Close
About
FAQ
Home
Collections
Login
USC Login
Register
0
Selected
Invert selection
Deselect all
Deselect all
Click here to refresh results
Click here to refresh results
USC
/
Digital Library
/
University of Southern California Dissertations and Theses
/
Integrated wireless piezoelectric ultrasonic transducer system for biomedical applications
(USC Thesis Other)
Integrated wireless piezoelectric ultrasonic transducer system for biomedical applications
PDF
Download
Share
Open document
Flip pages
Contact Us
Contact Us
Copy asset link
Request this asset
Transcript (if available)
Content
INTEGRATED WIRELESS PIEZOELECTRIC ULTRASONIC TRANSDUCER SYSTEM FOR
BIOMEDICAL APPLICATIONS
by
Jaehoon Lee
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
(ELECTRICAL ENGINEERING)
May 2024
Copyright 2024 Jaehoon Lee
ii
Acknowledgements
I would like to express my sincere gratitude to everyone who has played a pivotal role in
supporting me throughout my Ph.D. journey. First and foremost, immense thanks to my advisor,
Professor Eun Sok Kim, whose guidance and mentorship have been invaluable throughout the
years. He serves as an exceptional role model, demonstrating a never-ending enthusiasm and
meticulous approach to research that has inspired me to consistently strive for excellence and
explore diverse research projects. His unwavering support, insightful feedback, and
encouragement have been instrumental in shaping my research and academic growth. Even in the
face of the pandemic's constraints, he continued to prioritize the well-being and progress of his
students, ensuring that our academic pursuits remained on track. His support extended beyond
academic matters, as he demonstrated genuine concern for our overall well-being during these
trying times. I am profoundly grateful for Dr. Kim's guidance, without which the accomplishments
in my thesis would not have been possible.
I extend my sincere appreciation to the committee members of my PhD thesis defense, Prof.
Wei Wu and Prof. Ellis Meng, whose expertise and constructive criticism have greatly enriched
the quality of my work.
Collaborating with Prof. Jiang F. Zhong from School of Medicine at Loma Linda
University, Prof. Su-youne Chang from Mayo clinic, and Prof. Bruce Herring from Dornsife
School of Neurobiology at USC has been an enlightening experience, and I am grateful for the
diverse perspectives they brought to the table, which broadened my knowledge in neuroscience.
Additionally, I appreciate for the collaborative discussions and teamwork with both past and
current members of the USC MEMS Group. In particular, I am grateful for the mentorship and
iii
assistance provided by Dr. Yongkui Tang during the first year of my Ph.D. research. His guidance
was instrumental in shaping my early research endeavors, and his ongoing inspiration continues
to positively influence my academic journey. Dr. Tang not only served as a dedicated mentor but
has also become a cherished and enduring friend. I also thank to Matin, Akash, Kianoush, Junyi,
Hongxiang, Anik, Dr. Baptiste Neff, Dr. Kunfeng Wang, and Prof. Joung-hu Park for their help,
discussion, and friendship.
Furthermore, I want to send my heartfelt thanks to my parents in Korea. From the very
beginning of my life untill today, their encouragement and support have been endless. Every
success in my life is a proof of the love and guidance they have provided. I am deeply thankful for
their belief in me and the boundless love from my parents and family that has been instrumental
in every moment of my life.
Most importantly, I would like to express my deepest gratitude and love to my wife, Rihyun
Park. Your unwavering support, patience, and understanding have been essential elements that
sustained me during the challenges of pursuing a Ph.D. Your belief in me, sacrifices, and love have
been my constant motivation. I want to emphasize that without your love and support, none of my
achievements would have been possible. This achievement is as much yours as it is mine, and I
am profoundly grateful for the love and encouragement that you have showered upon me
throughout this journey.
iv
Table of Contents
Acknowledgements......................................................................................................................... ii
List of Table ................................................................................................................................. viii
List of Figures.............................................................................................................................. viii
Abstract...................................................................................................................................... xviii
Chapter 1. Introduction................................................................................................................... 1
1.1 Applications of Ultrasound in Biomedical............................................................................ 1
1.2 Overview of Acoustic Transducer......................................................................................... 2
1.3 Overview of Driving Acoustic Transducers.......................................................................... 5
1.4 Scope of the Thesis............................................................................................................... 8
Chapter 2. Design, Fabrication, and Characterization of Self-Focusing Acoustic
Transducers (SFAT) ...................................................................................................................... 10
2.1 Introduction......................................................................................................................... 10
2.2 Working Principle of SFAT................................................................................................. 11
2.3 Design parameters of SFATs............................................................................................... 14
2.3.1 Operating Frequency.................................................................................................... 14
2.3.2 Focal Diameter (Focal Size) ........................................................................................ 15
2.3.3 Focal Length ................................................................................................................ 16
2.3.4 Depth of Focus (Focal Depth)...................................................................................... 16
2.4 Simulation of Acoustic Pressure ......................................................................................... 17
2.5 Microfabrication Methods for SFATs ................................................................................. 19
2.5.1 Fabrication of SFAT with Patterned Electrode Rings.................................................. 19
2.5.2 Fabrication of SFAT with Air-Cavity Lens.................................................................. 20
2.6 Characterization of SFAT.................................................................................................... 21
2.6.1 Measurement of Electrical Impedance and Resonant Frequency................................ 21
2.6.2 Measurement of Acoustic Pressure.............................................................................. 23
Chapter 3. Design, Fabrication, and Characterization of electrically Controllable
Self-Focusing Acoustic Transducers (eCON SFAT)..................................................................... 26
3.1 Introduction......................................................................................................................... 26
3.2 Design and Working Principle of eCON SFAT................................................................... 26
3.2.1 Focal Length Controllability........................................................................................ 27
3.2.2 Focal Point Controllability........................................................................................... 28
3.3 Simulation of Focusing Controllability .............................................................................. 29
3.4 Microfabrication Methods of eCON SFATs........................................................................ 30
3.4.1 Fabrication of eCON SFAT Electrodes........................................................................ 30
3.4.2 Fabrication of Indium Solder Bump ............................................................................ 31
v
Chapter 4. Design, Fabrication, and Characterization of Spiral-arm Vortex-beam Acoustic
Transducer (SVAT) ....................................................................................................................... 38
4.1 Introduction......................................................................................................................... 38
4.2 Working Principle of SVAT ................................................................................................ 39
4.2.1 Acoustic Vortex-beam with A Sectored SFAT Design ................................................. 39
4.2.2 Acoustic Vortex-beam with A Spiral-Arm Electrode Design....................................... 40
4.3 Simulation of Acoustic Vortex ............................................................................................ 42
4.4 Microfabrication Methods of SVAT.................................................................................... 43
Chapter 5. Wireless Operating Integrated System Design for Acoustic Transducers................... 44
5.1 Introduction......................................................................................................................... 44
5.2 Power Source for Driving Acoustic Transducers................................................................ 45
5.2.1 Lithium-ion battery ...................................................................................................... 45
5.2.2 Supercapacitor.............................................................................................................. 47
5.2.3 Lithium-ion Capacitor.................................................................................................. 48
5.3 Electronic Circuit Design with An On-board Power Amplifier.......................................... 50
5.3.1 System Design ............................................................................................................. 50
5.3.2 Components Selection ................................................................................................. 51
5.3.3 PCB design .................................................................................................................. 54
5.3.4 Circuit Characterization............................................................................................... 54
5.4. Electronic Circuit Design with A Modified Z-source Inverter .......................................... 56
5.4.1 System Design ............................................................................................................. 56
5.4.2 Component selection.................................................................................................... 58
5.4.3 Circuit Simulation........................................................................................................ 62
5.4.4 PCB design .................................................................................................................. 64
5.4.5 Circuit Characterization............................................................................................... 65
5.5 Pulse operation using a Bluetooth Low Energy (BLE) ...................................................... 70
Chapter 6. Wireless and Stand-Alone Submarine Propeller Based on Acoustic Propulsion........ 73
6.1 Introduction......................................................................................................................... 73
6.2 Device Design for The Acoustic Propeller ......................................................................... 74
6.3 System Design .................................................................................................................... 75
6.4 Experimental Results.......................................................................................................... 78
6.4.1 Acoustic Propulsion with SFAT................................................................................... 78
6.4.2 Acoustic Propulsion with SVAT................................................................................... 81
6.5 Discussions ......................................................................................................................... 84
6.6 Summary............................................................................................................................. 85
Chapter 7. Levitating Acoustic Tweezers Based on SVATs.......................................................... 87
7.1 Introduction......................................................................................................................... 87
7.2 Principle of Acoustic Levitation and Trapping ................................................................... 88
vi
7.3 Device Design..................................................................................................................... 89
7.4 Experimental Results and Discussion................................................................................. 90
7.4.1 Trapping Capability ..................................................................................................... 90
7.4.2 Manipulation in a Three-Dimensional Space .............................................................. 92
7.4.3 Levitation ..................................................................................................................... 93
7.5 Summary............................................................................................................................. 94
Chapter 8. Non-invasive Focused Ultrasound Neuromodulation Using a SFAT.......................... 96
8.1 Introduction......................................................................................................................... 96
8.2 Device Design..................................................................................................................... 99
8.2.1 NeuroSFAT .................................................................................................................. 99
8.2.2 Control NeuroSFAT ................................................................................................... 101
8.2.3 Low-EMI NeuroSFAT ............................................................................................... 101
8.3 Experiment Design............................................................................................................ 104
8.3.1 Whole-cell Patch Clamp Setup .................................................................................. 104
8.3.2 Focused Ultrasound Stimulation Parameters............................................................. 105
8.4 Experiment Result............................................................................................................. 106
8.4.1 Electromagnetic Interference (EMI) Effect ............................................................... 107
8.4.2 Thermal Effect ........................................................................................................... 109
8.4.3 The Passive Properties of Cell Membrane................................................................. 110
8.4.4 The Excitability of Neuron .........................................................................................111
8.5 Discussions ....................................................................................................................... 113
8.6 Summary........................................................................................................................... 116
Chapter 9. Contactless Single Cell Extraction from Monolayer Cell Culture Using
Acoustic Droplet Ejector............................................................................................................. 117
9.1 Introduction....................................................................................................................... 117
9.2 Device Design................................................................................................................... 118
9.3 Experimental Setup and Method....................................................................................... 119
9.3.1 Preparing RPE Cell Culture ....................................................................................... 119
9.3.2 Experiment setup ....................................................................................................... 120
9.4 Experiment Result............................................................................................................. 122
9.5 Discussions ....................................................................................................................... 125
9.6 Summary........................................................................................................................... 127
Chapter 10. Summary and Future Directions ............................................................................. 129
10.1 Summary......................................................................................................................... 129
10.2 Future Work on Transducers........................................................................................... 131
10.2.1 Future Work on eCONSFAT.................................................................................... 131
10.2.2 Future Work on Low-EMI SFAT ............................................................................. 131
10.3 Future Work on Driving Circuits.................................................................................... 131
vii
10.4 Future Work on Focused Ultrasound Neuromodulation ................................................. 132
Reference ....................................................................................................................................133
viii
List of Tables
Table 3. 1 Description and Dimension Information of Symbols in Figure 3.7............................. 36
Table 5. 1 Feature comparison table of Lithium-ion battery, supercapacitor, and Lithium-ion
capacitor........................................................................................................................................ 50
Table 5. 2 Key Feature Comparison between On-Board Power Amplifier and Z-Source
Inverter.......................................................................................................................................... 69
Table 6. 1 Various Designs of Spiral-arm Vortex-beam Acoustic Transducers (SVATs).............. 82
Table 6. 2 Summary of Fabricated and Tested Underwater Acoustic Propellers.......................... 84
Table 7. 1 Summary of The Design Parameters for Single Arm and Multi-Arms SVAT ............. 89
Table 7. 2 Comparison of Trapping and Design Features Across Different Acoustic Vortex
Tweezers ....................................................................................................................................... 95
Table 8. 1 The Material Properties and Skin Depths of Aluminum, Copper, and Nickel........... 103
Table 8. 2 The Summary of Neuronal Inhibition and Excitation Induced by SFAT................... 112
Table 8. 3 Previous FUS Studies Related to TRP Channels....................................................... 114
Table 8. 4 Previous FUS Studies Related to K2P Channels ....................................................... 115
Table 9. 1 RPE Marker Genes with Their Populations............................................................... 126
Table 9. 2 The Results (p-value) of T-test Between the Acoustically Collected and the
Control Group ............................................................................................................................. 127
ix
List of Figures
Figure 2. 1 Conceptual view of focused ultrasound generated by SFAT, defining the focal size,
focal depth, focal length, half angle (θ), ring radius (Rn), and the width of the outermost ring
(ΔR)............................................................................................................................................... 12
Figure 2. 2 (a) The cross-sectional diagram (across the AA’ dashed line in (b)) showing a
typical SFAT based on a PZT substrate with patterned Fresnel annular-rings electrodes.
This illustration highlights how the Fresnel acoustic lens focuses ultrasound in the medium
by constructively interfering acoustic waves in-phase. (b) The top-view diagram of the same
SFAT, showing the relative positions of the top electrode, and soldering pad. ............................ 13
Figure 2. 3 (a) The cross-sectional diagram (across the AA’ dashed line in (b)) showing a
typical SFAT based on a PZT substrate with a Fresnel air-cavity acoustic lens made of
Parylene. This illustration highlights how the Fresnel acoustic lens focuses ultrasound in the
medium by obstructing destructively interfering acoustic waves (by the total reflection from
the air-cavity).(b) The top-view diagram of the same SFAT, illustrating the relative positions
of the top electrode, air-cavity rings, soldering pad, and Parylene-coated regions. ..................... 14
Figure 2. 4 Finite Element Method (FEM) analysis of the normalized acoustic pressure
distribution of SFAT, which operating frequency is 18.4Mhz, 15 rings, and focal length of
400µm, in the brain tissue media. The simulation result shows that the focal size of 46µm and
focal depth of 95µm...................................................................................................................... 18
Figure 2. 5 Fabrication steps of SFAT with a patterned rings;(a) PZT sheet sandwiched by
nickel(Ni) layers (b) Mask pattern for the Fresnel half wavelength band annular rings
(c) Wet-etching Ni electrodes and remove the photoresist mask and (d) Parylene D coating for
the water proof and biocompatibility............................................................................................ 20
Figure 2. 6 Fabrication steps of SFAT with an air-cavity lens;(a) PZT sheet sandwiched by
nickel(Ni) layers (b) Mask pattern for the top and bottom circular electrodes (c) Wet-etching
Ni electrodes and remove the photoresist mask (d) Parylene D coating as a protective layer (e)
Depositing photoresist sacrificial layer (f) Another Parylene D coating (g) RIE etching release
holes to open access to the sacrificial layer and (h) Depositing thick Parylene D layer for
insulation and biocompatibility..................................................................................................... 21
Figure 2. 7 Magnitude (black) and phase (red) of one-port S11 measurement of 2.32MHz SFAT.
Resonant frequency (or series resonance, fs) and anti-resonant frequency (or parallel resonance,
fp) are found at where the reactance values are zero, allowing to be estimated as 2.19MHz and
2.25MHz, respectively. ................................................................................................................. 22
Figure 2. 8 (a) Perspective view of a piezoelectric substrate (PZT) defining the area A and the
thickness d. (b) BVD equivalent model of SFAT showing motional inductance (Lm), motional
capacitance (Cm), motional resistance (Rm) and clamp capacitance (C0)................................... 22
x
Figure 2. 9 Finite element analysis (FEA) simulated normalized acoustic pressure (a) on the
lateral plane at z= 5 and (b) on the vertical plane along with the central axis. Measured peak
acoustic pressure using a hydrophone (c) on the lateral plane and (d) on the vertical plane. (e)
Illustration of the hydrophone measurement setup in the scanning tank...................................... 24
Figure 3. 1 (a) The top-view diagram of the eCON SFAT, showing the 8 sectors and 128
individual equal-width electrodes. (b) Photo of a fabricated eCON SFAT before a flip-chip
bonding packaging........................................................................................................................ 27
Figure 3. 2 Phase factor plot of 16 ring electrodes with different desired focal length from 5 to
15mm. The green color means activation and the red color means deactivation of the electrode.
....................................................................................................................................................... 28
Figure 3. 3 (a) FEM simulation model in a three-dimensional (3D) space and simulation
results of (b) a single sector activated with the focal length at 5mm and the focal point at
2mm right from the center (b) two adjacent sectors activated with the focal length at 5mm
and the two focal points at 2mm left and below from the center (c) two facing sectors activated
with the focal length at 7mm and the two focal points at 2mm above and below from the center
and (d) the focused ultrasound at the central axis with the focal length at 5mm by activating
three sectors with the same groups of electrodes at the same time............................................... 29
Figure 3. 4 A brief Fabrication steps of eCON SFAT;(a) PZT sheet sandwiched by nickel(Ni)
layers (b) Patterning and wet-etching the electrode with equal-with sectored rings (c) 7µm
thick Parylene D coating as a passivation layer (d)RIE-etching the holes for accessing
electrode and deposit the under-bump metallization (UBM) layer to define indium bump
position. (e) Depositing 1.5μm thick indium through the thermal evaporation on top of the
photoresist layer, followed by (f) the lift off process and the (g)reflow to form a spherical
shape of the bumps. (h) Flipping the device (i) Wet-etching Ni electrodes into circular shape
and remove the photoresist mask (j) Parylene D coating as a protective layer (k) Depositing
photoresist sacrificial layer (l) Another Parylene D coating (m) RIE etching release holes to
open access to the sacrificial layer and (n) Depositing thick Parylene D layer for insulation and
biocompatibility ............................................................................................................................ 31
Figure 3. 5 Indium bump after a reflow with different combination of lateral dimension of the
UBM (80μm) and the indium pattern (a) 150 µm, (b) 225 µm, (c) 500 µm, and (d) 600 µm. (e)
Graphical expression of analytical calculation and experimental measurement of the height of
indium bump after a reflow........................................................................................................... 33
Figure 3. 6 Reflown indium bumps under the different environments such as (a) air,
(b) nitrogen gas (N2), (c) hydrogen gas (H2), (d) liquid metal flux, and (e) formic acid
(HCOOH)...................................................................................................................................... 34
Figure 3. 7 Illustrated image of indium solder bump after the reflow, showing the dimension
and design parameter information (description and dimension is listed in Table 3.1) ................. 35
xi
Figure 3. 8 A brief fabrication process of indium bumps on the control board; (a) A printed
circuit board (PCB) with a patterned electroless nickel immersion gold (ENIG) pads (b) Spin
coating and patterning photoresist to define the position of indium bump (c) Depositing 1.5μm
thick indium through the thermal evaporation on top of the photoresist layer (d)Lifting- off
photoresist and (g)Reflowing to form a spherical shape of the bumps......................................... 37
Figure 4. 1 (a)The designs and (b)photos of the fabricated transducers of an 18-sectored SFAT,
generating an acoustic vortex. (c)The designs and (d)photos of the fabricated transducers of an
18-sectored SFAT with a center cavity electrode design. ............................................................. 40
Figure 4. 2 (a)Single-arm, (b)Two-arms, (c)Three-arms, and (d)Four-arms SVAT design with
parameters of a1 = 2.3 mm, a2 = 2.5 mm, b = 0.090, θ1 = 0˚, and θ2 = 12π ...................... 42
Figure 4. 3 (a) FEM simulation model in a three-dimensional (3D) space having a conical
shape and normalized acoustic pressure distribution simulation results of (b) a SVAT with
parameters of a1 = 0.5 mm, a2 = 0.7 mm, b = 0.090, θ1 = 0˚, and θ2 = 12π, and
(c) the 18-sectroed SFAT design. .................................................................................................. 42
Figure 4. 4 Normalized absolute acoustic pressure distributions over the xy plane at an
observation point (z= 7mm) produced by (a) 1-arm SVAT, (b) 2-arms SVAT, (c) 3-arms SVAT
and (d) 4-arms SVAT..................................................................................................................... 43
Figure 5. 1 An illustration that shows the structure of Lithium-ion battery and the movement of
Lithium ions and electrons during discharging and supplying power to the device (or load)...... 46
Figure 5. 2 An illustration depicting the structure of a supercapacitor with highlighted electric
double layers (dashed line) that store electrical energy within the supercapacitor....................... 48
Figure 5. 3 An illustration depicting the structure of a Lithium-ion capacitor (LIC), combining
the anode of Lithium-ion battery with the cathode of Supercapacitor.......................................... 49
Figure 5. 4 A system block diagram of SFAT driving electronic circuit using a power amplifier.
Black and blue lines represent the power and signal rails, respectively. A capacitor, denoted by
a red dashed line, is used to block the DC components of the PWM signal. .............................. 50
Figure 5. 5 Schematic diagrams of (a) single-ended and (b) differential topologies of
amplifying circuit.......................................................................................................................... 52
Figure 5. 6 Simulation results for the THS3095 in a differential topology revealing signal
performance under two conditions: (a) a 3ns and (b) a 22 ns time delay between the two input
PWM signals of the amplifier. ...................................................................................................... 53
Figure 5. 7 (a) A PCB layout design of the system, having dimension of 17mm × 28mm.
(b) A photo of test board PCB including headers, test points, and a SMA connect for testing
and debugging purposes................................................................................................................ 54
xii
Figure 5. 8 The plot showing the power conversion efficiency varying the load resistance (RL)
from 30Ω to 1.5kΩ. The maximum power efficiency of the system of 66% occurs at RL =130Ω
....................................................................................................................................................... 55
Figure 5. 9 The system bandwidth, showing the gain of the system over the frequency range
from 1kHz to 30MHz. The cutoff frequency (f − 3dB) is found at approximately 5MHz ......... 56
Figure 5. 10 Schematic diagrams of (a) a conventional boost converter, (b) a boost converter
with a modified Z-source network, and (c) an inverter with a modified Z-source network......... 57
Figure 5. 11 A Plot showing the minimum inductor value with varying the load (RL) and
the duty cycle (D), ensuring the Continuous Conduction Mode (CCM) operation. The red dots
indicate the minimum inductance value when the resistive load is 100Ω. ................................... 61
Figure 5. 12 Circuit schematic diagram for the simulation of the modified Z-source inverter
with BVD model of SFAT (red box) as a load.............................................................................. 62
Figure 5. 13 LTSPICE simulation results under the condition of ideal MOSFET (Q1) and ideal
diode (D1) with a PWM signal of 50% duty cycle (a) for 1-ms simulation time, and (b) the
magnified view of the output voltage and the PWM input during the last 10 cycles. The output
voltage, the PWM input signal, and the inductor current are represented with gray dashed line,
red solid line, and blue solid line, respectively. ............................................................................ 63
Figure 5. 14 LTSPICE simulation results under the condition of ideal MOSFET (Q1) and ideal
diode (D1) with a PWM signal of 47.5% duty cycle (a) for 1-ms simulation time, and (b) the
magnified view of the output voltage and the PWM input during the last 10 cycles. The output
voltage, the PWM input signal, and the inductor current are represented with gray dashed line,
red solid line, and blue solid line, respectively. ............................................................................ 63
Figure 5. 15 LTSPICE simulation results with EPC2001(Q1), its gating driver IC (LMG1205)
and RB048RSM10S diode (D1) with a PWM signal of 47.5% duty cycle (a) for 1-ms simulation
time, and (b) the magnified view of the output voltage and the PWM input during the last 10
cycles. (c) On-time of the PWM signal showing the rising and falling time induced by EPC2001
and LMG1205. The rising and falling are measured approximately 7.5ns and 3.8ns, respectively,
leading to the effective duty cycle of 47.0%. The orange dotted line indicates the threshold gate
voltage of EPC2001(Vth = 1.4V). The output voltage, the PWM input signal, and the inductor
current are represented with gray dashed line, red solid line, and blue solid line, respectively. .. 64
Figure 5. 16 A PCB layout design of the modified Z-source inverter system, having dimension
of 12mm × 24mm. Only top layer is shown because the bottom layer space is reserved for the
Bluetooth-Low Energy IC............................................................................................................. 65
Figure 5. 17 Photos of (a) the experiment setup utilizing LMG1205HBEVM, and (b) the
bottom view, and (c) the side-view of the modified Z-source network which is structured in
a 3D configuration. ....................................................................................................................... 66
xiii
Figure 5. 18 Waveform of (a) the output AC voltage along with a PWM signal and
(b) the output AC voltage and the input DC voltage, captured from the oscilloscope. ................ 66
Figure 5. 19 The power efficiency plot of the Z-source inverter reveals varying performance
under different resistive load conditions, with the peak efficiency reaching 91% at a load of
75 Ω.............................................................................................................................................. 67
Figure 5. 20 The analytically calculated (orange) and experimentally measured (blue) output
peak voltages with the varying duty cycle, when the input voltage is 4VDC and the SFAT is
connected at the load..................................................................................................................... 68
Figure 5. 21 The plot showing the gain of the system within the frequency range between
1.7MHz to 3.2MHz, without load (Black) and with a 75Ω load (Red). The bandwidth of the
system is limited to the frequency range of from1.7MHz to 2.4MHz.......................................... 69
Figure 5. 22 Conceptual illustration of the pulsed operation of acoustic transducer, showing the
key parameters such as the number of cycles (# of cycles), a pulse duration, a pulse repetitive
frequency (PRF), and the operating frequency of the transducer. ................................................ 70
Figure 5. 23 The system block diagram for the pulsed mode operation of the electronic circuit
with a Bluetooth-Low Energy (BLE), implemented on (a) the power amplifier design and
(b) the Z-source inverter design.................................................................................................... 71
Figure 5. 24 Waveforms of a pulsed output delivered to SFAT (Yellow), input voltage to the Zsource inverter (Skyblue), and a pulse control PWM signal from BLE (Blue), showing (a) three
pulses with a Pulse Repetition Frequency (PRF) of 10HZ and a Pulse Duration (PD) of 10ms
and (b) a single pulse with PD of 10ms. ....................................................................................... 72
Figure 6. 1 Conceptual illustration of generating acoustic propulsion with a Self-Focusing
Acoustic Transducer (SFAT) with air cavity reflector and Fresnel air-cavity lens....................... 74
Figure 6. 2 (a) Conceptual 3D schematic of the acoustic propeller and (b) photo of the
completed acoustic propeller. Lithium-ion capacitor (LIC) cannot be seen in the photo as it is
wrapped by the flexible printed circuit board (PCB).................................................................... 75
Figure 6. 3 (a) Schematic of the PCB layout showing the four main circuits sections and the
places where SFAT and LIC will be connected and (b) photo of the fabricated flexible PCB
before being folded into a 3D shape shown in Figure 6.2b. ......................................................... 76
Figure 6. 4 Functional block diagram of the system showing the power rails (black lines) and
the signal rails (blue lines). Wireless power receiver block is added on the input voltage,
so that it can supply power to the system and charge the LIC at the same time........................... 77
Figure 6. 5 Photos showing (a) wireless charging of LIC, while the SFAT in the propeller is
immersed in water to prevent thermal damage of the SFAT during the charging and (b) air gap
between transmission (TX) and receiving (RX) coils maintained by a pair of rigid wires and a
xiv
clamp holding TX and RX coils, respectively, with alignment between TX and RX coils
achieved by a 3-axis micromanipulation stage. ............................................................................ 77
Figure 6. 6 Photos of the acoustic propeller (a) floating close to the floor without acoustic
propulsion and (b) soaring up to the surface of sodium polytungstate (SPT) solution
(1.15g/cm3) when the SFAT generates the acoustic propulsion. ................................................. 78
Figure 6. 7 (a) Measured traveling distance (in mm) and (b) velocity (in mm/s) of the acoustic
propeller from the vertical propulsion shown in Fig. 8. The traveling distance is measured in
every 100ms and the averaged acceleration is calculated to be 1.16 mm/s2................................. 79
Figure 6. 8 Photo of the acoustic propeller floating horizontally on sodium polytungstate (SPT)
solution (1.20g/cm3)..................................................................................................................... 80
Figure 6. 9 The designs (left) and photos of the fabricated transducers (right) of (a) 18-sectored
Self-Focusing Acoustic Transducer (SFAT) based on annular Fresnel rings and (b) 1- arm
SVAT and SFAT with two rings in the center................................................................................ 81
Figure 6. 10 Photos of the propeller (the six-ring SFAT with 4.5 mm focal length) moving at
(a) t=0 and (b) t = 660ms.............................................................................................................. 83
Figure 6. 11 (a) Measured traveling distance (in mm) and (b) velocity (in mm/s) of the 4-arms
SVAT (A2 design) acoustic propeller. The traveling distance is measured in every frame (33ms)
and the averaged acceleration is calculated to be 238.8 mm/s2 ................................................... 83
Figure 6. 12 Photos of the propeller (1- arm SVAT(A0) + two- ring SFAT) moving at (a) t = 0,
(b) t = 330 ms, showing a twisting motion, (c) t = 660 ms, and (d) t = 1 s. ................................. 85
Figure 7. 1 Photos showing the filtering and sorting of particles with different size
(and density) by varying the applied voltage: (a) 28 Vpp trapping a large and light particle
(5mm in diameter and 1.03 g/cm3
), (b) 52 Vpp trapping a mid-size particle (2.96mm in
diameter and 1.05 g/cm3
) (c) 90 Vpp trapping a small and heavy particle (2.4 mm in diameter
and 1.30 g/cm3
)............................................................................................................................. 90
Figure 7. 2 Lifting a trapped microsphere (5 mm in diameter and 1.30 g/cm3 density) by
increasing the applied voltage, showing more vibration (indicated by the red arrow) of the
trapped particle with higher voltage; (a) trapped at z=1.1 cm without llinear vibration with 90
Vpp, (b) trapped at z= 1.5 cm with little rotation and vibration with 100 Vpp, and (c) trapped at
z=2.1 cm with a large rotation and vibration with 120 Vpp........................................................... 91
Figure 7. 3 Shooting a microsphere (5 mm in diameter and1.30g/cm3
density)bytheSVATactivatedwith140 Vpp, as the images captured in every 50 ms from a
recorded video with the z-positions of the particle being at (a) 2.1 cm, (b) 2.3 cm, (c) 2.8 cm,
and (d) 4.5 cm from the tweezers surface. (e) The particle has reached the liquid-air interface.. 92
xv
Figure 7. 4 Schematic of the experimental setup for manipulating the trapped particle in 3D
space with SVAT tweezers. ........................................................................................................... 92
Figure 7. 5 On-demand 1-cm (which can be longer) manipulation of a 5 mm (in diameter)
microsphere with density of 1.30 g/cm3 in 3-dimensional space in (a) lateral and (b) vertical
direction, as the tweezers is moved laterally and vertically, respectively..................................... 93
Figure 7. 6 Sequential photos showing the levitation process of a microsphere (density of 1.30
g/cm3 and diameter of 5 mm) from the surface of the tweezers and simultaneous trapping at z=
1.1 cm. The images are captured every 50 ms from a recorded video.......................................... 94
Figure 8. 1 (a) Top view of fabricated SFAT for neuromodulation (b) Bottom view of SFAT
with 0.5 mm thick polyester sheet and (c) SFAT with an IR light from the bottom to
demonstrate the translucent characteristic of 127 µm thick PZT ............................................... 100
Figure 8. 2 Illustrations of (a) positive Fresnel half wavelength band (FHBW) rings design with
circular electrode in the center, and (b) negative FHBW rings with no electrode in the center.
The electrodes area is indicated by black color. ......................................................................... 100
Figure 8. 3 Photos of (a) an air-cavity reflector made from three pieces of laser-machined
acetate sheets and (b) a control neuroSFAT with the air-cavity reflector attached, and packaged
by parylene coating..................................................................................................................... 101
Figure 8. 4 Photos of (a) neuroSFAT with a thermal lease tape mask at the center before
depositing Nickel, (b) the microscopic image of the thermal lease tape mask, (c) the fabricated
low-EMI neuroSFAT, and (d) the microscopic image validating the translucency of low-EMI
neuroSFAT .................................................................................................................................. 103
Figure 8. 5 (a) Conceptual illustration and (b) photo of the whole-cell patch clamp experiment
setup with the presence of the neuroSFAT.................................................................................. 104
Figure 8. 6 The recorded action potentials (black) along with the activation of focused
ultrasound stimulation (blue). FUS is on for 30 seconds, followed by 2 minutes of resting
period. ......................................................................................................................................... 106
Figure 8. 7 Photos of (a) acute slices from 3-4 weeks old Sprague Dawley rat, (b) hippocampus
(red box in (a)) region of the slice, (c) cultured hippocampal slices from an 8-days(P8) old
mouse, and (d) a magnified photo of single hippocampal slice from a P8 mouse. .................... 107
Figure 8. 8 (a)The illustration of the EMI recording setup in the patch clamp chamber, filled
with an ACSF solution. The recording electrode measures the EM, generated by the SFAT and
propagated through the ACSF solution. (b) The measured EMI with a low-EMI neuroSFAT
(Ch2, blue) and the driving voltage of the low-EMI neuroSFAT (Ch1, Yellow). The amplitude
of EMI is measured as about 8% of that of driving voltage. ...................................................... 108
xvi
Figure 8. 9 (a) The EMI level picked up during the patch clamp experiment varying the
amplitude of driving voltage. The EMI level exceeds the recording range when the driving
voltage is higher than 50Vpp. (b) The recorded EMI (large peak) along with the neuron’s
action potential (small peak), showing that the EMI level dominates the action potential,
hindering the recording of neuronal activity............................................................................... 108
Figure 8. 10 Temperature rise on the surface of the brain tissue with the activation of (a) the
active neuroSFAT and (b) the control neuroSFAT. ..................................................................... 110
Figure 8. 11 (a) The illustration of the Injected current profile (-50pA to 40pA for 200ms), the
measured membrane potential changes (b)without and (c) with the focused ultrasound
stimulation, and (d) I -V curve for both cases .............................................................................111
Figure 8. 12 (a) The inhibition effect of the Focused Ultrasound Stimulation (FUS) with the
parameters set at 40Vpp (ISPPA = 0.94W/cm2
), 45kCycles/pulse (PD =2.45ms) and a varied
PRF of 50-100-150-200Hz in order. (b) The excitation effect of FUS with the parameters
set at 120Vpp (ISPPA = 8.44W/cm2
), 45kCycles/pulse (PD =2.45ms) and a varied PRF of 10-
15-20-25-50Hz............................................................................................................................ 113
Figure 8. 13 Thermosensitive ion channel family with their activating temperature. Excitatory
and inhibitory neurons are represented by round and square shapes, respectively, and the color
indicates the gradient of their activation temperatures. .............................................................. 114
Figure 9. 1 (a) Top-view photo of a fabricated SFAT on PZT-4 substrate. Finite element
analysis (FEA) simulated normalized acoustic pressure (b) on the lateral plane at z= 5 and
(c) on the vertical plane along with the central axis ................................................................... 119
Figure 9. 2 Experiment setup for RPE cell extraction with SFAT mounted on the
laser-machined holder and immersed in PBS solution. A petri dish with RPE cells is held by
the movable stage connected to a 5-axis micro-manipulator and a collecting plate................... 121
Figure 9. 3 Microscope photos of 100% confluency human retinal pigment epithelium (RPE)
monolayer cells (a) before and (b) after an ejection of cells by SFAT. (c) Photos of the same
monolayer cells when the cells are re-cultured (for 4 days) after the cell ejection, new cells
filling the space and showing no damage on the edge of the ejection site. Top and bottom
photos are at low and high magnification, respectively.............................................................. 122
Figure 9. 4 Fluorescent microscope images showing the viability of the RPE monolayer (a)
before and (b) after the focused ultrasound cell ejection. A large ejected spot
(1-mm in diameter) on the monolayer is indicated with white arrow and the black shaded
areas are the targeted areas labeled with a permanent marker on the bottom face of the petri
dish.............................................................................................................................................. 123
Figure 9. 5 Bright-field microscope image (a) before and (b) after delivering focused
ultrasound (276.5Vpp, PRF of 10Hz and pulse width of 200µs) to extract RPE cells from a
spot of about 100µm in diameter. Fluorescent microscope images with the same acoustic
xvii
pulses (c) before and (d) after the cell ejection. White arrows indicate the ejected spots on
the monolayer.............................................................................................................................. 124
Figure 9. 6 Bright-field microscope images with two different sizes of ejected spot by
controlling the number of pulses delivered to the RPE monolayer: (a) with 55 pulses
(resulting in about 5 RPE extracted cells), (b) with 105 pulses (resulting in about 10 ejected
cells. (c) Two different sizes of the ejected spot on the same petri dish are made with SFAT
being applied with 55 pulses first (right arrow), being moved laterally 1-mm to the left, and
then being applied with 105 pulses (left arrow) to eject about 10 RPE cells. ............................ 124
Figure 9. 7 (a) Bright-field and (b) fluorescent microscope images of a series of single-cell
ejections (white arrows) by a single pulse (268Vpp and pulse width of 248µs). The location of
SFAT is changed by 127µm every one second, synchronized with the pulse repetitive
frequency (1 Hz) for clear evidence of single-cell ejection........................................................ 125
xviii
Abstract
Ultrasound technology has played significant roles in the field of biomedical engineering,
offering multipurpose applications in both diagnostics and therapeutics. While its current
prominence predominantly lies in diagnostics, there exists vast untapped potential for its
application in disease treatment, cell sorting, and manipulation. This thesis embarks on a
comprehensive exploration of an integrated wireless piezoelectric ultrasonic transducer system for
biomedical applications. The objective is to enhance the understanding of ultrasound's diverse
applications and the development of wireless ultrasound transducer system integrated with an
electronic driving circuit and wireless connectivity and wireless power transfers that push the
boundaries of what is possible in contemporary biomedical engineering.
The introductory chapter provides a panoramic view of ultrasound's journey in healthcare.
It begins by highlighting its pivotal role in diagnostic imaging. The discussion extends to
therapeutic ultrasound, where high-intensity focused ultrasound (HIFU) is employed to ablate
pathological tissues such as cancer cells. The emerging applications of ultrasound for drug delivery
and Low-intensity focused ultrasound (LIFU) non-invasive neuromodulation underscore the
technology's potential to revolutionize healthcare.
The subsequent chapters delve into the core of this thesis's research, commencing with an
exploration of acoustic transducers fabricated with microelectromechanical system (MEMS)
technology. These components, which generate ultrasound waves, are fundamental to ultrasound
systems. The scope of the thesis extends to the design, fabrication, and characterization of
specialized acoustic transducer systems. These systems include Self-Focusing Acoustic
Transducers (SFAT), electrically Controllable SFAT (eCON SFAT), and Spiral-arm Vortex-beam
Acoustic Transducers (SVAT). Each chapter provides a deep dive into the intricacies of these
xix
transducer systems, exploring their design parameters, fabrication techniques, and methods for
characterization.
The role of electronic circuits in driving and controlling acoustic transducers is examined
in detail. The thesis explores electronic systems designed for continuous and pulsed operation,
highlighting their significance in the proposed integrated wireless piezoelectric ultrasonic
transducer system for biomedical applications. For continuous operation with low operating
voltage (30Vppmax), a power amplifier and DC to DC converter are used, while Z-source network
with GaNFET (Gallium Nitride Field Effect Transistor) provides higher output voltage (60Vpp)
with a limited stability. The integration of Bluetooth Low Energy (BLE) technology for wireless
control is also addressed, emphasizing the adaptability and connectivity of the system.
Subsequent chapters shift the focus towards practical applications of the transducer systems.
The development of wireless integrated systems is explored, catering to scenarios such as
submarine propulsion, particle (or cell) levitating acoustic tweezers, contactless single-cell
extraction, and focused ultrasound neuromodulation. Experimental results and discussions for each
application highlight the potential of these systems in real-world scenarios.
The thesis concludes in a summary chapter, offering an overview of the research findings
and their implications for the future. It sets the stage for potential directions in the field of
ultrasound technology, emphasizing the ongoing innovation and interdisciplinary collaboration
that drive advancements in biomedical applications.
1
Chapter 1
Introduction
1.1 Applications of Ultrasound in Biomedical
An acoustic wave with a frequency above the hearing range (20–20,000 Hz) is referred to
as ultrasound. Ultrasound technology has been utilized to improve human lives in a variety of
sectors. One of the most prominent applications of the ultrasound technology is biomedical
application. The first use of ultrasound in a medical setting dates back to 1929, when Austrian
physicist Karl Dussik discovered that high-frequency sound waves could be used to visualize
the heart. Dussik's invention, called the "supersonic echolocation device," was the first
practical application of ultrasound in medicine [1]. Since then, ultrasound has become an
essential tool in a wide range of medical applications, including obstetrics and gynecology [2],
musculoskeletal imaging [5], vascular imaging [6], etc. However, ultrasound has found
applications beyond diagnostic imaging and is widely used in various biomedical fields,
including therapeutic and surgical applications. For example, in the realm of cancer treatment,
ultrasound plays a crucial role in high-intensity focused ultrasound (HIFU) therapy. This
innovative therapeutic approach utilizes focused ultrasound waves to target and ablate tumors
non-invasively [7]. HIFU has shown promise in treating certain types of cancers, such as
prostate cancer [8] and brain tumor [9], offering a less invasive alternative to traditional
surgical procedures. Additionally, ultrasound technology is increasingly integrated into
cardiovascular medicine [3][4], where it aids in the visualization of blood flow and assists in
procedures like echocardiography for assessing heart function. Furthermore, ongoing research
explores the use of ultrasound for drug delivery, harnessing its ability to enhance the
2
permeability of cell membranes, thereby improving the effectiveness of targeted therapies
[10][11]. Ultrasound can also deliver power and communicate wirelessly with sub-mm-sized
devices implanted deep in tissue [12]. Additionally, focused ultrasound is utilized in research
for non-invasive neuromodulation [13] and cell sorting and handling [14].
1.2 Overview of Acoustic Transducer
The acoustic transducer is a device that is designed to convert electrical energy into
acoustic waves or vice versa. The conversion can be achieved by electromagnetic, electrostatic,
and piezoelectric effects.
The electromagnetic acoustic transducer (EMAT) operates on the principles of
electromagnetic induction. EMAT is consist of a coil fed by a large dynamic current and an
electromagnet providing a static magnetic field. EMATs utilize a powerful electromagnetic
field to induce eddy currents within a conductive material, typically a metallic surface. These
induced eddy currents, in turn, act as sources of ultrasonic waves that traverse through the
material [15]. One of the key advantages of EMAT technology lies in its non-contact nature,
eliminating the need for direct coupling or special surface preparation during inspections. This
attribute makes EMATs particularly advantageous in non-destructive testing (NDT) application
across diverse industries [15]. In manufacturing, EMATs play a significant role in inspecting
the quality of materials and welds without compromising the integrity of the components [17].
EMATs are used in the aerospace industry to evaluate structural components and guarantee the
dependability and safety of vital aircraft elements [18]. Similar to this, EMATs may be used in
transportation to examine pipelines and railroad tracks, providing a non-invasive technique to
find defects and guarantee the lifespan of infrastructure [19]. While electromagnetic acoustic
3
transducers offer several advantages in non-destructive testing (NDT), they also come with
certain disadvantages such as limited sensitivity for the thin materials and limited depth of
penetration [20]. Furthermore, EMAT is effective with metallic materials due to their ability to
induce eddy currents in conductive substances and non-metallic materials can only be used to
a limited extent.
The electrostatic acoustic transducer, on the other hand, operates on the fundamental
principle of electrostatic induction. The electrostatic acoustic transducer consists of a
diaphragm with a conductive coating, positioned between charged plates. The application of
an electrical signal results in the modulation of electric fields, causing the diaphragm to vibrate
and produce sound waves [21]. The ability to reproduce high-fidelity audio, with accurate and
clear sound across a broad frequency range, is one of the most noticeable benefits of
electrostatic transducers. Additionally, they exhibit low distortion, making them desirable in
condenser microphone and high-quality speakers [22]. However, capacitive micromachined
ultrasonic transducers (CMUTs) emphasize the application of electrostatic acoustic transducer
in medical imaging and diagnostic sector by reducing in size and enabling array creation with
the aid of MEMS technology [23][24]. Also, electrostatic transducers present certain
challenges. They typically require a high voltage bias, which can impact their power efficiency.
Furthermore, they may have limited low-frequency response compared to other transducer
types [25].
Lastly, piezoelectric acoustic transducer operates on the foundational principle of the
piezoelectric effect. When the piezoelectric materials are subjected to an electrical field, these
materials undergo mechanical deformation, producing vibration. On the other hand, when they
undergo mechanical deformation, they produce electrical charges that result in an electrical
4
potential difference [26]. This bidirectional capability allows piezoelectric transducers to
seamlessly convert electrical signals into acoustic waves and vice versa. Piezoelectric acoustic
transducers come in various types, each tailored for specific applications and operational
requirements. One common type is the bulk thickness mode resonance transducer, where the
piezoelectric material is shaped into a thin disc with top and bottom electrodes. When an
electrical signal is applied to the electrodes, the disc vibrates in its thickness mode, producing
acoustic waves. Some piezoelectric materials such as quartz, lithium niobate (LiNbO3), lead
zirconate titanate (PZT), lead magnesium niobate-lead titanate (PMN-PT), and polyvinylidene
difluoride (PVDF) have the properties of low electrical and mechanical losses and high
electromechanical coupling coefficient, which make them to be desirable for bulk thickness
mode resonance transducers [27]-[30]. Especially, PZT transducers are commonly used due to
their excellent piezoelectric properties [31]. These transducers can operate in various modes,
including thickness mode [32]- [34] and radial mode [35][36], providing versatility in different
applications.
Another notable type is the Piezoelectric Micromachined Ultrasonic Transducer (PMUT).
PMUTs consist of small diaphragms that can flex in response to an applied voltage, generating
ultrasonic waves[37][38]. PMUTs utilize thin-film piezoelectric materials, including zinc
oxide (ZnO) [39], aluminum nitride (AlN) [40], and PZT [41], chosen for their compatibility
with microfabrication processes. Zinc oxide, recognized for its biocompatibility and
transparency, holds promise for applications in medical imaging [42], while aluminum nitride's
superior mechanical quality (Q) factor makes it ideal for high Q resonator applications [43].
Additionally, the versatility of PZT, a well-established piezoelectric material, ensures
adaptability across various operational modes [44]. The development of PMUTs using thin-
5
film piezoelectric materials emphasizes the continuous efforts to improve microscale
transducer technology, leading to innovation in industries such as high-frequency sensing and
medical diagnostics.
1.3 Overview of Driving Acoustic Transducers
Alongside the growth of transducer technology, there have been notable developments in
the history of driving electronics for acoustic transducers. In the early days, mechanical
systems were often employed to drive acoustic transducers, relying on mechanisms like tuning
forks or diaphragms connected to electrical circuits[45]. However, with the advent of electronic
components, the field transitioned towards more sophisticated methods. An important turning
point in the mid-1900s was the development of vacuum tube amplifiers, which allowed
electrical signals to be amplified and used to drive transducers with higher power and
efficiency [46]. The driving electronics were further revolutionized by the following invention
of integrated circuits (IC) and power amplifiers, which provided improved efficiency,
reliability, and compactness [47].
A power amplifier circuit driving acoustic transducers often needs to supply the high peak
pulse currents. A power amplifier can be classified by its circuit configuration and the operation
method, and can be categorized into two groups. The first group is the more conventional
amplifier classes including class A, B, AB, and C, which are derived from the traditionally
controlled conduction angle amplifiers. These traditional amplifier classes operate by
controlling the conduction angle of their output transistors, determining the portion of the input
signal cycle during which the transistor conducts [48]. The output transistor in Class A
amplifiers conducts throughout the entire input signal cycle. This continuous conduction
6
provides high linearity and low distortion, making Class A amplifiers ideal for applications
where audio quality is critical. However, the disadvantage of Class A amplifiers is low
efficiency, as they consume power constantly even when there is no input signal. At the same
time, they generate significant heat due to the constant current flow [49]. Class B amplifiers,
on the other hand, consist of two transistors and divide the input signal cycle into positive and
negative halves. Each output transistor conducts during only one half of the cycle, improving
efficiency compared to Class A. However, two transistors in class B amplifiers introduce
crossover distortion at the transition between positive and negative halves [50]. Class AB
amplifiers are designed to aim the balance of the benefits between Class A and Class B. They
allow a slight overlap in the conduction of both transistors, reducing crossover distortion while
maintaining improved efficiency compared to Class A. The conduction angle is typically
between 180 and 360 degrees [51]. Lastly, Class C amplifiers is designed to achieve the highest
efficiency among the traditional amplifier group by biasing the transistor in a cut-off region.
Class C amplifiers have a brief conduction angle, operating for less than half of the input signal
cycle which is typically around the 90 degrees. Class C amplifiers have an efficiency of around
80%, but because of their short conduction angle, they also cause a large amount of distortion
and the poor linearity [52][53].
The second group, on the other hand, includes the "switching" amplifier classes, such as
Class D, E, F, G, S, T, etc. These newer classes leverage digital circuits and pulse width
modulation (PWM) techniques to continually switch the signal between the extremes of "fullyON" and "fully-OFF." This dynamic switching mechanism propels the output forcefully into
the saturation and cut-off regions of the transistors. This switching mechanism allows to
achieve high efficiency by minimizing power dissipation, making them more energy-efficient
7
than traditional classes. Theoretically, they can achieve 100% power efficiency but, in practice,
80-90% of power efficiency is achieved due to the switching loss associated with their
switching frequency [54][55]. Among the mentioned amplifier classes, Class D is the most
popular and widely used, especially in the realm of audio amplification. Class D amplifiers
have gained significant popularity due to their high efficiency, making them suitable for a
variety of applications where power conservation and minimal heat generation are crucial
considerations [56]. Class S amplifiers, on the other hand, are similar to Class D in terms of
using a switch mode operation but they employ a sigma-delta modulation technique for more
precise control, improving the signal distortion [57]. While less common and not as widespread
as Class D, Class S amplifiers are also used in driving acoustic transducers and recently the
gain of 18.35dB at 25MHz has been achieved by using two stages of Class S amplifier [58] .
Lastly, another way of driving acoustic transducer is using an inverter which converts direct
current (DC) power into alternating current (AC). Inverters are often more energy-efficient
than linear power amplifiers, resulting in reduced heat generation and making them suitable
for applications where power conservation is crucial. Additionally, inverters provide precise
control over the frequency and amplitude of the output signal. However, inverters can introduce
harmonic distortion due to the nature of waveform conversion, potentially impacting the
destruction of the transducer [59]. Furthermore, the process of converting low DC voltage into
high AC voltage through inverters inherently demands a substantial input current, limiting the
power source and choice of passive components such as inductor and diode.
8
1.4 Scope of the Thesis
This thesis covers the design and microfabrication of piezoelectric acoustic transducers and
the development of electronic circuits driving the acoustic transducers, followed by their
applications in biomedical field (non-invasive focused ultrasound neuromodulation and single
cell ejection) and acoustic propulsion and manipulation.
The first chapter introduces an overview of the application of acoustic transducers in the
biomedical field and various types of acoustic transducers. In addition, the history and different
method of the electronic circuit to drive the acoustic transducers are described in the chapter
1, followed by the background and the motivation of this thesis.
Chapters 2 – 4 cover various designs of acoustic transducers with their working principles,
fabrication methods and design parameters for different applications. In chapter 2, the design,
fabrication, and characterization of Self-Focusing Acoustic Transducers (SFAT) has been
described. SFAT is able to generate a focused ultrasonic wave at a single point with two
different fabricating methods using the patterned electrodes and the acoustic lens. Chapter 3
introduces the electrically controllable SFAT, which is a variation of SFAT with multiple focal
points which directions and sizes can be controlled by activating electrode ring patterns.
Chapter 4 demonstrates the design, simulation, and fabrication of a Spiral-arm Vortex-beam
Acoustic Transducer (SVAT), which generates an acoustic vortex in the media.
Chapter 5 presents two types of wireless electronic driving circuit; one is continuous mode
operation with a commercially available power amplifier and the other is pulsed mode
operation designed by a discrete MOSFET to generate AC voltage with a DC input.
Chapters 6 - 9 describe various applications of acoustic transducers and their electronic
circuits. Chapter 6 demonstrates the acoustic propulsion application using SFAT and SVAT. In
chapter 7, the application of SVAT in a levitating acoustic tweezer is presented. Chapter 8
9
describes the use of SFAT in a non-invasive focused ultrasound neuromodulation application
to stimulate CA1 neurons in rat’s hippocampus region. Lastly, Chapter 9 shows another use of
SFAT in a single cell ejection application.
Finally, Chapter 10 concludes and summarized the thesis, followed by introducing the
remaining work in the future.
10
Chapter 2
Design, Fabrication, and Characterization of Self-Focusing Acoustic
Transducers (SFAT)
2.1 Introduction
Focused ultrasound has emerged as a promising tool in various biomedical applications for
its ability to selectively target tissues with high precision, allowing for both imaging and
therapeutic applications [60]. In the field of biomedicine, this technology has demonstrated
incredible promise in the treatment of a wide range of disorders, from neurological diseases[61]-
[63] to cancers [64]-[66]. Unlike traditional surgical procedures, focused ultrasound offers a noninvasive approach, minimizing the risks and recovery times associated with conventional
interventions. One of the most significant applications of focused ultrasound in biomedicine is
tumor ablation. Focused ultrasound transducers can be precisely directed to deliver thermal [67]
or mechanical energy [64] to cancerous tissues, leading to their destruction. This approach has
shown promise in treating solid tumors, providing a viable alternative or complement to traditional
cancer therapies. Focused ultrasound has also shown potential in the treatment of neurological
disorders, including epilepsy [68] and Parkinson's disease [69]. By precisely modulating neural
activities, focused ultrasound can alleviate symptoms and improve the quality of life for
individuals suffering from these conditions. The ability of focused ultrasound to temporarily
disrupt the blood-brain barrier or blood clogs has opened new avenues for targeted drug delivery
11
to the brain [70]. This technique enhances the effectiveness of therapeutic agents in treating
neurological conditions and central nervous system disorders.
Several methods are explored to generate focused ultrasound and each design has its unique
advantages and applications. The most straight forward method is using an acoustic lens, utilizing
curved surfaces to refract and converge ultrasound waves at a specific focal point. This lens-based
approach is effective for achieving high-intensity focused ultrasound (HIFU) but the fabrication
of curved lens is complex, causing the curvature errors [71]. Another approach involves phased
array transducers, consisting of multiple small elements that can be individually controlled to
produce and steer ultrasound beams [72]. This dynamic control enables rapid adjustments of the
focal point, making phased array transducers particularly versatile for applications requiring realtime adaptability. However, phased array transducers require multiple electrical connections to
each electrode, hence involved a complicated packaging technique such as a flip chip bonding [73].
Alternative method of achieving focused ultrasonic wave is using a Fresnel lens, which is
commonly used in optics to concentrate light [74]. Self-focusing Acoustic Transducer (SFAT) can
be microfabricated based on annular-ring Fresnel lens design to generate a single focus [75] or
multi-foci [76]. The most significant advantage of SFAT lies in its compact form factor, a feature
seamlessly compatible with MEMS fabrication processes, enabling efficient mass production.
Furthermore, SFAT offers a high degree of flexibility in beamforming and focusing capabilities,
allowing for precise control of ultrasound waves.
2.2 Working Principle of SFAT
A SFAT can be divided into two critical components: an ultrasonic sound source
responsible for the generation of ultrasound waves, and an acoustic lens designed for their precise
focalization. The sound source is constructed from a bulk sheet of piezoelectric material,
12
positioned between its top and bottom electrodes. Commonly employed substrate piezoelectric
materials include PZT, such as PZT-5A and PZT-4, chosen for their high piezoelectric coefficients
and electromechanical coupling coefficients. These materials exhibit high vibration amplitude
under an applied voltage, efficient conversion between electrical and acoustic energy, and can
tolerate a fabrication processing temperature up to their Curie temperature (about 350 °C). When
a sinusoidal electrical signal, whose frequency is matched with the thickness-mode resonance of a
PZT thin substrate, is applied between the top and bottom electrodes, the PZT vibrates and
generates acoustic waves by most efficiently converting electrical energy into acoustic energy [77].
Focusing, on the other hand, can be achieved by patterning the electrodes into annular Fresnel halfwavelength band (FHWB) rings so that all the acoustic waves produced by the voltage applied
between the patterned top and bottom electrodes can be collected at a focal length (F) with inphase (0 – 180 °) [78]. In other words, at each ring’s boundary is designed so that the distance to
the focal point is equal to integer multiples of the half wavelength (�/2) in the medium plus the
focal length (F) (Figure 2.1).
Figure 2. 1 Conceptual view of focused ultrasound generated by SFAT, defining the focal size, focal depth, focal
length, half angle (θ), ring radius (R!), and the width of the outermost ring (ΔR)
13
The radii of rings corresponding to a desired focal length (F) can be determined using the following
equations:
C�!
" + �" = � + � �/2 (� = 0, 1, 2,⋅⋅⋅) (2.1)
From this expression, we obtain:
�! = C� � × (� + � �/4) (� = 0, 1, 2,⋅⋅⋅) (2.2)
where �! is the radius of nth ring, � is a focal length, � is a wavelength of acoustic wave in the
media.
The Fresnel rings can easily be obtained through patterning the top and bottom electrodes
into Fresnel annular rings (Figure 2.2), but also be made out of Parylene-sealed air-cavity Fresnel
annular rings with the top and bottom electrodes patterned into a pair of circles (Figure 2.3). The
air-cavity lens exploits the high reflection coefficient (99%) between air and any solid material,
attributed to the substantial acoustic impedance difference between air (0.4MRayl) and solid (over
1MRayl) [79] so that all acoustic waves propagating through the air-cavity parts are reflected back
and do not contributes to the focusing.
Figure 2. 2 (a) The cross-sectional diagram (across the AA’ dashed line in (b)) showing a typical SFAT based on a
PZT substrate with patterned Fresnel annular-rings electrodes. This illustration highlights how the Fresnel acoustic
lens focuses ultrasound in the medium by constructively interfering acoustic waves in-phase. (b) The top-view diagram
of the same SFAT, showing the relative positions of the top electrode, and soldering pad.
14
Figure 2. 3 (a) The cross-sectional diagram (across the AA’ dashed line in (b)) showing a typical SFAT based on a
PZT substrate with a Fresnel air-cavity acoustic lens made of Parylene. This illustration highlights how the Fresnel
acoustic lens focuses ultrasound in the medium by obstructing destructively interfering acoustic waves (by the total
reflection from the air-cavity).(b) The top-view diagram of the same SFAT, illustrating the relative positions of the top
electrode, air-cavity rings, soldering pad, and Parylene-coated regions.
2.3 Design parameters of SFATs
2.3.1 Operating Frequency
A SFAT produces the largest acoustic waves when a sinusoidal voltage at the fundamental
thickness-mode resonance frequency of the piezoelectric substrate is applied between its top and
bottom electrodes. The thickness-mode resonance frequency, denoted as �!, is a function of the
piezoelectric substrate's thickness [79]:
�! = L
#
"
+ �M × $
% (� = 0, 1, 2,⋅⋅⋅) (2.3)
where � and � represent the sound velocity within the piezoelectric substrate material and the
thickness of the substate, respectively. Typically, the frequency is determined by the requirements
of the specific applications, and the appropriate thickness of the piezoelectric sheet is chosen to
ensure that its thickness-mode resonance frequency matched with the target frequency.
15
2.3.2 Focal Diameter (Focal Size)
Focal diameter or focal size is defined as twice the lateral distance from the center to the
point where the acoustic pressure decreases to -3dB of the peak pressure (Figure 2.1). Focal size
created by the Fresnel annular-ring lens is approximately equal to the width of the outermost ring
(Δ�!) [80]. From equation 2.2, we obtain
�!
" − �!
" = �� + (2� − 1)�"/4 (2.4)
Assuming that F is much longer than the wavelength of the acoustic wave in the medium (which
is true for most of the application), we can approximate the equation 2.4 as
�!
" − �!
" ≈ �� (2.5)
By substituting �! = �! − Δ� into the above equation, we get
�!
" − (�!
" − Δ�) = 2�!Δ� − (Δ�)" ≈ �� (2.6)
For n > 3 (the number of rings is greater than 3), the condition �! ≫ Δ� holds true, allowing the
term Δ�" being neglected, and
Δ� ≈ �� /2�! (2.7)
Upon revisiting the equation 2.2 and with � ≫ �, �! can be simplified as
�! ≈ √��� (2.8)
Finally, by comparing equation 2.7 and 2.8, we can derive the following equation for Δ� as a
function of the operating frequency:
Δ� ≈ �� /2√��� = C��/4� = C��/4�� (2.9)
In equation 2.9, we use the relationship � = �/ �, where c and f are the sound velocity in the
medium and operating frequency, respectively. As described earlier, as the focal size can be
approximated by the width of its outermost ring (Δ�), the focal size decreases with a higher
16
operating frequency (�), if the device is designed to have the same number of rings and focal
length and is operated in the same medium.
2.3.3 Focal Length
The focal length (F) can be derived as follows [81]. Since �� ≈ �!
"/�, equation 2.7 can be
expressed as
� ⋅ Δ� ≈ �� ≈ �!
"/� = �"/4� (2.10)
or
� ≈ 4� ⋅ Δ� (2.11)
where D is the diameter of Fresnel lens, which is equal to 2�! (� = 2�!). Thus, we obtain
� ≈ 2�!Δ�/� = � ⋅ Δ�/� (2.12)
Using equation 2.11, we turn equation 2.12 into
� ≈ 4� ⋅ (Δ�)"/� = 4� ⋅ (Δ�)" ⋅ �/� (2.13)
This relationship suggests that the focal length can be increased by increasing the operating
frequency, number of rings, and width of the outermost ring.
2.3.4 Depth of Focus (Focal Depth)
Numerical aperture (��) of a Fresnel lens is [81]:
�� ≡ sin � ≈ �!/� (2.14)
where � is the half angle measured form the central axis at the focal point back to the outermost
ring of the lens (Figure 2.1). From equation 2.7, we obtain:
�� ≈ �/2Δ� (2.15)
F-number (�#) of a lens, on the other hand, is defined as the focal length divided by the diameter
of the lens, and is
17
�# = �/� ≈ Δ�/� (2.16)
By comparing equations 2.15 and 2.16, we can obtain
�� ≈ 1/2�# (2.17)
Depth of focus (DOF) or focal depth is, then, defined as the axial length of two points
where their acoustic pressure decreases by 20% from the peak acoustic pressure at the center (Z=F)
(Figure 2.1) [81]
��� = Δ� = ± #
"
(
(*+)" (2.18)
Using equations 2.16 and 2.17, we obtain
��� = ±2(�#)"� = ±2(Δ�)"/� (2.19)
This relation implies that a lens with a larger numerical aperture has a short DOF and DOF is
proportion to the square of the outermost ring width and inversely proportional to the wavelength
of the acoustic wave in the medium.
2.4 Simulation of Acoustic Pressure
The distribution of acoustic pressure generated by an SFAT can be simulated using the
finite-element method (FEM) analysis [82]. This simulation can be performed in the frequency
domain using the Pressure Acoustics module of COMSOL Multiphysics 5.6 (COMSOL Inc.). The
simulation is modeled with a two- dimensional (2D) axial symmetry to significantly reduce the
computation time and memory, as SFAT has a symmetrical design. This approach allows modeling
only half of the volume cross-section and the complete simulation results can be reconstructed by
mirroring the simulated data along the central vertical axis (R = 0). Moreover, instead of modeling
piezoelectric material (PZT) and Fresnel air-cavity lens, only the volume of the medium where the
actual acoustic wave propagates is considered. The vibration of the piezoelectric sheet is modelled
as a normal displacement boundary condition applied to the Fresnel annular rings electrode. The
18
simulation employs a free triangular mesh, with the maximum element size set to be �/5 ,where
� is the wavelength of acoustic wave in the medium [83]. Additionally, the propagating medium
is assumed to possess isotropic and homogeneous properties for the sake of simplification and
spherical wave radiation boundary conditions is applied at the outer boundaries of the simulation
volume, where acoustic waves could pass through without reflection.
Figure 2.4 illustrates the Finite Element Method (FEM) analysis of the normalized acoustic
pressure distribution in brain tissue, where a sound velocity is approximately 1540 m/s [84]. The
SFAT is designed to operate at a frequency (�) of 18.4 MHz with a wavelength in the medium
calculating 84 µm. This specific SFAT design features a focal length (�) of 400 µm, 15 annular
rings (� = 15) and the width of the outermost ring (ΔR) is calculated as 42 µm. The simulation
result (Figure 2.4) indicates a focal size of 46µm and a focal depth of 95 µm. Upon comparison
with theoretical calculations from equations 2.9 and 2.19, the expected values for focal size and
focal depth are 47 µm and 84 µm, respectively.
Figure 2. 4 Finite Element Method (FEM) analysis of the normalized acoustic pressure distribution of SFAT, which
operating frequency is 18.4Mhz, 15 rings, and focal length of 400µm, in the brain tissue media. The simulation result
shows that the focal size of 46µm and focal depth of 95µm.
19
2.5 Microfabrication Methods for SFATs
As described in chapter 2.2, Fresnel annular-ring lenses can be fabricated through two
methods: one involves patterned electrodes, and the other utilizes an air-cavity Parylene lens. The
use of patterned electrode rings offers the advantage of straightforward and simple fabrication.
However, it’s peak acoustic pressure at the focal point is 2.6 times weaker compared to that
achieved with air-cavity lenses in one comparison study [85].
2.5.1 Fabrication of SFAT with Patterned Electrode Rings
Figure 2.5 illustrates the fabrication steps of SFAT with patterned electrode rings. The
fabrication process begins with a pre-deposited nickel layer on both the top and bottom of PZT
(Figure. 2.5a). Subsequently, AZ 5214 photoresist (PR) is spin-coated at 1,200 rpm to pattern the
Fresnel half-wavelength annular rings. The photoresist has a thickness of approximately 3 μm and
is developed using AZ400K developer (Figure. 2.5b). Front and back alignment are facilitated
through infrared (IR) imaging, made possible by the thin substrate thickness. For a thick PZT sheet
(over 200µm), front and back alignment roughly can be achieved by aligning one corner of PZT
sheet to the reference corner on the photomask for both top and bottom electrodes. Following this,
the nickel layers are wet-etched using a nickel etchant TFG from Transene (Figure 2.5c). To enable
operation in water and enhance biocompatibility, a thick layer of Parylene D is deposited as an
insulation layer (Figure 2.5d). The Parylene deposition involves two steps: initially, a 20 µm thick
layer is coated, covering the electrode solder pad areas with a Kapton tape. After the first deposition,
the Kapton tapes are removed, and wires are soldered to the pads, followed by a second Parylene
coating with a thickness of 8 µm.
20
Figure 2. 5 Fabrication steps of SFAT with a patterned rings;(a) PZT sheet sandwiched by nickel(Ni) layers (b) Mask
pattern for the Fresnel half wavelength band annular rings (c) Wet-etching Ni electrodes and remove the photoresist
mask and (d) Parylene D coating for the water proof and biocompatibility
2.5.2 Fabrication of SFAT with Air-Cavity Lens
Figure 2.6 demonstrates the fabrication steps of SFAT with an air-cavity lens. The
fabrication process also initiates with a pre-deposited nickel layer on both the top and bottom of
PZT sheet (Figure. 2.6a). Similar to the SFAT with patterned electrodes, AZ 5214 photoresist is
spin-coated at 1,200 rpm and subsequently wet-etched to pattern circular electrodes on the top and
bottom surfaces (Figure 2.6b-c). Front and back alignment procedures are achieved as detailed in
Chapter 2.5.1. Subsequently, a thin layer of Parylene D is coated as a protective layer (Figure 2.6d).
Another layer of AZ 5214 photoresist is deposited on top of the Parylene layer to pattern Fresnel
half-wavelength annular rings, serving as a sacrificial layer (Figure 2.6e). Following this, an
additional 4 µm thick layer of Parylene D is deposited (Figure 2.6f). The next step involves the
patterning of release holes through O2 reactive ion etching (RIE) to expose the photoresist
sacrificial layer, which is then removed using acetone (Figure 2.6g). Finally, a 20 µm thick layer
of Parylene D is deposited once again to seal the release holes and serve as insulation (Figure 2.6h).
21
Figure 2. 6 Fabrication steps of SFAT with an air-cavity lens;(a) PZT sheet sandwiched by nickel(Ni) layers (b) Mask
pattern for the top and bottom circular electrodes (c) Wet-etching Ni electrodes and remove the photoresist mask (d)
Parylene D coating as a protective layer (e) Depositing photoresist sacrificial layer (f) Another Parylene D coating (g)
RIE etching release holes to open access to the sacrificial layer and (h) Depositing thick Parylene D layer for insulation
and biocompatibility
2.6 Characterization of SFAT
2.6.1 Measurement of Electrical Impedance and Resonant Frequency
The electrical impedance and the resonant frequency of SFAT can be estimated by
measuring one-port S11 parameter, which can be converted into the impedance using below
relationship [86]:
� = 50 Ω ×
#-.##
#&.##
(2.20)
For a piezoelectric ceramic element, such as PZT, two sets of resonant frequencies exist;
the “anti-resonant frequency” or “parallel resonance” and the “resonant frequency” or “series
resonance” [87]. At both resonance and anti-resonance, the phase of the impedance is zero, which
gives only a real resistance value [88]. Figure 2.7 shows the magnitude and phase of S11 of SFAT
measure with a Network Analyzer (8753d, Hewlett Packard Inc.). SFAT used in the measurement
is fabricated on 1-mm-thick PZT (which thickness-mode resonant frequency is 2.32MHz based on
equation 2.3) with a focal length of 5-mm and 6 rings. The frequencies where the reactance is zero
22
are found in to be 2.19MHz and 2.25MHz, which are corresponding to series resonance (�/) and
parallel resonance (�0), respectively.
Figure 2. 7 Magnitude (black) and phase (red) of one-port S11 measurement of 2.32MHz SFAT. Resonant frequency
(or series resonance, fs) and anti-resonant frequency (or parallel resonance, fp) are found at where the reactance values
are zero, allowing to be estimated as 2.19MHz and 2.25MHz, respectively.
Near at its resonant frequency, PZT can be modeled with an equivalent circuit. Various
piezoelectric transducer circuit models have been proposed such as Mason's model [89],
Redwood's model [90] and the Butterworth-Van Dyke (BVD) model [91][92]. The BVD model is
one of the simple circuit models and is composed of motional inductance ( �1 ), motional
capacitance (�1), motional resistance (�1) (which represents a mechanical motion of SFAT) and
clamp capacitance (�2) [88] (Figure 2.8)
Figure 2. 8 (a) Perspective view of a piezoelectric substrate (PZT) defining the area A and the thickness d. (b) BVD
equivalent model of SFAT showing motional inductance (L$), motional capacitance (C$), motional resistance (R$)
and clamp capacitance (C%)
23
The clamp capacitance (�2) takes into consideration the presence of a dielectric material, though
piezoelectric, with dielectric permittivity (�.) and thickness (�). This material is sandwiched
between two electrodes over an area (�), resulting in a capacitance equal to:
�2 = �.
+
3 (2.21)
While it is theoretically possible to calculate �1 , �1 , and �1 , these parameters can also be
estimated by using the relationship with series (�/) and parallel (�0) resonant frequencies [88]:
�/ = #
"4 b #
5&6&
(2.22)
�0 = #
"4 b #
5&6&
+ #
5&6'
(2.23)
By combining 2.22 and 2.23, we obtain:
78
8(
= 8)&8(
8(
≈ 6&
"6*
(2.24)
2.6.2 Measurement of Acoustic Pressure
The acoustic pressure generated by the fabricated SFAT can be measured using a needle
hydrophone (HNP0200, Onda Corp.) with a tip diameter of 200 µm in the scanning tank system
(AIM III, Onda Corp.). The measured acoustic pressure is then compared with the simulation result
(refer to Figure 2c and 2d). To prevent the hydrophone from picking up electromagnetic
interference (EMI), a sinusoidal signal is periodically delivered to the SFAT, consisting of only
three cycles. This signal is delivered at a pulse repetitive frequency (PRF) of 1 kHz during the
hydrophone measurement. The hydrophone is connected to a 20-dB pre-amplifier (AH-2010, Onda
Corp.) and is securely fixed to a three-axis precision moveable stage. Meanwhile, the SFAT is
24
attached to a two-rotational-axis stage to align the beam axis between the hydrophone and the
SFAT center (Figure 2.9e).
Figure 2. 9 Finite element analysis (FEA) simulated normalized acoustic pressure (a) on the lateral plane at z= 5 and
(b) on the vertical plane along with the central axis. Measured peak acoustic pressure using a hydrophone (c) on the
lateral plane and (d) on the vertical plane. (e) Illustration of the hydrophone measurement setup in the scanning tank.
The acoustic signals picked up the hydrophone is visualized with the digital oscilloscope
(Picoscope 5000, Pico Technology Inc.) in a root-means-square (rms) voltage and converted into
the peak acoustic pressure. The minimum distance between the hydrophone and the SFAT is
limited to 3.0 mm (Figure 2.9d). The scanning resolutions for both xy-plane and xz-plane are 20
µm. Once the peak acoustic pressure, P0, at the focal point is measured, it can be used for
calculating the spatial peak pulsed average acoustic intensity, �.99+ [20],
�.99+ = 9'
"
":$ (2.25)
25
where � and � are the density of medium and the speed of sound in the medium,
respectively. By multiplying ISPPA with a pulse repetition frequency (PRF) over the pulse duration,
the spatial peak time average acoustic intensity, ISPTA, can further be calculated:
�.9;+ = ∫ �.99+
;
2 �� × ��� (2.26)
26
Chapter 3
Design, Fabrication, and Characterization of electrically
Controllable Self-Focusing Acoustic Transducers (eCON SFAT)
3.1 Introduction
Focused ultrasound with electrically tunable focal point and focal size can be obtained with
phased array transducers. Phased array acoustic transducers offer a distinct advantage in various
applications due to their ability to control and steer acoustic beams with precision. Unlike a single
focusing transducer that emit sound waves in a fixed direction, phased array transducers utilize an
array of individual elements that can be independently accessed and controlled with different
timing and amplitude. This allows for the formation of highly focused and steerable acoustic beams,
enabling improved imaging, localization, and inspection capabilities [93].
Although there has been attempt to fabricate SFAT with electrically controllable focal
length [94] and focal points [95] without a typical phased array, about 30% of the electrode area
has to be sacrificed only to access the 36 rings, while the controllability of the focal points was
limited to 2-Dimensional space [95]. To address these limitations, a phased array ultrasonic
transducer based on SFAT with 128 annular electrode rings is designed and explored, employing
a flip chip bonding technique to access each of the rings without sacrificing the electrode area.
This newly designed transducer, termed an electrically controllable SFAT or eCON SFAT, has the
capability to generate 8 different focal zones simultaneously in 3-dimensional space
3.2 Design and Working Principle of eCON SFAT
The design of an eCON SFAT consist of two parts: one is an equal-width ring design,
contributing to the controllability of focal length, while the other entails a sectored design, related
27
to the controllability of focal point. The eCON SFAT incorporates a total of 8 sectors, with each
sector containing 16 equal-width rings, resulting in a total of 128 rings (Figure 3.1)
Figure 3. 1 (a) The top-view diagram of the eCON SFAT, showing the 8 sectors and 128 individual equal-width
electrodes. (b) Photo of a fabricated eCON SFAT before a flip-chip bonding packaging
3.2.1 Focal Length Controllability
The electrical controllability of the focal length can be achieved by modifying the widths
of the ring electrodes of the SFAT to be identical to 200μm, instead of following equation 2.2.
Then a phase factor (��) of each ring is defined and calculated by [94]:
�� = sin g2�
<=+
"->"&>
( i (3.1)
Phase factor determines the contribution of acoustic waves generated from each ring. A positive
phase factor indicates that the activation of the ring contributes to constructive interference,
enhancing the signal at the focal point. Conversely, a negative phase factor suggests that the ring's
activation leads to destructive interference, resulting in a reduction of the signal at the focal point.
Figure 3.2 illustrates the sign of the phase factor of each ring (from 1 to 16) in one sector.
28
Figure 3. 2 Phase factor plot of 16 ring electrodes with different desired focal length from 5 to 15mm. The green color
means activation and the red color means deactivation of the electrode.
For example, to achieve a focal length of 5 mm with 16 rings, the group of rings (1, 4, 5, 6, 9, 10,
13, and 14) needs to be activated, as indicated by positive phase factors (shown in green in Figure
3.2), while the remaining electrodes must remain inactive (depicted in red in Figure 3.2). By
adjusting the selection of activated and non-activated electrode groups, control over the focal
length can be achieved along the direction perpendicular to the transducer surface.
3.2.2 Focal Point Controllability
The focal point in the plane parallel to the transducer surface is electrically controlled by
the sectored electrodes [95]. The sectored SFAT generates a tilted focused ultrasound wave due to
its asymmetrical design. Dividing a circular electrode into eight sectors (as illustrated in Figure
3.1) enables the alteration of the focal point in 9 different zones, including the focal point in the
central axis. Activating each sector produces a focal point at its uniquely designed location.
Consequently, simultaneous activation of all the sectors results in multiple focal points occurring
simultaneously.
29
3.3 Simulation of Focusing Controllability
Unlike an SFAT, the eCON SFAT is not axial symmetric, which means that FEM analysis
with a 2D axial symmetry is not applicable. Instead, a three-dimensional (3D) simulation model
could be designed to verify its capability of generating multiple tunable focal points. A simulation
space is designed to be semi-sphere with a diameter of 20mm (Figure 3.3a) to reduce the
computational complexity and the electrode patterns of eCON SFAT are modeled as a normal
displacement at the bottom surface of the semi-sphere [96]. The maximum mesh size for a 3D
simulation is also modified to decrease to �/4 , for the purpose of reducing computation time and
memory.
Figure 3. 3 (a) FEM simulation model in a three-dimensional (3D) space and simulation results of (b) a single sector
activated with the focal length at 5mm and the focal point at 2mm right from the center (b) two adjacent sectors
activated with the focal length at 5mm and the two focal points at 2mm left and below from the center (c) two facing
sectors activated with the focal length at 7mm and the two focal points at 2mm above and below from the center and
(d) the focused ultrasound at the central axis with the focal length at 5mm by activating three sectors with the same
groups of electrodes at the same time.
The simulation results at the focal plane are illustrated in Figure 3.3 for different groups of sectored
electrodes being activated. Activating a single sector produces focused ultrasound at 2mm to the
right from the center with a focal length of 5mm (Figure. 3.3b). When two sectors are activated
simultaneously, the results of activating two adjacent sectors and two facing sectors are depicted
30
in Fig. 3b and 3c, respectively. The focal lengths are 5 and 7 mm in Fig. 3.3c and 3.3d, respectively,
as the focal points are 2 mm apart from the center. Finally, when three sectors are activated all
together as shown in Figure. 3.3e, a single focusing on the center is generated instead of three
independent focused ultrasound beams, due to the constructive interference at the center.
3.4 Microfabrication Methods of eCON SFATs
The eCON SFAT is fabricated on a 1-mm thick PZT substrate with pre-deposited top and
bottom electrodes layers. The microfabrication process of the eCON SFAT can be divided into two
main parts. The first part involves fabricating 128 electrodes on the bottom side for the soldering
access points, along with the corresponding air-cavity lens on the top to minimize the fringing field.
The second part entails creating indium bumps on both the bottom electrodes of the eCON SFAT
and on the control board to drive electrical signals and bonding two devices together using a flipchip bonding packaging technology.
3.4.1 Fabrication of eCON SFAT Electrodes
Fabricating eCON SFAT requires both patterned electrodes on the bottom electrode (Figure
3.4 a-c) and air-cavity lens on the top electrode (Figure 3.4 h-n). The fabrication process follows
the very similar to the fabrication steps described in Chapter 2.5, with the exception that a 10µm
thick AZ4620 photoresist is used for patterning the air-cavity lens. As the resolution of the
electrode is as small as 50µm, precise front and back alignment is critical, meaning that the corner
alignment (described in Chapter 2.5) is not applicable. Additionally, using IR imaging is not
feasible due to the 1-mm thickness of the PZT substrate. Precise front and back alignment are
achieved using an advanced aligner (MA/BA6 Gen-4 from Suss MicroTec) equipped with a
bottom-view camera. As the typically thin photoresist may not be visible through the camera after
parylene deposition, a thicker photoresist layer is needed. The AZ4620 photoresist, with a higher
31
viscosity than AZ5214, enables the deposition of a 10 µm thick photoresist layer by spin-coating
at 2000 rpm.
Figure 3. 4 A brief Fabrication steps of eCON SFAT;(a) PZT sheet sandwiched by nickel(Ni) layers (b) Patterning
and wet-etching the electrode with equal-with sectored rings (c) 7µm thick Parylene D coating as a passivation layer
(d)RIE-etching the holes for accessing electrode and deposit the under-bump metallization (UBM) layer to define
indium bump position. (e) Depositing 1.5μm thick indium through the thermal evaporation on top of the photoresist
layer, followed by (f) the lift off process and the (g)reflow to form a spherical shape of the bumps. (h) Flipping the
device (i) Wet-etching Ni electrodes into circular shape and remove the photoresist mask (j) Parylene D coating as a
protective layer (k) Depositing photoresist sacrificial layer (l) Another Parylene D coating (m) RIE etching release
holes to open access to the sacrificial layer and (n) Depositing thick Parylene D layer for insulation and
biocompatibility
3.4.2 Fabrication of Indium Solder Bump
The most significant fabrication process in flip-chip bonding packaging is patterning and
generating solder bumps. While various materials such as Pb/Sn, Sn-Au, Sn-Ag, and Sn-Ag-Cu
alloy can be used as solder bumps, indium has been chosen for solder bump for its lower melting
temperature of 156°C compared to that of lead solder alloy (above 300 °C range) [97], compatible
to the baking procedures used in microfabrication processes of PZT. Fabricating indium solder
bump requires a passivation layer between the electrode and solder bumps and under-bump
metallization (UBM) layer [98]. 7µm thick Parylene D is deposited as a passivation layer (Figure
32
3.4c) and UBM layer, consisting of titanium, nickel, and gold in sequence (Ti-Ni-Au:10nm-50nm50nm), is deposited (Figure 3.4d). This UBM layer is utilized to define the positions of indium
bumps and enhance the wettability of indium bumps for the flip-chip bonding technique. Within
the UBM layer, titanium serves as the adhesion layer between parylene and nickel. Nickel is chosen
for its wettability to indium, and gold is employed not only to prevent the oxidation of nickel but
also to enhance wettability. Then, 3μm thick photoresist (AZ5214) is deposited and patterned to
define the lateral dimension of indium bump (Figure 3.4d). Over the photoresist layer, 1.5μm thick
indium is deposited through the thermal evaporation (Denton 502A) with the deposition rate of
5Å/s (Figure 3.4e). The deposited indium is then removed and patterned using the lift-off process
in acetone for 24hours (Figure 3.4f). Finally, the patterned indium is then reflowed to form a
spherical-shaped solder bump (Figure 3.4g).
The reliable connection between the transducer and the control board relies on the shape
and height of the indium bumps on both components. The shape of the indium bump is defined
during the reflow process and is influenced by factors such as the thickness of the deposited indium
layer, the lateral dimensions or diameter of the under-bump metallization (UBM) pattern, and the
lateral dimensions of the indium pattern. Additionally, when the diameter of the indium pattern is
small (less than 10µm), the thickness and the opening of the photoresist pattern for the liftoff
process also needs to be considered [99]. Furthermore, the temperature and humidity levels of the
ambient environment during the reflow process may also impact the shape of the solder bumps
[100]. The height of the indium bump can be estimated by considering the volume of a truncated
sphere, assuming that the volume of indium does not change before and after the reflow process
[101]:
� = � L
?,
@
+ ℎ A"
" M (3.2)
33
where �, ℎ and � are the volume of truncated sphere, height, and radius of bottom surface,
respectively. In the case of indium bump, ℎ represents the height of indium bump and � is the
radius of UBM pattern. If the lateral dimension of the indium pattern is significantly larger than
that of the UBM, the deposited indium around the UBM may not be entirely collected (over the
UBM) during reflow. This incomplete collection results in residual indium around the UBM after
reflow, causing the height of the indium bump to be lower than the theoretically calculated value
using equation 3.2. Conversely, when the indium pattern is not large enough, the shape of indium
bump may resemble more of a spherical cap, not a truncated sphere. For optimizing the dimensions
between the under-bump metallization (UBM) and indium bump pattern, various combinations are
tested, and the height of indium bumps is measured. For an 80 µm UBM diameter and 1 µm thick
deposited indium, the measured heights of indium bumps over 150 µm, 225 µm, 500 µm, and 600
µm diameters of the indium pattern are 5 µm, 13 µm, 26 µm, and 27 µm, respectively (Figure 3.5).
Additionally, different sizes of the UBM are tested, including diameters of 7 µm, 15 µm, 20 µm,
50 µm and 100 µm. The results indicates that a lateral dimension ratio of 1:2.5 between the UBM
and indium pattern provides the best wettability over the UBM with the smallest residuals [96].
Figure 3. 5 Indium bump after a reflow with different combination of lateral dimension of the UBM (80μm) and the
indium pattern (a) 150 µm, (b) 225 µm, (c) 500 µm, and (d) 600 µm. (e) Graphical expression of analytical calculation
and experimental measurement of the height of indium bump after a reflow.
34
The oxidation of indium typically impacts solderability and the reflow process, resulting
in indium residue after a reflow, and consequently, not all indium contributes to the formation of
a solder bump [102]. To eliminate indium oxide (In2O3), the reflow process was thoroughly
investigated under various environments, including air, nitrogen gas (N2), hydrogen gas (H2),
liquid metal flux (Superior 67Ds), and formic acid (HCOOH). Figure 3.6 illustrates the test results
with UBM and indium pattern diameters of 80μm and 200μm, respectively, with a deposited
indium thickness of 1.5μm.
Figure 3. 6 Reflown indium bumps under the different environments such as (a) air, (b) nitrogen gas (N2), (c)
hydrogen gas (H2), (d) liquid metal flux, and (e) formic acid (HCOOH)
In the reflow conducted in the regular solder oven (maximum temperature 275 °C for 7 minutes)
under the ambient air, there is no observable indium reflow or the formation of a spherical bump
(Figure 3.6a). When nitrogen gas (N2) is used in a Chemical Vapor Deposition (CVD) chamber
(maximum temperature 250°C for 10 minutes), which is a common inert gas to prevent oxygen
from reacting with indium during reflow, indium begins to reflow and forms a solder bump on the
UBM area but leaves a significant amount of residuals around the bump pattern (Figure 3.6b).
Considering the reported efficacy of hydrogen (H2) plasma in eliminating indium oxide during
reflow [103], reflow tests is conducted in a hydrogen gas environment within the CVD chamber
under identical heating conditions. In comparison to nitrogen gas, hydrogen gas results in fewer
35
residuals and taller bump heights (Figure 3.6c). Additionally, as liquid metal flux is commonly
used during a soldering step to remove any metal oxide and facilitate the soldering process, liquid
metal flux is tested during the reflow process. As depicted in Figure 3.6d, it completely cleans the
residuals and enables indium to form a perfectly spherical bump with a height of 16.8 µm. However,
it also weakens the adhesion of the UBM layer, resulting in the detachment of some components
after the cleaning process. On the other hand, formic acid is known to be capable of removing
metal oxide through a two-step chemical reaction [104]:
�� + 2����� → �(����)" + �"� (3.3)
�(����)" → � + 2��" + �" (3.4)
where �� and � are any metal oxide and metal, respectively. The reflow is conducted with 1 mL
of vaporized formic acid in the oven (220°C for 10 minutes), effectively removing indium oxide
and cleaning out the residuals (Figure 3.6e). The height of the indium bump after the reflow under
formic acid is 17.2 µm, which closely matched with the analytical value of 17.6 µm. In Figure 3.7
and Table 3.1, detailed information is provided regarding the dimensions of the indium solder
bump after the reflow process, along with the design parameters of eCON SFAT.
Figure 3. 7 Illustrated image of indium solder bump after the reflow, showing the dimension and design parameter
information (description and dimension is listed in Table 3.1)
36
Table 3. 1 Description and Dimension Information of Symbols in Figure 3.7
Symbol Description Dimension Symbol Description Dimension
A1 Diameter of UBM 80 µm A6 Minimum distance between two
solder bumps
280 µm
A2 Diameter of passivation opening hole 50 µm A7 Height of solder bump after a
reflow
17 µm
A3 Minimum distance between the rings 50 µm A8 Thickness of PZT substrate 1 mm
A4 Width of the ring 200 µm A9 Thickness of passivation layer 7µm
A5 Thickness of UBM (Ti-Ni-Au) 110 nm
(10 -50-50nm)
Indium solder bump can be fabricated on a control board which is designed and fabricated
on a printed circuit board (PCB) following the methodology described in Chapter 3.4.2. However,
the key difference is the absence of the need to deposit under bump metallization layers as they
can be replaced by electroless nickel immersion gold (ENIG) pads, which contains both nickel and
gold. The fabrication process of the control board is illustrated in Figure 3.8, which includes a
photo of the ENIG pads on the control board (Figure 3.8f). After fabricating indium solder bumps
on both transducer and the control board, they are connected through a flip chip bonder (Finetech
Fineplacer PICO A4).
37
Figure 3. 8 A brief fabrication process of indium bumps on the control board; (a) A printed circuit board (PCB) with
a patterned electroless nickel immersion gold (ENIG) pads (b) Spin coating and patterning photoresist to define the
position of indium bump (c) Depositing 1.5μm thick indium through the thermal evaporation on top of the photoresist
layer (d)Lifting- off photoresist and (g)Reflowing to form a spherical shape of the bumps.
38
Chapter 4
Design, Fabrication, and Characterization of Spiral-arm Vortexbeam Acoustic Transducer (SVAT)
4.1 Introduction
Acoustic vortex refers to the structural rotation of sound waves around the central axis.
This unique pattern of acoustic wave propagation has drawn considerable attention from the
researcher in various scientific and engineering domains. A key advantage of acoustic vortices is
their capability to transport and manipulate angular momentum within the sound field [105],
opening up new possibilities for biomedical applications such as an enhanced medical imaging
applications [106]- [108] and sonothrombolysis [109]. Through the manipulation of sound wave
angular momentum, acoustic vortex technology enables finer control over the spatial distribution
of ultrasound energy [105]. This leads to improved imaging resolution and the capacity to visualize
intricate details within biological tissues [107][108]. Beyond the medical applications, acoustic
vortex technology holds promise in various fields, including underwater communication
[110][111], acoustic propulsion[112] , and the trapping and manipulation of small-scale
objects[113]-[115] . Furthermore, recent study reveals that small solid objects, such as an ingestible
camera or a kidney stone, can be manipulated even when they are inside a living body [116]. As
the exploration of acoustic vortices progresses, their potential to revolutionize various industries
becomes increasingly evident, promising breakthroughs in fields as diverse as medicine,
telecommunications, and materials science.
A diverse method has been developed to generate acoustic vortices, each method exploiting
distinct principles to create the rotational patterns in sound waves. One commonly employed
39
technique involves the utilization of phased array transducers, such as an eCON SFAT described
in Chapter 3, where multiple individual transducer elements are controlled independently to create
constructive and destructive interference patterns. By carefully manipulating the phase and
amplitude of these transducers, acoustic vortices can be achieved with precise control of rotation
direction and beam steering [117]. Another approach involves the use of specially designed
acoustic lenses such as a hyperboloidal focused acoustic vortex (H-FAV) lens that shape the
acoustic vortex wave [118]. Furthermore, the concept of angular momentum transfer in acoustic
vortices has been explored by utilizing helical-shaped or spiral-shaped transducers[112][119]. The
spiral shaped electrode configurations introduce a twist to the propagating sound waves, imparting
angular momentum and resulting in the generation of acoustic vortices [120]. Finally, holographic
acoustic elements, inspired by optical holography, have emerged as a promising method, allowing
to encode complex phase patterns onto emitted sound waves and sculpt acoustic fields to generate
acoustic vortices [113].
In this chapter, acoustic vortex transducer designed based on a sectored SFAT is introduced,
followed by a Spiral-arm Vortex-beam Acoustic Transducer (SVAT) which generates acoustic
vortices using a single or multiple spiral arm(s) electrode design. The SVAT offers the advantage
of a straightforward fabrication process with a lens-less design, albeit with the limitation of
controlling the rotational direction or beam forming.
4.2 Working Principle of SVAT
4.2.1 Acoustic Vortex-beam with A Sectored SFAT Design
Acoustic vortex can be created with a sectored SFAT using the fact that the focal position
can be adjusted as explained in Chapter 3.2.2. To generate a vortex-shape using multiple focal
points, an 18-sectored SFAT with each sector is pie-shaped with angle of 20 ̊, is proposed (Figure
40
4.1a-b). The focal length (�) of each sector increased with the step of 100 µm starting from 2 mm
with the focal length (�B) of the �
%? sector being:
�B = 2 + 0.1 × � [��] (4.1)
The design ensures that each sector generates a focused ultrasound at focal points with 100µm
difference in vertically and 20˚ shifts in laterally, forming a vortex-like shape in the medium
starting from 2mm and ending at 3.8mm. Additionally, the same electrode design with leaving
empty space on the center (Figure 4.2c-d) for the observation near the central axis as the electrode
is generally not transparent.
Figure 4. 1 (a)The designs and (b)photos of the fabricated transducers of an 18-sectored SFAT, generating an acoustic
vortex. (c)The designs and (d)photos of the fabricated transducers of an 18-sectored SFAT with a center cavity
electrode design.
4.2.2 Acoustic Vortex-beam with A Spiral-Arm Electrode Design
Taking inspiration from an 18-sectored SFAT design, which has a spiral shape, a Spiralarm Vortex-beam Acoustic Transducer (SVAT) is designed with a logarithmic helix electrode
design. The equation defining a logarithmic helix in polar coordinate for a radius r is given by
[119]:
� = ��CD (4.2)
where a, b, and q are the initial radius, azimuth coefficient, and angle, respectively. The azimuth
coefficient b determines the growth rate of the radial distance from the center with angle increment.
41
A larger azimuth coefficient results in larger gaps between the spiral arms and wider width at the
outer electrode. A spiral arm electrode can be defined by two logarithmic helixes with initial radii
(�# and �"), termination angles (�# and �") and common azimuth coefficient (b) as follows:
�����#: � = �#�CD ��� �����": � = �"�CD, � ∈ [�# , �" ] (4.3)
The �# and �" determine the initial width of the arm (�" − �#), while the widest width can be
calculated as (�" − �#)�C(D"&D#) . The frequency bandwidth of SVAT is constrained by the
electrode width, allowing the generation of vortex beams within the wavelength range of
2�1B! and 2�1EF, where d represents the width of the electrode arm [119]. In other words, the
operating frequency range for the vortex-beam acoustic wave is limited to:
�1B! = $
"(E"&E#)G-(/"0/#) , �1EF = $
"(E"&E#) (4.4)
where c is the speed of sound in the medium. For example, with parameters �# = 2.3 ��, �" =
2.5 ��, � = 0.090, �# = 0˚, ��� �" = 12�, this design yields �1B! = 0.2 mm and �1EF = 5.95
mm from equation 4.4. In water, where the speed of sound is 1480 m/s, the operating frequency
range is from 124 kHz to 3.7 MHz, which includes the thickness-mode resonant frequency of 1-
mm-thick PZT, 2.32 MHz. Furthermore, SVAT design can be modified to have multiple spiral arms
by maintaining the termination angle × number of arms at 12� and rotating the initial position of
each arm by 180˚, 120˚, and 90˚ for the 2, 3, and 4-arms designs, respectively. The azimuth
coefficient of n-Arms SVAT is modified to �# × �, where �#is the azimuth coefficient of single
arm SVAT. (Figure 4.2).
42
Figure 4. 2 (a)Single-arm, (b)Two-arms, (c)Three-arms, and (d)Four-arms SVAT design with parameters of a2 =
2.3 mm, a3 = 2.5 mm, b = 0.090, θ2 = 0˚, and θ3 = 12π
4.3 Simulation of Acoustic Vortex
To validate the capability of the 18-sectored SFAT and SVAT to generate acoustic vortices,
finite element method (FEM) analysis in a three-dimensional space is performed, as the electrode
designs are not axial symmetric. Similar to the 3D simulation of eCON SFAT, a conical-shaped
medium space is defined with a 10-mm base diameter and a 15 mm height (Figure 4.3a), utilizing
a maximum mesh size of λ/4 to optimize computational efficiency and memory usage. The bottom
electrode is designed based on equation 4.2 and a normal displacement mimics the generation of
acoustic waves. Figure 4.3 illustrates 3D simulation results of SVAT (Figure 4.3b) with parameters
of �# = 0.5 ��, �" = 0.7 ��, � = 0.090, �# = 0˚, ��� �" = 12� , and 18 sectored SFAT
design (Figure 4.3c). As anticipated, the 18-sectored SFAT demonstrates the capability to generate
a partial acoustic vortex within the z-range of 2mm to 3.8mm, while the SVAT produces a
prolonged acoustic helix within the medium.
Figure 4. 3 (a) FEM simulation model in a three-dimensional (3D) space having a conical shape and normalized
acoustic pressure distribution simulation results of (b) a SVAT with parameters of a2 = 0.5 mm, a3 = 0.7 mm, b =
0.090, θ2 = 0˚, and θ3 = 12π, and (c) the 18-sectroed SFAT design.
43
Additionally, the FEM simulations with multi-arms SVATs, depicted in Figure 4.2, are
conducted to verify the capability of generating multiple acoustic helixes. Figure 4.4 demonstrates
the normalized acoustic pressure distributions over the xy-plane (which is parallel to the bottom
surface of cone) at z = 7mm produced by multi-arms SVATs.
Figure 4. 4 Normalized absolute acoustic pressure distributions over the xy plane at an observation point (z= 7mm)
produced by (a) 1-arm SVAT, (b) 2-arms SVAT, (c) 3-arms SVAT and (d) 4-arms SVAT.
4.4 Microfabrication Methods of SVAT
One of the advantages of SVAT is its simplified fabrication process, involving only the
patterning of the electrode without the need for a lens or complex packaging used in phased array
transducers. The complete fabrication process, including packaging with Parylene D, consists of
four steps. The fabrication process is the same as the patterned ring electrodes SFAT described in
Chapter 2.5.1.
44
Chapter 5
Wireless Operating Integrated System Design for Acoustic
Transducers
5.1 Introduction
The demand for wireless, integrated, and portable solutions has grown across various
industries with the technological innovation. This is also true in the field of acoustic transducers.
The conventional method of driving acoustic transducers requires dedicated laboratory spaces
equipped with specialized instruments, such as a power amplifier and a signal generator. Although
effective for research, these experimental setups are impractical for applications demanding
mobility, flexibility, and real-time operation. For example, one application that can greatly benefit
from portable acoustic systems is the field of ultrasonography medical diagnosis, particularly a
point-of-care ultrasound (POCUS)[124]. POCUS could provide clinicians or doctors the flexibility
to diagnose patients and perform an ultrasonography at the patient's bedside, in isolated areas, or
in emergency scenarios when prompt diagnostic information is necessary. Furthermore, portable
acoustic systems become invaluable tool in healthcare settings with limited resources or in rural
areas where access to centralized imaging facilities may be restricted.
Another notable example is the acoustic underwater propelling systems [125][126].
Utilizing acoustic propulsion for exploring underwater environments in scientific research,
surveillance, and industrial operations offers advantages such as low-noise and low-disturbance
operation, in contrast to traditional mechanical propellers [127]. However, operating with a wired
setup could be restrictive in such underwater scenarios, where maneuverability and real-time data
acquisition are essential. A wireless operating integrated system designed for acoustic transducers
45
provides a solution to overcome such challenges by enabling underwater exploration with
improved efficiency and adaptability.
This chapter explores the complex design considerations and technical innovations in the
development of wireless operating integrated systems for acoustic transducers. Firstly, various
power sources for a portable system are explored to drive acoustic transducers that require high
current. Subsequently, two different circuit designs are proposed: one utilizing a commercially
available power amplifier integrated circuit (IC), and the other employing an inverter design with
a Z-source network [128].
5.2 Power Source for Driving Acoustic Transducers
A portable ultrasound system has been explored with field programmable gate array (FPGA)
[121], an integrated power amplifier [122] or an application specific integrated circuit (ASIC)
[123]. However, these works focused on innovating the driving circuit without considering the
power source, which is critical for wireless system, as it determines the operation time and
maximum deliverable current. For wireless systems, three prominent energy storages are lithiumion batteries, supercapacitors, and lithium-ion capacitors, each exhibiting distinct characteristics,
particularly in terms of power density and energy density. Energy density, measured in watt-hours
per kilogram (�ℎ/��), indicates the amount of energy that power source can store relative to its
mass. Power density, on the other hand, is measured in watts per kilogram (�/��), representing
the amount of power which power source can generate relative to its mass [129] .
5.2.1 Lithium-ion battery
Lithium-ion battery (LIB) is one of the most common power sources that can be found in
many portable devices. LIB utilizes the movement of lithium ions (��-) between the cathode and
anode electrodes, filled with electrolyte. The electrolyte carries lithium ions through a separator to
46
charge and discharge the battery. The primary purpose of the separator is preventing the physical
contact between the cathode and anode and facilitating the ion transport within the battery cell
[130]. Typically, the cathode is composed of Lithium Metal Oxides (LiMxOy) such as Lithium
Cobalt Oxide (LiCoO2), Lithium Manganese Oxide (LiMn2O4) or Lithium Nickel Oxide (LiNiO2)
[131], while graphite is commonly used as anode material. However, as graphite anodes have the
problems in poor capacity and rise the safety concerns, recently alloy anodes, such as Aluminum
(Al), Tin (Sn), Magnesium (Mg), and Silver (Ag), are explored. When the load is connected to the
battery, lithium ions travel from the anode to the cathode, while electrons are released from the
anode and flow through the load, creating an electric current that can power devices (Figure 5.1).
The cell reaction at both cathode and anode can be simply express as [131]:
Cathode: ��" + ��- + �& → ����" (5.1)
Anode: ���@ → �@ + ��- + �& (5.2)
where M represents any metal, and a graphite anode is assumed.
Figure 5. 1 An illustration that shows the structure of Lithium-ion battery and the movement of Lithium ions and
electrons during discharging and supplying power to the device (or load)
Lithium-ion batteries have many advantages such as high energy densities, high coulombic
efficiencies, low self-discharge features. Among the, the success of lithium-ion batteries in the
portable electronics market can be attributed to the higher energy densities, enabling them to
47
provide a consistent power output over an extended duration [129]. However, for a small lithiumion battery, the power density is limited to lower than 0.1kW/kg, which is not suitable for driving
acoustic transducer due to the imperative high currents [127]. Another significant challenge of
LIBs is the reduction in capacity as the number of charge/discharge cycles increases, thereby
limiting the overall life cycle of the battery [135].
5.2.2 Supercapacitor
Supercapacitors (SC) or ultracapacitors, store electrical energy in an electrostatic form. A
typical supercapacitor consists of two porous electrodes, which is generally made of activated
carbon, two current collectors, an electrolyte, and a separator. The separator in a supercapacitor
has the same functions with that of lithium-ion battery, electrically isolating the two electrodes,
preventing potential short circuits, and facilitating the passage of ions. Supercapacitors store
electrical charge through the formation the electric double layers generated by the charge
separation at the interface between the electrodes and the electrolyte (Figure 5.2). This energy
storage mechanism involves no chemical reaction, unlike lithium-ion batteries, except for the fast
and reversible Faradaic reactions existing on the electrode surface [133]. This process enables
supercapacitors to charge and discharge rapidly, providing high power density (about 1kW/kg for
a small supercapacitor [127]). The lack of chemical reactions contributes to their long cycle life.
However, supercapacitors generally have lower energy density due to the limited double-layer
capacitance of activated carbons (<20 μF cm−2
) and their low density compared to their large
porosity [134]. In other words, supercapacitors have the capability to supply high current for
driving acoustic transducers, but only for a restricted duration (depending on their capacitance
value).
48
Figure 5. 2 An illustration depicting the structure of a supercapacitor with highlighted electric double layers (dashed
line) that store electrical energy within the supercapacitor.
5.2.3 Lithium-ion Capacitor
Lithium-ion capacitors (LIC) are hybrid energy storage power sources that combine the
energy storing mechanism of lithium-ion batteries and supercapacitors. By integrating features of
LIB and SC, they are capable of providing the advantages of two technologies, offering high
energy density and high power density at the same time. The structure of LIC consists of an anode
from LIB, and cathode from supercapacitor, filled with an electrolyte containing Lithium-ions
(Figure 5.3). In LICs, the charge/discharge process involves the adsorption/desorption of electrons
onto/from the cathode electrode (like supercapacitors), while the Lithium-ion accumulation/deaccumulation process takes place at the anode electrode, similar to LIB. As the adsorption/-
desorption reaction of electron at the cathode is non-Faradaic reaction which is much faster than
the accumulation/de-accumulation reaction of the Lithium-ion at the anode electrode, the power
density of LIC is generally determined by the anode electrode [136].
49
Figure 5. 3 An illustration depicting the structure of a Lithium-ion capacitor (LIC), combining the anode of Lithiumion battery with the cathode of Supercapacitor.
The energy (�) and the power (�) that can be stored in LIC can be calculated by:
� = ��"/2 (5.3a)
� = �"/4� (5.3b)
where �, �, and � are the capacitance, the cell potential, and the equivalent series resistance (ESR)
of LIC, respectively. Since both the energy and the power is proportional to the voltage square,
increasing the operational potential of LIC ensures higher energy and power density. The voltage
can be increased by using a different material for anode and cathode electrodes or different
electrolyte [136].
Finally, Table 5.1 shows the feature comparison among Lithium-ion battery, supercapacitor,
and Lithium-ion capacitor, noting that the provided values are typical values for small-sized battery
and capacitor [127].
50
Table 5. 1 Feature comparison table of Lithium-ion battery, supercapacitor, and Lithium-ion capacitor
Features Lithium-ion Battery Supercapacitor Lithium-ion Capacitor
Energy Density High (10 – 100Wh/kg) Low ( <10Wh/kg) Medium ( ~ 10Wh/kg)
Power Density Low (~ 0.1kW/kg) Medium ( ~ 1kW/kg) High (1k – 10kW/kg)
Operating Voltage 2.5 – 4.3V 0 - 3.0V 2.2 – 3.8V
Charge/Discharge Cycle 500 - 1000 100k 100k
Self-Discharge Very low Susceptible Very low
5.3 Electronic Circuit Design with An On-board Power Amplifier
5.3.1 System Design
An electronic circuit to drive acoustic transducer is explored using an on-board power
amplifier, which is composed of a DC/DC converter, an oscillator for generating a square wave
and a power amplifier for amplifying the square wave and driving the SFAT, as illustrated in Figure.
5.4.
Figure 5. 4 A system block diagram of SFAT driving electronic circuit using a power amplifier. Black and blue lines
represent the power and signal rails, respectively. A capacitor, denoted by a red dashed line, is used to block the DC
components of the PWM signal.
51
A 4V direct current (DC) voltage sourced from the Lithium-ion capacitor serves as the power
supply for the system. This voltage is converted to positive and negative 15V by a DC-to-DC
converter, which then powers the amplifier. The power amplifier is configured with a gain of
10V/V to amplify the 3Vpp square wave signal produced by an oscillator, achieving a 30Vpp
output to drive the SFAT.
5.3.2 Components Selection
The IC components are chosen with a primary focus on power efficiency, compact size,
and the ability to effectively drive the SFAT with a high-power density.
5.3.2.1 A DC/DC converter
TPS65131 (Texas Instruments, Inc.) is chosen as a dual DC/DC converter for its capability
to convert both positive and negative power from 4VDC with conversion efficiency of up to 89%
and 81%, respectively. It employs a boost converter topology for positive voltage conversion and
a buck-boost topology for negative voltage conversion. Both topologies operate with a 1.25MHz
pulse-width modulation (PWM) signal, contributing to the design of a system with a reduced
inductor size. It is capable of delivering up to 2A current to the load with the specified quiescent
current and shutdown current of 500µA and 1µA, respectively. Additionally, the compact size of
the IC at only 4 x 4 mm² contributes to the overall reduction in the total PCB size.
5.3.2.2 An Oscillator
Two types of oscillators have been considered: one is a LTC1799 (Analog Devices, Inc.)
and the other is LTC6904 (Analog Devices, Inc.), which frequencies of output PWM signal can be
controlled by an external resistor and digital signal through I2C interface, respectively. The
LTC1799 offers a frequency range from 1kHz to 33MHz, whereas the LTC6904 covers a broader
spectrum, ranging from 1kHz up to 68MHz. For the proof-of-concept design, LTCC6904 is
52
implemented as LTC6904 requires an external microcontroller (MCU) which occupies additional
space on the PCB. While a sinusoidal signal is typically applied to SFAT, past experiments have
revealed that driving SFAT with a PWM signal, whose frequency is matched with the thicknessmode resonant frequency of PZT, is capable of generating higher acoustic intensity at the focal
point. It is primarily due to the odd harmonics of resonant frequency, contributing to converting
electrical energy into acoustic energy. An extra capacitor is introduced at the output of the oscillator
(identified by a red box in Figure 5.4) to eliminate the DC component of the PWM signal. This
conversion transforms the 0-3V PWM into a range of -1.5V to +1.5V, enabling the amplifier to
provide an amplified rail-to-rail signal to the SFAT.
5.3.2.3 A Power Amplifier
The selection of a power amplifier is a critical and carefully chosen. It determines the
output peak-to-peak voltage and the maximum current delivered to the SFAT. Additionally, the
bandwidth of the entire system is highly dependent on the characteristics of the chosen power
amplifier. THS3095 is chosen because of it offers high voltage (up to ±15�) with 300mA output
current, high speed (slew rate of 6000V/µs), high bandwidth (Gain-Bandwidth Product, GBW of
305MHz) and low-noise (input voltage noise of 1.1nV/√��). The power amplifier is designed
with a gain of 10V/V to amplify 3Vpp square wave input signals to a 30Vpp output. Two topologies
are considered and simulated to achieve this gain: one is single-ended inverting topology (Figure
5.5a), and the other is differential topology (Figure 5.5b).
Figure 5. 5 Schematic diagrams of (a) single-ended and (b) differential topologies of amplifying circuit
53
For a single-ended topologies, �1 and �2 (in Figure 5.5a) are configured in a 1:10 ratio to
achieve the desired gain of 10V/V. However, due to the limited bandwidth of 30MHz (resulting
from a GBW of approximately 300MHz), this may not be sufficient for acoustic transducers
operating at frequencies higher than 30MHz, especially those designed with higher harmonics of
the thickness-mode resonant frequency of PZT. Therefore, a differential topology is being
considered. For the differential topology to generate 30Vpp output driving signals, the voltage gain
can be reduced to 5V/V, requiring a modified ratio between �# and �" (in Figure 5.5b) set to 1:5.
This adjustment is made to increase the bandwidth to 60MHz, while still achieving the desired
output. However, this configuration requires a NOT gate to invert the PWM signal, before feeding
it into the negative input of the amplifier, which occupies additional IC and space. Moreover, a
time delay, introduced by a NOT gate, potentially leads to distortion issues in the output signal.
For example, the circuit simulations, using a SPICE-based TINA (Texas Instruments Inc.), with
delay of 3 ns (Figure 3.6a) and 22 ns (Figure 3.6b) reveal the distortion of output signal even with
a small-time delay, induced by a NOT gate.
Figure 5. 6 Simulation results for the THS3095 in a differential topology revealing signal performance under two
conditions: (a) a 3ns and (b) a 22 ns time delay between the two input PWM signals of the amplifier.
Another specification considered for designing the power amplifier is its slew rate, given
the operating frequency can be as high as tens of MHz range. THS3095 offers a high slew rate of
54
6000V/µs, giving the rising time of 5ns to reach 30V from 0V, which is only less than 1% of the
period of 2.32MHz (which is the operating frequency of SFAT designed on 1-mm-thick PZT).
However, at higher frequencies like 20MHz, the rise time of 5 ns could represent a larger
proportion, potentially reaching 10% of the period.
5.3.3 PCB design
Followed by the selection of the components and topologies, a PCB has designed as shown
in Figure 5.7a. The dimension of the board is approximately 17 × 28 mm2
, including the SMA
connector. The Bill of Material (BOM) of this design includes three ICs (TPS65131, LTC1799,
and THS3095) and 50 passive components (26 resistors, 20 capacitors, 2 inductors and 2 Schottky
didoes). Additionally, a test board (Figure 5.7b), containing headers and test pins, has also been
developed to facilitate efficient debugging and testing processes.
Figure 5. 7 (a) A PCB layout design of the system, having dimension of 17mm × 28mm. (b) A photo of test board
PCB including headers, test points, and a SMA connect for testing and debugging purposes.
5.3.4 Circuit Characterization
Finally, the on-board power amplifier circuit is tested to characterize key parameters such
as power efficiency and the system bandwidth and operation time with a 100F LIC. The output
voltage, without a load, is measured to be 29Vpp, for the circuit designed for a closed-loop voltage
gain of 10. However, when the SFAT is connected to the load, a slight reduction to 28.2Vpp occurs
55
due to the loading effect. The power efficiency (�) is calculated under the different load condition
(�5) , measuring the input current (�B!) with a current probe (Tektronix’s TCO0030A).
� = 9*45
96+
= I*45,8&( " /=9
I6+×L6+
(5.4)
The maximum efficiency of the system is measured and calculated at 66% when the load resistance
is approximately 130Ω and it decreases rapidly as the load resistance increases (Figure 5.8)
Figure 5. 8 The plot showing the power conversion efficiency varying the load resistance (R:) from 30Ω to 1.5kΩ.
The maximum power efficiency of the system of 66% occurs at R: =130Ω
Additionally, the gain of the system is measured over the frequency range of 1kHz to
30MHz to determine the 3dB frequency (�&M3N), also known as the cutoff frequency, at which the
power output of the system decreases to half of its maximum value (-3dB). The measurement is
conducted without load and with a resistive load of 100Ω. The cutoff frequency is found at
approximately 5MHz as shown in Figure 5.9.
56
Figure 5. 9 The system bandwidth, showing the gain of the system over the frequency range from 1kHz to 30MHz.
The cutoff frequency (f;<=>) is found at approximately 5MHz
Finally, a 100F LIC is connected to the system, and the operating time is measured with a
SFAT connected to the load. The voltage level of the LIC is monitored, and the time is recorded
until its voltage drops to 2.5V, which is its minimum voltage level, resulting in 3 minutes and 10
seconds of operation.
5.4. Electronic Circuit Design with A Modified Z-source Inverter
5.4.1 System Design
To overcome the limitation of the drive circuit in the previous section such as the limited
output voltage, limited output current, and relatively large PCB size, another approach to drive the
acoustic transducers is explored, which is based on an inverter design.
The design process of the inverter starts with a conventional DC-to-DC boost converter
topology (Figure 5.10a). A boost converter steps up the input DC voltage by controlling the metal-
57
oxide semiconductor field-effect transistor (MOSFET) with a duty cycle of a PWM signal. The
output voltage delivered at the load is a function of the duty cycle [137]:
�OP% = #
#&Q
�B! (5.5)
where D is a duty cycle of PWM signal going into the gate of MOSFET, defined by:
� = ;*+
;*+-;*??
(5.6)
During the time MOSFET being turned on (�O!), the electrical energy coming from the input power
source is stored in the inductor (�#). When the MOSFET is turned off, the stored energy, combined
with the energy from the input power source, is delivered to the load. This results in a boosted
output voltage, proportionate to the on-time (�O!).
Figure 5. 10 Schematic diagrams of (a) a conventional boost converter, (b) a boost converter with a modified Z-source
network, and (c) an inverter with a modified Z-source network
58
Replacing an inductor (L1 in Figure 5.10a) with a modified Z-source network, composed
of two inductors, two capacitor, and one Schottky diode (Red box in Figure 5.10b) improves the
gain of the boost converter by [128]:
�OP% = #
#&"Q
�B! (5.7)
Theoretically, this design ensures infinite gain when the duty cycle of PWM is set to 50%. In this
topology, energy is stored in the inductors (�# and �") as current during the on-time, similar to a
conventional boost converter. However, when the MOSFET (Q1) is turned off, the stored energy,
along with the input source, not only delivers current to the load but also charges the capacitors
(�# and �") as voltage. In the following cycle, the stored voltage in the capacitors charges the
inductor, along with the input power source again, creating positive feedback to generate an infinite
output voltage.
Finally, the generation of AC voltage can be accomplished by removing the rectifier circuit
(�" and �M in Figure 5.10b), which converts the switching AC voltage to a DC voltage. In other
words, the load (acoustic transducer) is directly connected to the switching node whose peak-topeak voltage is equal to equation 5.7 (Figure 5.10c).
5.4.2 Component selection
One of the benefits of the modified Z-source inverter topology is its low parts counts,
consisting of one MOSFET with a gate driver IC and five passive components to form a modified
Z-source network. The selection of the components is carefully considered with the criteria of the
power efficiency, the power density, the bandwidth, and PCB size.
5.4.2.1 Switching FET
In a switching power topology, a choice of field-effect transistor (FET) is a critical factor
that can significantly impact the performance, including bandwidth, power delivery, efficiency,
59
and other key parameters. A wide-bandgap-semiconductor transistor, such as Gallium-Nitride
Field-Effect Transistor (GaNFET), has been used as an alternative to a silicon (Si)-based MOSFET
in the power electronics, as it offers a higher electron mobility, leading to lower on-state resistance,
and a higher breakdown voltage [139]. Furthermore, GaNFETs have very low capacitance,
compared to silicon MOSFETs, enabling the fast-switching operation. For example, the input
capacitance (�L..) and the output capacitance (�R..) of GaNFETs are approximately 4-5 times
lower than those of two comparable Si-based power MOSFETs, resulting in less charges involved
in turning FET on and off [139]. Furthermore, GaNFETs typically do not have an inherent body
diode, unlike silicon MOSFETs, eliminating the reverse recovery time and reverse conduction
concern such as a failure mode or a shoot-through current [140].
With the advantages of the GaNFETs over silicon MOSFETs, GaNFET EPC2001 from
Efficient Power Conversion Inc. has been chosen as the FET for the modified Z-source inverter.
The EPC2001 offers a maximum drain to source voltage (�Q.) of 100V with a drain current (�Q)
of 25A, while the on-state resistance (�Q.,R*) is only 5.6mΩ, and the input (�L..) and the output
capacitance (�R..) are 850 pF and 450 pF, respectively [145]. These characteristics make the
EPC2001 (of which the size is 4.1 x 1.6 mm2
) an attractive choice for high switching performance
with low conduction and switching loss in the Z-source inverter design.
5.4.2.2 Z-source network
The gain relationship of the modified Z-source boost converter (equation 5.7) is derived
based on the assumption that the converter is operated in the Continuous Conduction Mode (CCM).
This implies that the inductor current would not be zero during the whole switching period:
�5 − #
"
Δ�5 > 0 (5.8)
60
where Δ�5 is the inductor current ripple. Since the average inductor current (�5) is equal to the
average input current (�B!) and under the ideal power conversion without loss, the input power
must be equal to the output power, �B! = �OP%, we obtain
�5 = �B! = 96+
I6+
= 9*45
I6+
= I*45
"
"I6+=9
(5.9)
where �5 is the resistive load at the output. By substituting equation 5.7 into 5.9:
�5 = I6+
"(#&"Q)"=9
(5.10)
Additionally, the inductor ripple current (Δ�5) can be calculated by examining the relationship
between the inductor voltage and capacitance voltage during the off time [128]:
Δ�5 = (#&Q)QI6+
(#&"Q)58(
(5.11)
where � and �/ are the inductance and switching frequency, respectively. From 5.8, 5.10, and 5.11,
the condition that holds the CCM operation is:
� > Q(#&Q)(#&"Q)=9
"8(
(5.12)
Figure 5.11 shows the minimum inductance value required to ensure the CCM operation of the
circuit, while varying the load (�5) and the duty cycle (�). The switching frequency is set to
2.32MHz, corresponding to the thickness-mode resonant frequency of 1-mm-thick PZT. The red
dots highlight the minimum inductance value at a 100Ω load, representing a typical resistance of
the PZT substrate at resonance. The 4.7µH inductor has been chosen to guarantee the operation of
the circuit with the duty cycle in the range of 0.35 - 0.48, over which the gain is in the range of 2.5
– 25 V/V.
61
Figure 5. 11 A Plot showing the minimum inductor value with varying the load (R:) and the duty cycle (D), ensuring
the Continuous Conduction Mode (CCM) operation. The red dots indicate the minimum inductance value when the
resistive load is 100Ω.
The selection of passive components also depends on the maximum output voltage and
current. With a 100 Ω resistive load connected to the output and the duty cycle of 0.48, the output
voltage is 100Vpp (when the input voltage is 4VDC) and the root-mean-square output current is
0.7A. Neglecting power loss, the input current would be 17.5A, which imposes constraints on the
size of the inductor. For instance, both XAL5030-472 (Coilcraft Inc.) and XAL1010-472 (Coilcraft
Inc.) have 4.7µH. However, the saturation currents of these inductors are 6.6A and 25.4A, while
the size are 5 × 5 ��" and 10 × 11 ��", respectively [142]. Considering the design target for
the inverter is to achieve an output voltage of 60Vpp, corresponding to an input current of 6.3A, it
is desirable to limit the saturation current of the inductor to approximately 7.0A. This ensures that
the inductors can handle the required current without reaching saturation, thereby maintaining
stable and reliable operation of the inverter circuit.
62
5.4.3 Circuit Simulation
Circuit simulations are performed under various conditions using LTSPICE (Analog
Device Inc.). For all simulation, inductance, capacitance values are set to 4.7µH and 100nF.
Furthermore, the BVD model of SFAT (fabricated on 1-mm-thick PZT) is considered as a load, as
discussed in Chapter 2.6. By measuring its series and parallel resonant frequencies using a
Network Analyzer and using equations 2.21 - 2.24, a motional inductance (�1 ), a motional
capacitance (�1), a motional resistance (�1) and a clamp capacitance (�2) are obtained to be
512µH, 78Ω, 432pF, and 2.6nF, respectively. Figure 5.12 illustrates the circuit schematic diagram
used for following simulations.
Figure 5. 12 Circuit schematic diagram for the simulation of the modified Z-source inverter with BVD model of SFAT
(red box) as a load
Initially, the assumption of ideal MOSFET (�#) and ideal diode (�#)is made and a PWM
signal with a frequency of 2.32MHz and 50% duty cycle is applied to MOSFET. The simulation
runs for 1 ms, and as anticipated, the simulated output voltage and inductor ( �# ) current
consistently increase and reaches 200Vpp within 1 ms. Figure 5.13a illustrates the output voltage
and the PWM signal applied to the MOSFET gate throughout the entire simulation time.
Additionally, Figure 5.13b provides an overview of the waveform shapes over the last 10 cycles.
63
Figure 5. 13 LTSPICE simulation results under the condition of ideal MOSFET (Q2) and ideal diode (D2) with a
PWM signal of 50% duty cycle (a) for 1-ms simulation time, and (b) the magnified view of the output voltage and the
PWM input during the last 10 cycles. The output voltage, the PWM input signal, and the inductor current are
represented with gray dashed line, red solid line, and blue solid line, respectively.
Subsequently, the duty cycle is adjusted to 47.5%, corresponding to a gain of 20V/V from
equation 5.7, in the simulation under the same ideal condition. With an input voltage of 4VDC, the
expected zero-to-peak output voltage is 80V. The simulation results show that the peak output
voltage is about 85V, slightly higher than 80V, while the average current flowing through the
inductor in the Z-source network is 7.0A (Figure 5.14).
Figure 5. 14 LTSPICE simulation results under the condition of ideal MOSFET (Q2) and ideal diode (D2) with a
PWM signal of 47.5% duty cycle (a) for 1-ms simulation time, and (b) the magnified view of the output voltage and
the PWM input during the last 10 cycles. The output voltage, the PWM input signal, and the inductor current are
represented with gray dashed line, red solid line, and blue solid line, respectively.
64
Finally, in the simulation, the ideal MOSFET is replaced with the real EPC2001, along
with its gate driver IC, LMG1205 (Texas Instruments Inc.). Additionally, RB048RSM10S (Rohm
Semiconductor) has been chosen as the Schottky diode, offering a high breakdown voltage (�=) of
100V and a high average forward current (�>) of 8.0A. Due to the rise and fall time delay induced
by the EPC2001 (1.5 ns and 1.7ns, respectively) [143] and LMG1205 (7.0ns and 3.5ns,
respectively) [144], the time period of GaNFET fully on is reduced by 2% (4.2ns), as the gate
threshold voltage of EPC2001 is 1.4V [145]. This on-time reduction results in an effective duty
cycle of 47%, corresponding to a voltage gain of 16.7V/V. With an input voltage of 4VDC, the
expected zero-to-peak output voltage is approximately 67V. The simulation shows an output
voltage of about 70Vpp and an average inductor current of 6.5A (Figure 5.15a-b). Additionally, the
rising time and falling time are measured to be approximately 7.5 ns and 3.8 ns, respectively
(Figure 5.15c).
Figure 5. 15 LTSPICE simulation results with EPC2001(Q2), its gating driver IC (LMG1205) and RB048RSM10S
diode (D2) with a PWM signal of 47.5% duty cycle (a) for 1-ms simulation time, and (b) the magnified view of the
output voltage and the PWM input during the last 10 cycles. (c) On-time of the PWM signal showing the rising and
falling time induced by EPC2001 and LMG1205. The rising and falling are measured approximately 7.5ns and 3.8ns,
respectively, leading to the effective duty cycle of 47.0%. The orange dotted line indicates the threshold gate voltage
of EPC2001(V@A = 1.4V). The output voltage, the PWM input signal, and the inductor current are represented with
gray dashed line, red solid line, and blue solid line, respectively.
5.4.4 PCB design
The inverter is consisted of two ICs (a gate driver LMG1205 and a PWM signal generator
which can be replaced by a Bluetooth Low Energy IC), one GaNFET (EPC2001), and 7 passive
65
components (2 inductors, 3 capacitors, 1 resistor, and 1 diode). Compared to the parts counts of
the on-board power amplifier, which is 53 in total, the reduction is significant, leading to the PCB
area being reduced by 65%. Figure 5.16 shows an example of PCB layout of the Z-source inverter.
The bottom layer in the PCB is allocated for a PWM generator or a BLE IC for wireless
controllability, a topic to be discussed in the next chapter. The total dimension is 12 × 14 ��"
and further improvement can be achieved by choosing smaller-sized inductors and diodes,
especially when the system's output current is limited.
Figure 5. 16 A PCB layout design of the modified Z-source inverter system, having dimension of 12mm × 24mm.
Only top layer is shown because the bottom layer space is reserved for the Bluetooth-Low Energy IC
5.4.5 Circuit Characterization
The measurement and experiment are conducted utilizing the evaluation boards
(LMG1205HBEVM, Texas Instruments Inc.), containing both GaNFET (EPC2001) and its gate
driver (LMG1205). Since the LMG1205HBEVM is designed for an H-bridge topology, the highside GaNFET is forced to be turned off and only the low-side GaNFET is used, whose drain is
externally connected to the modified Z-source network structured in a 3D configuration having a
dimension of 5 × 10 × 2 ��M(� × � × �) as shown in Figure 5.17.
66
Figure 5. 17 Photos of (a) the experiment setup utilizing LMG1205HBEVM, and (b) the bottom view, and (c) the
side-view of the modified Z-source network which is structured in a 3D configuration.
5.4.5.1 Power Efficiency
First, the power efficiency is characterized using a 100Ω resistive load with the duty cycle
of 40%, corresponding to a voltage gain of 5. The output voltage is measured to be 20Vpp, when a
4VDC input voltage is supplied by a DC power supply (DP831A, Rigol Technology), using an
oscilloscope (DS4014, Rigol Technology). Interestingly, the rising and falling time delays have an
impact on the output waveform signals, causing them to look more like a sinusoidal signal than a
PWM signal. Figure 5.18a shows the output voltage (yellow) in comparison to a PWM input to
the GaNFET (blue), while Figure 5.18b demonstrates the output voltage (yellow) and the input
DC voltage (blue), noting that the output voltages (yellow) are recorded through an 10:1 attenuator.
Figure 5. 18 Waveform of (a) the output AC voltage along with a PWM signal and (b) the output AC voltage and the
input DC voltage, captured from the oscilloscope.
67
The average input supply current is measured with a current probe (Tektronix’s TCP0030A)
to be 580mA, leading to an input power �B! = �B! × �B! = 4� × 0.58� = 2.32�. The output
power, on the other hand, is calculated as �OP% = �OP%,A1/ " /�5 » L "2
√"I
M
"
/100� = 2.00�. Thus,
the power conversion efficiency is obtained as � = �OP%/�B! ≈ 86% . Similarly, the power
conversion efficiencies are calculated for various load conditions from measured data, resulting in
a maximum efficiency of 91% at a load condition of 75 Ω, as shown in Figure 5.19.
Figure 5. 19 The power efficiency plot of the Z-source inverter reveals varying performance under different resistive
load conditions, with the peak efficiency reaching 91% at a load of 75 Ω.
5.4.5.2 Maximum Output Voltage
Although the theoretical gain of the system is infinite based on the equation 5.7, the system
experiences energy losses due to the equivalent series resistances (ESR) of the capacitors and
inductors, on-resistance (�3/,O!) of the GaNFET, and switching loss during the switching time.
Moreover, the rising and falling time delays, induced by the GaNFET and the gate driver IC, limit
the on-time of the GaNFET, resulting in a limited output voltage. The maximum output voltage of
68
the system is measured by varying the duty cycle while connecting SFAT (fabricated on 1-mmthick PZT) at the load. The output peak voltage follows equation 5.7 until the duty cycle reaches
about 45% and saturates at around 58V as the duty is increased (Figure 5.20). In theory, the
maximum duty cycle is constrained to lower than 50%, but we experimentally increase the duty
cycle beyond 50%, as the effective duty cycle (due to the rising and falling time delays) is still
lower than 50%.
Figure 5. 20 The analytically calculated (orange) and experimentally measured (blue) output peak voltages with the
varying duty cycle, when the input voltage is 4VDC and the SFAT is connected at the load.
5.4.5.3 Bandwidth of The System
While the duty cycle is kept at 40%, corresponding to a gain of 5V/V, the system's gain is
measured by varying the frequency of PWM signals to GaNGET. It is observed that the system
bandwidth is quite limited, functioning only within the range of 1.7 - 2.4MHz (Figure 5.21). When
the frequency drops below 1.7MHz, the system transitions into the Discontinuous Conduction
Mode (DCM), resulting in a waveform shape that deviates from both PWM and sinusoidal patterns.
When the frequency is higher than 2.4MHz, the gain falls below the anticipated value.
69
Figure 5. 21 The plot showing the gain of the system within the frequency range between 1.7MHz to 3.2MHz, without
load (Black) and with a 75Ω load (Red). The bandwidth of the system is limited to the frequency range of from1.7MHz
to 2.4MHz.
Lastly, a 100F LIC is connected to the Z-source inverter with the SFAT load, and the
operating time is measured as the same method used in the on-board power amplifier. The
operating time with the output voltage of around 30Vpp is measured to be 5 minutes and 50 seconds,
an almost twofold improvement compared to the on-board power amplifier. Table 5.2 provides a
summary of the key feature comparison between the on-board power amplifier circuit (Chapter
5.3.4) and the modified Z-source inverter circuit (Chapter 5.4.5).
Table 5. 2 Key Feature Comparison between On-Board Power Amplifier and Z-Source Inverter
Features On-board Power Amplifier Z-source Inverter
Maximum Output voltage 26Vpp 58Vpp
Maximum Efficiency 66% (at 130Ω) 91% (at 75Ω)
Bandwidth �!"#$ = 5MHz 1.7 -2.4MHz
Time of Operation with a 100F LIC 3min 10 sec (190 sec) 5min 50 sec (350 sec)
70
5.5 Pulse operation using a Bluetooth Low Energy (BLE)
Pulsed operation of SFAT is used in various applications such as a droplet ejection [146],
cancer treatment [147], and neuromodulation [148], as it delivers an acoustic energy with a precise
control over the timing and the reduced heat generation, compared to continuous operation. A
series of acoustic pulses can be characterized by a pulse repetition frequency (PRF) and the number
of cycles per pulse, which has a relationship expressed as (Figure 5.22):
������ �� ������ = ��/ �/ = �� × �/ (5.13)
where ��, �/ and �/ are a pulse duration (PD), a period and a frequency of transducer driving
signal, respectively.
Figure 5. 22 Conceptual illustration of the pulsed operation of acoustic transducer, showing the key parameters such
as the number of cycles (# of cycles), a pulse duration, a pulse repetitive frequency (PRF), and the operating frequency
of the transducer.
Implementing a wireless connectivity through a BLE is explored for the pulse-mode
operation of the acoustic transducers. This is achieved by integrating a load switch IC between the
power source and the system described in Chapters 5.3 and 5.4, as depicted in Figure 5.23. When
the BLE IC, integrated with microcontroller unit (MCU), is used, it can generate PWM signals
with desired PRF and PD to control the load switch. Additionally, for the on-board power amplifier
71
design, it can serve as a host controller to modify the frequency of the digital oscillator (LTC6904),
replacing the use of an analog oscillator (LTC1799) (Figure 5.23a). In case of the Z-source inverter
design, the BLE IC can be used to generate two PWM signals, one for controlling the load switch
and the other for supplying the PWM signal to the gate driver IC in the inverter system (Figure
5.23b).
Figure 5. 23 The system block diagram for the pulsed mode operation of the electronic circuit with a Bluetooth-Low
Energy (BLE), implemented on (a) the power amplifier design and (b) the Z-source inverter design.
For preliminary study, a wireless chipset CYBLE 416045-02 (Infineon Technologies), a
BLE module with integrated antenna, is programmed to generate PWM signals and digital signals.
TPS22995 (Texas Instruments Inc.) is selected as the load switch device, which has a rising time
of 100 µs and a propagation delay of 100 µs. The demonstration is conducted with the Z-source
inverter and the pulse operation is controlled with PRF of 10Hz and pulse width of 10ms,
corresponding to approximately 23,000 cycles per pulse of the operating frequency of 2.32MHz
(Figure 5.24). Due to the propagation delay time of TPS22995, the output voltage is also delayed
by approximately 100µs. Furthermore, the capacitance value of the Z-source network affects the
rising and falling time of the pulsed output signal (Figure 5.24b).
72
Figure 5. 24 Waveforms of a pulsed output delivered to SFAT (Yellow), input voltage to the Z-source inverter
(Skyblue), and a pulse control PWM signal from BLE (Blue), showing (a) three pulses with a Pulse Repetition
Frequency (PRF) of 10HZ and a Pulse Duration (PD) of 10ms and (b) a single pulse with PD of 10ms.
73
Chapter 6
Wireless and Stand-Alone Submarine Propeller Based on Acoustic
Propulsion
6.1 Introduction
Underwater robotic systems play a significant role in diverse industrial, military and
biomedical applications [149], for exploration, salvaging, sophisticated delivery, and
environmental monitoring, a task that often requires intensive labor. In case of environmental
monitoring, the tasks include collecting the samples or monitoring the quality of water and other
environmental variables [150]. In case of application to security, an unmanned underwater vehicle
(UUV) robot monitors a designated area with the goal of identifying potential threats such as
intrusion [151]. The integration of miniaturized robots, especially if the size is on a micron (µm)
scale, show great promise in healthcare including drug delivery, capsule endoscopy, etc. These
microrobots can be made to carry imaging or surgical systems within the human body [152].
Traditional underwater propulsion methods, such as mechanical propelling fans or motors, are
susceptible to wear and tear, and potentially damage the surrounding environment with their
moving parts. For the environmental monitoring purposes, several researchers have developed a
robotic fish, mimicking fish movement with structures like fish fins for underwater propulsion, but
still requires moving parts [152].
On the other hand, acoustic propulsion offers an alternate solution which exploits the
energy of sound waves to generate a thrust underwater [125][126]. Unlike mechanical propellers,
acoustic propulsion systems do not require any moving parts or rotating components, leading to
the reduced risk of entanglement with the surrounding environment. This is critical not only in
74
industrial applications but also in drug delivery robots, where the surrounding environment may
include blood cells and tissues. Furthermore, eliminating the need of moving components reduces
maintenance requirements and increases the operational lifespan time.
In this chapter, a wireless and stand-alone underwater acoustic propelling system is
introduced, integrating a Self-Focusing Acoustic Transducers (SFAT) and on-board power
amplifier circuit (described in Chapter 5.3), powered by a 100F Lithium-ion capacitor (LIC). The
acoustic propulsion generated by this submarine system is characterized and compared with other
types of acoustic transducers such as different designs of Spiral-arm Vortex-beam Acoustic
Transducer (SVAT).
6.2 Device Design for The Acoustic Propeller
For the acoustic propeller, a SFAT is designed with a Fresnel air-cavity lens, focal length
of 5mm and 6 rings in total and fabricated on 1-mm- thick PZT, whose thickness-mode resonant
frequency is about 2.32MHz. Acoustic waves generated from SFAT are absorbed by the fluid as
the waves travel through the water, resulting in a reduction in the mean momentum flux [154]. The
acoustic pressure gradient provides a hydrodynamic drag force, associated with acoustic radiation
force and acoustic streaming [155]. On a fabricated SFAT, a laser- machined acrylic sheet (2 mm
thick) is glued to the opposite side of the Fresnel lens in order to form the air-cavity reflector, so
that there is a net propulsion force directed upward, as illustrated in Figure 6.1.
Figure 6. 1 Conceptual illustration of generating acoustic propulsion with a Self-Focusing Acoustic Transducer (SFAT)
with air cavity reflector and Fresnel air-cavity lens.
75
6.3 System Design
The SFAT-based acoustic propeller integrates four parts on a single stand-alone platform,
as shown in Figure 6.2: a SFAT for focused ultrasound (2.32MHz), an on-board power amplifier
driving electronics on a flexible printed circuit board, a lithium-ion capacitor (LIC) for power, and
a coil for wireless charging of the LIC. The flexible PCB wraps around the LIC located in the
center, while the wireless power-charge receiving (WPRX) coil and the SFAT are placed at the top
and bottom, respectively. The completed system (Figure 6.2b) is 18 x 18 x 38 mm3 in volume and
weighs 16.10 grams (which can be adjusted by filling the inside air volume with liquid or weight,
if desired), resulting in a mass density of 1.020 g/cm3.
Figure 6. 2 (a) Conceptual 3D schematic of the acoustic propeller and (b) photo of the completed acoustic propeller.
Lithium-ion capacitor (LIC) cannot be seen in the photo as it is wrapped by the flexible printed circuit board (PCB).
The drive electronics are designed based on the on-board power amplifier circuit design, described
in Chapter 5.3, and built on a flexible PCB (Figure 6.3).
76
Figure 6. 3 (a) Schematic of the PCB layout showing the four main circuits sections and the places where SFAT and
LIC will be connected and (b) photo of the fabricated flexible PCB before being folded into a 3D shape shown in
Figure 6.2b.
This drive electronics delivers 30Vpp 2.32MHz sinusoidal signal to the propeller when
powered by a 100F lithium-ion capacitor (Xeno Energy’s XLC-1030). The chosen 100F LIC is
capable of delivering enough power and energy to the acoustic propeller to generate the acoustic
propulsion in the water to make the propeller propagate stably for several minutes. The LIC is fully
discharged before assembling the components for the propeller, in order to prevent potential
damage to the SFAT (in air) due to the power from the LIC. To facilitate the recharging of the LIC,
a wireless charging feature is incorporated into the circuit design (Figure 6.4).
77
Figure 6. 4 Functional block diagram of the system showing the power rails (black lines) and the signal rails (blue
lines). Wireless power receiver block is added on the input voltage, so that it can supply power to the system and
charge the LIC at the same time.
After assembling all the components, 20μm thick Parylene is conformally deposited over all the
surface of the assembled acoustic propeller to make the unit immersible in liquid. The LIC is
wirelessly charged (with Taida Century Technology’s T3168) through a wireless charging receiver
(WPRX) coil (3- layers, 26μH, 1.1A, 520mΩ max) on the top of the assembled propeller unit. The
current flowing through the wireless charging coils is controlled not to exceed 700mA to avoid
potential damage to the Parylene layer due to excessive heat when the air-gap between the
transmitting and receiving coils is less than 1mm (Figure 6.5).
Figure 6. 5 Photos showing (a) wireless charging of LIC, while the SFAT in the propeller is immersed in water to
prevent thermal damage of the SFAT during the charging and (b) air gap between transmission (TX) and receiving
(RX) coils maintained by a pair of rigid wires and a clamp holding TX and RX coils, respectively, with alignment
between TX and RX coils achieved by a 3-axis micromanipulation stage.
78
6.4 Experimental Results
6.4.1 Acoustic Propulsion with SFAT
For testing and quantifying the acoustic propulsion, the mass density of liquid solution is
adjusted to be 1.15g/cm3 by adding sodium polytungstate (SPT) powder into deionized water (DI
water) at a weight ratio of 1:4.89 (SPT:water) such that the acoustic propeller sinks down close to
the floor without generating any propulsion but still floats above the floor (Figure 6.6a). The
acoustic propeller floating in the liquid experiences the gravitational force downward as well as
the upward buoyancy, �N, which can be calculated by:
�N = ��� (6.1)
where ρ, g and V are the mass density of liquid, gravitational constant, and object volume,
respectively. As buoyancy is a function of the density of liquid, the position of the acoustic
propeller can be adjusted by changing the mass density of SPT solution. When the SFAT generates
acoustic propulsion upward, the propeller soars up to the surface of the liquid (Figure 6.6b). With
a continuous sinusoidal signal driving the propeller, the signal amplitude decreases gradually after
being at 30 Vpp for 1.5 minutes, due to the LIC running out of energy, and when the amplitude
becomes lower than 18 Vpp, the acoustic propulsion stops, after which the propeller still moves
due to inertia for a while.
Figure 6. 6 Photos of the acoustic propeller (a) floating close to the floor without acoustic propulsion and (b) soaring
up to the surface of sodium polytungstate (SPT) solution (1.15g/cm3) when the SFAT generates the acoustic propulsion.
79
The acceleration of the vertical propulsion is obtained through measuring the traveling
distance as a function of time (Figure 6.7). From the acceleration, the acoustic propulsion force is
calculated to be 18.68μN, when the LIC is fully charged, ignoring the friction and drag forces that
resist the motion of the acoustic propeller
Figure 6. 7 (a) Measured traveling distance (in mm) and (b) velocity (in mm/s) of the acoustic propeller from the
vertical propulsion shown in Fig. 8. The traveling distance is measured in every 100ms and the averaged acceleration
is calculated to be 1.16 mm/s2
The electrical power PE consumed by the SFAT can be calculated with:
�U = I8&(
"
|W|" × � (6.2)
where �A1/, �, and � are the root-mean-square (rms) voltage, SFAT’s impedance, and the real part
of the impedance, respectively. The impedance of SFAT is measured using a one-port reflection
coefficient S11, as explained in Chapter 2.6.1. The impedance magnitude |Z| and resistance R of
the SFAT at its resonant frequency, which is measured as 2.27MHz, are 321.61 and 319.89 Ω,
respectively. However, at the 2.32MHz which the electronic circuit generates and delivers to the
SFAT, |Z| and R are measured as 256.09 and 209.10Ω, respectively. Thus, the electrical power is
to be 358.70mW, and the propulsion force per electrical power consumption by the SFAT is 52.07
μN/W.
80
To move the acoustic propeller horizontally, SPT powder is added to DI water at a weight
ratio of 1:3.54, resulting in the SPT solution’s mass density being 1.20g/cm3 which leads to the
acoustic propeller floating on the top surface of the solution while the SFAT part is still immersed
in the solution as shown in Figure 6.8.
Figure 6. 8 Photo of the acoustic propeller floating horizontally on sodium polytungstate (SPT) solution (1.20g/cm3).
Since some portion of the acoustic propeller is exposed to the air, it experiences more drag force
than the vertical movement. Also, as it is tilted with an angle of 27° on the surface, only 89%
(cos27°) of acoustic propulsion contributes to the lateral motion.
Another way of estimating the acoustic propulsion is using a drag force. An object moving
at a velocity, �, in a fluidic media experiences a drag force �3:
�3 = #
" ��"��3 (6.3)
where �, � and �3 are the liquid density, cross-sectional area, and drag constant, respectively.
When the acoustic propeller travels with the constant velocity, the acoustic propulsion force is
equal to the drag force so that there is zero net force, leading to no acceleration. With the measured
steady-state velocity of the acoustic propeller moving horizontally at 6 mm/s, the drag force Fd is
calculated to be 11.51 μN. The calculated acoustic propulsion in the horizontal motion using the
steady-state drag is smaller than that of the vertical propulsion because the propeller receives lower
81
voltage and power from the LIC by the time when the acoustic propeller reaches the steady-state
velocity.
6.4.2 Acoustic Propulsion with SVAT
To improve the acoustic propulsion per electrical power consumed, various SVAT designs
for the acoustic propeller are explored, as an acoustic vortex could provide a more pressuregradient in the medium. Table 6.1 shows the various design of (single arm and multi-arms) SVATs
used in the measurement. Other than SVATs, an 18-sectored SFAT (Chapter 4.2.1) and a new
design of combining SVAT and SFAT (A0 SVAT embracing a two-rings SFAT in its the center
cavity area) are also explored for the acoustic propulsion (Figure 6.9).
Figure 6. 9 The designs (left) and photos of the fabricated transducers (right) of (a) 18-sectored Self-Focusing
Acoustic Transducer (SFAT) based on annular Fresnel rings and (b) 1- arm SVAT and SFAT with two rings in the
center.
82
Table 6. 1 Various Designs of Spiral-arm Vortex-beam Acoustic Transducers (SVATs)
Characterizing the acoustic propulsion by SVATs is conducted with a wired operation,
powered by the RF Power Amplifier (Amplifier Research’s 75A250). Acoustic propulsion force
generated by the transducer is measured by quantifying the acceleration of a pendulum swing, as
the transducer is suspended with flexible wires. With 100Vpp sinusoidal wave applied
(continuously) to the transducer, the propelling motion is observed and captured with a video
camera of 30 frames per second (fps), and its traveling distance, velocity, and acceleration are
obtained every frame (33ms) by a fine precision ruler immersed in the water, as shown in Figure.
6.10.
Design Parameter
(�%, �&, �%) 1-Arm SVAT 2-Arms SVAT 3-Arms SVAT 4-Arms SVAT
A0 Design
(2.3mm, 2.5mm, 0.090)
A1 Design
(0.5mm, 0.7mm, 0.090)
A2 Design
(0.2mm, 0.3mm, 0.083)
83
Figure 6. 10 Photos of the propeller (the six-ring SFAT with 4.5 mm focal length) moving at (a) t=0 and (b) t = 660ms
Figure 6.11 shows the measured traveling distance and velocity of the 4-arms SVAT (A2 design in
Table 6.1) acoustic propeller. By multiplying the acceleration with the transducer mass (2.86 grams
in average including the mass of the air-cavity reflector on one side), the acoustic propulsion forces
are obtained and summarized in Table 6.2.
Figure 6. 11 (a) Measured traveling distance (in mm) and (b) velocity (in mm/s) of the 4-arms SVAT (A2 design)
acoustic propeller. The traveling distance is measured in every frame (33ms) and the averaged acceleration is
calculated to be 238.8 mm/s2
84
Table 6. 2 Summary of Fabricated and Tested Underwater Acoustic Propellers
Features SFAT 1-Arm SVAT
(A0 Design)
1-Arm SVAT
(A2 Design)
4-Arm SVAT
(A2 Design)
18- Sectored
SFAT
SVAT+SFAT
(Fig. 6.9b)
Total Weight [g] 2.88 2.86 2.84 2.88 2.93 3.05
Acceleration [mm/s2
] 74.28 98.50 218.72 238.78 112.69 170.30
Activated area [mm2
] 45.00 45.95 77.02 75.60 56.75 55.90
Propulsion per Power [µN/W] 53.68 70.69 155.87 172.56 82.56 130.34
Propulsion per Power per Area
[µN/W/mm2]
1.19 1.53 2.02 2.28 1.46 2.33
6.5 Discussions
Experimental results among the 1-arm SVATs show that the one with a larger transducer
area (A2 design in Table 6.1) generates higher propulsion force than other designs. Also, as the
number of arms increases, higher acoustic propulsion force is obtained. For example, the acoustic
propulsion force from the 4-arms A2-design SVAT (Figure 6.11) is measured to be 711.6 μN, while
the 1- arm SVAT (same design) generates 621.2 μN, even though the active electrode area of the
4-arms SVAT (75.60 mm2) is smaller than that of the 1-arm SVAT (77.02 mm2). Since the acoustic
pressure at the focal point depends on the total active electrode area of the transducer, the acoustic
propulsion forces per electrical power per transducer area also are compared in Table 6.2. Although
4-arms A2-design SVAT generates the highest acoustic propulsion per electrical power, 172.6
μN/W which is 3.2 times larger than that of the SFAT (53.68 μN/W), the highest acoustic
propulsion force per power per transducer area is obtained from the combined design of the SVAT
(A0 design) with SFAT (2.33 μN/W/mm2), and is almost double of that of the SFAT-based
propeller (1.19 μN/W/mm2). As can be seen in Figure 6.12, the SVAT/SFAT combination produces
much stronger propulsion force than the SFAT alone (Figure 6.10). However, the vortex beam from
85
the SVAT produces forces that are not perpendicular to the transducer surface, resulting in a
twisting motion as the transducer propels itself in the direction perpendicular to the transducer
surface. This phenomenon is observable in other SVAT- based propellers, limiting their
implementation on the flexible PCB.
Figure 6. 12 Photos of the propeller (1- arm SVAT(A0) + two- ring SFAT) moving at (a) t = 0, (b) t = 330 ms, showing
a twisting motion, (c) t = 660 ms, and (d) t = 1 s.
6.6 Summary
This chapter introduces a wireless and stand-alone underwater acoustic propelling system
for underwater robotic systems. Acoustic propulsion, using a focused ultrasound waves to generate
thrust, is presented with the design of a Self-Focusing Acoustic Transducer (SFAT) having a
Fresnel air-cavity lens for the acoustic propeller, highlighting its advantages over traditional
methods. The system integrates the SFAT with an on-board power amplifier, a lithium-ion
capacitor (LIC) for power, and a wireless charging coil. Experimental results demonstrate the
system's ability to generate acoustic propulsion and to move in both vertical and horizontal
direction in sodium polytungstate (SPT) solution with mass density of 1.15 - 1.20 g/cm3
. The
driving electronics integrated in the system provides a 2.32MHz sinusoidal signal whose amplitude
is about 30Vpp, powered by a single 100F LIC and operates for about 90 seconds. When the 30 Vpp
86
is applied to the propeller, it generates the maximum acoustic propulsion of 18.68μN, while the
SFAT is measured to consume 358.70mW.
Additionally, various designs of Spiral-arm Vortex-beam Acoustic Transducers (SVATs)
are explored for improved acoustic propulsion, comparing their performance to the SFAT. The
chapter concludes with discussions on the experimental results and the potential of combining
SVAT and SFAT for enhanced propulsion force. The acoustic propulsion force generated by the
SVAT is measured to be up to 3.2 times larger than that by the SFAT-based propeller, while the
highest acoustic propulsion force per electrical power per transducer area is obtained with the
combined design of SVAT and SFAT. From the wired experiments, the 4-arms A2-design SVAT is
measured to produce 711.6 μN acoustic propulsion force, 3.35 times larger than the SFAT-based
propeller
87
Chapter 7
Levitating Acoustic Tweezers Based on SVATs
7.1 Introduction
Trapping and manipulation of a particle without contact, heat or label is highly desired
when the particle, such as a live cell, is prone to get damaged by physical contact or heat.
Contactless trapping of a particle can be achieved by various technologies such as magnetic,
optical, or acoustic tweezers. Magnetic tweezers utilizes magnetic fields, involving manipulating
magnetic particles with external magnetic fields. The target particles are magnetic beads or
particles coated with magnetic materials so that they are responsive to the external magnetic fields
[156]. Magenetic tweezers are often used for the study of the living cells using ferrofluids which
are magnetic nanoparticles in aqueous suspensions [157]. The introduction of ferrofluids into cells
through endocytosis enables the examination of cytoplasmic and organelle structures, with a
magnetic force being largest on the areas where the magentic particles accumulate [157]. However,
magnetic tweezers require magnetic materials as a label or force enhancer. Optical tweezers uses
a highly focused laser to exert the force around the particle to trap[156], having advantages of high
precision and high speed manipulation. However, the concentrated light during the trapping
generates a localized heating on the trapped particle, which may damage the particles, especially
live cells [159]. On the other hand, acoustic tweezers can offer contactless, heatless, and label-free
trapping of a particle, and has been explored with both active lens and passive lens, as acoustic
trapping can be achieved by various pressure patterns. Similar to optical tweezers, a combination
of focused acoustic waves with different focal points generates an energy well that can trap a
88
particle in a fluid [161][162]. Acoustic Bessel-beam also has been shown to be capable of
entrapment and manipulation of particles [113].
In this chapter, levitating, trapping, and manipulating a large and heavy particle (2.4 mm
in diameter and 1.30 g/cm3) using Spiral-arm Vortex-beam Acoustic Transducers (SVATs), the A2
design (Table 6.2), are demonstrated. The experimental results show that the SVAT produces a
levitation force of up to 21.3μN and is capable of trapping and manipulating the particles in a
three-dimensional (3D) space. Particles of varying sizes and densities are trapped in water by
changing the driving voltage. Additionally, the particle can be lifted from the surface of the
transducer and be ejected out of the medium after being trapped by adjusting the driving voltage.
7.2 Principle of Acoustic Levitation and Trapping
Acoustic radiation force (ARF) is generally assumed to play a key role in particle trapping.
However, a recent work on a large particle (5mm in diameter with 1.03 g/cm3 density) assumed
to be trapped by ARF showed a limited manipulation of the trapped particle in 3D space when the
trapping is heavily influenced by a standing wave [162]. Acoustic trapping is also possible with an
acoustic vortex beam that produces a levitational force through ARF and acoustic streaming. In
this case, the trapped particle in liquid experiences a zero net force among the gravitational force,
buoyancy, and acoustic vortex streaming. A levitation force induced by acoustic vortex is then
calculated by subtracting the buoyancy from the gravitational force:
�X = (� − ��)� (7.1)
where �W represents the levitational force by acoustic vortex streaming in z-axis, �, �, � and � are
the mass and the volume of a particle, the mass density of the surrounding fluid, and gravitational
constant, respectively. A recent study showed a levitation force of 5.2μN by an acoustic vortex
beam and a large particle (1.5 mm in diameter with 1.3 g/cm3 density) was successfully trapped
89
and manipulated with a 4.1 mm thick piezoelectric transducer (38 mm in diameter) generating
500kHz acoustic vortex beam along with air bubbles in the liquid [113].
7.3 Device Design
For the SVAT tweezers application, A2 design of SVATs in Table 6.1 are utilized which
design parameters are summarized in Table 7.1. The n-arms SVATs are designed with rotating a
single arm by 2�/ � ,where n is the number of arms, while the azimuth coefficient is adjusted by
0.083 × �. Additionally, the termination angle is determined by �" = 12�/� to keep the length
of the outmost electrode same, regardless of the number of arms. For example, designing a four –
arms SVAT requires the azimuth coefficient of 0.332 and the rotation angle of 90 ̊, while the
termination angle is reduced to 3�. Increasing the number of the arms along the central axis
increases the number of acoustic helixes, making the particle more stably trapped inside the vortex.
Followed by the design modification, all SVATs are fabricated with a lens-less process, providing
the advantage of the simplicity in the fabrication steps.
Table 7. 1 Summary of The Design Parameters for Single Arm and Multi-Arms SVAT
SVAT Design Initial radii
(�%, �&)
Azimuth
coefficient
(�')
Initial angle
(�%)
Terminating angle
(�&) Rotation
1-Arm SVAT (A2) 0.2 mm, 0.3 mm 0.083 0 12� 0
2-Arms SVAT(A2) 0.2 mm, 0.3 mm 0.165 0 6� �
3-Arms SVAT(A2) 0.2 mm, 0.3 mm 0.244 0 4� 2
3 �
4-Arms SVAT(A2) 0.2 mm, 0.3 mm 0.332 0 3� 1
2 �
90
7.4 Experimental Results and Discussion
7.4.1 Trapping Capability
The trapping capabilities of the SVAT tweezers are investigated with the tweezers placed
at the bottom of a beaker filled with deionized water facing upward. To minimize the reflection
from the liquid-air interface at the water surface, an acoustic absorber (AptFlex F28 from Precision
Acoustics) is placed on the liquid-air interface above the top of the tweezers. Microspheres are
released from the top near the water- absorber level close to the center of the tweezers, while the
SVAT is operated with a continuous sinusoidal voltage with a frequency of 2.32 MHz, matched to
the 1-mm thick PZT’s thickness-mode resonance frequency, generated by a function generator and
amplified by a power amplifier. The SVAT tweezers traps distinct microspheres at different
locations in Z-axis. The voltage required for trapping also varies with the size and the density of
particles (Figure. 7.1).
Figure 7. 1 Photos showing the filtering and sorting of particles with different size (and density) by varying the applied
voltage: (a) 28 Vpp trapping a large and light particle (5mm in diameter and 1.03 g/cm3
), (b) 52 Vpp trapping a midsize particle (2.96mm in diameter and 1.05 g/cm3
) (c) 90 Vpp trapping a small and heavy particle (2.4 mm in diameter
and 1.30 g/cm3
)
For example, a large and low-density particle (5 mm in diameter with 1.03g/cm3 density) can be
trapped at 5 cm above from the surface of the tweezers by 28V peak-to- peak (Vpp) voltage (Fig.
4a), while a small-size and higher- density (2.4-mm in diameter with 1.30g/cm3 density) particle
91
is trapped at Z= 1.1cm with 90Vpp (Figure 7.1c). The maximum acoustic streaming force with a
SVAT tweezers is obtained with a small-size and high-density particle (Figure 7.1c) and is
calculated (by equation 7.1) to be 21.3 μN based on the measurement. The dependency of the
trapping voltage and location on the mechanical properties of a particle can be exploited for sorting
applications.
The trapped particle experiences both rotating and linearly-vibrating motion around and
along the z-axis, respectively. However, the linear vibration can be reduced by decreasing the
applied voltage, and stable trapping without linear vibration can be obtained. Interestingly, the zlocation of the trapped particle can also be adjusted by varying the applied voltage at the cost of
stability of the trapped particle. In other words, although an increased voltage introduces more
linear vibration, it enables to manipulate the z-location of a trapped particle without having to
move the tweezers in z-direction (Figure 7.2). The red arrows in Figure 7.2 indicate the trapped
particle’s vibration length, which increases with the voltage up to 120Vpp which adjusts the
trapping position as well. Additionally, if the applied voltage is increased further, the particle is
pushed all the way up to the surface of the water (Figure. 7.3).
Figure 7. 2 Lifting a trapped microsphere (5 mm in diameter and 1.30 g/cm3 density) by increasing the applied voltage,
showing more vibration (indicated by the red arrow) of the trapped particle with higher voltage; (a) trapped at z=1.1
cm without llinear vibration with 90 Vpp, (b) trapped at z= 1.5 cm with little rotation and vibration with 100 Vpp, and
(c) trapped at z=2.1 cm with a large rotation and vibration with 120 Vpp.
92
Figure 7. 3 Shooting a microsphere (5 mm in diameter and1.30g/cm3 density)bytheSVATactivatedwith140 Vpp, as the
images captured in every 50 ms from a recorded video with the z-positions of the particle being at (a) 2.1 cm, (b) 2.3
cm, (c) 2.8 cm, and (d) 4.5 cm from the tweezers surface. (e) The particle has reached the liquid-air interface.
7.4.2 Manipulation in a Three-Dimensional Space
The SVAT tweezers is also capable of moving its trapped particle with it when the tweezers
is moved in XYZ directions through a 5-axis micro-manipulation stage as illustrated in Figure 7.4.
Figure 7. 4 Schematic of the experimental setup for manipulating the trapped particle in 3D space with SVAT tweezers.
The particle is shown to follow the movement of the tweezers in both lateral (Figure 7.5a)
and vertical directions (Figure 7.5b). This capability along with the electrical controllability of the
Z-position of the particle (Figure 7.2) allows for trapping and moving a particle to a desired
position and then extracting the particle at a given position on water surface by increasing the
93
voltage to shoot the particle out of the water surface, which provides the possibility of the pick and
placement applications.
Figure 7. 5 On-demand 1-cm (which can be longer) manipulation of a 5 mm (in diameter) microsphere with density
of 1.30 g/cm3 in 3-dimensional space in (a) lateral and (b) vertical direction, as the tweezers is moved laterally and
vertically, respectively.
7.4.3 Levitation
The SVAT tweezers is also shown to be capable of levitating the particle from the surface
of the transducer. When the SVAT is off, a microsphere is placed on the SVAT’s surface close to
the center of the spiral electrode. Upon turning on the SVAT, the particle is levitated and trapped
at a location (Figure 7.6); the levitation and trapping location are repeatable. For levitating the
particle, the applied voltage needs to be about 5% higher than the trapping voltage, since a moving
object in liquid experiences a drag force �Q (equation 6.3), directed in the opposite direction of the
motion. The drag coefficient (�Q) of a 0.47, for a rigid sphere object at the Reynolds number (Re)
of 1000 [163], is used in the calculation.
94
Figure 7. 6 Sequential photos showing the levitation process of a microsphere (density of 1.30 g/cm3 and diameter of
5 mm) from the surface of the tweezers and simultaneous trapping at z= 1.1 cm. The images are captured every 50 ms
from a recorded video.
When the particle is lifted up, it is trapped with a little linear vibration for a short while,
similar to Figure 7.2, and stably trapped by reducing the applied voltage. If the applied voltage
(with which the SVAT is turned on) is too high, the particle is pushed out of the trapping area, as
it fails to meet the trapping condition. On the other hand, an applied voltage that is too small
produces acoustic vortex streaming force which is smaller than the summation of the gravitational
force and the drag force, resulting in lifting up the particle a short distance, followed by the particle
falling back on the SVAT’s surface but not necessarily to the starting point.
7.5 Summary
In this chapter, a lens-less acoustic tweezers based on a spiral-arm vortex-beam acoustic
transducer (SVAT) is introduced for trapping large and heavy particles. The experimental results
show that the tweezers is capable of trapping and manipulating different particles with various
sizes and densities. The advantages of SVAT-based tweezers include small size, high acoustic
streaming force, and simple fabrication process. The acoustic tweezers generate the levitating force
95
of up to 21.3μN, which is four times higher than the previous study [113]. Table 7.2 summarizes
and emphasizes the advantages of SVAT tweezers in comparison to the other acoustic vortex
tweezers [113][118].
Table 7. 2 Comparison of Trapping and Design Features Across Different Acoustic Vortex Tweezers
Feature SVAT Tweezers Acoustic Vortex Tweezer [113] Acoustic Vortex Tweezer [118]
Operating Frequency 2.32MHz 500kHz 1MHz
Transducer Dimension 18 × 18 × 1��" 38mm in diameter
4.1mm in thickness
50mm in diameter
Acoustic Vortex Generation Spiral-arm patterned
electrodes
PDMS Lens with a
phase delay mapping
Hyperboloidal Lens
(R = 50mm F=121mm)
Fabrication Lens-less design
(4 fabrication steps)
PDMS molding process for the
lens fabrication
3D printed Lens on top of a
commercial transducer
Levitation from the surface Yes No No
Trapping Force 21.3 µN 5.2 µN ~ 1 nN
In addition to trapping, the SVAT acoustic tweezers is shown to be able to levitate the
particle (to be trapped or not) and eject it out of the medium when applied with proper voltage
levels. This feature opens up the possibility of SVAT tweezers in the application of pick and
placement of devices or biological tissues and cells.
96
Chapter 8
Non-invasive Focused Ultrasound Neuromodulation Using a SFAT
8.1 Introduction
Neuromodulation is the technique that can be used to treat neurological disorders such as
Parkinson’s disease (PD) [164] and epilepsy [165], and is also used to investigate how brain works.
Most clinical modulation is based on invasive electrode probes, necessitating neurosurgery for the
implantation of microelectrodes into the brain, as in Deep Brain Stimulation (DBS). However, it
has been reported that implanted electrodes can cause the gradual loss of neurons within
approximately 50-70μm of the microelectrodes [166]. Various alternative neurostimulation
techniques have been explored, such as optical stimulation [167], chemical stimulation [168],
transcranial magnetic stimulation (TMS) [169], transcranial electrical stimulation (TES) [170],
thermal stimulation [171], and ultrasound stimulation [172]. Each of these techniques presents its
own advantages and disadvantages.
Optical stimulation, also known as ontogenetic stimulation, is a genetic technique designed
to activate or inhibit the neural activity using a light source. Optogenetic uses light-sensitive
channels or opsin which can be virally introduced into neurons. Activating these channels enables
the manipulation of neuronal activities with precision through light exposure. Optical stimulation
offers spatial resolution ranging from 100µm to 1mm using optical fibers, commonly employed in
optogenetic tests, and achieves temporal resolution in the order of milliseconds [173]. One notable
advantage of optogenetics is its ability to provide cell-type-specific stimulation within a tissue,
achieved through selective genetic activation and inactivation of light-sensitive ion channels [174].
97
However, optogenetic stimulation requires injection of virus and has limited penetration depth
[175].
Chemogenetic method is similar to optogenetic in that it also requires genetic modifications
of tissue, but chemical stimulation employs chemically engineered molecules or ligands instead of
light or light-sensitive channels. Designer Receptor Exclusively Activated by Designer Drugs (or
DREADDs) are commonly in chemogenetic, providing the spatial selectivity [176], but chemical
stimulation suffers a low temporal resolution with a range from minutes to hours [177]. While both
optical and chemical neuromodulation techniques have been influential in understanding brain
activity, they share a limitation in requiring genetic modification of the tissue, which hinders their
translation to clinical applications. For this reason, electrode implantation remains the preferred
method within a clinical environment [178].
TES and TMS are non-invasive techniques that stimulate nerve cells in the brain using
electromagnetic waves and magnetic fields, respectively. However, the efficacy of the current
delivered to neurons through TES is not as potent as that of TMS in eliciting an action potential
[179]. Additionally, TES faces limitations in spatial resolution, while TMS can be focused on a
small volume. Prefrontal TMS therapy is widely employed for treating major depressive disorder
(MDD) by depolarizing cerebral neurons, and it has received approval from the US Food and Drug
Administration (FDA) [180]. Despite being considered a safe tool, TMS requires a large coil to
generate the magnetic field, which is attached to the patient's head and may be accompanied by a
few side effects such as recurrent headaches and a tingling sensation on the face [181].
In contrast, focused ultrasound stimulation (FUS) offers spatial focusing as small as 10µm
in diameter [182], with a high temporal resolution as low as 1.5ms [183]. Unlike the magnetic
stimulation, FUS does not necessitate large equipment to deliver sufficient acoustic energy for
98
eliciting neuronal activities; the drive electronics can be miniaturized and made portable, as
described in Chapter 5. Ultrasonic waves have been used in clinical settings for imaging and tumor
ablation [184]. Moreover, the FDA has recently approved a focused ultrasound device for treating
patients with advanced Parkinson’s disease who experience symptoms such as mobility issues,
rigidity, or dyskinesia [185]. This underscores the growing acceptance and application of focused
ultrasound in therapeutic interventions.
In 1958 William Fry, et al. reported a temporary neural inhibition in a cat’s eye when it was
stimulated with a focused ultrasonic beam onto the lateral geniculate nuclei [186]. Since then,
ample efforts have been made for focused ultrasound neural modulation in both the central and
peripheral nervous systems. Some studies indicate that low-intensity ultrasound (ISPTA =
23mW/cm2
, ISPPA = 2.9W/cm2
) at low frequencies (0.44 - 0.66MHz) can increase Na+ and Ca2+
currents [172]. Transcranial pulsed ultrasound (ISPTA = 84.32mW/cm2
, f = 0.25 MHz) has been
reported to increase spike frequency by triggering TTX-sensitive neuronal activity without a
significant rise in brain temperature (<0.01˚C) [187]. Moreover, continuous ultrasonic stimulation
at 43 MHz for 200 ms is reported to activates Piezo1 channels in human embryonic kidney (HEK)
cells [188], as well as TWIK-related acid-sensitive K + (TASK) channels in pyramidal cells of the
hippocampus CA1 region [189]. Applying 10 MHz ultrasound (ISPTA = 4.9W/cm2
) for 20 ms with
a 1kHz pulse repetition frequency (PRF) modulates and increases Nav1.5 channel current [190].
Furthermore, ultrasonic wave has been reported to alter the passive properties of cell membranes,
including conductance and capacitance [189][191]. More recent reports suggest that FUS can both
stimulate and inhibit neurons by tuning the pulse repetition frequency (PRF) and pulse duration
(PD) [192]-[195].
99
Although various potential mechanisms of neural stimulation and inhibition by focused
ultrasound have been suggested, the underlying mechanism is still not clear and well understood.
One reason is the technical difficulty associated with a patch clamp experiment that can provide
the ionic current changes and its contribution to the neuronal firing, as electromagnetic interference
and thermal effects interfere with the patch clamp experiment. In this chapter, the focused
ultrasound neuromodulation using a self-focusing acoustic transducer (SFAT) operating at
18.4MHz is explored. This investigation involves conducting whole-cell patch clamp experiments
on neurons located in the CA1 region of hippocampal slices obtained from rodents.
8.2 Device Design
8.2.1 NeuroSFAT
A conventional patch clamp experiment is hindered by the presence of the optically-opaque
acoustic transducer in the pathway between the light source and the tissue on which a cell is to be
patched. Consequently, various approaches have been used, such as a blind patch clamp [195],
usage of reflector cones to redirect acoustic waves [193], or emulation of acoustic effect with a
mechanical poking on a cell with a glass pipette [196]. To address these challenges, a special SFAT
for neuromodulation (neuroSFAT) is designed and fabricated on only a 127-µm thick PZT
substrate, operating at 18.4MHz. Notably, the SFAT’s annular electrodes which absorb light are
designed such as there is no electrode in the center of the SFAT (Figure 8.1) so that light may pass
through the SFAT over the region where the ultrasound is focused during the patch clamp
experiment. The high frequency is to achieve a small focal point and enhance spatial resolution for
neurostimulation. The focal length is designed to 400 µm according to the thickness of a rat brain
tissue prepared for the patch clamp experiment. This focal length ensures that the ultrasonic wave
can be focused near the top surface of the patched tissue. The transducer is designed to have 15
100
constructive Fresnel rings so that the focal size to be around 100 µm and the total dimension of
SFAT to be about 3.1 mm in diameter (Figure 8.1a).
Figure 8. 1 (a) Top view of fabricated SFAT for neuromodulation (b) Bottom view of SFAT with 0.5 mm thick
polyester sheet and (c) SFAT with an IR light from the bottom to demonstrate the translucent characteristic of 127 µm
thick PZT
The design of Fresnel half-wave band (FHWB) rings can be made with a “positive source” with a
circular electrode in the center (Figure 8.2a) or a “negative source” with no electrode in the center
(Figure 8.2b). For the patch clamp experiments, the negative source is used to allow light to pass
through the SFAT center where there is no opaque electrode. Figure 8.1c shows the translucent
characteristic of neuroSFAT when exposed to infrared (IR) light emitted from the bottom of the
transducer.
Figure 8. 2 Illustrations of (a) positive Fresnel half wavelength band (FHBW) rings design with circular electrode in
the center, and (b) negative FHBW rings with no electrode in the center. The electrodes area is indicated by black
color.
101
8.2.2 Control NeuroSFAT
Acoustic transducers not only generate an acoustic wave, but also emit the electromagnetic
interference signals and the heat, which affect neuronal activity during the patch clamp experiment.
To eliminate the impact of other variables, a control device has been designed and fabricated with
identical ring patterns (15 rings and a 400µm focal length) and operating frequency (18.4MHz).
To block the acoustic wave, an air-cavity reflector (similar to one shown in Figure 6.1) is created
through laser machining, comprising three stacked acetate sheets, each with a thickness of 127µm.
The combined thickness of the acetate sheets results in the air-cavity lens measuring about 390µm
(Figure 8.3a). These acetate sheets are attached to the SFAT using super glue and sealed with an
additional layer of parylene coating (Figure 8.3b).
Figure 8. 3 Photos of (a) an air-cavity reflector made from three pieces of laser-machined acetate sheets and (b) a
control neuroSFAT with the air-cavity reflector attached, and packaged by parylene coating
8.2.3 Low-EMI NeuroSFAT
Acoustic transducers not only generate acoustic waves but also emit electromagnetic
interference (EMI) energy into the surrounding medium. While this EMI is generally not
problematic for most applications, it can pose an issue when recording electrical signals from
102
neurons immersed in the same medium and positioned above the SFAT. Given that the amplitude
of the action potential (AP) generated by neurons is approximately 100mV, it is crucial to control
and minimize the EMI induced by SFAT to prevent it from overwhelming the AP signals. Typically,
tissues act as EMI absorbers, known to absorb EMI more effectively than water (about 5%) [197].
When the size of brain tissue is large, it effectively absorbs most of the EMI. However, in the case
of a smaller tissue size, typically prepared from a young mouse, the tissue may not absorb enough
EMI, leading to the dominance of EMI in the electrical recording.
To tackle the EMI issue, a low-EMI neuroSFAT is developed by depositing additional metal
layer on top of neuroSFAT. Any metal serves as an effective EMI shielding material to block EMI
generated from the transducer since electromagnetic waves propagating through metal attenuate
very rapidly. The amplitude of EMI at traveling inside the metal can be expressed as:
�(�) = �2�&3/Y (9.1)
where �2, �, and � are the initial amplitude, the traveling distance, and the skin depth of the metal.
The amplitude of EMI is reduced to approximately 37% at the traveling distance of skin depth.
The skin depth is also a function of the frequency of EMI signal which can be obtained from:
� = b ":
Z [8['
(9.2)
where �, �A, �2 and � are the resistivity, the relative magnetic permeability of the metal, the
magnetic permeability of vacuum (�2 = 4� × 10&\ �/�), and the angular frequency (� = 2��)
of the EMI signal. Aluminum (Al), Copper (Cu), and Nickel (Ni) are commonly utilized metals
for EMI shielding [198], and their material properties along with skin depths at a frequency of
18.4MHz are listed in Table 8.1.
103
Table 8. 1 The Material Properties and Skin Depths of Aluminum, Copper, and Nickel
Metals Aluminum Copper Nickel
Resistivity [µΩ•cm] 2.65 1.68 6.84
Relative Permeability 1.00 0.99 600
Skin Depth [µm] 19.12 15.20 1.25
Nickel is selected due to its lowest skin depth, due to its high relative magnetic permeability.
A 1.25µm thick Nickel layer is deposited on top of the SFAT with 100nm thick Titanium (Ti) layer
acting as an adhesion layer. Before the deposition, a thermal release tape (Semiconductor
Equipment Crop.), shaped in a circle with a 1.5-mm diameter, is manually positioned at the center
of SFAT (Figure 8.4a and b). This serves as a mask to protect the central area, preserving the
translucency of SFAT. Then it is released by elevating the temperature to 90˚C in the oven.
Figure 8. 4 Photos of (a) neuroSFAT with a thermal lease tape mask at the center before depositing Nickel, (b) the
microscopic image of the thermal lease tape mask, (c) the fabricated low-EMI neuroSFAT, and (d) the microscopic
image validating the translucency of low-EMI neuroSFAT
104
8.3 Experiment Design
8.3.1 Whole-cell Patch Clamp Setup
The whole-cell patch clamp technique is a method of recording the electrical activity of
individual neurons (or cells) in neuroscience and cellular physiology research, contributing to the
understanding in areas such as neuronal signaling, synaptic transmission, and the roles of specific
ion channels [199]. The experiment setup of focused ultrasound neuromodulation with SFAT and
the whole cell patch clamp is illustrated in Figure 8.5.
Figure 8. 5 (a) Conceptual illustration and (b) photo of the whole-cell patch clamp experiment setup with the presence
of the neuroSFAT
The SFAT is placed at the bottom of a transparent chamber with a continuous flow of
artificial cerebrospinal fluid (ACSF) at a rate of 2-3mL/min and rat’s brain slice is positioned on
top of SFAT (Figure 8.5). The hippocampal brain slices (350-400µm in thickness) are prepared
from Sprague Dawley rat (3-4weeks old and 250-300g body weight) and transferred into a prechamber filled with ACSF (aerated with 95% O2 / 5% CO2) composed of (in mM) : 124 NaCl, 2.5
KCl, 1.25 KH2PO4, 1 MgSO4, 2 CaCl2, 26 NaHCO3, and 11 D-glucose. Then the slices are
incubated in ACSF at 35˚C for the first 15 minutes and kept in ACSF at room temperature for at
105
least 1hour before recording the neuronal activity. For a whole-cell patch clamp, a borosilicate
glass capillary is pulled with a horizontal puller (P-1000, Sutter Instrument) to build micropipette
recording electrodes. Inside the micropipette, the patch pipette solution is filled with (in mM) 125
K-gluconate, 10 KCl, 10 HEPES, 1 EGTA, 2 MgCl2, 0.1 CaCl2, and 4 ATP-Na2. The recording
electrode is connected to the preamplifier (MultiClamp 700B Microelectrode Amplifier from
Molecular Devices) and the recording acquisition system with a sampling rate of 10kHz. The SFAT
is driven with pulsed sinusoidal signals by the power amplifier (75A250, Amplifier Research Corp.)
and the signal generator (AFG-3252, Tektronix Inc.).
8.3.2 Focused Ultrasound Stimulation Parameters
First, three peak-to-peak voltage values are chosen representing the low, mid, and high
voltage ranges: 40, 65, and 115Vpp, with corresponding ISPPA values calculated by equation 2.25 as
0.93, 2.48, and 7.75 W/cm2
, respectively. Then, various pulse repetition frequencies (PRFs) are
tested, ranging from 1Hz to 200Hz, with the number of cycles per pulse varying between 5k cycles
and 65k cycles per pulse. Utilizing equations 2.26, ISPTA values are computed, guiding the
experimental conditions, spanning from 1 mW/cm2 to 1.03 W/cm2
. Adhering to safety guidelines
from the US Food and Drug Administration (FDA) for diagnostic ultrasound (ISPPA <190W/cm2
and ISPTA <720mW/cm2) [200], the ultrasound stimulation parameters for neurostimulation
provide room for an increase in peak-to-peak voltage applied to SFAT and are within the safe limits
for ISPTA in most cases. However, higher peak to peak voltage generates more heat from SFAT,
posing a risk of damage to either the transducer or the neurons. The limit of the driving voltage is
also influenced by the number of cycles per pulse and PRF, as stress and heat on and from SFAT
accumulate over the pulse duration.
106
Ultrasound stimulation is applied for 30 seconds, followed by a 2-minute resting period to
allow cells to recover from any potential thermal effects induced by the ultrasound. In some cases,
positive or negative current is injected into the cell to adjust the baseline membrane potential. For
example, assessing the effect of FUS on neurons with a low resting membrane potential (without
spontaneous action potentials) can be challenging. Injecting a positive current depolarizes the
baseline membrane potential, enabling a clearer comparison of the number of action potentials
before/during/after FUS. Conversely, in neurons with a high baseline membrane potential that
exhibit bursting spikes, applying negative current inhibits bursting spikes and stabilizes the
neuronal firing pattern, facilitating analysis of the FUS effect. Figure 8.6 illustrates the recorded
action potential alongside the time profile of FUS, where the stimulation is active for 30 seconds
followed by a 2-minutes resting period.
Figure 8. 6 The recorded action potentials (black) along with the activation of focused ultrasound stimulation (blue).
FUS is on for 30 seconds, followed by 2 minutes of resting period.
8.4 Experiment Result
The neuronal activities in CA1 region of hippocampus are recorded from the hippocampal
slices, prepared from two types of rodents: one is an 8-days old (P8) mouse and cultured in the
incubator for 7 days, and the other is a 3-4 weeks old Sprague Dawley rat. The size of hippocampal
slice from 3-4weeks old Sprague Dawley rat is approximately 5 × 5 ��" (Figure 8.7a and b),
107
whereas the size of the tissue from the P8 mouse is about 1 × 2 ��" (Figure 8.7c and d). The
cultured slices from P8 mouse are used to validate the low-EMI neuroSFAT, as the EMI absorption
by the tissue is minimal in P8 mouse. For all other experiment results in the subsequent chapters,
the acute slices of 3-4 weeks old Sprague Dawley rat are utilized.
Figure 8. 7 Photos of (a) acute slices from 3-4 weeks old Sprague Dawley rat, (b) hippocampus (red box in (a)) region
of the slice, (c) cultured hippocampal slices from an 8-days(P8) old mouse, and (d) a magnified photo of single
hippocampal slice from a P8 mouse.
8.4.1 Electromagnetic Interference (EMI) Effect
First, the low-EMI neuroSFATs are tested in the patch clamp experiment chamber, filled
with ACSF solution, without the brain tissue. The recording electrode and the ground electrode are
immersed in ACSF solution, measuring the EMI signal generated by SFAT (Figure 8.8a). A pulsed
18.4MHz sinusoidal signal with an amplitude of 30Vpp is delivered to SFAT with PRF of 10Hz
and 10,000 cycles per pulse. The EMI picked up by the recording electrode in ACSF solution is
about 4Vpp and 240mVpp with a neuroSFAT and a low-EMI neuroSFAT, respectively, resulting in
an approximately 94% reduction in EMI (Figure 8.8b).
108
Figure 8. 8 (a)The illustration of the EMI recording setup in the patch clamp chamber, filled with an ACSF solution.
The recording electrode measures the EM, generated by the SFAT and propagated through the ACSF solution. (b) The
measured EMI with a low-EMI neuroSFAT (Ch2, blue) and the driving voltage of the low-EMI neuroSFAT (Ch1,
Yellow). The amplitude of EMI is measured as about 8% of that of driving voltage.
Next, a neuron in a hippocampal slice from an 8-day-old mouse is patched with the
recording micropipette electrode, and its neuronal activities are recorded while gradually
increasing the amplitude of the driving voltage. The EMI picked up by the recording electrode
during patching exponentially increases at a driving signal amplitude of 40Vpp. When the
amplitude is further increased to 50Vpp, the EMI amplitude surpasses the recording range (< 1V)
and becomes dominant over the action potentials, limiting the range of the amplitude of driving
voltage (Figure 8.9). A high level of EMI causes the damages on the patched neuron, leading to
the loss of the patch.
Figure 8. 9 (a) The EMI level picked up during the patch clamp experiment varying the amplitude of driving voltage.
The EMI level exceeds the recording range when the driving voltage is higher than 50Vpp. (b) The recorded EMI (large
peak) along with the neuron’s action potential (small peak), showing that the EMI level dominates the action potential,
hindering the recording of neuronal activity.
109
In contrast, patching a neuron from the Sprague Dawley rat does not present the EMI issues,
as the tissue is large enough to absorb most of the electromagnetic energy. The amplitudes of EMI
picked up during the patch are 0.73mV, 2.04mV, and 5.43mV, when a regular neuroSFAT (without
a metal shielding layer) is driven with the amplitude of 40Vpp, 65Vpp, and 115Vpp, respectively.
8.4.2 Thermal Effect
Acoustic neuromodulation often accompanies the thermal elevation in the tissue. A recent
study in normal intact mice indicated that acoustic energy stimulates neuronal activity without a
significant rise in brain temperature (<0.01˚C) [187]. However, conflicting findings suggest that
heat is the primary effector of ultrasound neuromodulation, proposing that modulation of neuronal
activities with FUS is observed only with a pulse duration long enough to generate heating
exceeding 2˚C [190]. Whether thermal effects are the primary source of FUS neuromodulation or
not, it is crucial to control and measure temperature changes in the neuron with the focused
ultrasound stimulation.
To explore the thermal effects induced by FUS, temperature changes on the surface of the
brain slice are measured without patching the cell, using a miniature k-type thermocouple with an
800µm diameter tip from a digital cataloging thermometer (HH506RA, Omega Engineering Inc.).
Various driving parameters with ISPTA, ranging from 5 to 500 mW/cm2
, are tested, and the
temperature rise varies from 0.3 to 3.4˚C from the active device and 0.2 to 2.7˚C from the control
device. Temperature changes on the tissue surface exhibit a linear relationship with the ISPTA of
FUS (Figure 8.10). Considering heat is also conducted to the tissue by the conduction heat
generated by the transducer itself, not just by ultrasound, the effective thermal rise due to the
focused ultrasound beam would be less than 1˚C.
110
Figure 8. 10 Temperature rise on the surface of the brain tissue with the activation of (a) the active neuroSFAT and
(b) the control neuroSFAT.
8.4.3 The Passive Properties of Cell Membrane
It has been reported that FUS can modulate the passive properties of the cell membrane,
including both the membrane resistance (Rm) [191] and the membrane capacitance (Cm) [189]. To
measure the resistance and capacitance changes, a set of currents from -50 to 40pA with 10pA
increments is injected (Figure 8.11a), and the membrane potential is measured with and without
FUS (Figure 8.11b and c). Using Ohm’s law, the membrane resistance is calculated by plotting the
I-V curve as shown in Figure 8.11d, revealing that Rm decreases from 230 to 210MΩ with FUS,
representing about a 10% reduction. To measure capacitance changes, the time constant (τ) is
defined and calculated as the time when the membrane potential reaches 63% of the steady-state
value. Then, the capacitance is calculated by substituting the resistance value into � = ��, leading
to approximately 20% reduction from 0.205 to 0.157pF with FUS.
111
Figure 8. 11 (a) The illustration of the Injected current profile (-50pA to 40pA for 200ms), the measured membrane
potential changes (b)without and (c) with the focused ultrasound stimulation, and (d) I -V curve for both cases
8.4.4 The Excitability of Neuron
The direction of the excitability (excitation/inhibition) of neuron by the focused ultrasound
stimulation is explored. . Though evidence of the inhibitory effect of acoustic stimulation has been
more prevalent [172][189], recent studies indicate that FUS can modulate the firing rate of the APs
in both direction by changing the PRF and the pulse duration [192]-[195].
We, in collaboration with a Patch Clamp expert at Mayo Clinic, have designed and
conducted a number of experiments with varying the PRF and the number of cycles per pulse are
designed and conducted to find the optimal parameters that inhibit or excite the neuron’s activities.
For example, one experiment is designed with a fixed ISPPA (2.48W/cm2
) and a fixed pulse duration
(45k cycles), while the PRFs are varied within the range of 5 - 200Hz. This experiment may suggest
a dependency of neuron excitability on PRF. The experiments have been systemically conducted
with varied ultrasound parameters on the different cells and different devices, culminating in a
comprehensive dataset across 328 distinct experimental configurations. The active devices,
generating the focused ultrasound, have been utilized in 281 experiments out of 328, while the
control devices are employed in remaining 42 experiments. The determination of neuronal activity
inhibition or excitation is predicated on changes in the baseline membrane potential. Specifically,
a depolarization exceeding 2% during FUS was deemed indicative of excitation, while
hyperpolarization surpassing 2% was considered representative of inhibition. Table 8.2
112
summarizes the number of inhibitions and excitations for each device, along with their respective
percentages.
Table 8. 2 The Summary of Neuronal Inhibition and Excitation Induced by SFAT
Device Type Inhibition Excitation No Response Total
Total 99 (30%) 48 (15%) 175 (55%) 328
Active Device 92 (33%) 42 (15%) 147 (52%) 281
Control Device 7 (16%) 6 (14%) 29 (70%) 42
Figure 8.12 demonstrates two clear examples of the inhibition and excitation (Figure 8.12b)
effects of the FUS from SFAT. The inhibition effect (Figure 8.12a) is observed with the ultrasound
parameters set at the 40Vpp (ISPPA = 0.94W/cm2
), 45kCycles/pulse (PD =2.45ms) and a varied
PRF of 50-100-150-200Hz, with corresponding ISPTA values of 114 - 229 – 344 - 458mW/cm2
. The
firing rate of action potential is clearly decreased during the ultrasound stimulation. Whereas, the
excitation effect (Figure 8.12b) is not very frequent but observed with the ultrasound parameters
set at the 120Vpp (ISPPA = 8.44W/cm2
), 45kCycles/pulse (PD =2.45ms) and a varied PRF of 10-
15-20-25-50Hz, with corresponding ISPTA values of 206 - 309 – 412 – 515-1031mW/cm2
. However,
it is also detected that even with the same ultrasound parameters, neurons could exhibit either
excitation or inhibition by FUS, depending on its cell type.
113
Figure 8. 12 (a) The inhibition effect of the Focused Ultrasound Stimulation (FUS) with the parameters set at 40Vpp
(ISPPA = 0.94W/cm2
), 45kCycles/pulse (PD =2.45ms) and a varied PRF of 50-100-150-200Hz in order. (b) The
excitation effect of FUS with the parameters set at 120Vpp (ISPPA = 8.44W/cm2
), 45kCycles/pulse (PD =2.45ms) and
a varied PRF of 10-15-20-25-50Hz.
8.5 Discussions
Based on the observation and findings from the patch clamp experiments, we hypothesize that
FUS can influence neurons through two primary mechanisms: mechanical stimulation caused by
ultrasound waves and thermal stimulation generated by heat. Considering that thermosensitive ion
channels, which contain thermal receptors, are mostly associated with excitatory neurons [201],
the excitatory effect of FUS might be induced by the heat generated by the transducer and the
focused ultrasound beam. The most popular thermosensitive ion channel families are Transient
Receptor Potential (TRP) channels and Two-Pore Potassium (K2P) channels, with their varying
114
activating temperature. In contrast to TRP channels, it's important to note that not all K2P channels
are thermosensitive. Figure 8.13 specifically includes thermally influenced K2P ion channels and
TRP ion channels [201].
Figure 8. 13 Thermosensitive ion channel family with their activating temperature. Excitatory and inhibitory neurons
are represented by round and square shapes, respectively, and the color indicates the gradient of their activation
temperatures.
Indeed, numerous studies have explored the excitatory effect of FUS in relation to TRP
channels. Recently, Yoo et al. reported that FUS-induced opening of TRP channels leads to an
increased Ca2+ concentration, subsequently activating Ca2+-gated Na+ channels and depolarizing
the neuron [202]. Table 8.3 summarizes additional studies on FUS with respect to TRP channels,
outlining their mechanisms and specific blockers for each channel.
Table 8. 3 Previous FUS Studies Related to TRP Channels
115
• Excitatory effect of FUS Neurostimulation
† No significant effect of FUS Neurostimulation
x Not cortical neurons
In addition to TRP channels, K2P channels are known for their responsiveness to various
stimuli, including heat, chemical, and mechanical factors. They play a role in facilitating outward
K+ ion movement, resulting in hyperpolarization of the membrane potential and subsequent
neuron inhibition [205]. The hypothesis proposes that K2P ion channels are responsible for that
the inhibitory effect of FUS. K2P ion channels encompass various subfamilies, including TWIK,
TREK, TASK, TRESK, and others. However, research on FUS has primarily focused on TREK
channels, as shown in Table 8.4.
Table 8. 4 Previous FUS Studies Related to K2P Channels
Subfamily Subtype Primary Mechanism Blocker Previous Study
TWIK TWIK1/2 Voltage-gated Quinidine* None
TREK
TREK1/TREK2
Mechanical Stimuli/ Temperature
BaCl2 (10mM)* [190]Kubanek et. al.(2016)
TRAAK
TRAAK Unknown [206]Sorum et. Al.(2020)
TASK TASK1/2/3 Extracellular pH (Acidic) /
Mechanical Stimuli Bupivacaine None
TALK TALK Extracellular pH (Alkaline) Quinidine* None
THIK THIK Voltage-gated Quinidine* None
TRESK TRESK Intracellular Ca2+/Temperature /GPCR Quinidine* None
• BaCl2/Quinidine blocks all potassium channels (Including Voltage gated potassium channel)
Subfamily Subtype Primary Mechanism Blocker Previous Study
TRPC
TPRC1 G-protein coupled receptor
(GPCR) / Mechanical Stimuli
GsMTx4 (10 μM) [202] Yoo et. al. (2021)*
TRPC1 Unknown [203] Burk et. al. (2019)*X
TRPV
TRPV1 Temperature (Heat) Ruthenium Red
(RR, 1 μM) [202] Yoo et. al.(2021)†
TRPV2 Heat/ Memebrane Stretch
TRPV4 Heat/ Memebrane Stretch
TRPM
TRPM2 Intracellular Calcium In-vivo [204] Yang et. al.(2023)*
TRPM7 Intracellular Calcium sgRNA [202] Yoo et. al.(2021)†
TRPA TRPA1 Temperature(Cold) and
Mechanical Stimuli
HC030031
(40 µM) [196] Oh et al.(2020)*
TRPP TRPP1/TRPP2 Mechanical Stimuli sgRNA [202] Yoo et al.(2021)*
116
Among these subfamilies, the TRESK channel emerges as a particularly promising candidate
potentially associated with focused ultrasound neuromodulation. This is due to the fact that the
activation mechanism of TRESK involves intracellular calcium concentration, a factor that has
been reported to increase with FUS [202].
8.6 Summary
A Focused Ultrasound Stimulation (FUS) technique using a Self-Focusing Acoustic
Transducer (SFAT) for the patch clamp experiment is explored for its neuromodulatory effects on
neurons, particularly in the CA1 region of hippocampal slices obtained from rodents. The
neuroSFAT is specially designed and fabricated on a 127-µm thick PZT to ensure the light passing
through the substrate. This neuroSFAT operates at 18.4MHz, allowing for a small focal point
(97µm) and enhanced spatial resolution during the patch clamp experiment. Furthermore, a lowEMI neuroSFAT is examined by adding a metal (Ni) shielding layer on top of the neuroSFAT to
reduce the electromagnetic energy generated by SFAT. The patch clamp experiment results show
that FUS reduces the passive property such as the membrane resistance and membrane capacitance
by 10% and 20%, respectively. Moreover, the excitability neurons induced by FUS is investigated
in both excitatory and inhibitory directions. Finally, a proposed hypothesis suggests TRP channels
contribute to the excitatory effect, while K2P channels, particularly the TRESK channel, play a
role in FUS-induced inhibition.
117
Chapter 9
Contactless Single Cell Extraction from Monolayer Cell Culture
Using Acoustic Droplet Ejector
9.1 Introduction
Contactless and damage-less cell extraction is an unmet need for assessing cell quality,
especially in a monolayer, before employing the monolayer for healthcare. It allows the isolation
of individual cells without damaging the remaining cells at the edge of the extraction spot. Human
retinal pigment epithelium (RPE) cells located behind the photoreceptor cell of the eye degrade
due to age-related macular degeneration (AMD) which is a leading cause of vision loss for people
over the age of 50 [207]. A promising treatment technique for AMD is stem cell therapy, involving
the replacement of the damaged RPE cells with a healthy RPE monolayer obtained through
culturing stem cells [208]- [210]. To verify the quality and maturity of the RPE monolayer before
transplanting it into human retina, it is necessary to extract out a single cell or a few cells from
various spots over the monolayer without damaging the monolayer.
There are several methods of extracting cells from a monolayer culture. For example, cells
can be extracted by physically scraping the surface with a micro-pipette tip [211]. Although this
method is simple and inexpensive, it damages the remaining cells at the tip edges, and results in
scars when the remaining cells are regrown. A laser microdissection for single cell extraction has
been used, but there is a risk of damaging the remaining cells due to the heat associated with high
intensity light which limits the survival rate of the ejected cell to be 25.3% [212]. On the other
hands, focused ultrasound with its inherent large mechanical force (unlike light) is capable of
extracting a single (or a few) cell without any associated heat.
118
In this chapter, an acoustic droplet ejector based on a Self-Focusing Acoustic Transducer
(SFAT) for cell(s) extraction from RPE monolayer is explored and presented. The number of cells
ejected from a monolayer RPE cell culture can be varied by varying the SFAT operating parameters.
The viability of the acoustically ejected cell and the remaining cells in the monolayer are assessed
with three different fluorescent dyes (CFDA-SE, NucBlue reagent, and Propidium iodide dyes) in
addition to regrowing the cells to fill up the ejected spot on the monolayer to confirm that the
remaining cells are not damaged by the cell ejection. The genes of the extracted RPE cells are
studied with the real-time reverse transcription Polymerase Chain Reaction (RT-PCR).
9.2 Device Design
The SFAT for the RPE cell ejection is designed with an air-cavity Fresnel lens and
fabricated on a 1-mm-thick lead zirconate titanate (PZT-4) substrate, operating with 20.1 MHz
sinusoidal electrical signal for its 9th harmonic of thickness-mode resonance. A PZT-4 substrate is
chosen over PZT-5A for its lower tangent loss [64]. A short focal length of 5mm is selected to
minimize the acoustic loss in the medium, as the transducer is operated at a relatively high
frequency of 20.1 MHz for a small focal size (to extract a single cell) of approximately 100µm. A
Finite-Element Method (FEM) simulation results verify the focusing capability at focal point of
5mm, focal size of 100µm, comparable to RPE cell size [213], and focal depth of 790µm (Figure
9.1)
119
Figure 9. 1 (a) Top-view photo of a fabricated SFAT on PZT-4 substrate. Finite element analysis (FEA) simulated
normalized acoustic pressure (b) on the lateral plane at z= 5 and (c) on the vertical plane along with the central axis
9.3 Experimental Setup and Method
9.3.1 Preparing RPE Cell Culture
Retinal Pigmented Epithelial (RPE) cells derived from human embryonic stem cell line
(H14) are prepared in the Clegg Lab in the University of California Santa Barbara, and cultured
on a petri dish which has been coated with 4-µm-thick hydrophobic Parylene layer to reduce the
adhesion between the grown cells and the dish, followed by treatment with Geltrex matrix (Thermo
Fisher Scientific Inc). The RPE cells are incubated in X-VIVO™ 10 Media (Lonza) with 10uM
rock inhibitor Y-27632 (Tocris Bioscience) until the cells reach the 100% confluency with
homogenous cell shape. The cultured RPE monolayer cells are moved to phosphate-buffered saline
(PBS) solution along with the fluorescent dyes of three different colors for the purpose of
visualization of cell’s viability. CFDA-SE (5(6)-Carboxyfluorescein diacetate, succinimidyl ester)
is a membrane-permeable dye binding to intracellular proteins and emitting green fluorescence
[214], and the VybrantTM CFDA-SE dye (InvitrogenTM) is used to label the proliferating cells and
the ejected cell(s). NucBlue reagent (Thermo Fisher Scientific Inc.) is a membrane-permeable
DNA stain that emits blue when it enters the cells and binds to the DNA of the live cells, while
Propidium iodide (PI) dye (Thermo Fisher Scientific Inc ) is a membrane- impermeable dye, only
120
binding with DNA of the dead cells [215]. Both NucBlue and PI fluorescent dyes are utilized to
determine the viability of the surrounding and the ejected cells, after experiencing the focused
acoustic energy for cell extraction.
9.3.2 Experiment setup
The transducer is mounted onto a laser-machined acrylic holder submerged in deionized
water (DI water). Another laser-machined acrylic holder to position a petri dish containing RPE
cultured monolayer is controlled by a 5-axis micromanipulator and placed at 4.8 mm above the
transducer which corresponds to the focal length of the transducer (Figure 9.2). Before ejecting
the cell, a circular bulls-eye surface leveler (BMI 680040-D) is used to level the position of the
holder with a dummy petri dish without RPE cells. After the leveling process, a dummy dish is
filled with 1 ml of DI water to confirm the proper generation of droplets by adjusting the voltage,
pulse width and pulse repetitive frequency (PRF) of the driving electrical signal used to operate
the transducer (Figure 9.2). Once the pre-alignment is completed, the petri-dish containing RPE
cultured monolayer immersed in a shallow layer (5mL) of PBS solution is placed on the acrylic
holder, and the target area (which is labeled on the outer bottom of the dish with a permanent
marker) is manually aligned to the center of the transducer. The SFAT is driven with a pulsed 20.1
MHz sinusoidal electrical signals at 10 Hz PRF by a power amplifier (75A250, Amplifier Research
Corp.) which amplifies the pulsed signal from the signal generator (AFG-3252, Tektronix Inc.).
121
Figure 9. 2 Experiment setup for RPE cell extraction with SFAT mounted on the laser-machined holder and immersed
in PBS solution. A petri dish with RPE cells is held by the movable stage connected to a 5-axis micro-manipulator and
a collecting plate.
The cell extraction experiments are designed according to the number of cells to be
extracted. For a large number (more than 10) of cells, the SFAT is moved not only side-to-side
around the target areas but also up and down. For a small number (1 - 10) of cells, the operating
time is controlled to 5 and 10 seconds, while the position of SFAT is fixed. The single cell
extraction from various spots is achieved by synchronizing the lateral movement of SFAT with
PRF so that each single pulse may be delivered to different locations of the monolayer. The ejected
cell(s) out of the monolayer is collected on a microscope cover glass attached to the laser-machined
acrylic lid (Figure 9.2), and the cell viability is verified with PI fluorescent stains. Also, PI
fluorescent dye is added into the RPE monolayer petri-dish to confirm no damages of the cells
around the ejected spot. A single cell RT-PCR is used for validate the expression of RPE-specific
genes and T-test is performed to compare the control RPE cell group (cells manually collected
using a micropipette) with the acoustically collected RPE cell group.
122
9.4 Experiment Result
The droplet-assisted particle ejection based on SFAT was previously verified with 10-µm
diameter Polystyrene microspheres embedded in agarose gel [216]. The distance between the petridish and SFAT is matched to the focal length of SFAT, 4.8 mm. The minimum required voltages
and the number of cycles for ejecting water droplets out of the petri dish (Figure 9.2) are measured
to be 268Vpp and 3,000 cycles per pulse (which is corresponding to 149-µs pulse width or 0.15%
duty cycle with PRF of 10Hz), respectively. The calculated ISPPA and ISPTA, using equation 2.25
and 2.26, are 17.7 and 26.5 mW/cm2
, respectively, which are substantially lower than the FDA
safety limits for diagnostic ultrasound (ISPPA < 190 W/cm2
, ISPTA < 720 mW/cm2
) [217].
First, a clear isolation of cells from the RPE monolayer is obtained by ejecting the cells
with the SFAT (Figure 9.3b). When the monolayer RPE cells (after the SFAT-based acoustic
ejection out of a monolayer) is re-cultured, the cells can be regrown without any scar, as can be
seen in Figure 9.3c. This result confirms that SFAT-based cell isolation (from an RPE monolayer
grown on a solid scaffold) leaves no damage around the edge where cell extraction is performed,
unlike a knife-based cell isolation.
Figure 9. 3 Microscope photos of 100% confluency human retinal pigment epithelium (RPE) monolayer cells (a)
before and (b) after an ejection of cells by SFAT. (c) Photos of the same monolayer cells when the cells are re-cultured
(for 4 days) after the cell ejection, new cells filling the space and showing no damage on the edge of the ejection site.
Top and bottom photos are at low and high magnification, respectively.
123
A large number of RPE cells are initially ejected out of the monolayer for easy observation
by dynamically adjusting the position of SFAT both vertically and laterally, while a pulsed
sinusoidal signal with 289 Vpp and 200 µs pulse width is continuously delivered to the SFAT. The
ejected spot on the RPE monolayer is confirmed under the fluorescent microscope with three
different stains (Figure 9.4) and the diameter of ejected spot is measured about 1 mm.
Figure 9. 4 Fluorescent microscope images showing the viability of the RPE monolayer (a) before and (b) after the
focused ultrasound cell ejection. A large ejected spot (1-mm in diameter) on the monolayer is indicated with white
arrow and the black shaded areas are the targeted areas labeled with a permanent marker on the bottom face of the
petri dish.
The number of ejected cells is controlled by fixing the position of SFAT at an optimal
ejection point, which is about 4.8 mm from the bottom of the petri-dish, and controlling the time
of operation to adjust the total acoustic energy delivered to the RPE monolayer. The sinusoidal
signal applied to SFAT is 276.5 Vpp with 10 Hz PRF and 0.2% duty cycle (200 µsec on time). In
other words, each pulse is on for 200 µs and is off for 99.8 ms, so that the SFAT delivers to the
focal point ISPTA of 37.6 mW/cm2 per each pulse. When SFAT is operated for 10 seconds in a fixed
position, the ejected area is about 100 µm in diameter, which is closely matched with the size of
the focal diameter (Figure 9.5 a-b). Also, the viability of the surrounding RPE cells is verified with
fluorescent dyes, including staining the monolayer again with PI (which binds only with dead
cells), after the acoustic droplet-assisted cell extraction (Figure 9.5 c-d).
124
Figure 9. 5 Bright-field microscope image (a) before and (b) after delivering focused ultrasound (276.5Vpp, PRF of
10Hz and pulse width of 200µs) to extract RPE cells from a spot of about 100µm in diameter. Fluorescent microscope
images with the same acoustic pulses (c) before and (d) after the cell ejection. White arrows indicate the ejected spots
on the monolayer.
The number of pulses delivered to the SFAT determines the number of ejected cells. For
example, 5.5 seconds of continuous operation, equivalent to approximately 55 pulses delivered
to SFAT, results in about five RPE cells extraction. However, when the pulse count is elevated to
105 pulses (or 10.5 seconds of operation time), SFAT exhibits its capability to extract as many as
ten RPE cells with notable consistency (Figure 9.6 a-b). This phenomenon has been consistently
and reliably validated across multiple trials in eight separate petri dishes. Moreover, by changing
the position of the SFAT by 1 mm in a lateral direction and operating the SFAT for different
durations, two distinct extraction sizes can be achieved (Figure 9.6 c).
Figure 9. 6 Bright-field microscope images with two different sizes of ejected spot by controlling the number of pulses
delivered to the RPE monolayer: (a) with 55 pulses (resulting in about 5 RPE extracted cells), (b) with 105 pulses
(resulting in about 10 ejected cells. (c) Two different sizes of the ejected spot on the same petri dish are made with
SFAT being applied with 55 pulses first (right arrow), being moved laterally 1-mm to the left, and then being applied
with 105 pulses (left arrow) to eject about 10 RPE cells.
125
Single-cell extraction is realized by delivering a single pulse to RPE monolayer by driving
SFAT with 268 Vpp and 248 µs pulse width. However, locating the ejected spot under the
microscope is extremely challenging, when only a single cell is ejected out of the monolayer. Thus,
instead of delivering a single pulse, a series of single pulses at 1 Hz PRF are applied to different
locations of the monolayer by synchronizing the movement of the SFAT with the PRF. In other
words, SFAT’s position is changed by approximately 127 µm every one second, as SFAT is driven
to deliver a single pulse at each spot, leaving the sequential ejected spots on a targeted area (Figure
9.7).
Figure 9. 7 (a) Bright-field and (b) fluorescent microscope images of a series of single-cell ejections (white arrows)
by a single pulse (268Vpp and pulse width of 248µs). The location of SFAT is changed by 127µm every one second,
synchronized with the pulse repetitive frequency (1 Hz) for clear evidence of single-cell ejection.
9.5 Discussions
The ejected single cells are collected with a microscope cover slip (with a droplet of PBS
solution) which is suspended about 2 mm above the RPE monolayer (Figure 9.2). The viability of
the extracted cell is checked by adding PI dye, after the ejection and collection, under a fluorescent
microscope. Furthermore, the collected single cell is proceeded with uniform whole transcriptome
amplification on each single cell using a Qiagen REPLI-g WTA Single Cell Kit (Qiagen). The
amplified complemented DNA (cDNA) is quantified and verified with the real-time RT-PCR test
for both housekeeping genes ((Glyceraldehyde 3-phosphate dehydrogenase (GAPDH) and �-
126
actin) and RPE cell specific-expressed genes with a PerfeCTa SYBR® Green SuperMix
(Quantabio), in Applied Biosystems OnePlus™ Real-Time PCR System (Thermo Fisher Scientific
Inc). Microphthalmia-associated transcription factor (MITF) gene, pigment epithelium-derived
factor (PEDF) gene, and Premelanosome protein (PMEL) are chosen as RPE marker genes based
on their normalized transcripts per million (nTPM) expression levels in RPE cells with low,
medium, and high population densities, respectively [218].
Table 9. 1 RPE Marker Genes with Their Populations
RPE marker gene Mature RPE nTPM Atrophy RPE nTPM Remarks
MITF 56-120 <56 Low Population
PEDF 3,959-7,861 <3959 Medium Population
PMEL17 7,639-13,776 <7639 High Population
The threshold cycle (Ct) values for the RPE specific genes and the housekeeping genes are
measured three times each from both the acoustically ejected sample and control sample (manually
collected sample) and the difference between the averaged Ct values (∆Ct) of the RPE specific
gene and the housekeeping gene is calculated on every sample [219]. Then the relative gene
expression of the ejected single cells (n = 8), compared with the control sample (n = 3) which is
manually collected with a micropipette from the RPE monolayer cells, is tested with T-Test, a
statistical method used to determine if there is a significant difference between the means of two
groups [220]. The T-test is conducted on the ∆Ct values of three RPE marker genes, measured with
respect to two housekeeping genes (GAPDH and β-actin). The calculated p-values for each gene,
with respect to the housekeeping genes, are summarized in Table 9.2. The results show that most
of the p-values are greater than 0.05, which suggests that there are no significant differences in the
127
expression level of the RPE specific genes between the acoustically collected group and the
manually collected group (control group) [220].
Table 9. 2 The Results (p-value) of T-test Between the Acoustically Collected and the Control Group
Housekeeping
genes
RPE genes
GAPDH β-actin
MITF 0.55 0.27
PEDF 0.58 0.79
PMEL17 0.14 0.38
9.6 Summary
Acoustic droplet-based single cell ejection is made possible with a self-focusing acoustic
transducer (SFAT) which generates a single focal point at a focal length of 5 mm over a focal
diameter of 97 µm. Operating SFAT with a pulsed sinusoidal signal of 20.1 MHz, the 9th harmonics
of the thickness-mode resonance of the PZT used for the SFAT and varying the amplitude, PRF
and duty cycle (along with the movement of the SFAT) results in the different numbers of cells
being ejected. When a single pulse of 268 Vpp with 248-µs pulse width is applied to the SFAT, a
single cell is extracted from the RPE monolayer due to the focused ultrasound from the SFAT. The
viability of the remaining monolayer cells and the ejected cell(s) are visualized by the
fluorescence-based cell viability assay involving a CFDA-SE, a NucBlue reagent, and a Propidium
iodide (PI) dye, which selectively bind to specific cellular components and emit the different
fluorescent colors. And both remaining cells in the monolayer and extracted cell are confirmed to
be intact without damage. Furthermore, the acoustically ejected single cell is collected and put
through the real time RT-PCR to detect two housekeeping genes and three RPE specific genes.
The T-tests on the ∆Ct’s of the three RPE specific genes with respect to two housekeeping genes
128
show no significant difference between the conventionally collected cells and the acoustically
ejected cells.
129
Chapter 10
Summary and Future Directions
10.1 Summary
This thesis introduces an integrated wireless piezoelectric ultrasonic transducer system
designed for biomedical applications, specifically the design, fabrication, and applications of three
distinct types of acoustic transducers. Additionally, two designs of electronic circuits driving an
acoustic transducer are explored with the circuit simulation. The power efficiency, the system
bandwidth, and the operation time are characterized using a 100F Lithium-ion Capacitor (LIC).
A Self-Focusing Acoustic Transducer (SFAT) is the acoustic transducer generating a single
focused ultrasonic wave at the desired focal points by using a Fresnel half-band wavelength
(FHBW) annular ring. The SFAT is typically fabricated on the lead zirconate titanate (PZT)
substrate with varying thickness, leading to a different operating frequency tailored to specific
application. The focusing capability of SFAT is verified with a Finite Element Method (FEM)
analysis and the characterization and the measurement of focused ultrasound are conducted using
a hydrophone in the scanning tank system. Two fabrication methods for Fresnel annular rings are
presented: one involves a straightforward patterned electrode design, while the other utilizes an
air-cavity parylene lens. Different designs of SFAT are explored in various applications such as
acoustic propulsion, single cell ejection, and the focused ultrasound neuromodulation.
An electrically controllable SFAT, abbreviated as eCONSFAT, represents a modified design
of the SFAT, having an equal width of 128 individual electrodes and featuring a phased array
acoustic transducer. The design, FEM simulation and fabrication process of eCONSFAT are
130
studied, with a particular emphasis on the fabrication of indium solder bumps essential for the flip
chip bonding packaging.
A Spiral-arm Vortex-beam Acoustic Transducer (SVAT) design is introduced which
generates an acoustic helix with a simple logarithmic spiral patterned electrode. The acoustic
vortex pressure distribution in the media is confirmed with a three- dimensional (3D) FEM
simulation. The thesis further explores various designs, including multi-arms SVAT and 18-
sectored SFAT, comparing their performance in applications such as acoustic propulsion force for
underwater acoustic propellers and levitation force in acoustic tweezers.
In addition, two different designs of driving electronic circuits are compared. The first
design is an on-board power amplifier, containing a DC-to-DC converter, an oscillator, and a power
amplifier. This circuit generates square wave signals with the amplitude of 30Vpp to the transducer.
When powered by a 100F LIC, it can continuously deliver the electrical energy supply to the SFAT
for about 3minutes and 10 seconds until the LIC voltage drops to 2.5V. The maximum power
conversion efficiency is measured at approximately 66% under a 130Ω load, and the system
bandwidth (�&M3N) is approximately 5MHz.
The second circuit design is the modified Z-source inverter, incorporating a Gallium
Nitride Field Effect Transistor (GaNFET), a gate driver IC, and a small microcontroller or
Bluetooth Low Energy IC. The Z-source inverter offers a high gain (of up to 19.5) and high power
efficiency of 91% at a 75Ω load. When powered by the same 100F LIC, it operates for 5 minutes
and 50 seconds. However, it is constrained by a limited operating frequency range, spanning from
1.7 to 2.4MHz.
131
10.2 Future Work on Transducers
10.2.1 Future Work on eCONSFAT
While the design and fabrication of eCONSFAT have been thoroughly examined, the
packaging process including flip chip bonding remains incomplete. Future work on eCONSFAT
will include development of its packaging process and assessment of its ability to generate multiple
focal points simultaneously at different positions and to shape the acoustic wave as desired by
electrical signals applied to the sectored electrodes.
10.2.2 Future Work on Low-EMI SFAT
The low-EMI SFAT successfully demonstrates a 94% reduction in EMI levels during the
patch clamp experiment of a young mouse tissue by using a thin Nickel layer. Nevertheless, the
recording electrodes still pick up a substantial amount of EMI, saturating at a specific level of
applied voltage (around 50Vpp), even with the low-EMI SFAT. Consequently, future studies need
to explore alternative EMI shielding materials or study the sudden increase of EMI beyond a
specific applied voltage.
10.3 Future Work on Driving Circuits
The two proposed driving circuit designs exhibit their own sets of advantages and
disadvantages. While the Z-source inverter is intended to increase the voltage gain and reduce the
size, its design is not currently optimized, leaving room for further improvements, especially in
increasing system bandwidth and reducing the size. The limited system bandwidth is potentially
caused by the prolonged rising and falling time of GaNFET, likely due to its input and output
capacitances. Additionally, the capacitors in the modified Z-source network impacts the overall
rising and falling time, thereby constraining the duty cycle and the frequency of pulse width
modulation (PWM) signals. Optimizing the choice of capacitors and GaNFET can enhance the
132
system bandwidth. Additionally, the overall PCB size can be reduced by limiting the input current,
as inductors in the Z-source network occupy a large area in the PCB. Furthermore, employing a
smaller BLE module can further reduce the PCB size.
10.4 Future Work on Focused Ultrasound Neuromodulation
To validate the proposed hypothesis regarding the underlying mechanism of focused
ultrasound neuromodulation (FUS), further experiments are necessary. The experiment design
involves the use of specific channel blockers along with both active and control SFAT to verify the
involvement of Transient Receptor Potential (TRP) and Two-pore Potassium (K2P) ion channels
in FUS. For example, using a concentration of 52µM Hydroxyl-α-sanshool, which is reported to
effectively block the TRESK channel [221] may provide insights, as TRESK channels (among
K2P ion channel family) are known to play a role in the regulation of seizures in the hippocampus
region [222]. On the other hand, the use of 1µM Ruthenium Red (RR), known to block TRPV ion
channels [202], can serve to validate the association between FUS and TRPV channels.
133
References
[1] Marc A. Shampo and Robert A. Kyle, “Karl Theodore Dussik—Pioneer in Ultrasound”, MAYO Clinic
Proceedings, vol. 70, no. 12, pp. 1136, 1995, doi: 10.4065/70.12.1136
[2] M. Hansmann, B. J. Hackelöer, A. Staudach, “Ultrasound Diagnosis in Obstetrics and Gynecology”, Springer,
2012
[3] Claude R. Joyner, John M. Reid, “Applications of ultrasound in cardiology and cardiovascular physiology”,
Progress in Cardiovascular Diseases, vol. 5, no. 5, pp.482-497, 1963, doi: 10.1016/S0033-0620(63)80011-0
[4] Villeman et. al., “Ultrafast Ultrasound Imaging in Pediatric and Adult Cardiology: Techniques, Applications, and
Perspectives”, J Am Coll Cardiol Img, vol.13, no.8, pp. 1771-1791, 2020, doi: 10.1016/j.jcmg.2019.09.019
[5] Cristi R. Cook, “Ultrasound Imaging of the Musculoskeletal System”, Veterinary Clinics: Small Animal Practice,
vol.46, no.3, pp. 355- 371, 2016, doi: 10.1016/j.cvsm.2015.12.001
[6] Spyretta Golemati and Demosthenes D. Cokkinos, “Recent advances in vascular ultrasound imaging technology
and their clinical implications”, Ultrasonics, vol. 119, 2022, doi: 10.1016/j.ultras.2021.106599
[7] Y. Tang et. al., “In Vivo Non-Thermal, Selective Cancer Treatment With High-Frequency Medium-Intensity
Focused Ultrasound”, IEEE Access, vol.9, pp.122051 – 122066, doi: 10.1109/ACCESS.2021.3108548
[8] S. Guillaumier et al., “A Multicentre Study of 5-year Outcomes Following Focal Therapy in Treating Clinically
Significant Nonmetastatic Prostate Cancer,” Eur. Urol., vol. 74, no. 4, pp. 422–429, Oct. 2018, doi:
10.1016/j.eururo.2018.06.006.
[9] Z. Ram et al., “Magnetic resonance imaging-guided, high-intensity focused ultrasound for brain tumor therapy.,”
Neurosurgery, vol. 59, no. 5, pp. 949–55; discussion 955-6, Nov. 2006, doi:
10.1227/01.NEU.0000254439.02736.D8.
[10]F. P. Curra and L. A. Crum, “Therapeutic ultrasound: Surgery and drug delivery,” Acoust. Sci. Technol., vol. 24,
no. 6, pp. 343–348, 2003, doi: 10.1250/ast.24.343.
[11]S. Mitragotri, “Healing sound: The use of ultrasound in drug delivery and other therapeutic applications,” Nat.
Rev. Drug Discov., vol. 4, no. 3, pp. 255–260, Mar. 2005, doi: 10.1038/nrd1662
[12]Neely RM, Piech DK, Santacruz SR, Maharbiz MM, Carmena JM. Recent advances in neural dust: towards a
neural interface platform. Curr Opin Neurobiol. 2018 Jun;50:64-71. doi: 10.1016/j.conb.2017.12.010.
[13]H. Baek, K. Pahk, H. Kim, “A review of low-intensity focused ultrasound for neuromodulation”, Springer, 2017
pp. 135-142.
[14]T.J. Matula, “Ultrasound-based cell sorting with microbubbles: A feasibility study”, J. Acoust. Soc. Am. vol.144,
no.1 pp. 41–52, 2018
[15]P. Xin, Lianwen Yang and Yongke Li, "Mechanism and application of EMAT technology based on NDT," 2014
China International Conference on Electricity Distribution (CICED), Shenzhen, China, 2014, pp. 100-102, doi:
10.1109/CICED.2014.6991672.
134
[16]Hirao, M., Ogi, H., “Field Applications of EMATs. In: Electromagnetic Acoustic Transducers.”, Springer Series
in Measurement Science and Technology. Springer, Tokyo, 2017, pp.347-372, doi: 10.1007/978-4-431-56036-
4_18
[17]Baillie, I, Griffith, P, Jian, X, and Dixon, S, “Implementing an ultrasonic inspection system to find surface and
internal defects in hot, moving steel using EMATs”, Insight - Non-Destructive Testing and Condition Monitoring,
vol. 49, no.2, pp. 87-92, 2007, doi: 10.1784/insi.2007.49.2.87
[18]Michal Kubinyi, Radislav Smid and Marcel Kreidl, “EMAT Applied in the Aviation Industry”, The 26th Congress
of ICAS and 8th AIAA ATIO, 2008, doi: 10.2514/6.2008-8870
[19]J. Aron et. al., “DEVELOPMENT OF AN EMAT IN-LINE INSPECTION SYSTEM FOR DETECTION,
DISCRIMINATION, AND GRADING OF STRESS CORROSION CRACKING IN PIPELINES”. United States,
2005, doi:10.2172/840950.
[20]R. Ribichini, F. Cegla, P. B. Nagy and P. Cawley, "Study and comparison of different EMAT configurations for
SH wave inspection," IEEE Transactions on Ultrasonics, Ferroelectrics, and Frequency Control, vol. 58, no. 12,
pp. 2571-2581, 2011, doi: 10.1109/TUFFC.2011.2120.
[21]F. V. Hunt, Electroacoustics: The analysis of transduction, and its historical background, 2nd Ed. Cambridge,
MA: Harvard University Press, 1982.
[22]Qin Zhou, A. Zettl, “Electrostatic graphene loudspeaker”, Appl. Phys. Lett., vol.102, no.22, 2013, doi:
10.1063/1.4806974
[23]A. Caronti et. al., “Capacitive micromachined ultrasonic transducer (CMUT) arrays for medical imaging”,
Microelectronics Journal, vol.37, no.8, 2006, doi: 10.1016/j.mejo.2005.10.012.
[24]D. M. Mills, "Medical imaging with capacitive micromachined ultrasound transducer (cMUT) arrays," IEEE
Ultrasonics Symposium, 2004, Montreal, QC, Canada, 2004, pp. 384-390 Vol.1, doi:
10.1109/ULTSYM.2004.1417744.
[25]A. S. Ergun, G. G. Yaralioglu, O. Oralkan, and B. T. Khuri-Yakub, “MEMS/NEMS techniques and applications
of capacitive micromachined ultrasonic transducers,” in MEMS/NEMS: Handbook Techniques and Applications,
C. T. Leondes, Ed. Boston, MA: Springer US, 2006, pp. 553–615.
[26]Ahmad Safari, E. Koray Akdogan, “Piezoelectric and Acoustic Materials for Transducer Applications”, Springer,
2008
[27]H. Jaffe and D. A. Berlincourt, "Piezoelectric transducer materials," in Proceedings of the IEEE, vol. 53, no. 10,
pp. 1372-1386, Oct. 1965, doi: 10.1109/PROC.1965.4253.
[28]Wallace A. Smith, “New opportunities in ultrasonic transducers emerging from innovations in piezoelectric
materials”, Proceedings, vol. 1733, 1992, doi: 10.1117/12.130585
[29]Warren P. Mason, “Physical Acoustics: Principles and Methods”, Academic Press, 2013, pp.169
[30] J. G. Gualtieri, J. A. Kosinski and A. Ballato, "Piezoelectric materials for acoustic wave applications," in IEEE
Transactions on Ultrasonics, Ferroelectrics, and Frequency Control, vol. 41, no. 1, pp. 53-59, Jan. 1994, doi:
10.1109/58.265820.
[31]G. L. Smith et. al., “PZT-Based Piezoelectric MEMS Technology”, J. Am. Ceram. Soc., vol.95, no.6, 2012, doi:
10.1111/j.1551-2916.2012.05155.x
135
[32]Hayward et al, “A systems model of the thickness mode piezoelectric transducer”, J. Acoust. Soc. Am., vol. 76,
no.2, 1984, doi: 10.1121/1.391138
[33]J. F. Gelly and F. Lanteri, "Comparison of piezoelectric (thickness mode) and MEMS transducers," IEEE
Symposium on Ultrasonics, 2003, Honolulu, HI, USA, 2003, pp. 1965-1974 Vol.2, doi:
10.1109/ULTSYM.2003.1293302.
[34]Dauchy, F., Dorey, R.A, “Thickness mode high frequency MEMS piezoelectric micro ultrasound transducers”,
J. Electroceram, vol. 19, pp.383–386, 2007, doi: 10.1007/s10832-007-9317-x
[35]Ayodele Sanni and Antonio Vilches, “Powering low-power implants using PZT transducer discs operated in the
radial mode”, Smart Mater. Struct, vol.22, 2013, doi: 10.1088/0964-1726/22/11/115005
[36]Vamshi Krishna Chillara, Cristian Pantea, Dipen N. Sinha, “Low-frequency ultrasonic Bessel-like collimated
beam generation from radial modes of piezoelectric transducers”, Appl. Phys. Lett., vol. 6, pp.110, 2017, doi:
10.1063/1.4975800
[37]Qiu, Yongqiang, James V. Gigliotti, Margeaux Wallace, Flavio Griggio, Christine E. M. Demore, Sandy Cochran,
and Susan Trolier-McKinstry. 2015. "Piezoelectric Micromachined Ultrasound Transducer (PMUT) Arrays for
Integrated Sensing, Actuation and Imaging", Sensors, vol.15, no. 4, pp.8020-8041,2015, doi:10.3390/s150408020
[38]Nazemi, Haleh, Jenitha Antony Balasingam, Siddharth Swaminathan, Kenson Ambrose, Muhammad Umair
Nathani, Tara Ahmadi, Yameema Babu Lopez, and Arezoo Emadi. 2020. "Mass Sensors Based on Capacitive
and Piezoelectric Micromachined Ultrasonic Transducers—CMUT and PMUT", Sensors, vol.20, no. 7, 2020, doi:
10.3390/s20072010
[39]Li, Junhong, Wei Ren, Guoxiang Fan, and Chenghao Wang. 2017. "Design and Fabrication of Piezoelectric
Micromachined Ultrasound Transducer (pMUT) with Partially-Etched ZnO Film", Sensors, vol. 17, no. 6,
pp.1381, 2017, doi:10.3390/s17061381
[40]Bingzhang Chen, Futong Chu, Xingzhao Liu, Yanrong Li, Jian Rong, Huabei Jiang, “AlN-based piezoelectric
micromachined ultrasonic transducer for photoacoustic imaging”, Appl. Phys. Lett., vol. 15, pp.103, 2013,
doi:10.1063/1.4816085
[41]Un-Hyun Lim, Jin-Hee Yoo, Vijay Kondalkar, Keekeun Lee, “Development of High Frequency pMUT Based on
Sputtered PZT”, Journal of Electrical Engineering & Technology, vol.13, no.6, pp.2434-2440, 2018.
[42]Gopikrishnan, R., Zhang, K., Ravichandran, P. et al., “Synthesis, characterization and biocompatibility studies of
zinc oxide (ZnO) nanorods for biomedical application”, Nano-Micro Lett., vol. 2, pp. 31–36, 2010,
doi:10.1007/BF03353614
[43]Zang, Junbin, Zheng Fan, Penglu Li, Xiaoya Duan, Chunsheng Wu, Danfeng Cui, and Chenyang Xue. 2022.
"Design and Fabrication of High-Frequency Piezoelectric Micromachined Ultrasonic Transducer Based on an
AlN Thin Film", Micromachines, vol.13, no. 8, pp.1317, 2022, doi: 10.3390/mi13081317
[44]The development of PMUTs using thin-film piezoelectric materials emphasizes the continuous efforts to improve
microscale transducer technology, leading to innovation in industries such as high-frequency sensing and medical
diagnostics.
[45]Halliwell, Martin J, Jones, B. E, “Pressure Transducers Using Resonating Double-Ended Tuning Forks”,
Doctoral dissertation, The University of Manchester (United Kingdom) ProQuest Dissertations Publishing,
1987. U047403.
[46]J. X. Qiu et al., "Vacuum tube amplifiers," in IEEE Microwave Magazine, vol. 10, no. 7, pp. 38-51, Dec. 2009,
doi: 10.1109/MMM.2009.934517.
136
[47]J. Borg and J. Johansson, "An Ultrasonic Transducer Interface IC With Integrated Push-Pull 40 Vpp, 400 mA
Current Output, 8-bit DAC and Integrated HV Multiplexer," IEEE Journal of Solid-State Circuits, vol. 46, no. 2,
pp. 475-484, 2011, doi: 10.1109/JSSC.2010.2096113.
[48]N. O. Sokal, "RF power amplifiers-classes A through S," Proceedings of Electro/International, pp. 335-400, 1995,
doi:10.1109/ELECTR.1995.471024.
[49]Kyounghoon Yang, G. I. Haddad and J. R. East, "High-efficiency class-A power amplifiers with a dual-biascontrol scheme," IEEE Transactions on Microwave Theory and Techniques, vol. 47, no. 8, pp. 1426-1432, 1999,
doi: 10.1109/22.780390.
[50]G. R. Walker, "A Class B switch-mode assisted linear amplifier," IEEE Transactions on Power Electronics, vol.
18, no. 6, pp. 1278-1285, 2003, doi: 10.1109/TPEL.2003.818825.
[51]A. Far, "Class AB amplifier with noise reduction, speed boost, gain enhancement, and ultra low power," 2018
IEEE 9th Latin American Symposium on Circuits & Systems (LASCAS), Puerto Vallarta, Mexico, 2018, pp. 1-4,
doi: 10.1109/LASCAS.2018.8399961.
[52]F. E. Terman and W. C. Roake, "Calculation and Design of Class C Amplifiers," Proceedings of the Institute of
Radio Engineers, vol. 24, no. 4, pp. 620-632, 1936, doi: 10.1109/JRPROC.1936.227364.
[53]Choi, Hojong, "Class-C Linearized Amplifier for Portable Ultrasound Instruments" Sensors, vol. 19, no. 4, pp.
898, 2019, doi: 10.3390/s19040898
[54]S. . -A. El-Hamamsy, "Design of high-efficiency RF Class-D power amplifier," IEEE Transactions on Power
Electronics, vol. 9, no. 3, pp. 297-308, 1994, doi: 10.1109/63.311263.
[55]L. Fanori and P. Andreani, "Class-D CMOS Oscillators," IEEE Journal of Solid-State Circuits, vol. 48, no. 12,
pp. 3105-3119, Dec. 2013, doi: 10.1109/JSSC.2013.2271531.
[56]D. Dapkus, "Class-D audio power amplifiers: an overview," 2000 Digest of Technical Papers. International
Conference on Consumer Electronics. Nineteenth in the Series (Cat. No.00CH37102), Los Angeles, CA, USA,
2000, pp. 400-401, doi: 10.1109/ICCE.2000.854703.
[57]A. Wentzel, C. Meliani and W. Heinrich, "RF class-S power amplifiers: State-of-the-art results and potential,"
2010 IEEE MTT-S International Microwave Symposium, Anaheim, CA, USA, pp. 812-815,2010, doi:
10.1109/MWSYM.2010.5517402.
[58]You, Kiheum, and Hojong Choi, "Wide Bandwidth Class-S Power Amplifiers for Ultrasonic
Devices" Sensors, vol.20, no. 1, pp.290, 2020, doi: 10.3390/s20010290
[59]Pentz et al., “Driving an ultrasonic transducer with a multicell inverter.”, ECCE Asia Downunder (ECCE Asia),
pp. 976-980, 2013, doi: 10.1109/ECCE-Asia.2013.6579225.
[60]Bachu, V.S., Kedda, J., Suk, I. et al., “High-Intensity Focused Ultrasound: A Review of Mechanisms and Clinical
Applications.”, Ann Biomed Eng, vol.49, pp.1975–1991, 2021, doi: 10.1007/s10439-021-02833-9
[61]Baek, H., Pahk, K.J. & Kim, H., “A review of low-intensity focused ultrasound for neuromodulation.”, Biomed.
Eng. Lett. , vol.7, pp.135–142, 2017, doi: 10.1007/s13534-016-0007-y
[62]Darrow, D.P. “Focused Ultrasound for Neuromodulation.”, Neurotherapeutics vol.16, pp.88–99, 2019,
doi:10.1007/s13311-018-00691-3
[63]Kubanek, J., “Neuromodulation with transcranial focused ultrasound.”, Neurosurgical Focus FOC, vol. 44, no.2,
pp.E14, 2018, doi: 10.3171/2017.11.FOCUS17621
137
[64]Y. Tang, L. -Y. Chen, A. Zhang, C. -P. Liao, M. E. Gross and E. S. Kim, "In Vivo Non-Thermal, Selective Cancer
Treatment With High-Frequency Medium-Intensity Focused Ultrasound,", IEEE Access, vol. 9, pp. 122051-
122066, 2021, doi: 10.1109/ACCESS.2021.3108548.
[65]Osama Al-Bataineh, Jürgen Jenne, Peter Huber, “Clinical and future applications of high intensity focused
ultrasound in cancer”, Cancer Treatment Reviews, vol.38, no.5, pp.346-353, 2012, doi:
10.1016/j.ctrv.2011.08.004.
[66]Hsiao YH, Kuo SJ, Tsai HD, Chou MC, Yeh GP., “Clinical Application of High-intensity Focused Ultrasound in
Cancer Therapy.”, J Cancer, vol.7, no.3, pp.225-331, 2016, doi: 10.7150/jca.13906. PMID: 26918034; PMCID:
PMC4747875.
[67]B. Karaboce, "Investigation of thermal effect by focused ultrasound in cancer treatment," IEEE Instrumentation
& Measurement Magazine, vol. 19, no. 5, pp. 20-64, October 2016, doi: 10.1109/MIM.2016.7579066.
[68]Lin, Yicong, and Yuping Wang. “Neurostimulation as a promising epilepsy therapy.” Epilepsia open, vol. 2, no.4,
pp.371-387, 2017, doi:10.1002/epi4.12070
[69]Schuepbach et. al. ,“Neurostimulation for Parkinson's disease with early motor complications.” N Engl J Med.,
vol.368, no.7, pp.610-622, 2013, doi: 10.1056/NEJMoa1205158.
[70]Phenix, C. P., Togtema, M., Pichardo, S., Zehbe, I., & Curiel, L. , “High Intensity Focused Ultrasound Technology,
its Scope and Applications in Therapy and Drug Delivery.”, Journal of Pharmacy & Pharmaceutical
Sciences, vol. 17, no.1, pp. 136–153, 2014, doi: 10.18433/J3ZP5F
[71]Y. Kim, A. D. Maxwell, T. L. Hall, Z. Xu, K. -W. Lin and C. A. Cain, "Rapid prototyping fabrication of focused
ultrasound transducers," IEEE Transactions on Ultrasonics, Ferroelectrics, and Frequency Control, vol. 61, no.
9, pp. 1559-1574, 2014, doi: 10.1109/TUFFC.2014.3070.
[72]Lean, H.Q., Zhou, Y., “Acoustic Field of Phased-Array Ultrasound Transducer with the Focus/Foci Shifting.”, J.
Med. Biol. Eng. Vol.39, pp. 919–931, 2019, doi: 10.1007/s40846-019-00464-z
[73]J. Lee and E. S. Kim, "Phase Array Ultrasonic Transducer Based on a Flip Chip Bonding with Indium Solder
Bump," 2021 IEEE International Ultrasonics Symposium (IUS), Xi'an, China, 2021, pp. 1-4, doi:
10.1109/IUS52206.2021.9593776.
[74]Davis, A. and Kühnlenz, F., “Optical Design using Fresnel Lenses.”, Optik & Photonik, vol.2, no.4, pp. 52-55,
2011, doi: 10.1002/opph.201190287
[75] Jae Wan Kwon, Qiang Zou and Eun Sok Kim, "Directional ejection of liquid droplets through sectoring halfwave-band sources of self-focusing acoustic transducer," Technical Digest. MEMS 2002 IEEE International
Conference. Fifteenth IEEE International Conference on Micro Electro Mechanical Systems (Cat.
No.02CH37266), Las Vegas, NV, USA, 2002, pp. 121-124, doi: 10.1109/MEMSYS.2002.984219.
[76]Y. Choe, J.W. Kim, K.K. Shung, and E.S. Kim, “Ultrasonic Microparticle Trapping by Multi-Foci Fresnel Lens,”
Joint Conference of the IEEE International Frequency Control Symposium and European Frequency and Time
Forum, San Francisco, CA, May 1-5, 2011, Digital Object Identifier: 10.1109/FCS.2011.5977900.
[77]Chia-Che Wu, Cheng-Chun Lee, G.Z. Cao, I.Y. Shen, “Effects of corner frequency on bandwidth and resonance
amplitude in designing PZT thin-film actuators”, Sensors and Actuators A: Physical, vol. 125, no.2, pp. 178-185,
2006, doi:10.1016/j.sna.2005.07.007.
[78]K. Yamada and H. Shimizu, “Planar-Structure Focusing Lens for Acoustic Microscope,” Proc. IEEE Int.
Ultrason. Symp. (IUS), 1985, pp. 755–758, doi: 10.1109/ULTSYM.1985.198612.
138
[79]K. K. Shung, Diagnostic ultrasound: Imaging and blood flow measurements, 1st ed. Boca Raton, FL, USA: CRC
Press, 2005.
[80]H. H. Barrett and F. A. Horrigan, “Fresnel zone plate imaging of gamma rays; theory.,” Appl. Opt., vol. 12, no.
11, pp. 2686–702, Nov. 1973, doi: 10.1364/AO.12.002686.
[81]Attwood, D. (1999). SOFT X-RAY MICROSCOPY WITH DIFFRACTIVE OPTICS. In Soft X-Rays and
Extreme Ultraviolet Radiation: Principles and Applications (pp. 337-394). Cambridge: Cambridge University
Press. doi:10.1017/CBO9781139164429.010
[82]O. C. Zienkiewicz, R. L. Taylor, and J. Z. Zhu, The Finite Element Method: Its Basis and Fundamentals, 7th
edition. Oxford, UK: Butterworth-Heinemann, 2013.
[83]COMSOL Multiphysics Acoustics Module User's Guide. COMSOL AB, pp.125-149 Stockholm, Sweden. 2017.
[84]J.W. Wladimiroff, I.L. Craft, D.G. Talbert, “In vitro measurements of sound velocity in human fetal brain tissue”,
Ultrasound in Medicine & Biology, vol.1, no.4, 1975, doi: 10.1016/0301-5629(75)90125-8.
[85]Y. Tang, S. Liu, and E. S. Kim, “MEMS Focused Ultrasonic Transducer with Air-Cavity Lens Based on
Polydimethylsiloxane (PDMS) Membrane,” in IEEE 33rd Int. Conf. on Micro Electro Mech. Syst. (MEMS), Jan.
2020, pp. 58–61, doi: 10.1109/MEMS46641.2020.9056313.
[86]J. A. Dobrowolski, Scattering, “Parameters in RF and Microwave Circuit Analysis and Design”, Norwood, MA,
USA: Artech House, 2016.
[87]C. J. Steinem, Andreas, “Piezoelectric Sensors”, New York, Springer, 2007.
[88]Kim ES. Fundamentals of Microelectromechanical Systems (MEMS). First edition. McGraw-Hil Education; 2021.
[89]Mason, W.P.,"An electromechanical representation of a piezoelectric crystal used as a transducer," Proceedings
of the Institute of Radio Engineers, Vol. 23, No. 10, pp. 1252-1263, 1935.
[90]Redwood, M.: "Experiments with the electrical analog of a piezoelectric transducer," J. Acoust. Soc. Amer., Vol.
36, No. 10, pp. 1872-1880, October 1964.
[91]Butterworth, S.: "On a nuU method of testing vibration galvanometers," Proc. Phys. Soc., London, Vol. 26, pp.
264-273, 1913.
[92]Van Dyke, K.S.: "The piezo-electric resonator and its equivalent network," Proceedings of the Institute of Radio
Engineers, Vol. 16, No. 6, pp. 742- 764, 1928.
[93]K. K. Shung and M. Zippuro, "Ultrasonic transducers and arrays," in IEEE Engineering in Medicine and Biology
Magazine, vol. 15, no. 6, pp. 20-30, Nov.-Dec. 1996, doi: 10.1109/51.544509.
[94]L. Zhao and E. S. Kim, "Focused ultrasound transducer with electrically controllable focal length," 2018 IEEE
Micro Electro Mechanical Systems (MEMS), Belfast, UK, 2018, pp. 245-248, doi:
10.1109/MEMSYS.2018.8346530.
[95]L. Zhao and E. S. Kim, "Focused Ultrasonic Transducer with Electrically Controllable Focal-Point Location,"
2018 IEEE International Ultrasonics Symposium (IUS), 2018, pp. 1-3, doi: 10.1109/ULTSYM.2018.8580054..
[96]J. Lee and E. S. Kim, "Phase Array Ultrasonic Transducer Based on a Flip Chip Bonding with Indium Solder
Bump," 2021 IEEE International Ultrasonics Symposium (IUS), Xi'an, China, 2021, pp. 1-4, doi:
10.1109/IUS52206.2021.9593776.
139
[97]Datta M., “Manufacturing processes for fabrication of flip-chip micro-bumps used in microelectronic packaging:
An overview”, Journal of Micromanufacturing., vol.3, no.1, pp:169-183, 2020, doi:10.1177/2516598419880124
[98]Ch. Broennimann, F. Glaus, J. Gobrecht, S. Heising, M. Horisberger, R. Horisberger, H.C. Kästli, J. Lehmann, T.
Rohe, S. Streuli, “Development of an Indium bump bond process for silicon pixel detectors at PSI”, Nuclear
Instruments and Methods in Physics Research Section A: Accelerators, Spectrometers, Detectors and Associated
Equipment, Vol. 565, no. 1, pp. 303-308, 2006
[99]Manasson, Alexander, Mohamed A. Bah, Brian Desmarais, David Douglass, Craig Outten, Joshua Schumacher,
Matthew Robinson, and Chen Zhang. "Indium bump deposition for flip-chip micro-array image sensing and
display applications." In Micro-and Nanotechnology Sensors, Systems, and Applications X, vol. 10639, pp. 425-
437. SPIE, 2018.
[100] Fan, Cui, Xue Li, Xiumei Shao, Zhijiang Zeng, Hengjing Tang, Tao Li, and Haimei Gong. "Study on reflow
process of SWIR FPA during flip-chip bonding technology." In Infrared Technology and Applications XLII, vol.
9819, pp. 342-348. SPIE, 2016.
[101] Kuo-Ning Chiang and Chang-An Yuan, "An overview of solder bump shape prediction algorithms with
validations," in IEEE Transactions on Advanced Packaging, vol. 24, no. 2, pp. 158-162, May 2001, doi:
10.1109/6040.928749
[102] Kim, J., Schoeller, H., Cho, J. et al., “Effect of Oxidation on Indium Solderability.”, J. Electron.
Mater., vol.37, pp.483–489, 2008, doi:10.1007/s11664-007-0346-7
[103] Young-Ho Kim, Jong Hwa Choi, Kang-Sik Choi, Hee Chul Lee, Choong- Kim, "New reflow process for
indium bump," Proc. SPIE 3061, Infrared Technology and Applications XXIII, (13 August 1997); doi:
10.1117/12.280315
[104] Ning-Cheng Lee, “Reflow soldering processes and troubleshooting:SMT, BGA, CSP and Flip chip
technologies” Chapter3, 1st edition, Newnes, 2002
[105] Zhang, Likun and Marston, Philip L., “Angular momentum flux of nonparaxial acoustic vortex beams and
torques on axisymmetric objects”, Phys. Rev. E vol.84, no.6, pp.065601, 2011,doi: 10.1103/PhysRevE.84.065601
[106] W. -C. Lo, Y. -L. Huang, C. -H. Fan and C. -K. Yeh, "3-D Ultrafast Ultrasound Imaging of Microbubbles
Trapped Using an Acoustic Vortex," in IEEE Transactions on Ultrasonics, Ferroelectrics, and Frequency Control,
vol. 68, no. 12, pp. 3507-3514, Dec. 2021, doi: 10.1109/TUFFC.2021.3095241.
[107] Matthew D. Guild, Jeffrey S. Rogers, Charles A. Rohde, Theodore P. Martin, Gregory J. Orris, "Far-field
superresolution imaging using shaped acoustic vortices," Proc. SPIE 10600, Health Monitoring of Structural and
Biological Systems XII, 1060015 (27 March 2018), doi: 10.1117/12.2296671
[108] N. Jiménez, J. M. Benlloch and F. Camarena, "A new elastographic technique using acoustic vortices," 2020
IEEE International Ultrasonics Symposium (IUS), Las Vegas, NV, USA, 2020, pp. 1-4, doi:
10.1109/IUS46767.2020.9251417.
[109] Shifang Guo, Zhen Ya, Pengying Wu, Lei Zhang, Mingxi Wan, “Enhanced Sonothrombolysis Induced by
High-Intensity Focused Acoustic Vortex”, Ultrasound in Medicine & Biology,vol. 48, no. 9, pp. 1907-1917, 2022,
doi:10.1016/j.ultrasmedbio.2022.05.021.
[110] Xinjia Li, Yuzhi Li, Qingyu Ma, Gepu Guo, Juan Tu, Dong Zhang; Principle and performance of orbital
angular momentum communication of acoustic vortex beams based on single-ring transceiver arrays. J. Appl.
Phys., vol. 127, no.12, pp.124902, 31 March 2020, doi: 10.1063/1.5135991
140
[111] Chengzhi Shi and Marc Dubois and Yuan Wang and Xiang Zhang, “High-speed acoustic communication
by multiplexing orbital angular momentum”, Proceedings of the National Academy of Sciences, vol.114, no.28,
pp.7250-7253, 2017, doi: 10.1073/pnas.1704450114
[112] J. Lee, K. S. Esfahani and E. S. Kim, "Vortex-Beam Acoustic Transducer for Underwater Propulsion," 2023
IEEE 36th International Conference on Micro Electro Mechanical Systems (MEMS), Munich, Germany, 2023,
pp. 977-980, doi: 10.1109/MEMS49605.2023.10052504.
[113] Li, J., Crivoi, A., Peng, X. et al. Three dimensional acoustic tweezers with vortex streaming. Commun Phys,
vol. 4, no. 113, 2021, doi: 10.1038/s42005-021-00617-0
[114] M. Baudoin and J.-L. Thomas, “Acoustic Tweezers for Particle and Fluid Micromanipulation”, Annual
Review of Fluid Mechanics, vol. 25, pp.205-234, 2020, doi: 10.1146/annurev-fluid-010719-060154
[115] J. Lee, K. Sadeghian Esfahani, M. Barekatain and E.S. Kim, “Lens-Less Acoustic Tweezers Based on SpiralArm Vortex-Beam Transducers Capable of Levitating, Trapping, and Manipulating Large and Heavy Particles,”
Transducers '23, The 22nd International Conference on Solid-State Sensors, Actuators and Microsystems, Kyoto,
Japan, June 25 - 29, 2023, pp. 405 - 408.
[116] Mohamed A. Ghanem et. al., “Noninvasive acoustic manipulation of objects in a living body”. Proceedings
of the National Academy of Sciences, vol. 117, no.29, pp.16848-16855, 2020, doi:10.1073/pnas.2001779117.
[117] Jhon F. Pazos-Ospina, Joao L. Ealo, Ediguer E. Franco, “Characterization of phased array-steered acoustic
vortex beams.”, J. Acoust. Soc. Am., vol.142, no.1, pp.61-71, 2017, doi: 10.1121/1.4985194
[118] Chen-chen Zhou, Pei-xia Li, Ning Ding, Shi-fu Pu, Ge-pu Guo, Yu-zhi Li, Qing-yu Ma, “Performance
improvement of focused acoustic-vortex tweezers constructed by a hyperboloidal acoustic lens and a circular
array”, Applied Acoustics, vol.200, 2022, 109503, 2022, doi: 10.1016/j.apacoust.2022.109053.
[119] Han Zhang and Yang Gao., “Acoustic Vortex Beam Generation by a Piezoelectric Transducer Using Spiral
Electrodes, Chinese Phys. Lett., vol.36, no.11, 2019, doi: 10.1088/0256-307X/36/11/114302
[120] Gong Z, Baudoin M. Three-Dimensional Trapping and Dynamic Axial Manipulation with Frequency-Tuned
Spiraling Acoustical Tweezers: A Theoretical Study. Physical review applied. 2021;16(2).
doi:10.1103/PhysRevApplied.16.024034
[121] G. -D. Kim et al., "A single FPGA-based portable ultrasound imaging system for point-of-care
applications," in IEEE Transactions on Ultrasonics, Ferroelectrics, and Frequency Control, vol. 59, no.
7, pp. 1386-1394, July 2012, doi: 10.1109/TUFFC.2012.2339.
[122] J. M. Baran and J. G. Webster, "Design of low-cost portable ultrasound systems: Review," 2009
Annual International Conference of the IEEE Engineering in Medicine and Biology Society, Minneapolis,
MN, USA, 2009, pp. 792-795, doi: 10.1109/IEMBS.2009.5332754.
[123] Juin-Jet Hwang, J. Quistgaard, J. Souquet and L. A. Crum, "Portable ultrasound device for battlefield
trauma," 1998 IEEE Ultrasonics Symposium. Proceedings (Cat. No. 98CH36102), Sendai, Japan, 1998, pp. 1663-
1667 vol.2, doi: 10.1109/ULTSYM.1998.765266.
[124] Yanick Baribeau, Aidan Sharkey, Omar Chaudhary, Santiago Krumm, Huma Fatima, Feroze Mahmood,
Robina Matyal, “Handheld Point-of-Care Ultrasound Probes: The New Generation of POCUS”, Journal of
Cardiothoracic and Vascular Anesthesia, vol.34, no.11, pp.3139-3145, 2020, doi:10.1053/j.jvca.2020.07.004.
141
[125] Deqing Kong, Yuan Qian, Minoru Kuribayashi Kurosawa, Manabu Aoyagi, “Evaluation method for acoustic
underwater propulsion systems”, J. Acoust. Soc. Am., vol.150, no.2, pp.1156-1164, 2021, doi:
10.1121/10.0005657
[126] Kong, Deqing, Takumi Hirata, Yimeng Wang, Fei Li, Minoru Kuribayashi Kurosawa, and Manabu Aoyagi.
"Acoustic Underwater Propulsion System Based on Ultrasonic Disc PZT Transducer." Sensors and Actuators A:
Physical, vol. 359, no. 1, 2023,doi: 10.1016/j.sna.2023.114502.
[127] J. Lee and E. S. Kim, “Wireless and Stand-Alone Submarine Propeller Based on Acoustic Propulsion,” SolidState Sensor and Actuator Workshop, Hilton Head Island, SC, June 5 - 9, 2022, pp. 230 - 233.
[128] Liqiang Yang, Dongyuan Qiu, Bo Zhang, Guidong Zhang and Wenxun Xiao, "A modified Z-source DC-DC
converter," 2014 16th European Conference on Power Electronics and Applications, Lappeenranta, 2014, pp. 1-
9, doi: 10.1109/EPE.2014.6910751.
[129] Manthiram, Arumugam., "An outlook on lithium ion battery technology.", ACS central science, vol. 3, no.
10, pp.1063-1069, 2017, doi: 10.1021/acscentsci.7b00288
[130] Jang J, Oh J, Jeong H, Kang W, Jo C., “A Review of Functional Separators for Lithium Metal Battery
Applications.”, Materials (Basel), vol.13, no.20m, 2020, doi: 10.3390/ma13204625.
[131] Kim, Taehoon ; Song, Wentao ; Son, Dae-Yong ; Ono, Luis K ; Qi, Yabing , “Lithium-ion batteries: outlook
on present, future, and hybridized technologies”, Journal of materials chemistry, vol. 7, no.7, pp.2942-2964, 2019
[132] P.U. Nzereogu, A.D. Omah, F.I. Ezema, E.I. Iwuoha, A.C. Nwanya, “Anode materials for lithium-ion
batteries: A review”, Applied Surface Science Advances, vol. 9, 2022, doi: 10.1016/j.apsadv.2022.100233.
[133] Zhang, L., Hu, X., Wang, Z., Sun, F. and Dorrell, D.G., “A review of supercapacitor modeling, estimation,
and applications: A control/management perspective.”, Renewable and Sustainable Energy Reviews, vol. 81,
pp.1868-1878, 2018, doi: 10.1016/j.rser.2017.05.283
[134] Goubard-Bretesché, N., Crosnier, O., Favier, F. and Brousse, T., “Improving the volumetric energy density
of supercapacitors.”, Electrochimica Acta, vol. 206, pp.458-463, 2016, doi: 10.1016/j.electacta.2016.01.171
[135] Kim, Hee-Je, TNV Krishna, Kamran Zeb, Vinodh Rajangam, Chandu V. V. Muralee Gopi, Sangaraju
Sambasivam, Kummara Venkata Guru Raghavendra, and Ihab M. Obaidat. 2020. "A Comprehensive Review of
Li-Ion Battery Materials and Their Recycling Techniques" Electronics, vol.9, no. 7, pp.1161, 2020,
doi.org/10.3390/electronics9071161
[136] Soltani, M. and Beheshti, S.H., “A comprehensive review of lithium ion capacitor: development, modelling,
thermal management and applications.”, Journal of Energy Storage, vol.34, 2021, doi: 10.1016/j.est.2020.102019
[137] Raghavendra, Kummara Venkat Guru, Kamran Zeb, Anand Muthusamy, T. N. V. Krishna, S. V. S. V
Prabhudeva Kumar, Do-Hyun Kim, Min-Soo Kim, Hwan-Gyu Cho, and Hee-Je Kim. "A Comprehensive Review
of DC–DC Converter Topologies and Modulation Strategies with Recent Advances in Solar Photovoltaic
Systems" Electronics, vol. 9, no. 1, 2020, doi:10.3390/electronics9010031
[138] Mehta, N., 2015. GaN FET module performance advantage over silicon. Texas Instruments, pp.1-7.
[139] S. Khoshzaman and I. Hahn, "A Performance Comparison of GaN FET and Silicon MOSFET," 22nd IEEE
International Conference on Industrial Technology (ICIT), Valencia, Spain, 2021, pp. 127-133, doi:
10.1109/ICIT46573.2021.9453693.
[140] Sun, Bingyao. "Does GaN have a body diode?-understanding the third quadrant operation of
GaN." Application Report SNOAA36; Texas Instruments: Dallas, TX, USA (2019).
142
[141] Coilcraft, XAL5030, “Shielded Power Inductors – XAL5030”, Datasheet, revised 2023
[142] Coilcraft, XAL1010, “Shielded Power Inductors – XAL1010”, Datasheet, revised 2021
[143] Ahn, J.H., Lee, B.K. and Kim, J.S., “Comparative Performance Evaluation of Si MOSFET and GaN FET
Power System.”, The Transactions of the Korean Institute of Power Electronics, vol.19, no.3, pp.283-289.,2014,
doi: 10.6113/TKPE.2014.19.3.283
[144] Texas Instruments, LMG1205, “LMG1205 100-V, 1.2-A to 5-A, Half Bridge GaN Driver with Integrated
Bootstrap Diode”, Datasheet, 2017
[145] Efficient Power Conversion, EPC2001, " Enhancement Mode Power Transistor”, Datasheet, 2013
[146] Y. Tang and E. S. Kim, "Nozzleless Acoustic Droplet Ejector With Electrically Tunable Droplet Size for
Picking and Placing Semiconductor Chips," in Journal of Microelectromechanical Systems, vol. 30, no. 2, pp.
262-270, April 2021, doi: 10.1109/JMEMS.2020.3046564.
[147] Blana A, Walter B, Rogenhofer S, Wieland WF., “High-intensity focused ultrasound for the treatment of
localized prostate cancer: 5-year experience.” Urology (Ridgewood, NJ)., vol. 63, no.2, pp. 297-300, 2004,
doi:10.1016/j.urology.2003.09.020
[148] Darrow, D.P., “Focused Ultrasound for Neuromodulation.”, Neurotherapeutics, vol.16, pp.88–99, 2019, doi:
10.1007/s13311-018-00691-3
[149] J. Yuh & M. West, “Underwater robotics” , Advanced Robotics, vol. 15, no.5, pp.609-639, 2001, doi:
10.1163/156855301317033595
[150] Bogue, R., "Robots for monitoring the environment", Industrial Robot, Vol. 38 No. 6, pp. 560-566., 2011,
doi:10.1108/01439911111179066
[151] Terracciano, D., Bazzarello, L., Caiti, A. et al., “Marine Robots for Underwater Surveillance.”, Curr Robot
Rep, vol. 1, pp.159–167, 2020, doi: 10.1007/s43154-020-00028-z
[152] Jang D, Jeong J, Song H, Chung SK., “Targeted drug delivery technology using untethered microrobots: a
review.”, Journal of micromechanics and microengineering., vol.29, no.5, 53002, 2019, doi:10.1088/1361-
6439/ab087d
[153] R. Du, Z. Li, K. Youcef-Toumi, and P. Valdivia y Alvarado, eds. “Robot fish: Bio-inspired fishlike underwater
robots,” Springer, 2015.
[154] Allison, E.M., Springer, G.S. and Van Dam, J., “Ultrasonic propulsion.”, Journal of propulsion and
power, vol.24, no.3, pp.547-553, 2008, doi:10.2514/1.30044.
[155] Doinikov AA., “Acoustic radiation pressure on a compressible sphere in a viscous fluid.” Journal of fluid
mechanics, vol. 267, pp.1-22, 1994, doi:10.1017/S0022112094001096
[156] Iwijn De Vlaminck, Cees Dekker, “Recent Advances in Magnetic Tweezers”, Annual Review of Biophysics,
vol. 41, no.1 ,pp. 453-473, 2012, doi: 10.1146/annurev-biophys-122311-100544
[157] Monica Tanase, Nicolas Biais, Michael Sheetz, “Magnetic Tweezers in Cell Biology”, Methods in Cell
Biology, vol. 87, pp.473-493, 2007, doi: 10.1016/S0091-679X(07)83020-2
[158] P.A. Valberg, D.F. Albertini, “Cytoplasmic motions, rheology, and structure probed by a novel magnetic
particle method”, J. Cell Biol., vol. 101, pp. 130-140, 1985, doi: 10.1083/jcb.101.1.130
143
[159] Philip H. Jones, Onofrio M. Maragò, Giovanni Volpe, Optical Tweezers: Principles and Applications,
Cambridge University Press, 1995.
[160] Y. Liu, D.K. Cheng, G.J. Sonek, M.W. Berns, C.F. Chapman, B.J. Tromberg, “Evidence for localized cell
heating induced by infrared optical tweezers”, Biophysical Journal, Vol.68, no,5, Pages 2137-2144, 1995,
doi:10.1016/S0006-3495(95)80396-6
[161] Y. Tang and E. S. Kim, "Ring-Focusing Fresnel Acoustic Lens for Long Depth-of-Focus Focused Ultrasound
With Multiple Trapping Zones," in Journal of Microelectromechanical Systems, vol. 29, no. 5, pp. 692-698, Oct.
2020. doi:10.1109/JMEMS.2020.3000715
[162] K. Sadeghian Esfahani, Y. Tang, J. Lee, M. Barekatain, and E. S. Kim, “Underwater Acoustic Tweezers
Capable of Trapping Large and Heavy Particles,” Solid-State Sensor and Actuator Workshop, Hilton Head Island,
SC, June 5 - 9, 2022, pp. 43 – 4.
[163] Goossens WRA., “Review of the empirical correlations for the drag coefficient of rigid spheres”. Powder
technology. Vol. 352, pp. 350-359, 2019, doi:10.1016/j.powtec.2019.04.075
[164] Carmona-Torre F, Martinez-Urbistondo D, Del Pozo JL, et al. Neurostimulation for Parkinson’s Disease with
Early Motor Complications. The New England journal of medicine., vol. 368, no.21, pp. 2037-2038, 2013,
doi:10.1056/NEJMc1303485
[165] Lin, Yicong, and Yuping Wang. “Neurostimulation as a promising epilepsy therapy.” Epilepsia open vol. 2,
no.4, pp. 371-387. 23 Aug. 2017, doi:10.1002/epi4.12070
[166] D. Mccreery, V. Pikov, and P. R. Troyk, “Neuronal loss due to prolonged controlled-current stimulation with
chronically implanted microelectrodes in the cat cerebral cortex”, Journal of Neural Engineering, vol. 7, no. 3, p.
036005, Nov. 2010, doi:10.1088/1741-2560/7/3/036005
[167] Beltramo R., D'Urso G., Dal Maschio M. D., Farisello P., Bovetti S., Clovis Y., Lassi G., Tucci V., De Pietri
Tonelli D. and Fellin T., “Layer-specific excitatory circuits differentially control recurrent network dynamics in
the neocortex”, Nat. Med. , vol. 16, pp.227-234, 2013,doi:10.1038/nn.3306
[168] Armbruster B. N., Li X., Pausch M. H., Herlitze S. and Roth B. L., “Evolving the lock to fit the key to create
a family of G protein-coupled receptors potently activated by an inert ligand”, Proc. Natl. Acad. Sci., vol. 104,
pp. 5163-5168, 2007, doi:10.1073/pnas.0700293104
[169] K. R. Mills, “Transcranial magnetic stimulation,” Oxford Medicine Online, 2016.
[170] Grossman N., Bono D., Dedic N., Kodandaramaiah S. B., Rudenko A., Suk H. J., Cassara A. M., Neufeld E.,
Kuster N., Tsai L. H. et al., “Noninvasive deep brain stimulation via temporally interfering electric fields”, Cell,
vol. 169, pp. 1029-1041, 2017, doi:10.1016/j.cell.2017.05.024
[171] Wang Y, Guo L., “Nanomaterial-enabled neural stimulation.”, Frontiers in neuroscience., vol. 10, pp.69-69,
2016, doi:10.3389/fnins.2016.00069
[172] Tyler WJ, Tufail Y, Finsterwald M, Tauchmann ML, Olson EJ, Majestic C., “Remote excitation of neuronal
circuits using low-intensity, low-frequency ultrasound.”, PloS one., vol. 3, no. 10, pp.e3511-e3511, 2008
doi:10.1371/journal.pone.0003511
[173] Mahmoudi P, Veladi H, Pakdel FG. “Optogenetics, Tools and Applications in Neurobiology.”, Journal of
medical signals and sensors., vol. 7, no. 2, pp. 71-79, 2017 doi:10.4103/2228-7477.205506
144
[174] Beltramo R, D’Urso G, Dal Maschio M, et al., “Layer-specific excitatory circuits differentially control
recurrent network dynamics in the neocortex.”, Nature neuroscience., vol. 16, no. 2, pp. 227-234. 2013,
doi:10.1038/nn.3306
[175] Jorfi M, Skousen JL, Weder C, Capadona JR., “Progress towards biocompatible intracortical microelectrodes
for neural interfacing applications.”, J Neural Eng., vol. 12, no. 1, 2015, doi: 10.1088/1741-2560/12/1/011001.
[176] Dobrzanski, G., Kossut, M., “Application of the DREADD technique in biomedical brain
research.”, Pharmacol. Rep, vol. 69, pp. 213–221, 2017, doi:10.1016/j.pharep.2016
[177] Whissell PD, Tohyama S, Martin LJ. “The Use of DREADDs to Deconstruct Behavior.”, Front Genet., 2016
May 17;7:70. doi: 10.3389/fgene.2016.00070
[178] Soloperto A, Boccaccio A, Contestabile A, Moroni M, Hallinan GI, Palazzolo G, Chad J, Deinhardt K,
Carugo D, Difato F. “Mechano-sensitization of mammalian neuronal networks through expression of the bacterial
large-conductance mechanosensitive ion channel.”, J. Cell Sci. 2018 Mar 8;131(5):jcs210393. doi:
10.1242/jcs.210393.
[179] Radman T, Ramos RL, Brumberg JC, Bikson M. Role of cortical cell type and morphology in subthreshold
and suprathreshold uniform electric field stimulation in vitro. Brain Stimul. 2009 Oct;2(4):215-28, 228.e1-3. doi:
10.1016/j.brs.2009.03.007.
[180] Garnaat SL, Yuan S, Wang H, Philip NS, Carpenter LL. Updates on Transcranial Magnetic Stimulation
Therapy for Major Depressive Disorder. Psychiatr Clin North Am. 2018 Sep;41(3):419-431. doi:
10.1016/j.psc.2018.04.006. PMID: 30098655; PMCID: PMC6979370.
[181] Rizvi, Sukaina, and Ali M Khan. “Use of Transcranial Magnetic Stimulation for Depression.” Cureus, vol.
11, no. 5, pp. e4736, 2019, doi:10.7759/cureus.4736
[182] C. -Y. Lee, H. Yu and E. S. Kim, "Harmonic Operation of Acoustic Transducer for Droplet Ejection
Application," TRANSDUCERS 2007 - 2007 International Solid-State Sensors, Actuators and Microsystems
Conference, Lyon, France, 2007, pp. 1283-1286, doi: 10.1109/SENSOR.2007.4300372.
[183] Soloukey S, Vincent AJPE, Satoer DD, et al. Functional Ultrasound (fUS) During Awake Brain Surgery: The
Clinical Potential of Intra-Operative Functional and Vascular Brain Mapping. Frontiers in neuroscience.
2020;13:1384-1384. doi:10.3389/fnins.2019.01384
[184] Krishna V, Sammartino F, Rezai A. A Review of the Current Therapies, Challenges, and Future Directions of
Transcranial Focused Ultrasound Technology: Advances in Diagnosis and Treatment. JAMA Neurol. 2018 Feb
1;75(2):246-254. doi: 10.1001/jamaneurol.2017.3129. PMID: 29228074.
[185] Eisenberg HM, Krishna V, Elias WJ, et al. “MR-guided focused ultrasound pallidotomy for Parkinson’s
disease: Safety and feasibility.”, Journal of neurosurgery., vol. 135, no. 3, pp.792-798, 2021,
doi:10.3171/2020.6.JNS192773
[186] FRY FJ, ADES HW, FRY WJ. Production of reversible changes in the central nervous system by ultrasound.
Science. 1958 Jan 10;127(3289):83-4. doi: 10.1126/science.127.3289.83. PMID: 13495483.
[187] Tufail Y, Matyushov A, Baldwin N, et al. Transcranial Pulsed Ultrasound Stimulates Intact Brain
Circuits. Neuron (Cambridge, Mass). 2010;66(5):681-694. doi:10.1016/j.neuron.2010.05.008
[188] Prieto ML, Firouzi K, Khuri-Yakub BT, Maduke M. Activation of Piezo1 but Not NaV1.2 Channels by
Ultrasound at 43 MHz. Ultrasound in medicine & biology. 2018;44(6):1217-1232.
doi:10.1016/j.ultrasmedbio.2017.12.020
145
[189] Prieto ML, Madison DV, Khuri-Yakub BT, Maduke M. Focused Ultrasound Activates Task Potassium
Channels, Increases Membrane Capacitance, and Modulates Action Potential Waveform and Firing Properties in
Hippocampal Brain Slices. Biophysical journal. 2018;114(3):669a-669a. doi:10.1016/j.bpj.2017.11.3609
[190] Kubanek J, Shi J, Marsh J, Chen D, Deng C, Cui J. Ultrasound modulates ion channel currents. Scientific
reports. 2016;6(1):24170-24170. doi:10.1038/srep24170
[191] Lin JW, Yu F, Müller WS, Ehnholm G, Okada Y. Focused ultrasound transiently increases membrane
conductance in isolated crayfish axon. Journal of neurophysiology. 2019;121(2):480-489.
doi:10.1152/jn.00541.2018
[192] Yu K, Niu X, Krook-Magnuson E, He B. Intrinsic functional neuron-type selectivity of transcranial focused
ultrasound neuromodulation. Nature communications. 2021;12(1):2519-2519. doi:10.1038/s41467-021-22743-7
[193] Manuel TJ, Kusunose J, Zhan X, et al. Ultrasound neuromodulation depends on pulse repetition frequency
and can modulate inhibitory effects of TTX. Scientific reports. 2020;10(1):15347-15347. doi:10.1038/s41598-
020-72189-y
[194] Asan AS, Kang Q, Oralkan Ö, Sahin M. Entrainment of cerebellar Purkinje cell spiking activity using pulsed
ultrasound stimulation. Brain stimulation. 2021;14(3):598-606. doi:10.1016/j.brs.2021.03.004
[195] Prieto ML, Firouzi K, Khuri-Yakub BT, Madison DV, Merritt Maduke. Spike-frequency dependent inhibition
and potentiation of neural activity by ultrasound. bioRxiv. Published online 2020. doi:10.1101/2020.06.01.128710
[196] Oh SJ, Lee JM, Kim HB, et al. Ultrasonic Neuromodulation via Astrocytic TRPA1. Current biology.
2019;29(20):3386-3401.e8. doi:10.1016/j.cub.2019.08.021
[197] Hounsfield GN. The E.M.I. Scanner. Proceedings of the Royal Society of London Series B, Biological
sciences. 1977;195(1119):281-289. doi:10.1098/rspb.1977.0008
[198] Rachit Pandey, Sravya Tekumalla, Manoj Gupta, “MI shielding of metals, alloys, and composites”, Materials
for Potential EMI Shielding Applications, pp.341-355, 2020, doi: 10.1016/B978-0-12-817590-3.00021-X.
[199] Neher, Erwin, and Bert Sakmann. “The Patch Clamp Technique.” Scientific American, vol. 266, no. 3, pp.
44–51, 1992, doi:10.1038/scientificamerican0392-44
[200] Nelson TR, Fowlkes JB, Abramowicz JS, Church CC. Ultrasound Biosafety Considerations for the Practicing
Sonographer and Sonologist. Journal of ultrasound in medicine. 2009;28(2):139-150.
doi:10.7863/jum.2009.28.2.139
[201] Lamas et. al. “Ion Channels and Thermosensitivity: TRP, TREK, or Both?”, Int. J. Mol. Sci. 2019, 20, 2371;
doi:10.3390/ijms20102371
[202] Yoo S, Mittelstein DR, Hurt RC, Lacroix JJ, Shapiro MG. Focused ultrasound excites neurons via
mechanosensitive calcium accumulation and ion channel amplification. bioRxiv. Published online 2020.
doi:10.1101/2020.05.19.101196
[203] Burks SR, Lorsung RM, Nagle ME, Tu TW, Frank JA. Focused ultrasound activates voltage-gated calcium
channels through depolarizing TRPC1 sodium currents in kidney and skeletal muscle. Theranostics.
2022;12(3):1341-1341. doi:10.7150/thno.69908
[204] Yang, Y., Yuan, J., Field, R.L. et al. Induction of a torpor-like hypothermic and hypometabolic state in rodents
by ultrasound. Nat Metab ,vol.5, pp. 789–803,2023 doi: 10.1038/s42255-023-00804-z
146
[205] Feliciangeli S, Chatelain FC, Bichet D, Lesage F. The family of K2P channels: salient structural and
functional properties. The Journal of physiology. 2015;593(12):2587-2603. doi:10.1113/jphysiol.2014.287268
[206] Sorum B, Rietmeijer RA, Gopakumar K, Adesnik H, Brohawn SG. Ultrasound activates mechanosensitive
TRAAK K+ channels directly through the lipid membrane. bioRxiv. Published online 2020.
doi:10.1101/2020.10.24.349738
[207] H Gao; J G Hollyfield, “Aging of the human retina. Differential loss of neurons and retinal pigment epithelial
cells”, Investigative Ophthalmology & Visual Science, vol. 33, pp.1-17, 1992, doi: 10.3389/fnagi.2022.778404.
[208] Amanda-Jayne F. Carr, Matthew J.K. Smart, Conor M. Ramsden, Michael B. Powner, Lyndon da Cruz, Peter
J. Coffey, “Development of human embryonic stem cell therapies for age-related macular degeneration”, Trends
in Neurosciences, vol. 36, no.7, pp.385-395, 2013, doi: 10.1016/j.tins.2013.03.006
[209] Maeda T, Sugita S, Kurimoto Y, Takahashi M., “Trends of Stem Cell Therapies in Age-Related Macular
Degeneration.”, Journal of Clinical Medicine, vol. 10, no.8, 2021, doi:10.3390/jcm10081785
[210] Hossein Nazari, Li Zhang, Danhong Zhu, Gerald J. Chader, Paulo Falabella, Francisco Stefanini, Teisha
Rowland, Dennis O. Clegg, Amir H. Kashani, David R. Hinton, Mark S. Humayun, “Stem cell based therapies
for age-related macular degeneration: The promises and the challenges”, Progress in Retinal and Eye Research,
vol. 48, pp. 1-39, 2015, doi:10.1016/j.preteyeres.2015.06.004
[211] Heng, B., Liu, H., Ge, Z. and Cao, T., “Mechanical dissociation of human embryonic stem cell colonies by
manual scraping after collagenase treatment is much more detrimental to cellular viability than is trypsinization
with gentle pipetting”, Biotechnology and Applied Biochemistry, vol. 47, pp. 3- 37, 2007, doi:
10.1042/BA20060151
[212] P. Liang, B. Liu, Y. Wang, K. Liu, Y. Zhao, W. E. Huang, and B. Li, “Isolation and Culture of Single Microbial
Cells by Laser Ejection Sorting Technology”, Applied and Environmental Microbiology, vol. 88, no. 3, 2022.
doi:10.1128/aem.01165-21
[213] Austin Roorda, Yuhua Zhang, Jacque L. Duncan, “High-Resolution In Vivo Imaging of the RPE Mosaic in
Eyes with Retinal Disease.”, Invest. Ophthalmol. Vis. Sci. vol. 48, no.5, pp. 2297-2303, 2007,
doi:10.1167/iovs.06-1450.
[214] Wang XQ, Duan XM, Liu LH, Fang YQ, Tan Y., “Carboxyfluorescein diacetate succinimidyl ester
fluorescent dye for cell labeling.”, Acta Biochim Biophys Sin, vol. 37, no. 6, pp.379-385, 2005, doi:
10.1111/j.1745-7270.2005.00051.x. PMID: 15944752.
[215] Ye PP, Brown JR, Pauly KB. “Frequency Dependence of Ultrasound Neurostimulation in the Mouse
Brain.”, Ultrasound in medicine & biology., vo. 42, no. 7, pp. 1512-1530, 2016
doi:10.1016/j.ultrasmedbio.2016.02.012
[216] Y. Tang and E.S. Kim, "Acoustic Droplet-Assisted Particle Ejection through and from Agarose-Gel-Filled
Petri Dish," 2019 IEEE International Ultrasonics Symposium (IUS), Glasgow, UK, 2019, pp. 64-67, doi:
10.1109/ULTSYM.2019.8925932.
[217] Nelson TR, Fowlkes JB, Abramowicz JS, Church CC., “Ultrasound Biosafety Considerations for the
Practicing Sonographer and Sonologist.”, Journal of ultrasound in medicine., vol. 28, no. 2, pp. 139-150, 2009,
doi:10.7863/jum.2009.28.2.139
[218] Smith EN, D’Antonio-Chronowska A, Greenwald WW, et al. ,“Human iPSC-Derived Retinal Pigment
Epithelium: A Model System for Prioritizing and Functionally Characterizing Causal Variants at AMD Risk
Loci.”, Stem cell reports., vol. 12, no.6, pp.1342-1353, 2019, doi:10.1016/j.stemcr.2019.04.012
147
[219] Livak KJ, Schmittgen TD., “Analysis of relative gene expression data using real-time quantitative PCR and
the 2(-Delta Delta C(T)) Method. Methods. 2001 Dec;25(4):402-8. doi: 10.1006/meth.2001.1262.
[220] Kim, Tae Kyun. "T test as a parametric statistic." Korean journal of anesthesiology, vol. 68, no. 6, pp. 540-
546, 2015, doi: 10.4097/kjae.2015.68.6.540
[221] Nicoll RA, Tsuruda PR, Milstein AD, et al. Pungent agents from Szechuan peppers excite sensory neurons
by inhibiting two-pore potassium channels. Nature neuroscience. 2008;11(7):772-779. doi:10.1038/nn.2143
[222] Huang W, Ke Y, Zhu J, Liu S, Cong J, Ye H, Guo Y, Wang K, Zhang Z, Meng W, Gao TM, Luhmann HJ,
Kilb W, Chen R. TRESK channel contributes to depolarization-induced shunting inhibition and modulates
epileptic seizures. Cell Rep. 2021 Jul 20;36(3):109404. doi: 10.1016/j.celrep.2021.109404. PMID: 34289346.
Abstract (if available)
Abstract
Ultrasound technology has played significant roles in the field of biomedical engineering, offering multipurpose applications in both diagnostics and therapeutics. While its current prominence predominantly lies in diagnostics, there exists vast untapped potential for its application in disease treatment, cell sorting, and manipulation. This thesis embarks on a comprehensive exploration of an integrated wireless piezoelectric ultrasonic transducer system for biomedical applications. The objective is to enhance the understanding of ultrasound's diverse applications and the development of wireless ultrasound transducer system integrated with an electronic driving circuit and wireless connectivity and wireless power transfers that push the boundaries of what is possible in contemporary biomedical engineering.
The introductory chapter provides a panoramic view of ultrasound's journey in healthcare. It begins by highlighting its pivotal role in diagnostic imaging. The discussion extends to therapeutic ultrasound, where high-intensity focused ultrasound (HIFU) is employed to ablate pathological tissues such as cancer cells. The emerging applications of ultrasound for drug delivery and Low-intensity focused ultrasound (LIFU) non-invasive neuromodulation underscore the technology's potential to revolutionize healthcare.
The subsequent chapters delve into the core of this thesis's research, commencing with an exploration of acoustic transducers fabricated with microelectromechanical system (MEMS) technology. These components, which generate ultrasound waves, are fundamental to ultrasound systems. The scope of the thesis extends to the design, fabrication, and characterization of specialized acoustic transducer systems. These systems include Self-Focusing Acoustic Transducers (SFAT), electrically Controllable SFAT (eCON SFAT), and Spiral-arm Vortex-beam Acoustic Transducers (SVAT). Each chapter provides a deep dive into the intricacies of these transducer systems, exploring their design parameters, fabrication techniques, and methods for characterization.
The role of electronic circuits in driving and controlling acoustic transducers is examined in detail. The thesis explores electronic systems designed for continuous and pulsed operation, highlighting their significance in the proposed integrated wireless piezoelectric ultrasonic transducer system for biomedical applications. For continuous operation with low operating voltage (30Vppmax), a power amplifier and DC to DC converter are used, while Z-source network with GaNFET (Gallium Nitride Field Effect Transistor) provides higher output voltage (60Vpp) with a limited stability. The integration of Bluetooth Low Energy (BLE) technology for wireless control is also addressed, emphasizing the adaptability and connectivity of the system.
Subsequent chapters shift the focus towards practical applications of the transducer systems. The development of wireless integrated systems is explored, catering to scenarios such as submarine propulsion, particle (or cell) levitating acoustic tweezers, contactless single-cell extraction, and focused ultrasound neuromodulation. Experimental results and discussions for each application highlight the potential of these systems in real-world scenarios.
The thesis concludes in a summary chapter, offering an overview of the research findings and their implications for the future. It sets the stage for potential directions in the field of ultrasound technology, emphasizing the ongoing innovation and interdisciplinary collaboration that drive advancements in biomedical applications.
Linked assets
University of Southern California Dissertations and Theses
Conceptually similar
PDF
Piezoelectric ultrasonic and acoustic microelectromechanical systems (MEMS) for biomedical, manipulation, and actuation applications
PDF
Transducers and signal processing techniques for simultaneous ultrasonic imaging and therapy
PDF
Additive manufacturing of piezoelectric and composite for biomedical application
PDF
High-resolution data acquisition with neural and dermal interfaces
PDF
Development of implantable Parylene-based MEMS technologies for cortical applications
PDF
Fabrication of ultrasound transducer and 3D-prinitng ultrasonic device
PDF
Array transducers for high frequency ultrasound imaging
PDF
Design and development of ultrasonic array transducers for specialized applications
PDF
High frequency ultrasonic phased array system and its applications
PDF
A wireless implantable MEMS micropump system for site-specific anti-cancer drug delivery
PDF
Battery-less detection and recording of tamper activity along with wireless interrogation
PDF
Acoustic ejector employing lens with air-reflectors and piezoelectrically actuated tunable capacitor
PDF
Parylene C bioMEMS for implantable devices with electrochemical interfaces
PDF
Magnetic spring in electromagnetic vibration energy harvester and applications of focused ultrasonic transducer
PDF
Microfluidic cell sorting with a high frequency ultrasound beam
PDF
Development of high frequency focused transducers for single beam acoustic tweezers
PDF
Parylene-based implantable interfaces for biomedical applications
PDF
Strategies for improving mechanical and biochemical interfaces between medical implants and tissue
PDF
The electrochemical evaluation of Parylene-based electrodes for neural applications
PDF
Parylene-based biomems sensors for multiple physiological systems
Asset Metadata
Creator
Lee, Jaehoon
(author)
Core Title
Integrated wireless piezoelectric ultrasonic transducer system for biomedical applications
School
Viterbi School of Engineering
Degree
Doctor of Philosophy
Degree Program
Electrical Engineering
Degree Conferral Date
2024-05
Publication Date
01/30/2024
Defense Date
12/18/2023
Publisher
Los Angeles, California
(original),
University of Southern California
(original),
University of Southern California. Libraries
(digital)
Tag
Acoustic MEMS,Acoustic transducer,Acoustics,biomedical,bioMEMS,integrated circuits,MEMS,neuromodulation,Portable Wireless System
Format
theses
(aat)
Language
English
Contributor
Electronically uploaded by the author
(provenance)
Advisor
Kim, Eun Sok (
committee chair
), Meng, Ellis Fan-Chuin (
committee member
), Wu, Wei (
committee member
)
Creator Email
jaehoon.lee13@gmail.com,lee172@usc.edu
Permanent Link (DOI)
https://doi.org/10.25549/usctheses-oUC113814340
Unique identifier
UC113814340
Identifier
etd-LeeJaehoon-12638.pdf (filename)
Legacy Identifier
etd-LeeJaehoon-12638
Document Type
Thesis
Format
theses (aat)
Rights
Lee, Jaehoon
Internet Media Type
application/pdf
Type
texts
Source
20240130-usctheses-batch-1123
(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
Tags
Acoustic MEMS
Acoustic transducer
biomedical
bioMEMS
integrated circuits
MEMS
neuromodulation
Portable Wireless System