important parameter of the grid matrixes is the pitch between
the channels on the material side and circuit side. For this design, the array elements are spaced
at 480 µm and the circuit array elements are spaced at 900 µm. This is the first time we have
used this structure in a project, as the larger circuit spacing reduces the cost of circuit fabrication.
The model of the intermediary layer was designed by a CAD tool and manufactured by a ProJet
3500 HD printer (3D Systems, Inc., Rock Hill, SC, USA) in XHD mode using VisiJet M3
Crystal acrylate material as shown in Fig. 4-3(A). The printer was able to achieve 34 µm in the X
and Y dimensions and 16 µm resolution in the Z direction only (layer thickness). In fact, the final
dimensions of the smallest structures are typically 60 to 80 µm in the X and Y dimensions. This
is still sufficient for the design of primary 2D arrays and for the fabrication of 2D arrays with
A B C
Figure 4-2. (A)Composite dicing in 1st direction. (B) composite dicing in both directions.
(C)Composite after both sides sputtered.

43
denser cell matrices.
To fill the channels of the intermediary layer with electronic solder, we designed and printed a
frame (Fig. 4-3(B)) slightly larger than the size for each printed grid. This frame will act as a
holder and a dam for the E-solder during the filling channels process of the interposer and is
typically designed to be about 15 mm taller than the interposer.
SLA 3D printing often results in print material remaining in the channel when printing tiny
channels, so the last step before filling is the blockage detection of the interposer. Printing highdensity channels can be a challenging task for 3D printers. The condition of the printer can
significantly affect the quality and yield of the mesh matrix. Due to the high price of E-solder,
securing the channels is critical. However, the inspection process is simple and efficient: the
print channels are first observed visually through a microscope for blockages and breakage, and
if no problems are found, fill the interposer with water. Pushing the interposer into the water or
alcohol is an effective way to achieve this goal, and it is important to ensure that all channels are
sealed with liquid. Next, use compressed air from one side to blow the liquid out of the channels.
Check if there are still water channels. If some channels are blocked, we can clean the channels
with a tiny electrical cord. However, this action may also damage the walls and cause a short
A B
Figure 4-3. (A)3D printed interposer grids before filling; (B) 3D printed dams.

44
circuit between channels. Letting go of the blocked interposers would be the best way to go.
After confirming that the yield of the grid is 100%, we can start the filling process using the
dam. The dam was designed with three walls and one side of ‘door’ (Fig. 4-3(B)) for easy
removal and was secured to the glass with 15-minute epoxy to ensure its bottom was sealed. ESolder will be poured into the frame after blending and mixing. Due to the inherent viscosity of
the E-solder, air bubbles are created during the mixing process and are difficult to remove
completely by vacuuming alone. Therefore, we use centrifugation to remove the air bubbles from
the mixed E-solder. And as discussed earlier, we do not want the backing layer to absorb too
many acoustic waves in the array used for stimulation, so a backing layer with a higher acoustic
attenuation coefficient after high-speed centrifugation is a poor choice. In this case we set the
speed of the centrifuge at 900-1000 rpm for 5 minutes. As a matter of practice, the E-solder does
not delaminate at such a setting and discharges most of the air bubbles very well.
The next step is to press the clean and cleared un-filled interposer into the dam (Fig. 4-
4(A)and (B)). Due to the blockage of the dam, the E-solder will be pressed into the empty
channels. The filling process is complete when we observe the E-solder spilling out of the top
channel. In the next step, we will cut off the dam on the ‘door’ side and remove the filled
interposer to come. In practice we have also found that removing the E-solder from the surface
of the interposer in this step will seriously affect the quality of the finished product as the Esolder will shrink to some extent after curing. Here, we move the filled frame away from the
‘door’ side to ensure that excess E-solder remains on both sides of the channel. The filled
interposer is cured at room temperature for 24 hours, after which it is placed in a 45 °C oven for
a post-cure. We will sand the finished cured middle layer with sandpaper to clean the surface and
ensure the flatness of both sides of the channel (Fig. 4-4(C)and (D)).

45
Finally, we also need to make the alignment marks on the side where the interposer is
connected to the piezoelectric ceramic (Fig. 4-4(E)). Since we will not be able to confirm the
channel position after fixing the piezoelectric ceramic to the interposer in the next step, we use a
dicing saw at this step to dice kerfs of about 100-200 µm in depth on the upper surface of the
interposer and fill it with dyed epoxy as an alignment mark for subsequent dicing.
4.2.5 Array Assembly Process
The assembly of the array is mainly divided into two parts: the assembly of the piezoelectric
material to the interposer and the assembly of the circuit board and to the interposer. The key to
assembly is flatness and alignment. To achieve that, we designed a customized fixture with a 2-
axis stage combined with a rotation stage (Fig. 4-5). The bottom stage is built by combining two
stages and allows the adjustment of the location of the object on the top of it. The sample holder
can be taken out from the fixture, and usually used to hold the top part in the assembly. A USB
microscope (Dino-Lite, Taiwan) was used to monitor the alignment process.
Assembling the material onto the interposer is a relatively simple process. We start by fixing
the fabricated interposer to the glass with wax, and then fix the glass to the assembly table. The
A B C D E
Figure 4-4. Interposer filling and alignment marks process (A)3D printed dam.
(B) interposer filled with E-solder. (C) cured and lapped interposer on the transducer side.
(D) cured and lapped interposer on the circuit side. (E)alignment marks dicing and dyeing.

46
piezoelectric material will be fixed to another smaller piece of glass with wax. Since the material
we prepared is larger than the actual area of the array, this step is easy to calibrate, just make
sure that the piezoelectric material completely covers the array elements area of the 2D array
(Fig. 4-6(A)). Next, we apply a layer of E-solder between the material and the interposer and
press the material to it. The amount of E-solder is not an issue in this step. A larger amount of
solder will ensure a good electrical connection between the material and the interposer.
After the E-solder has cured, we will proceed to the next step: dice out the array elements
along the alignment marks. By the time the material assembly is finished, all the channels are
shorted together by an e-solder. To create independent elements, we diced through the material
and the e-solder layer and filled the kerf with epoxy. Again, due to the fragility of the ceramic,
we need to cut it in two ways and fill the cut kerfs with epoxy (EPO-TEK 301) between the twodicing process (Fig.4-6(B) and (C)). During this step, the epoxy will also provide a GND
connection structure. In the process of designing the interpose, GND channels are added at the
four corners. In this step, we planned to short the upper surfaces of the piezoelectric ceramics
together and connect them to the GND elements of the circuit. We did this by sputtering gold
Figure 4-5. The lab-made array assembly stage.

47
electrode Cr/Au (50/100nm) on the upper surface of the 2D array after the epoxy cured. Since
the epoxy provides a gentle downhill structure at the edge of the array, the sputtered gold
electrodes can smoothly short the array elements together and connect to the GND channel (Fig.
4-6(D)).
The connection of the interposer to the circuit board is more challenging than the assembly of
piezoelectric materials. The assembly of piezoelectric materials is on the upper surface of the
interposer, and even if it fails, there is a possibility of remediation. In contrast, the circuit
connection area is in the middle part of the structure, and we need to ensure that all 64 arrays are
connected to the circuit at the same time, with no shorts or breaks and in only one go. Before
that, we need to dice the channels where the interposer is connected to the circuit into separate
pins with a dicing saw (Fig. 4-7(D)). Here the kerf width and depth of cut need to be set at 100-
200 μm. Too wide or deep will result in too fragile pins and too thin or shallow will result in a
short circuit during the subsequent assembly process.
After the dicing work is completed, we will cover the pins with a small amount of E-solder
evenly and connect them to the circuit. Since too much E-solder will inevitably lead to a short
circuit, we use a ‘stamping’ process to ensure that the circuit is connected. We first applied a 50-
60 μm thick copper sheet to a piece of glass and applied a small amount of solder in the middle
of the sheet. After that, we use a hard and horizontal object, such as a razor blade or a copper rod
A B C D
Figure 4-6. The piezo material assembly process:
(A) Assemble the material onto the interposer. (B) First dicing and filling process.
(C) Dicing and filling process finished. (D) GND connection finished.

48
to scrape across the surface of the copper sheet, which will leave a uniform and flat layer of Esolder in the middle of the copper sheet (Fig. 4-7(A), (B) and (C)).
Next, we will stamp the interposer onto the thin layer of E-solder. This process ensures that
only the right amount of E-solder adheres to the pins of the interposer to form the solder bumps.
Finally, we will use a USB microscope (Dino-Lite, Taiwan) to look at the four directions of
the fixture and press the interposer onto the board while ensuring that the acoustic stack is
aligned with the board perfectly (Fig. 4-7(E) and (F)).
In both assemblies, the whole fixture remains locked during the curing and post-curing (45℃,
2 hours). When the post-curing is finished, the whole fixture will be heated to 65℃, which is the
melting point of the wax, and the array with the circuit will be removed from the fixture and the
array fabrication is finished (Fig. 4-8).
d e f
Figure 4-7. The acoustic stack assembly process: (A) E-solder layer before stamping. (B) E-solder
layer after stamping. (C) Close view of the E-solder after stamping. (D) Dicing of the pins
finished. (E) Close view of the acoustic stack assembly. (F) Top view after assembly finished.

49
4.2.6 Array Performance Test
The performance of the 2D array was evaluated by pulse-echo (P/E) testing and hydrophone
testing. For the P/E test, a quartz cube was used as the target and placed in a deionized (DI)
water tank at a depth of approximately 20 mm, which is the focal length used in the following
tests and experiments. The array is firmly attached to a controlled stage with a micrometer that
allows for in-plane adjustment of the position in both directions. During the test, one channel at a
time is opened and each element is interrogated in turn. The frequency of the single-cycle pulse
is set to the operating frequency of the array. To ensure signal consistency of the P/E results, the
array tilt is manually corrected by monitoring the time of flight of each element and adjusting the
physical orientation of the array.
From the P/E results, the bandwidth and sensitivity of the individual array arrays can be
derived. We have selected one array element made of composite material and one array element
of optical material array for comparison. From Fig. 4-9, we can clearly see that the experimental
results are in good agreement with the simulation results (Fig. 4-1). P/E response waveform of
the optical material shows a longer pulse duration corresponding to a narrow bandwidth of -6 dB.
This is also evident in the fast Fourier transform (FFT) of the P/E waveform.
Figure 4-8. The Finished 3-MHz 64-element 2D array

50
At the same time, we generated sensitivity maps using the P/E results. We plotted the
normalized echo amplitude as a sensitivity map according to the channel location of its array
element. As shown in Fig. 4-10, this map represents the yield of the array and the degree of
uniformity of the array element performance. Arrays with too low a yield will not be worthy of
subsequent testing. The main reason for the insensitive arrays in the figure is the breakage caused
by the lack of flatness of the components during the assembly process. As shown in Fig. 4-10(B),
the yield rate of the array made with bulk material is over 95%. After testing the wire connecting
the ultrasound system (Vantage 256, Verasonics, Inc., Kirkland, WA, USA), we found that one
of the two damaged elements was broken due to a cable problem.
A B
C D
Figure 4-9. The P/E results comparison of 3-MHz 2D array: (A) P/E impulse response of composite array
element. (B) P/E impulse response of bulk material array element. (C) Electrical input impedance of
composite array element. (D) Electrical input impedance of bulk material array element.

51
The acoustic fields of the arrays were calibrated and mapped using a hydrophone (HGL-0085,
ONDA Co, Sunnyvale, CA, USA) and a 3D scanning system (SGSP33-200, OptoSigma
Corporation, Santa Ana, CA, USA), which was controlled by the PC to perform a full scan the
opening window area parallel to the array. A function generator (AFG 3252 C, Tektronix,
Beaverton, OR, USA) was used to synchronize the Verasonics system with the digitizer card in
the PC. The drive signal’s frequency was set at 3 MHz, and the output voltage is 5V VP-P. The
voltage applied during the measurements was much lower than the voltage level during the
experiments to protect the hydrophone. During the testing, the focal depth was set at 5 mm. For
both arrays, we measured its acoustic field with different steering parameters. The hydrophone
results are shown in Fig.4-11, pixel size in the figures is 180 μm by 180 μm:
From the shape of the sound field of the hydrophone results we can clearly see that the 2D
array made of bulk material has a concentrated focus area and a weaker side flap. In addition to
this, the hydrophone measurements also provide data on the pressure of the sound field at the
A B
Figure 4-10. The P/E magnitude mapping for (A) 3-MHz composite array and
(B) 3-MHz bulk material array.

52
focal point. At an output voltage of 10 VPeak, the pressure at the focal point of the 2D array made
of composite material is 1.1 MPa, while the array made of bulk material has 2 MPa, which is a
very strong pulse, even higher than the sound field pressure of many single element transducers.
4.3 2D Array Based Ultrasound Retina Prosthesis Experiment
It is demonstrated that US stimulation on either normal-sighted or retinal degenerative blind rats
in vivo can evoke neuronal activities in the visual pathway to the brain, demonstrating ultrasound
as an effective neuromodulation approach. The 3-MHz 8 × 8 ultrasound 2D array was used to
dynamically stimulate the retina.
The schematic diagram of the experimental setup with ultrasound stimulation and the actual
experiment photo is shown in Fig. 4-12. We plan to use an angular spectrum (AS) algorithm to
calculate the amplitude and phase distribution required to hit different patterns with an
ultrasonic two-dimensional array and insert electrodes in the corresponding regions of the rat
brain to receive and restore the electrical signal patterns.
Normalized Intensity
Intensity
Normalized
Figure 4-11. The hydrophone results in XY plane of 3-MHz 2D array made with
(A)composite. (B)bulk material.
A B

53
However, in our experiments, we found that the array area of only 64 array elements was too
small to find the location of the rat retina. Also, the number of array elements limits the area
and shape of the pattern, and it is difficult to capture the electrical signal and restore the shape
of the pattern in the brain. Therefore, we plan to make the next generation of ultrasound 2D
arrays with more elements and larger areas for this experiment.
4.4 Discussion and Conclusion
In this chapter, the fabrication of linear arrays is briefly introduced and the process of
designing and manufacturing ultrasound 2D arrays using a pitch-shifted interposer is described
in detail. The fabricated arrays showed very close performance to the simulated results in the
tests.
As a first attempt to use a pitch-shifted interposer and use it for a stimulation 2D array, the
array design is conservative and the fewer array elements reduce the difficulty of assembly. As
a comparison, the array using the bulk material has a larger sound field pressure and is more
suitable for stimulation. After the 2D arrays are fabricated, pulse/echo tests and hydrophone
A B
Figure 4-12. (A) Schematic diagram of the experiment setup and (B) Photo of the ongoing
experiment. (A)a) The retina stimulation part. A focused single-element transducer was used to
generate ultrasound waves targeting the retina. Retinal neurons were excited and generated neural
signals transmitting through optic nerve to the brain. b) The brain recording part. A multielectrode
array (MEA) was inserted to the contralateral superior colliculus (SC) or visual cortex (VC)

54
tests are required to evaluate their performance. Any defect in the fabrication process will result
in reduced array performance. A step-by-step approach must be taken to ensure a high yield
rate.
In the process of production, we have also accumulated many valuable experiences that
cannot be gained without practice. For example, we suffered a lot in the process of printing the
interposer. The interposer often blocked channels after printing, which were difficult to
distinguish with the eyes. When we tried to melt the clogged wax in the channels with heat, we
also found that the interposer was distorted. After filling the E-solder, the interposer needs to
be checked for pathways, otherwise it is easy to end up with short circuits and broken circuits.
The use of alignment marks is also a process that we have worked out in practice.
As the first attempt to use a pitch-shifted interposer, the success of this 2D array has allowed
us to gain experience and confidence in the feasibility of this structure. In the next sections, we
will present our work on building 2D arrays with different frequencies and for different
purposes using the same PCB board.

55
Chapter 5 Design and Fabrication of 4-MHz 256-elements
stimulation 2D array
5.1 4-MHz Ultrasound 2D Array Design
As mentioned in Chapter 3, we plan to make 2D arrays with more array elements and larger
areas for experiments on rat fundus retinal stimulation using a similar process. However, as a
prototype, the advantage of a pitch-shifted interposer that can greatly reduce the circuit
requirements is not fully demonstrated. In this design, we will also take advantage of the pitchshifted interposer, using the same circuit board to drive 2D arrays with different designs thus
reducing cost. Based on the above considerations, we have taken the lead in fabricating a 2D
array with 256 arrays at a center frequency of 4 MHz. The array aperture had a 16 by 16 matrix,
the element pitch of the array was designed to 750 μm. As in the previous chapter, no matching
layer is added, only a layer of parylene was plated as a waterproof layer. An illustration of the
transducer design and a photo of the fabricated 2D array are shown in Fig. 5-1. The details of
the array design can be found in Table. 5-1
Figure 5-1. (a) The layered structure of an element in the array. (b) The photo of the fabricated
array together with interposer and PCB board.

56
Table 5-1. Parameters of the 2 & 4 MHz stimulation array
5.2 Circuit Design
High-density interconnect printed circuit boards (HDI-PCB) were used to route the
elements to the system cable (Fig. 5-2). These industry-standard substrates have a 2+N+2
architecture, 3 mils (75 m) trace and space and are capable of being fabricated by general
manufacturers. Compared with conventional PCBs, HDI-PCBs are smaller in size and provide
different via types instead of through holes only, which are advantageous in reducing routing
complexity and increasing the density of the pad matrix. The PCB used for array connection
has a 16  16 pad matrix with a 635 μm pad pitch in the central area, which will be used for
array assembly in the following process.
Parameters Values
Piezo material PZT-DL48 bulk ceramic
Design Center Frequency 4 MHz
Number of Elements 256
Pitch 770 µm
Element Size 620 µm * 620 µm
Kerf 150 µm
Volume Fraction 75 %
Array Size 14 * 14 mm
Pad matrix
Figure 5-2. High-density interconnect printed circuit board for the 2D array.

57
5.3 Interposer Design and Fabrication
The transducer aperture was a 16  16 element matrix with a 770 μm element pitch. Arrays
with 2 pitch demonstrate grating lobes during steering, yet a larger aperture benefits acoustic
field pressure and provides increased penetration depth. The interposer architecture can be
decomposed into three parts. The interposer was designed with an isometric magnification
structure and is mainly formed with parallel channels where the pitch increased from 635 μm
(at the PCB side) to 770 μm (at the transducer side). The top and bottom were straight channel
regions each with a 16  16 matrix with matching pitch to the pad matrix on the PCB and the
aperture. The thickness of the walls between the channels was 235 μm and 150 μm for the
PCB side matrix and the transducer side matrix respectively. The wall thickness to pitch ratio
was optimized differently for the PCB vs. the transducer matrix. On the transducer side, a
thinner wall was preferred for a more active area of the interposer backing for better acoustic
absorption. The thicker walls on the circuit side reduce the risk of short circuits between
channels during the assembly process. In the center region, the channels were scaled
isometrically and the distance between the channels changed from 635 μm to 770 μm.
The 3D Systems HD 3500 plus (3D Systems, Rock Hill, SC) with a minimum resolution of 32 μm was
selected for this interposer. The fabrication process is like the array we discussed in chapter 3, only a brief
description is given here, and process improvements are highlighted.
Dams for containing the silver epoxy filling material were fabricated to match the outer dimension of the
interposers and were glued to a supporting glass substrate for easy handling (Fig. 5-3(A)). The dams were
filled with the silver epoxy and the interposers were carefully lowered into the dam-encapsulated epoxy.
An important consideration in this regard is that the dams need to fit snugly with the interposer outer
dimension with minimal gaps between the two.
The fabricated dam/glass substrate structures were filled with conducting silver epoxy paste (E-Solder

58
3022, Von Roll Isola, New Haven, CT, USA) and centrifuged at 1000 rpm for 5 minutes. The interposers
were pushed into the dam-confined silver epoxy slowly and with evenly applied high pressure to ensure
that the paste fills every channel (Fig. 5-3(B)). After the interposer was pushed all the way to the glass, the
dam was cut, and the interposer was released (Fig. 5-3(C)). Filled interposers were left in a dry box for 24
hours for curing. After curing, the silver paste in the channels hardens and forms a conductive pathway.
The cured interposers were then lapped at the top and bottom surface to remove the extra E-solder and
create a flat surface on each side (Fig. 5-3(D)). Surface co-planarity is critical for both the subsequent
transducer fabrication steps as well as the assembly to the electronic substrate. We therefore incorporated
surface measurements into the post-fabrication testing criteria with a goal of co-planarity below +/-10m.
The last step in the fabrication process was dicing on the surface to create fiducial marks which are critical
for precise alignment for the subsequent steps in the overall acoustic module fabrication process (Fig. 5-
3(E)).
5.4 Array Fabrication
The fabrication process for the acoustic stack (Fig. 5-4) was similar to the arrays in chapter 3.
Hard PZT material (DL-48, DeL Piezo Specialties, LLC, West Palm Beach, FL, USA) was used
for array fabrication. It had a pitch width of 770 μm and kerf width of 100-μm diced by a highspeed dicing saw (Tcar 864-1, Thermocarbon, Casselberry, FL, USA). The finished piezoelectric
material was lapped down to a thickness of 350 μm to achieve the designed center frequency (4
MHz). Then the sample was then mechanically polished and sputtered with a Cr/Au (50/100 nm)
electrode using a sputtering system (NSC-3000 Sputter Coater, Nano-Master, Inc., Austin, TX,
A B C D E
Figure 5-3. The interposer filling process.
(A)The dam and the unfilled interposer. (B) Interposer filling process. (C) Interposer after filling
(D) Interposer cured and lapped. (E) Alignment marks finished.

59
USA). The piezoelectric sample was then bonded to the interposer backing using a thin layer of Esolder. After the assembly had cured, the piezoelectric material was diced along both azimuthal
and elevational directions with the alignment marks to create 2-D array elements, resulting in a
total of 16 × 16 2-D array elements. It had a pitch width of 770 μm and kerf width of 150 μm diced
by a high-speed dicing saw (Tcar 864-1, Thermocarbon, Casselberry, FL, USA). The kerfs were
filled with EPO-TEK 301 (Epoxy Technology, Billerica, MA, USA) and allowed to cure. This was
followed by sputtering a Cr/Au (50/100 nm) electrode to create the top electrode linking all the
elements as the ground connection. The finished acoustic stack had 256 active elements in total.
Next step is to flip the array over to dice the connector pins out.
After the fabrication was completed, the acoustic stack was bonded with the FPC using Esolder 3022 and previous ‘stamping’ process. The FPC also comprised two connectors for
interface with two modified Philips/ATL L7-4 cables, which were then connected to the
ultrasound system (Vantage 256, Verasonics, Inc., Kirkland, WA, USA).
Note that in this production, we noticed that the array elements at the outermost part of the
A B C D
E F G H
Figure 5-4. The array fabrication process. (A)Bulk DL-48 assembled onto the interposer.
(B) First dicing work finished. (C)Element dicing finished. (D) GND connection sputtered.
(E) Connector pins dicing. (F) Connector pins dicing with thinner blades.
(G) Connector pins dicing finished. (H) ‘stamping’ process with USB microscope.

60
array in previous tests had poor performance, so we deliberately kept a circle of dummy elements
outside the outermost array elements as a control when cutting the array elements.
5.5 Array Performance Test
The P/E results for the 2D array were obtained by a similar experiment as in Chapter 3.
Sensitivity plots were produced from the data, and the measured combined yield was 95% (Fig. 5-
5)
The main reason for the non-functional elements was blocked interposer channels. The HD
3500 plus printer uses wax to support the structure. Due to the interposers’ porous structure,
complete wax removal was difficult to achieve, with the potential for any residual wax to end up
blocking the interposer channels. The blocked channels could not provide a continuous
conducting path due to the resulting failure of the e-solder extrusion process at that element.
Printers which use resin are prone to the presence of residual resin in the channels, however, the
removal process for the resin is usually easier and has a higher yield.
The assembly of the 2D array is more successful from the results. Due to the large number of
array elements, there are some array elements at the edges of the array due to unevenness
resulting in poor contact.
A B
Figure 5-5. Array performance maps: (A) Sensitivity map. (B) Bandwidth map.

61
From the hydrophone measurement (Fig. 5-6), 60 mV voltage was received at the focal point
with a 17 V driving voltage. And the pressure of the acoustic field was calculated to be 1.4 MPa.
The driving voltage was limited by design during this measurement to protect the hydrophone. In
the actual neural stimulation experiment, the Vantage system with HIFU configuration was used
to provide up to 90 Vpeak output voltage. Therefore, given linear gain in the acoustic field pressure
with drive voltage, the required pressure can be easily achieved.
5.6 2D-array Based Ultrasonic Retinal Prosthesis Experiment
The schematic diagram of the experimental setup with ultrasound stimulation is shown in Fig.
5-7. A function generator (AFG3252C, Tektronix, Beaverton, OR, USA) was implemented in this
study to synchronize the ultrasound stimulation system (Vantage 256, Verasonics, Inc., Kirkland,
WA, USA) and multi-channel electrophysiological data acquisition Lablynx system (Neuralynx,
Bozeman, MT, USA).
Figure 5-6. Acoustic field measured by hydrophone.

62
During the experiment, the US 2D array was connected to a 3D-printed collimator, which
controls the distance between the array and eyeball to be the exact 10 mm. The collimator was
filled with de-gassed ultrasound gel. The 2D array was controlled by the Verasonics Ultrasonic
system, which can control the amplitude and phase of each element in the 2D array. By applying
Figure 5-7. Schematic diagram of ultrasound retina stimulation system. Top: The workflow of 2Darray ultrasound retina stimulation with dynamic patterns. Bot: Verasonics ultrasound system. The
structure of one array element. The ultrasound evoked neuron responses with a “C” pattern.
Figure 5-8. The distribution of different letter patterns. (a) Amplitude distribution of 2D array
for “C” pattern. (b) Phase distribution of 2D array for “C” pattern. (c) Simulated generated “C”
at the pattern plane (20 mm away from the array surface). (d) Mapped neuron responses at SC of
“C” pattern. (e) Mapped neuron responses at SC of “V” pattern. (f) Mapped neuron responses at
SC of “S” pattern.

63
the correct amplitude and phase distribution, an ultrasound 2D array can generate arbitrary desired
patterns. Here we demonstrated that different letters can be generated using our 256-elements 4.5-
MHz 2D array. The pattern plane is 20 mm away from the surface of the array so they can be
applied on the rat retina during the experiment. The “C” pattern is used as a representative example.
Fig. 5-8(a)&(b) show the calculated amplitude and phase distribution of the 2D array to generate
pattern “C”. The generated pattern at a 20 mm distance is shown in Fig. 5-8(c). Then, this pattern
is used to stimulate the retina, and the neuron responses at SC were mapped using MEA. Response
mapping is shown in Fig. 5-8(d). Gray color points indicate the location of electrodes. We further
showed letters “V” and “S” are successfully achieved in rat brains.
5.7 Discussion and Conclusion
In this work a 2D ultrasound array with a center frequency of 4 MHz was designed, fabricated
and tested. The circuit design cost was saved by using pitch-shifted interposer in the array
design. In the testing of the array performance, we can see that the stimulated array has enough
power to meet the experimental requirements. In the subsequent rat fundus retinal stimulation
experiments, the designed pattern was successfully struck in the rat fundus by changing the
amplitude and phase of the 2D array elements, and the signal of the same pattern was received at
the electrodes inserted into the rat brain.
We have gained valuable experience and achieved good results for stimulated 2D arrays. In
the future, we will continue to explore stimulated 2D arrays with larger dimensions and aim to
apply them in the research of diseases related to visual impairment.

64
Chapter 6 Design and Fabrication of
15-MHz 256-elements Imaging 2D Array
6.1 2D Array and Interposer Design
In Chapter 5, we have mentioned the plan to use the same board to drive a 2D array of different
designs to. This chapter will describe the design, fabrication, and testing of a high frequency
ultrasound 2D array for imaging based on a pitch-shifted interposer. Unlike the 2D arrays used for
stimulation in the previous chapters, the acoustic part of the imaging high frequency 2D array
requires two matching layers and a very small array element size.
The materials selected for the 2D array in this chapter is 1-3 composite made from soft PZT
material (DL-53, DeL Piezo Specialties, LLC, West Palm Beach, FL, USA) and epoxy as kerf
filler (EPO-TEK 301, Epoxy Technology, Billerica, MA, USA). Two matching layers (2-3 μm
silver epoxy and ABS plastic) were applied. Details of the materials and parameters used in this
two-dimensional array design can be found in the Table 6-1.
Table 6-1. Parameters of the 15MHz image array
Parameters Values
Piezo Material 1-3 composite of Soft PZT ceramic
kt 0.6
First Matching Layer (1st ML) Material 2-3 μm silver epoxy
1st ML Sound Speed 1961 m/s
1st ML Acoustic Impedance 7.84 MRayl
1st ML Thickness 31 μm
Second Matching Layer (2nd ML) Material ABS plastic
2nd ML Sound Speed 1850 m/s
2nd ML Acoustic Impedance 2.2 MRayl
2nd ML Thickness 29 μm
Backing Layer Material E-solder with 3D printed acrylic interposer
Design Center Frequency 15 MHz
Number of Elements 256
Pitch 200 µm
Element Size 140 µm * 140 µm
Kerf 60 µm
Array Size 14 * 14 mm

65
Since the array element size of 2D arrays is very small, single crystal material may be the
optimal solution in this case. However, based on previous experience, single-crystal materials are
very fragile, and it is difficult to achieve high yields with such small processing sizes. Therefore,
we chose materials with relatively high dielectric constants aiming to obtain more stable outputs.
As shown in Fig. 6-1, a 16-by-16 matrix is lined up at the effective aperture of the array. Due to
process constraints and the trade-off between field-of-view area and imaging quality, the array's
element pitch was designed to be 2 (200 µm). When conditions allow, the array pitch can be
reduced by increasing the number of elements, thus reducing the grating lobes to obtain an array
with the same field of view and higher image quality. The pitch-shifted, which serves as the
Figure 6-1. Top: Conceptual drawings for image 2D array matrix and elements’ structure.
Bottom: Perspective View of the pitch-shifted interposer.

66
connection between the piezoelectric material and the circuit, is designed as an isometric
amplification structure consisting of 263 parallel channels (256 array and 8 grounded channels).
The apertures at each end of the interposer are matched to the part to which they are attached. The
top acoustic side matrix channel has a side length of 140 µm and a wall thickness of 60 µm. The
bottom circuit side matrix channel has a side length of 400 µm and a wall thickness of 235 µm.
BioSono KLM 2.0 (Fremont, CA, USA) was implemented to evaluate the acoustic
performance of the array with the design parameters. The pulse-echo simulation result was
presented in Fig. 6-2. With the two matching layers design, the bandwidth of the elements would
reach 90%.
In practice, we have found that 3D printing longer channels at small size can easily lead to
overexposure and lead to blocked channels. After discussion, we shortened the channel of the top
acoustic part as much as possible, and selected 2mm as the channel length that can be used for
effective sound energy attenuation on the back of the acoustic part in the simulation results. For
imaging arrays, the effective acoustic attenuation of the back can reduce ringdown and thus
greatly improve the axial resolution.
Figure 6-2. Pulse-Echo impulse response simulation for 15-MHz 2D array element.

67
6.2 Interposer and Array Fabrication
Unlike low-frequency arrays, the size of the interposer used for high-frequency array is very
small, placing high demands on the 3D printing resolution. The smallest printed structure in the
interposer is designed to be 60 μm, which makes the printing task more challenging for the printer.
Here we set the channel length at 4 mm because in our previous attempts we had printed interposers
with a length of 10 mm to reduce ringdown, but the printed interposers had channel shorts (Fig.6-
3 (A)). After discussion, we decided to print long channels at small sizes where the fragile structure
is prone to errors, so we shortened the length. In fact, the minimum resolution of the 3D printer
can only be achieved under certain conditions. Therefore, in order to successfully print this
interposer, a printer that offers finer resolution than the minimum structure size is required. For
this project, we used a DLP 3D printer (Kudo, Dublin, CA, USA) with a minimum resolution of
15 μm to produce this design, figuring out how to print the interposer in a trial-and-error process.
After printing, the interposer was immersed in alcohol and placed in an ultrasonic cleaner for 15-
20 minutes. Due to the porous structure of the interposer, this step was critical to remove any
residual resin from the channels to avoid blockage of the channels. The interposers are not fully
cured after printing, so they are further cured in an oven at 45°C for one more hour after cleaning.
Each interposer is checked again to ensure a clean channel yield after curing.
The filling process is similar to the steps in the previous chapters. Dam containing silver epoxy
filling material is fabricated to match the outer dimensions of the interposer and bonded to the
supporting glass substrate for easy handling. The dam is filled with silver epoxy and the interposer
is carefully placed into the dam secured with epoxy. An important consideration in this regard is
that the dam needs to fit closely to the outer dimensions of the interposer with a minimum gap
between them. If the gap is not large enough, the interposer cannot be lowered due to friction with

68
the wall, while too large a gap can lead to leakage during filling and waste of material due to
inadequate channel filling. Here, we use Kapton tape to wrap around the interposer to adjust it to
the right size.
The fabricated dam/glass substrate structures were filled with conducting silver epoxy paste
(E-Solder 3022, Von Roll Isola, New Haven, CT, USA) and centrifuged at 1000 rpm for 5
minutes. Centrifuging is important for consolidating the silver epoxy particles to increase the
acoustic attenuation of the backing material. After this process, the paste surface was flat, and
any air bubbles created during the mixing process were removed. Uniformity of the backing
material with no air bubbles is important to maintain high-quality acoustic attenuation without
any significant echoes back to the transducer composite. The interposers were pushed into the
dam-confined silver epoxy slowly and with evenly applied high pressure to ensure that the paste
fills every channel. Yield in the fabrication of 2D arrays is multiplicative, and every step must
have a target yield of 100% such that the final yield of the overall process is high. Therefore, we
optimized the fabrication process of the interposers to achieve the highest yield possible and
further validated the electrical functioning of every channel in the completed structures before
continuing the overall transducer fabrication process.
After the interposer was pushed to the glass, the dam was cut, and the interposer was released.
Filled interposers were left in a dry box for 24 hours for curing. Extra silver paste on the surface
of each active site of the interposers was desired since the paste shrinks during curing and can
lead to dents and holes in the channels. Filling with a given interposer can only be done one time
as an attempt to rework an incompletely filled channel invariably leads to an air gap that isolates
the previously filled paste and new paste in the channels, and this reduces the yield of the
conductive channels significantly. After curing, the silver paste in the channels hardens and

69
forms a conductive pathway. The cured interposers were then lapped at the top and bottom
surface to remove the extra e-solder and create a flat surface on each side (Fig.6-3 (B)). Surface
co-planarity is critical for both the subsequent transducer fabrication steps as well as the
assembly to the electronic substrate. We therefore incorporated surface measurements into the
post-fabrication testing criteria with a goal of co-planarity below +/-10m. The last step in the
fabrication process was dicing on the surface to create isolated assembly pins with air gaps
between them. We further diced the surface to create fiducial marks which are critical for precise
alignment for the subsequent steps in the overall acoustic module fabrication process (Fig.6-3
(C)).
The fabrication process (Fig.6-4) for the acoustic stack was similar to the ring arrays except
that the imaging array had the first matching layer attached before assembly. The first matching
layer was a mixture of Insulcast 501 (Insulcast, Montgomeryville, PA) and 2-3 μm silver
powder. The mixture was cast on the composite while placed in a dam and centrifuged. After
curing, it was lapped to the design thickness of 31 μm. Once the piezo material was ready, it was
bonded on the transducer side of the interposer using a thin layer of E-solder. The E-solder paste
was applied to the interposer surface first, then the material was aligned and pressed down on the
paste (Fig.6-5 (A)). The extra paste squeezed out and surrounded the piezo material. This is by
A B C
Figure 6-3. Interposer fabrication for 15-MHz 2D array. (A) Long interposer with shorted
channels. (B) Interposer filled with E-solder. (C) Interposer with alignment marks.

70
design, as the channels connected to the ground were designed to surround the central channels.
The extra paste connects with those channels and is used for ground connection in subsequent
fabrication steps. The assembly was left in a dry box overnight for curing. Once cured, the piezo
material was diced to create independent elements. The marks created in previous steps were
used for the alignment of the dicing saw to the interposer grid. This process was applied for both
directions of the matrix and EPO-TEK 301 was used as kerf filler to support the elements. In this
step, we found that because the elements of the array element were too small, the dicing process
could easily lose the connection at the interface between the matching layer and the gold layer or
the interface between the gold layer and the piezo material. To solve this problem, we diced and
filled the epoxy resin in two times in the subsequent dicing, which ensured that the contact area
between the array element and the matching layer was large enough to avoid the problem of
matching layer falling off (Fig.6-5 (B) & (C)).
Figure 6-4. Schematic of Acoustic Stack Fabrication, the most left and right pillars are
connected to GND: a: Creating electrodes on the circuit side of the interposer; b: Assembling
Piezo material (with first matching layer for imaging array); c: Creating separated elements and
filling kerf; d: Sputtering gold; e: Attaching second matching layer; f: Finished acoustic stack.

71
It is critical to avoid dicing along the edge of the material, as in this case a short circuit would
be formed between the elements on the edges and ground. We, therefore poured additional kerf
filler in these kerfs to build a smooth bridge from the element surface to the E-solder that covers
the ground. After the curing process of EPO-TEK 301, the array was sputtered with a Cr/Au
(50/100 nm) electrode on the surface of the matrix. By doing so, the top electrodes of all the
elements and the extra E-solder connected to the ground were connected. The “bridge”
mentioned earlier is necessary to make this connection possible and its shape directly influences
the integrity of this connection. The above-described process steps fully describe the fabrication
of the stimulation array. For the imaging array, an additional step is added to the fabrication
process which is a lamination of the second matching layer.
Finished acoustic stacks were next assembled to the PCBs. The PCB was designed at USC
using the Altera CAD tool and fabricated by an external vendor (MKT Electronic Co., Ltd.,
Guangdong, China). The “stamping” process introduced in our previous work was used for
assembly. The circuit side of the interposer was first stamped on a thin layer of E-solder and then
the acoustic stack was aligned and assembled on the matrix area of the HDI-PCB. After curing,
the array fabrication process was completed.
A B C
Figure 6-5. The array fabrication process of the 15-MHz 2D array. (A) Piezo ceramic attached.
(B) Matching layer detaching issue. (C) Dicing with 2 pitch to prevent detachment.

72
6.3 Performance Testing and Imaging Results
A similar process was undertaken to acquire the imaging array’s P/E results. The sensitivity
map was created based on the data and the resulting measured overall yield was 82% (Fig.6-6).
Most of the disconnected elements are located at the upper left corner. We guess this is due to
the unflattens of the interposer during the fabrication process leading to poor contact with the
circuit at one corner of the assembly. In addition, wax is needed to fix the array on a small-sized
piece of glass during the assembly of the circuit. Due to the very small size of the array aperture,
it is difficult to ensure the flatness of each contact surface during this process.
A typical element’s pulse-echo was presented in Fig. 6-7. The average bandwidth of the
working array elements is 46%, which is significantly lower than the simulation result. We
believe that there are two main reasons for this problem. First, the backing layer is short and not
centrifuged, resulting in a relatively low acoustic attenuation coefficient of the backing. In
addition, because the 3D printed interposer has a slight deformation during the cleaning and
fabrication process, we can only ensure that the backing of the array element located in the
middle is in the right size, and near the edge, the backing layer has deformation, so the array
element at the edge position will appear to be partly on the electronic solder and partly on the
3D-printed acrylic.
A B
Figure 6-6. 15 MHz array performance maps: (A) Sensitivity map. (B) Bandwidth map.

73
In addition, due to the very small size of the array elements, a real 1-3 composite material that
can reach the correct vibration mode requires a kerf width of less than 20 microns. The 1-3
composites we purchased had difficulty meeting this requirement, so the actual material in each
array element did not achieve the correct vibration mode for the composite.
The array was next used to image the wire targets. A target with five micro-scale steel wires
was used as the phantom to evaluate the imaging array. The wires were arranged in a trapezoidal
pattern with a vertical spacing of 1.5 mm and horizontal spacing of 0.5 mm (Fig. 6-8).
Coherent plane-wave compounding was implemented as the beamforming technique. Each
imaging cycle generated 9 angled plane waves in the X-Z plane and Y-Z plane respectively.
Imaging views for XZ, YZ, and XY planes were acquired by the Verasonics system in real-time
Figure 6-7. Typical pulse-echo response and spectrum of the 15-MHz image array.
Figure 6-8. Experiment set up and wire target with 5 micro-scale stainless steel strings.
The strings were mounted on the stepwise structure of the frame tightly.

74
(Fig. 6-9). XZ and YZ planes correspond to the cross-section that crossed the y-axis and x-axis
respectively, and the XY plane was located at Z = 20 wavelengths. Two sets of images were
recorded. The target was rotated by 90° in the second image set, which can be seen by the
orientation of the wires in XY images.
6.4 Discussion and Conclusion
In this chapter, the design, fabrication, and imaging performance of the interposer based 15-
MHz 2D array was introduced. The array presented in this work was an example to show the
promising potential of implementing pitch-shifting interposer architectures and 3D printing
technology in ultrasound array fabrication. Pitch-shifting interposers were designed following
the same idea in our previous work: filling the channels with a conductive epoxy to create
conducting channels that act as a backing layer and it is successfully applied to arrays with
different center frequencies. The pitch-shifted architecture is a straightforward modification of
C
D E F
A B
Figure 6-9. Images of the wires target. A-C were acquired with the wires parallel to the xdirection, and D-F were acquired after the target was rotated by 90 degrees. A, D: XY plane; B,
E: XZ plane; C, F: YZ plane.

75
the original idea of interposers: by applying isometric scaling of the pitch between channels and
the diameter of the channels, this design methodology allows us to adapt the fixed pad matrix on
the circuit board to any number of different pitch array matrices. The transformation that is
inherent in pitch-shifting also introduces new and potentially far-reaching topological
possibilities provided by interposers. Specifically, the deformation achieved in such interposers
is not limited to pitch scaling alone and can be further used for reshaping the geometry of the
aperture. For example, non-rectangular 2D sparse arrays, such as spiral array and concentric
circular arrays with irregular geometry [108-111] could be implemented using this technique. In
general, any re-mapping of the transmission and receive element location that was previously
constrained by the ASIC pad array definition, can now be realized as long as the geometry of the
circuit electrodes provides a one-to-one mapping with the target element matrix, and the
structure is supported by the 3D printer used to create the interposer.
Fabrication of filled interposers as described here is a hands-on process that is limited to
prototyping and therefore not readily amenable to volume production. For the quick-turn
realization of novel element mapping geometries, this prototyping capability is invaluable to
rapidly test and validate design concepts. In the future, this process could be adapted to volume
production using injection-molded interposer die parts and industrial epoxy filling extrusion
machines.
In such a volume process it would be critical to identify important yield issues including
bubbles formed in the epoxy fill which reduce the effectiveness of the acoustic backing. In
addition, the yield of electrical connections is critically dependent on issues such as pinholes
formed in the walls of the channels which can potentially lead to shorting of neighboring
elements. Finally, coplanarity of the top and bottom surface of the interposers is a particularly

76
important yield criterion because it has significant effects on the assembly of both the top side
transducer array as well as the bottom side electronic substrate connections. All these issues must
be taken into consideration if a viable production process is to be implemented.
In addition, the active material (soft PZT) applied in this project was not suitable for 2D array
fabrication. The limitation of the dicing machine led to the “bulk material” issue, which was
described in the array performance section, and significantly influenced the performance of the
array. Considering the array size, the single crystal material produced by the etching process
would be promising for this application.
In summary, by using advanced 3D printing techniques, we have implemented novel 3D printed
pitch-shifted interposers that enable an adaptive 2D array fabrication process where the circuit pad
topology does not limit the geometry of the transducer matrix. With refinement of the
manufacturing process and improvements in 3D printing technology, interposers with continuous
deformation of the channel matrix will open new possibilities for array design and address circuit
design challenges for dense 2D array implementations with low cost and rapid prototyping
capabilities.

77
Chapter 7 Design and Fabrication of
17.5 MHz 256-elements Imaging 2D Array
7.1 2D Array and PCB Design
In the previous chapters, we have designed and built ultrasound 2D arrays with pitch-shifted
interposer structure and different central frequencies. This chapter will introduce the design and
fabrication process of an ultrasound 2D array with higher central frequency and a new interposer
design. In chapter 5, we built a 2D array with a pitch of 200 μm which is also the finest pitch we
can reach with commercial 3D printer. In this chapter, we tried to print interposers with 150 μm
pitch but the channels yield is always below 85%. Thus, I changed the design back to 200 μm
pitch but the active array element size is increased to 160 µm * 160 µm. During the fabrication
process, it was found that if the previous pitch-shifted interposer structure was utilized, the
channels would all have different length to travel. For example, the channels in the middle area
will all have straight routes but the channels at corners will have sloped routes. Since the sloped
routes are longer than straight channels and uncured E-solder has a high viscosity, it would be
impossible for the E-solder to pass through all the channels for future production of higher
frequency ultrasound 2D arrays with finer pitch size. Therefore, we developed the interposer
structure into all-straight channels structure. The interposer is supported by BMF with no cost
(Boston Micro Fabrication, Maynard, Massachusetts, USA). The central frequency of the array
was designed to 17.5 MHz; thus, the array’s element has a pitch of ~2.3 (200 µm).
As a result of the change in the structure of the interposer, the design of the PCB returned to the
original problem: The PCB to control the array element with too small a pitch is difficult and
expensive to fabricate. We solved this problem by making a 16-layer 1*16 stacked PCB and
exposing the sides of the circuit with a dicing saw and connecting the arrays with 16*16

78
electrode matrix on the exposed side of the stacked circuit. The schematic diagram of the circuit
is shown in Figure 7-1.
Figure 7-1. Schematic diagram of the 16*16 stacked PCB[112].
The materials selected for the 2D array in this chapter is 1-3 composite made from soft PZT
material (DL-50, DeL Piezo Specialties, LLC, West Palm Beach, FL, USA). The kerf is filled
with epoxy (EPO-TEK 301, Epoxy Technology, Billerica, MA, USA) and two same matching
layers (2-3 μm silver epoxy and ABS plastic) were applied. Table 7-1 shows the details of the
materials and parameters used in this 2D array design.
Table 7-1. Parameters of the 15MHz image array
Parameters Values
Piezo Material 1-3 composite of Soft PZT ceramic
kt 0.64
First Matching Layer (1st ML) Material 2-3 μm silver epoxy
1st ML Sound Speed 1961 m/s
1st ML Acoustic Impedance 7.84 MRayl
1st ML Thickness 25 μm
Second Matching Layer (2nd ML) Material ABS plastic
2nd ML Sound Speed 1850 m/s
2nd ML Acoustic Impedance 2.2 MRayl
2nd ML Thickness 23 μm
Backing Layer Material E-solder with 3D printed acrylic interposer
Design Center Frequency 17.5 MHz
Number of Elements 256
Pitch 200 µm
Element Size 160 µm * 160 µm
Kerf 40 µm
Array Size 3.2 * 3.2 mm

79
We implemented the online tool BioSono KLM 2.0 (Fremont, CA, USA) to evaluate the
acoustic performance of the array with the design parameters. Parameters of the composite we
used to simulate are: Longitudinal velocity 3591 m/s, Density 4360 kg/m3
and Dielectric
constant of 429.The pulse-echo simulation result is showed in Figure. 7-2. The bandwidth of the
array elements would reach 92% with the designed parameters presented in Table 7-1.
7.2 2D Array and PCB Fabrication and Preparation
In chapter 7.1, we changed the structure of the interposer from pitch-shifted to all straight
channels. The interposer is a consist of a matrix of channels embedded in a solid light-cured 3D
printed material. We filled the interposer with E-solder with the injection molding method
mentioned in the previous chapters and the interposer thereby plays both as acoustic backing and
interconnection. Slightly different from the previous process, the size of the interposer mold was
intentionally increased when the interposer was designed to consider the need for a larger
operating area and fault tolerance during the subsequent placement with the piezoelectric
ceramics. In addition, after the E-solder filling process, all arrays on one side were shortFigure 7-2. Pulse-Echo impulse response simulation for 17.5-MHz 2D array element.

80
circuited and tested for conductivity to confirm that the yield of the interposer elements were at
least 90%. The rest of the acoustic stack assembly steps, as shown in Figure 7-3, including the
attachment of the piezoelectric ceramics, the first and the second matching layers were the same
as the process described in chapter 6. In this step, the array lost 2 rows and 1 column of array
elements from two edges in the result due to the small size of the piezoelectric ceramics ordered,
which moved about 300 micrometers due to errors during the bonding process with the
interposer.
Figure 7-3. The 3D printed interposer fabrication and testing process (A) The 3D Perspective
View of the all straight channels interposer. (B) Conductivity test after filling process. (C)
Cleaned all the surfaces and alignment marks are diced. (D) The final size of the interposer.
A
B
C
D

81
At the end of the assembly, we removed the excess of the sample to get the actual desired
dimensions. At this point, we have obtained an acoustic stack of 256 array elements with a pitch
of 200 µm.
The next thing we want to tackle is the hardest part of this new assembly method, realizing
PCB with small array element spacing. My lab colleague Robert Wodnicki helped me design and
ordered the stacked PCB as shown in the Figure 7-4.
Figure 7-4. Schematic view of the stacked PCB board to drive the 2D array[112].
The stacked PCB can be printed with commertially available PCB fabrication process. It can
also be printed with flexible circuit or with rigid and flexible sections. The stacked PCB board
cmay further comprise multiple intervening copper layers between routing layers for electrical
shielding to reduce crosstalk.
In practice, we ordered two 8*16 circuit boards from PCBWay (Shenzhen, China). First, the
edges of the boards were removed with a dicing saw (Thermocarbon, Casselberry, FL) to expose
the electrodes needed to assemble the 2D arrays. Then, using the positioning holes on the boards,
we secured the two boards to a vertical support fixture and added shims in between to ensure that
the distance between the two boards was also exactly 200 µm. The next step is to spread epoxy
EPO-TEK 301 (Epoxy Technology, Billerica, MA, USA) evenly over the cut surfaces, siphoning
the resin to fill the gaps between the two boards and the shims and curing it to make the structure

82
into one piece. Finally another cut is made to remove the excess resin from the surface and to
make the middle electrode section protrude 200 to 300 microns higher than the surrounding
structure (Figure 7-5).
However, we found a new problem at this step. As shown in Figure 7-6, due to the lack of
precision in the fabrication of the board, the exposed electrodes are not sufficient for the
assembly. Here we need to reroute the electrode arrays to match the 200 µm pitch
interconnection points on the interposer. Our solution to this is: first, sputter the exposed surface
with Cr/Au electrodes with a gold sputtering machine so that the surfaces are all short-circuited
together; in the next step, draw a grid with a spacing of 200 µm according to the pattern we need,
cover the entire surface of the electrodes with the grid, and look for a pattern that puts each
Figure 7-5. The stacked PCB preparation work. (A) The dicing instruction to expose the
assembly pads. (B) The dicing work after exposing the assembly pads. (C) The vertical support
fixture with PCB sandwich holded. (D) The photo of the PCB sandwich structure. (E) The
photo of epoxy application to bond the sandwich together. (F) Dicing work as the final step to
expose the assembly pads after lamination.

83
electrode into a grid and dice the surface according to the pattern. Finally the electrode surfaces
can be redistributed to obtain the electrode matrix we need with a pitch of 200 µm.
In a further embodiment, a 3D-printed metal grid of pins may be applied to the interposer
connection points and patterned to reroute signals to match the interconnection points on the
routing substrate.
7.3 2D Array Assembly Process
At this point, we have completed the acoustic stack and the control circuit, respectively. The
next challenge is to perfectly combine both sides of the 256 array elements within an area of only
Figure 7-6. The reroute of the electrodes. (A) The photo of the actual electrode pads. (B)
schematic draw to reroute the electrodes matrix. (C) The photo of the rerouted elements. (D)
The photo of final sputtered and patterned electrodes matrix.

84
3.2 mm by 3.2 mm. In order to solve this problem, we built an assembly platform as shown in
Figure 7-7.
Figure 7-7. Schematic draw of the assembly platform for the high frequency 2D array [112].
This self-made assembly platform has 6 degrees of freedom, including freedom to move in the
direction of the x, y, and z right-angle coordinate axes and freedom to rotate around these three
axes. Since both sides of the completed acoustic stack module are structured with a high center
and low sides, we 3D printed a small platform as shown in the Figure 7-7 and fixed the
completed acoustic stack module to one side of the assembly platform with wax.
In practice, we utilize two digital microscopes to achieve accurate array assembly. Above the
platform we used a fixed-view digital microscope (Long Stand Digital Microscope, Hayve,
Shenzhen, China) to ensure foolproof assembly in the vertical direction, and a portable USB
digital microscope (Dino-Lite, Taiwan) connected to a computer on the side to ensure accurate
assembly in the horizontal direction.
After ensuring that the vertical and two horizontal assemblies are calibrated, we can assume
that all 256 array elements are calibrated and ready for the final assembly step. Since the
electrodes and acoustic stack are machined with a dicing saw to form the pillars for assembly,

85
and the cured E-solder is softer than the metal electrodes, the stacked PCB and acoustic stack
modules can be tightly affixed together with an assembly platform after the arrays are aligned,
the electrodes will be submerged into the interposer pillars by a few tens of microns. Finally, the
structure will be fixed by the addition of epoxy resin (EPO-TEK 301, Epoxy Technology,
Billerica, MA, USA) at the interface of the joints after the calibration is confirmed correctly
under the microscope. The epoxy resin will penetrate along the gaps at the interface between the
two parts, and once cured, it will provide sufficient robustness for the bonding of the stacked
PCB and the acoustic stack. Photographs of the assembly process are shown in Figures 7-8.
Figure 7-8. The assembly process of the array. (A) Schematic drawing of the interface between
the Stacked PCB and the interposer assembly. (B) Schematic drawing of the diced pillars ready
for the assembly. (C) The photo of the stacked PCB. (D) The photo of the acoustic stack. (E)
The photo of the ongoing assembly process with two digital microscopes. (F) Top view during
the assembly process. (G) (H)Two side views during the assembly process. (I) Epoxy filling as
the final assembly step.

86
The final step is to ground the complete array. We did this by removing the second matching
layer of the complete 2D array with a dicing saw to remove a corner exposing the first matching
layer, after which it was connected to the circuit board's ground with E-solder. Since the first
matching layer is conductive, the upper surfaces of the piezoelectric ceramics will be shorted
together to form the ground connection.
7.4 Performance Testing and Imaging Results
Performance testing of the ultrasound high frequency 2D imaging array completed in this
chapter is essentially the same as the work done in Chapter 5. We first obtained the P/E results of
the imaging array using the Verasonics system (Vantage 256, Verasonics, Inc., Kirkland, WA,
USA) and plotted the sensitivity map in conjunction with the array element map of the 2D array,
resulting in a measured overall yield of 75% as shown in Figure 7-9. Noting that 2 rows and 1
column of array elements were lost during fabrication due to the calibration error of the
piezoelectric ceramics mentioned in section 7.2, the yield after removing these array elements
was about 93%.
Besides, the existing assembly step is better calibrated than the previous chapters due to the
additional vertical digital microscope, but the flatness requirement is still high. In addition, the
Figure 7-9. Array performance maps: (A) Sensitivity map. (B) Bandwidth map.
(C) The photo of array elements with missing rows and columns.

87
yield of the interposer has not reached 100%, so there is still room for the overall 2D array yield
to increase.
Figure 7-10 illustrates the pulse echo of a typical 2D array element. The average bandwidth of
the entire 2D array element is about 40%, again much lower than the simulation results. The
reason for this is that the problems mentioned in the previous chapter are still not solved, since
the channel size of the interposer is still very narrow, so we cannot use the centrifuged E-solder
to increase its acoustic attenuation coefficient. In addition, the length of the interposer was
shortened to 3 mm due to filling difficulties. The dicing process of the composite material was
also difficult to ensure that the kerfs was all over the epoxy, so the final vibration pattern may be
different than expected.
Next, we used the array to image the same wire targets as in Chapter 5. The four wires were
arranged in an equally spaced trapezoidal arrangement with a horizontal spacing of 0.5 mm and a
vertical spacing of 1.5 mm. Imaging was still performed using the coherent plane wave
composite beamforming technique and averaged over 6 angular plane waves generated in the
XZ, YZ plane at each imaging cycle. After recording the first set of images, the sample was
rotated by 90 degrees and a second set of images was recorded. The build photographs and
Figure 7-10. Typical pulse-echo response and spectrum of the 17.5-MHz image array.

88
imaging results of the imaging experiments are shown in Figures 7-11. The XZ and YZ planes
correspond to the Y- and X-axis cross-sections, respectively, and the XY plane is located at the
nearest wire to the array, approximately 3 mm away.
Three lines and three points in the corresponding directions can be clearly seen in the graph of
the imaging results, proving that the fabrication of this 2D array is feasible.
7.5 Discussion and Conclusion
In this chapter, we present the design, fabrication, and imaging performance of a highfrequency ultrasound 2D array using a novel approach. The advantages include avoiding the
slanted-channel structure of the interposer used in previous chapters, enabling higher-precision
3D printing, and solving the design cost of the control circuit with laminated circuit boards,
demonstrating the potential to fabricate higher-frequency ultrasound 2D arrays using the same
Figure 7-11. Images of the wires target. Top 3 were acquired with the wires parallel to the x-direction,
and bottom 3 were acquired after the target was rotated by 90 degrees.

89
3D printing technique. The downside, however, is that the design of the circuits returns to the
previous paradigm where different ultrasonic arrays require specially customized circuit boards.
Nevertheless, the cost is kept under control compared to ASIC pad arrays and flexible circuit
board solutions.
There are still some problems and room for improvement in the methodology expressed in this
chapter. For example, how to solve the accuracy problem of stacked circuits, a matrix can be
considered to be fabricated by 3D printing to directly realize the patterning of electrodes; in
addition, the fabrication of acoustic modules for high-frequency ultrasound arrays still has
unresolved difficulties, such as the balance of acoustic attenuation performance and thickness of
backing layer, and the dicing and calibration process of composites still need to be mapped out.
In conclusion, the fabrication process for ultrasound 2D arrays provided in this chapter is
proved to be feasible. With the development of 3D printing technology, we may be able to
directly fabricate the backing layer with both conductive and acoustic attenuation properties
through multi-material 3D printing in the future; moreover, we can incorporate photolithography
in Micro-Electro-Mechanical systems technology to achieve micron-level accuracy in the
fabrication process of ultrasound 2D arrays at higher frequencies.

90
Chapter 8 Summary and Future Work
8.1 Summary and Discussion
In the previous chapters, several ultrasound transducers including single element transducers
and 2D array designed with interposers were fabricated, tested, and implemented in various
applications.
In the field of ultrasound single element transducers, we demonstrate the use of textured PMNPZT ceramics with high piezoelectric constant and large electromechanical coupling coefficients
for the construction of high frequency and high sensitivity transducers for biomedical ultrasound
imaging. The fabricated needle transducer has a high center frequency of 15 MHz, an effective
electromechanical coupling coefficient k_eff of 0.69, a fractional bandwidth of 67% at -6 dB, and
an insertion loss of 21 dB. These preliminary results are promising and point to the possible future
clinical use of texturized PMN-PZT ceramics for next-generation high-frequency ultrasound
devices requiring high sensitivity.
We also designed and fabricated 2D arrays for ultrasound stimulation as a practice of the
interposer. After the successful fabrication of a 64-element stimulation ultrasound 2D array, we
designed and fabricated a 16×16 4-MHz 2D array and utilized it in an experiment to restore vision
in blind rats. It was successfully demonstrated that the 2D array could dynamically generate
arbitrary patterns and stimulate the retina, as well as evoke corresponding visual pattern signals in
the visual circuits of the brain. In addition, the 2D array can provide sufficient guidance images to
localize the retina and guide the pattern focus prior to stimulation. The safety of ultrasound retinal
prostheses was further investigated using a combined imaging and histological approach.
The higher tolerance and more geometry options with the help of interposer made the array
fabrication a smooth, efficient, and easy iterative process, providing a solid foundation for

91
successful experiments. This was also an important test of the feasibility of pitch-shifted interposer
in 2D array fabrication, which became the cornerstone of the subsequent project.
Constrained by stringent requirements for component scale, high operating frequency 2D arrays
rely on high-resolution printed circuit technology or ASICs, which can be both costly and timeconsuming. In addition, circuits with defined structures, whether PCBs or ASICs, can only be
implemented for a specific array design. A new design implies a new circuit design, which is both
expensive and time-consuming. Introducing pitch-shifted interposer into array design and
fabrication can effectively address both challenges. Thanks to the rapid development of 3D
printing technology, interposers with sophisticated and complex structures can be produced
quickly and at an affordable cost. The cost of redesign and trial-and-error is significantly reduced
compared to circuit design. As a prototype, a 15 MHz fully sampled 2D array was fabricated and
tested. The array was tested and used for volumetric imaging of line targets. The variant structure
made it possible to build this array using a PCB with standard features. In the future, the same
circuit can be used for multiple circuit designs by adjusting the ratio of size variations.
8.2 Future Work
In recent years, 3D printers that support objects with minimum structure from 10μm to
submicron were gradually available and affordable for research groups. Besides the 10MHz
array’s interposer presented in this paper, a prototype pitch-shifted interposer for a 20MHz
1.75D array was introduced in our previous work [113] (Fig. 8-1). This interposer was printed
by Printer P130 (Boston Micro Fabrication, Maynard, Massachusetts, USA). On the transducer
side, it provided an 8 by 16 matrix and 50 μm by 130 μm element size with a 20 μm wall. The
filling and further processing for this tiny interposer were challenging, yet not impossible. In the
future, with the help of an advanced 3D printer, we could push on these technologies to fabricate

92
2D arrays with even higher operating frequency. Additionally, the development of novel 3D
printers that support multi-material printing and printing materials also brings us possibilities in
optimizing the interposer’s fabrication process, e.g., building the interposer in one shot with the
multi-material printers. By avoiding the filling process and applying new material as a backing
layer to improve the yield and performance of the array.
Another advantage that 3D printing technology has is the diversity of materials. In the future,
with more choices of printing materials, we can apply different materials to the manufacturing of
interposers. In the current design, the interposer is filled with channels by E-solder. However, Esolder is costly and hard to fill. In fact, for the interposer, it is not necessary that every channel is
fully conductive, but only the part in contact with the piezoelectric material is acoustically
attenuated and partially conductive to meet the requirements. Therefore, in the future, 3D
printers with high resolution that can print conductive materials can produce integrated molding,
without the need to fill the interposers of the process will be very helpful for the design of
ultrasound 2D arrays. In addition, there are still many problems that need to be solved in today's
2D array fabrication process. For example, the matching layer falls off, the interposer is difficult
Figure 8-1. Interposer for a 20MHz 1.75D array. a) Before filling; b) After filling.

93
to fill, and the flatness is difficult to achieve. We hope to find more suitable processes and tools
to solve these problems in the future.
In addition to this, it is also possible to combine Micro-Electro-Mechanical systems (MEMS)
techniques to realize 2D array processes with smaller dimensions. The bottleneck for 2D arrays
at this stage is the backing layer at tens of micrometer size. Backing needs to achieve the
acoustic attenuation and conductivity in the smaller size is almost impossible through the current
3D printing combined with E-solder method to promote, so the precision in the single-digit
micron-level MEMS technology if used may be able to greatly improve the center frequency of
ultrasound 2D arrays. For example, one could make an entire backing layer that only provides
acoustic attenuation, the material of which could be any mixture of metal powder and epoxy
resin, and machine it into a flat backing module. The backing module could similarly be made to
function as both acoustic attenuation and electrical conductivity by inscribing micrometer-scale
circuits on the backing template via a MEMS mask plate and photolithography. The difficulty
with this approach is that the space between the array elements is very tight, and it is difficult to
route between a large number of array elements to achieve independent connections for hundreds
of array elements.
For the application area of ultrasound 2D arrays, we demonstrate the concept of 2D
stimulation arrays as visual prostheses and its favorable prospects for clinical applications.
Although we have demonstrated the feasibility of ultrasound retinal prostheses, the design and
development of wearable devices remains difficult. Our ultimate goal is to integrate an
ultrasound retinal prosthesis into a portable wearable device that generates electrical signals to
stimulate the user's retina in real time by capturing the view in front of the user for signal
conversion. There is still a long way to go to achieve this goal, and I think there are several

94
difficulties that need to be solved:
The first is the medium in which the ultrasound is delivered. Since ultrasound cannot use air as
a medium, ultrasound gel must be provided as a delivery medium between the ultrasound
transducer and the eye. This can be achieved by customizing the mold, but the specific
implementation also needs to take into account the loss and contamination of water or ultrasound
gel.
The next issue is the efficiency of real-time signal processing. At this stage, what can be
realized is just edited signal pattern imaging experiments, to do the analysis and processing of
real-time image signals, the signal processing capability of wearable devices is very demanding.
Finally, the portability of the device has to be taken into account. In addition to the signal
processing module mentioned above, the ultrasound system itself requires a power supply and a
complete set of acoustic modules, the size and weight may be up to several kilograms.
In summary, there are many issues that need to be addressed to achieve the ultimate goal of a
portable ultrasound retinal prosthesis. Much engineering work must be devoted to developing
and optimizing the final prosthetic device.

95
BIBLIOGRAPHY
[1] F. S. Foster, G. Lockwood, L. Ryan, K. Harasiewicz, L. Berube, and A. Rauth, "Principles and
applications of ultrasound backscatter microscopy," IEEE transactions on ultrasonics, ferroelectrics, and
frequency control, vol. 40, no. 5, pp. 608-617, 1993.
[2] A. Ng and J. Swanevelder, "Resolution in ultrasound imaging," Continuing Education in
Anaesthesia Critical Care & Pain, vol. 11, no. 5, pp. 186-192, 2011.
[3] E. Alasaarela and J. Koivukangas, "Evaluation of image quality of ultrasound scanners in medical
diagnostics," Journal of ultrasound in medicine, vol. 9, no. 1, pp. 23-34, 1990.
[4] S.-E. Park and T. R. Shrout, "Characteristics of relaxor-based piezoelectric single crystals for
ultrasonic transducers," IEEE Transactions on Ultrasonics, Ferroelectrics, and Frequency Control, vol.
44, no. 5, pp. 1140-1147, 1997.
[5] K. Rajan, M. Shanthi, W. Chang, J. Jin, and L. Lim, "Dielectric and piezoelectric properties of [0
0 1] and [0 1 1]-poled relaxor ferroelectric PZN–PT and PMN–PT single crystals," Sensors and Actuators
A: Physical, vol. 133, no. 1, pp. 110-116, 2007.
[6] C. S. Desilets, J. D. Fraser, and G. S. Kino, "The design of efficient broad-band piezoelectric
transducers," IEEE Transactions on sonics and ultrasonics, vol. 25, no. 3, pp. 115-125, 1978.
[7] Q. Zhou, K. H. Lam, H. Zheng, W. Qiu, and K. K. Shung, "Piezoelectric single crystal ultrasonic
transducers for biomedical applications," Progress in materials science, vol. 66, pp. 87-111, 2014.
[8] S.-K. Hong and Y. G. Ohr, "Ultrasonic speckle pattern correlation interferometry using a pulseecho method," Journal of Physics D: Applied Physics, vol. 31, no. 11, p. 1392, 1998.
[9] T. Morita, "Piezoelectric materials synthesized by the hydrothermal method and their
applications," Materials, vol. 3, no. 12, pp. 5236-5245, 2010.
[10] A. S. Karapuzha, N. K. James, H. Khanbareh, S. Van der Zwaag, and W. Groen, "Structure,
dielectric and piezoelectric properties of donor doped PZT ceramics across the phase diagram,"
Ferroelectrics, vol. 504, no. 1, pp. 160-171, 2016.
[11] T. Gururaja, W. A. Schulze, L. E. Cross, R. E. Newnham, B. A. Auld, and Y. J. Wang,
"Piezoelectric composite materials for ultrasonic transducer applications. Part I: Resonant modes of
vibration of PZT rod-polymer composites," IEEE Trans. Sonics Ultrason, vol. 32, no. 4, pp. 481-498,
1985.
[12] X. Geng and Q. Zhang, "Resonance modes and losses in 1-3 piezocomposites for ultrasonic
transducer applications," Journal of applied physics, vol. 85, no. 3, pp. 1342-1350, 1999.
[13] L. Jiang et al., "Flexible piezoelectric ultrasonic energy harvester array for bio-implantable
wireless generator," Nano energy, vol. 56, pp. 216-224, 2019.
[14] Y. Zeng et al., "3D-printing piezoelectric composite with honeycomb structure for ultrasonic
devices," Micromachines, vol. 11, no. 8, p. 713, 2020.

96
[15] X. Jiang, K. Snook, A. Cheng, W. Hackenberger, and X. Geng, "Micromachined PMN-PT single
crystal composite transducers--15–75 MHz PC-MUT," in 2008 IEEE ultrasonics symposium, 2008: IEEE,
pp. 164-167.
[16] L. Jiang et al., "Fabrication of a (K, Na) NbO3-based lead-free 1-3 piezocomposite for highsensitivity ultrasonic transducers application," Journal of Applied Physics, vol. 125, no. 21, p. 214501,
2019.
[17] L. Jiang et al., "Ultrasound‐induced wireless energy harvesting for potential retinal electrical
stimulation application," Advanced Functional Materials, vol. 29, no. 33, p. 1902522, 2019.
[18] FAMUS. "Ultrasound Theory Module." http://famus.org.uk/modules/ultrasound-theory-module
(accessed.
[19] D. G. Wildes and L. S. Smith, "Advanced ultrasound probes for medical imaging," in AIP
Conference Proceedings, 2012, vol. 1430, no. 1: American Institute of Physics, pp. 801-808.
[20] J. A. Jensen, "Medical ultrasound imaging," Progress in biophysics and molecular biology, vol.
93, no. 1-3, pp. 153-165, 2007.
[21] K. K. Shung, Diagnostic ultrasound: Imaging and blood flow measurements. CRC press, 2015.
[22] C. M. Otto, The practice of clinical echocardiography. Elsevier Health Sciences, 2007.
[23] M. H. Mozaffari and W.-S. Lee, "Freehand 3-D ultrasound imaging: a systematic review,"
Ultrasound in medicine & biology, vol. 43, no. 10, pp. 2099-2124, 2017.
[24] G. M. Treece, A. H. Gee, R. W. Prager, C. J. Cash, and L. H. Berman, "High-definition freehand
3-D ultrasound," Ultrasound in medicine & biology, vol. 29, no. 4, pp. 529-546, 2003.
[25] G. M. Treece, J. E. Lindop, A. H. Gee, and R. W. Prager, "Freehand ultrasound elastography with
a 3-D probe," Ultrasound in medicine & biology, vol. 34, no. 3, pp. 463-474, 2008.
[26] P. Yan, J. C. Cheeseborough III, and K. C. Chao, "Automatic shape-based level set segmentation
for needle tracking in 3-D TRUS-guided prostate brachytherapy," Ultrasound in medicine & biology, vol.
38, no. 9, pp. 1626-1636, 2012.
[27] R. Ohbuchi and H. Fuchs, "Incremental volume rendereing algorithm for interactive 3D
ultrasound imaging," in Biennial International Conference on Information Processing in Medical
Imaging, 1991: Springer, pp. 486-500.
[28] R. Ohbuchi, D. Chen, and H. Fuchs, "Incremental volume reconstruction and rendering for 3-D
ultrasound imaging," in Visualization in Biomedical Computing'92, 1992, vol. 1808: International Society
for Optics and Photonics, pp. 312-323.
[29] C.-S. Huang et al., "Whole-breast ultrasound for breast screening and archiving," Ultrasound in
medicine & biology, vol. 43, no. 5, pp. 926-933, 2017.
[30] O. H. Gilja, T. Hausken, S. Olafsson, K. Matre, and S. Ødegaard, "In vitro evaluation of threedimensional ultrasonography based on magnetic scanhead tracking," Ultrasound in medicine & biology,
vol. 24, no. 8, pp. 1161-1167, 1998.

97
[31] H. G. Lim, H. H. Kim, C. Yoon, and K. K. Shung, "A one-sided acoustic trap for cell
immobilization using 30-MHz array transducer," IEEE Transactions on Ultrasonics, Ferroelectrics, and
Frequency Control, vol. 67, no. 1, pp. 167-172, 2019.
[32] C. Wong et al., "20-MHz phased array ultrasound transducer for in vivo ultrasound imaging of
small animals," Ultrasonics, vol. 126, p. 106821, 2022.
[33] J. Hung et al., "3D echocardiography: a review of the current status and future directions,"
Journal of the American Society of Echocardiography, vol. 20, no. 3, pp. 213-233, 2007.
[34] J. Yen and R. Wodnicki, "Orthogonal Bowtie-Shaped 2D array for Real-time 3D Imaging," in
2020 IEEE International Ultrasonics Symposium (IUS), 2020: IEEE, pp. 1-4.
[35] I. E. Timor-Tritsch and L. D. Platt, "Three-dimensional ultrasound experience in obstetrics,"
Current Opinion in Obstetrics and Gynecology, vol. 14, no. 6, pp. 569-575, 2002.
[36] J. Zhu et al., "3D super-resolution US imaging of rabbit lymph node vasculature in vivo by using
microbubbles," Radiology, vol. 291, no. 3, pp. 642-650, 2019.
[37] C. Errico et al., "Ultrafast ultrasound localization microscopy for deep super-resolution vascular
imaging," Nature, vol. 527, no. 7579, pp. 499-502, 2015.
[38] J. A. Jensen et al., "Three-dimensional super-resolution imaging using a row–column array,"
IEEE transactions on ultrasonics, ferroelectrics, and frequency control, vol. 67, no. 3, pp. 538-546, 2019.
[39] S. Harput et al., "3-D super-resolution ultrasound (SR-US) imaging with a 2-D sparse array,"
arXiv preprint arXiv:1902.01608, 2019.
[40] R. Wodnicki et al., "Tiled large element 1.75 D aperture with dual array modules by adjacent
integration of PIN-PMN-PT transducers and custom high voltage switching ASICs," in 2019 IEEE
International Ultrasonics Symposium (IUS), 2019: IEEE, pp. 1955-1958.
[41] A. T. Fernandez, K. L. Gammelmark, J. J. Dahl, C. G. Keen, R. C. Gauss, and G. E. Trahey,
"Synthetic elevation beamforming and image acquisition capabilities using an 8/spl times/128 1.75 D
array," IEEE transactions on ultrasonics, ferroelectrics, and frequency control, vol. 50, no. 1, pp. 40-57,
2003.
[42] J. O. Fiering, P. Hultman, W. Lee, E. D. Light, and S. W. Smith, "High-density flexible
interconnect for two-dimensional ultrasound arrays," IEEE transactions on ultrasonics, ferroelectrics,
and frequency control, vol. 47, no. 3, pp. 764-770, 2000.
[43] J. Hansen-Shearer, M. Lerendegui, M. Toulemonde, and M.-X. Tang, "High contrast Ultrafast 3D
Ultrasound Imaging using Row Column specific Frame Multiply and Sum," arXiv preprint
arXiv:2102.07746, 2021.
[44] K. Latham, C. Samson, J. Woodacre, and J. Brown, "A 30-MHz, 3-D Imaging, Forward-Looking
Miniature Endoscope Based on a 128-Element Relaxor Array," IEEE Transactions on Ultrasonics,
Ferroelectrics, and Frequency Control, vol. 68, no. 4, pp. 1261-1271, 2020.

98
[45] K. Latham, C. Ceroici, C. A. Samson, R. J. Zemp, and J. A. Brown, "Simultaneous azimuth and
Fresnel elevation compounding: A fast 3-D imaging technique for crossed-electrode arrays," IEEE
transactions on ultrasonics, ferroelectrics, and frequency control, vol. 65, no. 9, pp. 1657-1668, 2018.
[46] H. Kang et al., "2-D Array Design and Fabrication With Pitch-Shifting Interposer at Frequencies
From 4 MHz up to 10 MHz," IEEE Transactions on Ultrasonics, Ferroelectrics, and Frequency Control,
vol. 69, no. 12, pp. 3382-3391, 2022.
[47] C. Chen et al., "A Front-End ASIC With Receive Sub-array Beamforming Integrated With a
$32\times 32$ PZT Matrix Transducer for 3-D Transesophageal Echocardiography," IEEE Journal of
Solid-State Circuits, vol. 52, no. 4, pp. 994-1006, 2017.
[48] D. Wildes et al., "4-D ICE: A 2-D array transducer with integrated ASIC in a 10-Fr catheter for
real-time 3-D intracardiac echocardiography," IEEE transactions on ultrasonics, ferroelectrics, and
frequency control, vol. 63, no. 12, pp. 2159-2173, 2016.
[49] P. Guo et al., "A 1.2 mW/channel 100µm-Pitch-Matched Transceiver ASIC with BoxcarIntegration-Based RX Micro-Beamformer for High-Resolution 3D Ultrasound Imaging," in 2022 IEEE
International Solid-State Circuits Conference (ISSCC), 2022, vol. 65: IEEE, pp. 496-498.
[50] Z. Yu, M. A. Pertijs, and G. C. Meijer, "A programmable analog delay line for Microbeamforming in a transesophageal ultrasound probe," in 2010 10th IEEE international conference on
solid-state and integrated circuit technology, 2010: IEEE, pp. 299-301.
[51] H. G. Kang et al., "Column-based micro-beamformer for improved 2D beamforming using a
matrix array transducer," in 2015 IEEE Biomedical Circuits and Systems Conference (BioCAS), 2015:
IEEE, pp. 1-4.
[52] S. Blaak et al., "Design of a micro-beamformer for a 2D piezoelectric ultrasound transducer," in
2009 IEEE International Ultrasonics Symposium, 2009: IEEE, pp. 1338-1341.
[53] H. Kang et al., "2-D ultrasonic array-based optical coherence elastography," IEEE Transactions
on Ultrasonics, Ferroelectrics, and Frequency Control, vol. 68, no. 4, pp. 1096-1104, 2020.
[54] R. Wodnicki et al., "Co-integrated PIN-PMN-PT 2-D array and transceiver electronics by direct
assembly using a 3-D printed interposer grid frame," IEEE transactions on ultrasonics, ferroelectrics, and
frequency control, vol. 67, no. 2, pp. 387-401, 2019.
[55] T. Mohanraj, S. Shankar, R. Rajasekar, N. Sakthivel, and A. Pramanik, "Tool condition
monitoring techniques in milling process—a review," Journal of Materials Research and Technology,
vol. 9, no. 1, pp. 1032-1042, 2020.
[56] S. A. M. Tofail, E. P. Koumoulos, A. Bandyopadhyay, S. Bose, L. O’Donoghue, and C.
Charitidis, "Additive manufacturing: scientific and technological challenges, market uptake and
opportunities," Materials Today, vol. 21, no. 1, pp. 22-37, 2018, doi: 10.1016/j.mattod.2017.07.001.
[57] N. Shahrubudin, T. C. Lee, and R. Ramlan, "An overview on 3D printing technology:
Technological, materials, and applications," Procedia Manufacturing, vol. 35, pp. 1286-1296, 2019.
[58] B. Mueller, "Additive manufacturing technologies–Rapid prototyping to direct digital
manufacturing," Assembly Automation, 2012.

99
[59] Z.-X. Low, Y. T. Chua, B. M. Ray, D. Mattia, I. S. Metcalfe, and D. A. Patterson, "Perspective on
3D printing of separation membranes and comparison to related unconventional fabrication techniques,"
Journal of membrane science, vol. 523, pp. 596-613, 2017.
[60] B. C. Gross, J. L. Erkal, S. Y. Lockwood, C. Chen, and D. M. Spence, "Evaluation of 3D printing
and its potential impact on biotechnology and the chemical sciences," ed: ACS Publications, 2014.
[61] J. F. Destino et al., "3D printed optical quality silica and silica–titania glasses from sol–gel
feedstocks," Advanced Materials Technologies, vol. 3, no. 6, p. 1700323, 2018.
[62] M. Li, Y. Yang, F. Iacopi, J. Nulman, and S. Chappel-Ram, "3D-printed low-profile singlesubstrate multi-metal layer antennas and array with bandwidth enhancement," IEEE Access, vol. 8, pp.
217370-217379, 2020.
[63] J. E. Huddy, M. S. Rahman, A. B. Hamlin, Y. Ye, and W. J. Scheideler, "Transforming 3Dprinted mesostructures into multimodal sensors with nanoscale conductive metal oxides," Cell Reports
Physical Science, vol. 3, no. 3, p. 100786, 2022.
[64] D. Hong et al., "Binder-jetting 3D printing and alloy development of new biodegradable Fe-MnCa/Mg alloys," Acta biomaterialia, vol. 45, pp. 375-386, 2016.
[65] P. Kunchala and K. Kappagantula, "3D printing high density ceramics using binder jetting with
nanoparticle densifiers," Materials & Design, vol. 155, pp. 443-450, 2018.
[66] S. A. Skoog, P. L. Goering, and R. J. Narayan, "Stereolithography in tissue engineering," Journal
of Materials Science: Materials in Medicine, vol. 25, no. 3, pp. 845-856, 2014.
[67] J. Wang, A. Goyanes, S. Gaisford, and A. W. Basit, "Stereolithographic (SLA) 3D printing of
oral modified-release dosage forms," International journal of pharmaceutics, vol. 503, no. 1-2, pp. 207-
212, 2016.
[68] F. Yang, M. Zhang, and B. Bhandari, "Recent development in 3D food printing," Critical reviews
in food science and nutrition, vol. 57, no. 14, pp. 3145-3153, 2017.
[69] W. J. Choi et al., "Rapid development of dual porous poly (lactic acid) foam using fused
deposition modeling (FDM) 3D printing for medical scaffold application," Materials Science and
Engineering: C, vol. 110, p. 110693, 2020.
[70] M. P. Chae, W. M. Rozen, P. G. McMenamin, M. W. Findlay, R. T. Spychal, and D. J. HunterSmith, "Emerging applications of bedside 3D printing in plastic surgery," Frontiers in surgery, vol. 2, p.
25, 2015.
[71] A. J. D. Shaikeea, H. Cui, M. O’Masta, X. R. Zheng, and V. S. Deshpande, "The toughness of
mechanical metamaterials," Nature materials, vol. 21, no. 3, pp. 297-304, 2022.
[72] Z. Xu et al., "Vat photopolymerization of fly-like, complex micro-architectures with dissolvable
supports," Additive Manufacturing, vol. 47, p. 102321, 2021.
[73] F. P. Melchels, J. Feijen, and D. W. Grijpma, "A review on stereolithography and its applications
in biomedical engineering," Biomaterials, vol. 31, no. 24, pp. 6121-6130, 2010.

100
[74] V. Karageorgiou and D. Kaplan, "Porosity of 3D biomaterial scaffolds and osteogenesis,"
Biomaterials, vol. 26, no. 27, pp. 5474-5491, 2005.
[75] S. J. Hollister, R. Maddox, and J. M. Taboas, "Optimal design and fabrication of scaffolds to
mimic tissue properties and satisfy biological constraints," Biomaterials, vol. 23, no. 20, pp. 4095-4103,
2002.
[76] Z. Weng, Y. Zhou, W. Lin, T. Senthil, and L. Wu, "Structure-property relationship of nano
enhanced stereolithography resin for desktop SLA 3D printer," Composites Part A: Applied Science and
Manufacturing, vol. 88, pp. 234-242, 2016.
[77] Y. Yan et al., "Giant energy density in [001]-textured Pb (Mg1/3Nb2/3) O3-PbZrO3-PbTiO3
piezoelectric ceramics," Applied physics letters, vol. 102, no. 4, p. 042903, 2013.
[78] R. Wodnicki et al., "PIN-PMN-PT single crystal composite and 3D printed interposer backing
for ASIC integration of large aperture 2D array," in 2017 IEEE International Ultrasonics Symposium
(IUS), 2017: IEEE, pp. 1-4.
[79] M. D. Menz et al., "Radiation force as a physical mechanism for ultrasonic neurostimulation of
the ex vivo retina," Journal of Neuroscience, vol. 39, no. 32, pp. 6251-6264, 2019.
[80] W. Lee et al., "Transcranial focused ultrasound stimulation of human primary visual cortex,"
Scientific reports, vol. 6, no. 1, pp. 1-12, 2016.
[81] K. K. Shung, J. Cannata, and Q. Zhou, "Piezoelectric materials for high frequency medical
imaging applications: A review," Journal of Electroceramics, vol. 19, no. 1, pp. 141-147, 2007.
[82] D. Zhou et al., "Lead-free piezoelectric single crystal based 1–3 composites for ultrasonic
transducer applications," Sensors and Actuators A: Physical, vol. 182, pp. 95-100, 2012.
[83] M. Matsubara, T. Yamaguchi, K. Kikuta, and S.-i. Hirano, "Sintering and piezoelectric properties
of potassium sodium niobate ceramics with newly developed sintering aid," Japanese journal of applied
physics, vol. 44, no. 1R, p. 258, 2005.
[84] D. Lv, R. Zuo, and S. Su, "Reactive templated grain growth and anisotropic electrical properties
of (Na 0.5 K 0.5) NbO 3 ceramics without sintering aids," Journal of Materials Science: Materials in
Electronics, vol. 23, no. 7, pp. 1367-1372, 2012.
[85] D. Lin, K. W. Kwok, K. H. Lam, and H. Chan, "Phase structure and electrical properties of K0.
5Na0. 5 (Nb0. 94Sb0. 06) O3-LiTaO3 lead-free piezoelectric ceramics," Journal of Physics D: Applied
Physics, vol. 41, no. 5, p. 052002, 2008.
[86] Y. Zhang et al., "Fabrication of angle beam two-element ultrasonic transducers with PMN–PT
single crystal and PMN–PT/epoxy 1–3 composite for NDE applications," Sensors and Actuators A:
Physical, vol. 168, no. 1, pp. 223-228, 2011.
[87] Q. Yue et al., "Design and fabrication of relaxor-ferroelectric single crystal PIN–PMN–PT/epoxy
2–2 composite based array transducer," Sensors and Actuators A: Physical, vol. 234, pp. 34-42, 2015.

101
[88] R. Chen et al., "Eco-friendly highly sensitive transducers based on a new KNN–NTK–FM leadfree piezoelectric ceramic for high-frequency biomedical ultrasonic imaging applications," IEEE
Transactions on Biomedical Engineering, vol. 66, no. 6, pp. 1580-1587, 2018.
[89] J. M. Cannata, T. A. Ritter, W.-H. Chen, R. H. Silverman, and K. K. Shung, "Design of efficient,
broadband single-element (20-80 MHz) ultrasonic transducers for medical imaging applications," IEEE
transactions on ultrasonics, ferroelectrics, and frequency control, vol. 50, no. 11, pp. 1548-1557, 2003.
[90] Z. Zhang et al., "Design and comparison of PMN-PT single crystals and PZT ceramics based
medical phased array ultrasonic transducer," Sensors and Actuators A: Physical, vol. 283, pp. 273-281,
2018.
[91] M. Bah et al., "Ultrasonic transducers based on undoped lead-free (K0. 5Na0. 5) NbO3
ceramics," Ultrasonics, vol. 63, pp. 23-30, 2015.
[92] X. Yan et al., "Correspondence: Lead-free intravascular ultrasound transducer using BZT-50BCT
ceramics," IEEE transactions on ultrasonics, ferroelectrics, and frequency control, vol. 60, no. 6, pp.
1272-1276, 2013.
[93] Y. Yan, J. E. Zhou, D. Maurya, Y. U. Wang, and S. Priya, "Giant piezoelectric voltage coefficient
in grain-oriented modified PbTiO 3 material," Nature communications, vol. 7, no. 1, pp. 1-10, 2016.
[94] R. H. Silverman, "Focused ultrasound in ophthalmology," Clinical Ophthalmology (Auckland,
NZ), vol. 10, p. 1865, 2016.
[95] W. L. Wong et al., "Global prevalence of age-related macular degeneration and disease burden
projection for 2020 and 2040: a systematic review and meta-analysis," The Lancet Global Health, vol. 2,
no. 2, pp. e106-e116, 2014.
[96] S. Ferrari, E. Di Iorio, V. Barbaro, D. Ponzin, F. S Sorrentino, and F. Parmeggiani, "Retinitis
pigmentosa: genes and disease mechanisms," Current genomics, vol. 12, no. 4, pp. 238-249, 2011.
[97] G. Lu et al., "Ultrasound retinal stimulation: a mini-review of recent developments," IEEE
Transactions on Ultrasonics, Ferroelectrics, and Frequency Control, 2022.
[98] M. D. Menz, Ö. Oralkan, P. T. Khuri-Yakub, and S. A. Baccus, "Precise neural stimulation in the
retina using focused ultrasound," Journal of Neuroscience, vol. 33, no. 10, pp. 4550-4560, 2013.
[99] O. Naor, Y. Hertzberg, E. Zemel, E. Kimmel, and S. Shoham, "Towards multifocal ultrasonic
neural stimulation II: design considerations for an acoustic retinal prosthesis," Journal of neural
engineering, vol. 9, no. 2, p. 026006, 2012.
[100] Q. Jiang et al., "Temporal neuromodulation of retinal ganglion cells by low-frequency focused
ultrasound stimulation," IEEE Transactions on Neural Systems and Rehabilitation Engineering, vol. 26,
no. 5, pp. 969-976, 2018.
[101] A. Fomenko, C. Neudorfer, R. F. Dallapiazza, S. K. Kalia, and A. M. Lozano, "Low-intensity
ultrasound neuromodulation: An overview of mechanisms and emerging human applications," Brain
stimulation, vol. 11, no. 6, pp. 1209-1217, 2018.

102
[102] H. A. S. Kamimura, A. Conti, N. Toschi, and E. Konofagou, "Ultrasound neuromodulation:
mechanisms and the potential of multi-modal stimulation for neuronal function assessment," Frontiers in
physics, vol. 8, p. 150, 2020.
[103] M. E. Downs et al., "Toward a cognitive neural prosthesis using focused ultrasound," Frontiers in
neuroscience, vol. 11, p. 607, 2017.
[104] H. Jeong et al., "A pilot clinical study of low-intensity transcranial focused ultrasound in
Alzheimer's disease," Ultrasonography (Seoul, Korea), 2021.
[105] X. Qian et al., "Noninvasive Ultrasound Retinal Stimulation for Vision Restoration at High
Spatiotemporal Resolution," BME Frontiers, vol. 2022, 2022.
[106] Y. Qu et al., "Acoustic radiation force optical coherence elastography of corneal tissue," IEEE
Journal of Selected Topics in Quantum Electronics, vol. 22, no. 3, pp. 288-294, 2016.
[107] Y. Qu et al., "Quantified elasticity mapping of retinal layers using synchronized acoustic
radiation force optical coherence elastography," Biomedical optics express, vol. 9, no. 9, pp. 4054-4063,
2018.
[108] H. Yoon and T.-K. Song, "Sparse rectangular and spiral array designs for 3D medical ultrasound
imaging," Sensors, vol. 20, no. 1, p. 173, 2020.
[109] S. Harput et al., "3-d super-resolution ultrasound imaging with a 2-d sparse array," IEEE
transactions on ultrasonics, ferroelectrics, and frequency control, vol. 67, no. 2, pp. 269-277, 2019.
[110] A. Ramalli et al., "Real-time 3-D spectral doppler analysis with a sparse spiral array," IEEE
Transactions on Ultrasonics, Ferroelectrics, and Frequency Control, vol. 68, no. 5, pp. 1742-1751, 2021.
[111] E. Roux, A. Ramalli, P. Tortoli, C. Cachard, M. C. Robini, and H. Liebgott, "2-D ultrasound
sparse arrays multidepth radiation optimization using simulated annealing and spiral-array inspired
energy functions," IEEE transactions on ultrasonics, ferroelectrics, and frequency control, vol. 63, no.
12, pp. 2138-2149, 2016.
[112] Robert Wodnicki, Y.S., Qifa Zhou, "MODULAR ACOUSTIC SENSOR ASSEMBLIES," United
States Patent Appl. 63/536,056, 2023.
[113] R. Wodnicki, H. Kang, Y. Sun, L. Jiang, H. Lu, and Q. Zhou, "High Frequency 1.75 D array
using a 3D printed pitch-changing interposer backing," in 2020 IEEE International Ultrasonics
Symposium (IUS), 2020: IEEE, pp. 1-4.

Abstract (if available)

Abstract Ultrasound medical imaging is an entrenched and powerful tool for medical diagnosis, it demonstrates high efficiency, reduced radiation exposure, and non-invasive characteristics. The ultrasound transducers can be classified according to their morphology into single-element transducers, linear arrays and 2D arrays.

The process of fabricating ultrasound single element transducers has matured over decades; the research efforts have focused on ceramic materials with better performance. Ultrasound linear arrays are arrays of ultrasound transducers with multiple array elements that fire the individual elements of the array in a specific sequence to adjust the phase to direct the ultrasound waves in a specific direction, thus achieving a larger and controllable imaging area than that of a single element transducer. Ultrasound 2D arrays are widely used for elastography as they can generate highly focused acoustic beams that can be tuned in 3D space. At the same time, ultrasonic 2D arrays enable real-time volumetric imaging, and thus have been designed, fabricated, and applied to transesophageal echocardiography and intracardiac echocardiography.

Despite all these advantages, the technical obstacle of ultrasound 2D array mainly lies in its manufacturing process. Due to its strict limitations on element density and element size, the fabrication process of linear arrays is difficult to be applied to 2D arrays. High-frequency ultrasound 2D arrays require the realization of the highest possible number of array elements and array element spacing up to 50 m level to achieve high-resolution and high-frequency operation. Existing assembly methods to date have generally relied on vertical flexure integration, but the vias width of electronic substrates at this stage is at least 50 m, and circuit structures that do not use vias have difficulty in achieving the required number of array elements, making the process of standard 2D arrays difficult to realize. For fully sampled 2D arrays, the connection and fabrication of circuits as fine as a few hundred microns is extremely difficult and expensive. In recent years, with the flourishing development of 3D printing technology, it is finally possible to print complex and fine 3D structures in a relatively short period, which also provides a solution for fabricating ultrasound 2D arrays.

The work in this dissertation proposal investigates the feasibility of an ultrasound transducer based on a new textured ceramic material, the fabrication process of ultrasound 2D arrays using 3D printing technology, and its applications. In our work on ultrasound 2D arrays based on 3D printing technology, we have continued to explore and optimize the fabrication process successfully upgrading the 64 array elements to 256 array elements and the center frequency from 3-MHz to 15-MHz. the fabricated ultrasound 2D arrays have successfully acquired 3D reconstructed images of wire targets and have been applied in experiments on ultrasound-stimulated rat visual prostheses. These studies demonstrate the feasibility and broad application prospects of ultrasound 2D arrays fabricated using 3D printing technology. Moreover, our future research will continue to focus on improving the performance and fabrication success of ultrasound 2D arrays and applying them to multiple areas of ultrasound stimulation and imaging.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

2D ultrasonic transducer array’s design and fabrication with 3D printed interposer and applications

PDF

Fabrication of ultrasound transducer and 3D-prinitng ultrasonic device

PDF

Development of high-frequency (~100mhz) PZT thick-film ultrasound transducers and arrays

PDF

Highly integrated 2D ultrasonic arrays and electronics for modular large apertures

PDF

Array transducers for high frequency ultrasound imaging

PDF

Design and development of ultrasonic array transducers for specialized applications

PDF

High-frequency ultrasonic transducers for photoacoustic applications

PDF

Novel beamforming with dual-layer array transducers for 3-D ultrasound imaging

PDF

Development of novel 1-3 piezocomposites for low-crosstalk high frequency ultrasound array transducers

PDF

Quantification of cellular properties using ultra-high frequency single-beam acoustic tweezer

PDF

Development of front-end circuits for high frequency ultrasound system

PDF

High frequency ultrasound array for ultrasound-guided breast biopsy

PDF

High resolution elastography in ophthalmology

PDF

High frequency ultrasonic phased array system and its applications

PDF

Additive manufacturing of piezoelectric and composite for biomedical application

PDF

High frequency ultrasonic imaging for the development of breast biopsy needle with miniature ultrasonic transducer array

PDF

Miniature phased-array transducer for colorectal tissue characterization during TEM robotic surgery; and, Forward-looking phased-array transducer for intravascular imaging

PDF

Configurable imaging platform for super-harmonic contrast-enhanced ultrasound imaging

PDF

Single-cell analysis with high frequency ultrasound

PDF

Development of high frequency focused transducers for single beam acoustic tweezers

Asset Metadata

Creator Sun, Yizhe (author)

Core Title Design and fabrication of ultrasound transducers: from single element to high frequency 2D array

Contributor Electronically uploaded by the author (provenance)

School Andrew and Erna Viterbi School of Engineering

Degree Doctor of Philosophy

Degree Program Biomedical Engineering

Degree Conferral Date 2024-05

Publication Date 03/30/2025

Defense Date 02/28/2024

Publisher Los Angeles, California (original), University of Southern California (original), University of Southern California. Libraries (digital)

Tag Biomedical ultrasonic imaging,ceramic composites,OAI-PMH Harvest,piezoelectric,ultrasound imaging.,ultrasound transducer

Format theses (aat)

Language English

Advisor Zhou, Qifa (committee chair), Chen, Yong (committee member), Shen, Keyue (committee member)

Creator Email yizhesun@usc.edu

Permanent Link (DOI) https://doi.org/10.25549/usctheses-oUC113865218

Unique identifier UC113865218

Identifier etd-SunYizhe-12743.pdf (filename)

Legacy Identifier etd-SunYizhe-12743

Document Type Dissertation

Format theses (aat)

Rights Sun, Yizhe

Internet Media Type application/pdf

Type texts

Source 20240403-usctheses-batch-1134 (batch), University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the author, as the original true and official version of the work, but does not grant the reader permission to use the work if the desired use is covered by copyright. It is the author, as rights holder, who must provide use permission if such use is covered by copyright.

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA

Repository Email cisadmin@lib.usc.edu