The Pennsylvania State University

The Graduate School

College of Engineering

ACOUSTIC TWEEZERS: MANIPULATING CELLS IN MICROFLUIDICS

A Dissertation in

Engineering Science and Mechanics

Xiaoyun Ding

 2013 Xiaoyun Ding

Submitted in Partial Fulfillment of the Requirements for the Degree of

Doctor of Philosophy

August 2013

The dissertation of Xiaoyun Ding was reviewed and approved* by the following:

Tony Jun Huang Professor of Engineering Science and Mechanics Dissertation Advisor Chair of Committee

Bernhard R. Tittmann Schell Professor of Engineering Science and Mechanics

Corina S. Drapaca Assistant Professor of Engineering Science and Mechanics

Qiming Zhang Distinguished Professor of Electrical Engineering

Judith A. Todd Professor P. B. Breneman Head of the Department of Engineering Science and Mechanics

*Signatures are on file in the Graduate School

iii

ABSTRACT

Techniques that can noninvasively and dexterously manipulate cells and other bioparticles (such as organisms, DNAs, proteins, and viruses) in a compact system are invaluable for many applications in life sciences and medicine. Historically, optical tweezers have been the primary tool used in the scientific community for bioparticle manipulation. Despite the remarkable capability and success, optical tweezers have notable limitations, such as complex and bulky instrumentation, high equipment costs, and potential damage to cells. To overcome the limitations of optical tweezers and other particle manipulation methods, we have developed a series of acoustic-based, on-chip devices called acoustic tweezers that can manipulate cells and other bioparticles using sound waves. The power density required by our acoustic device is

10,000,000 times less than that of optical tweezers, which renders the technique noninvasive and amenable to miniaturization. Cell viability, proliferation, and apoptosis assays are also conducted to confirm the non-invasiveness of our technique. The simple structure/setup of these acoustic tweezers can be integrated with a small radio-frequency power supply and basic electronics to function as a fully integrated, portable, and inexpensive cell-manipulation system.

In this dissertation, we demonstrated an acoustic tweezers platform that can achieve the following functions: 1) trap and dexterously manipulate single microparticle, cell, and entire organism (i.e., C. elegans) along a programmed route in two-dimensions within a microfluidic chip; 2) dynamically tune the cell patterning in microfluidic channel; 3) precisely sort cell into five separate outlets in one step, rendering it particularly desirable for multi-type cell sorting; and

4) achieve high-efficiency (>95%) separation of human leukemia cells (HL-60) and human breast cancer cells (MCF-7) from human blood cells.

TABLE OF CONTENTS

List of Figures ...... vi

List of Tables ...... xii

Acknowledgements ...... xiii

Chapter 1 Introduction and Background ...... 1

1.1 Motivation ...... 2 1.2 Fundamentals of SAW ...... 5 1.2.1 Generation of SAW ...... 5 1.2.2 SAW-induced streaming ...... 7 1.2.3 Formation of SSAWs ...... 10 1.2.4 Primary acoustic radiation force in SSAWs ...... 12 1.3 Literature Review of SAW microfluidics ...... 15 1.3.1 Fluid mixing ...... 15 1.3.2 Fluid translation...... 16 1.3.3 Jetting and atomization ...... 23 1.3.4 Particle/cell concentration ...... 25 1.3.5 Reorientation of nano-objects ...... 26 1.3.6 Phononic Crystal-Assisted SAWs ...... 30 1.3.7 Microfluidic Technologies Enabled by SSAWs ...... 32 1.4 Goals and organization ...... 38

Chapter 2 Single particle, cell, and organism manipulation using acoustic tweezers ...... 40

2.1 Introduction ...... 40 2.2 Working mechanism ...... 42 2.3 Methods ...... 45 2.3.1 Materials ...... 45 2.3.2 Device fabrication ...... 46 2.3.3 Experimental Setup ...... 49 2.3.3 Cell Viability and Proliferation Assays ...... 50 2.4 Results and discussion ...... 51 2.4.1 Characterization of the acoustic tweezers ...... 51 2.4.2 Two-Dimensional Manipulation of Single Particles, Cells, and Organisms ... 53 2.4.3 Massive manipulation of particles ...... 58 2.5 Conclusion ...... 61

Chapter 3 Multichannel cell sorting using acoustic tweezers ...... 62

3.1 Introduction ...... 62 3.2 Working mechanism ...... 66 3.3 Methods ...... 68 3.4 Results and discussion ...... 69 3.5 conclusion ...... 73

Chapter 4 Tunable patterning of particles and cells using acoustic tweezers ...... 74

4.1 Introduction ...... 74 4.2 Working mechanism ...... 77 4.3 Methods ...... 81 4.3.1 Device design and fabrication ...... 81 4.3.2 System setup ...... 82 4.4 Results and discussion ...... 82 4.4.1 Tunable 1D patterning ...... 82 4.4.2 Tunable 2D patterning ...... 87 4.5 conclusion ...... 90

Chapter 5 High efficiency cell separation using acoustic tweezers ...... 92

5.1 Introduction ...... 92 5.2 Working mechanism ...... 94 5.3 Results and discussion ...... 97 5.3.1 Particle separation based on size ...... 97 5.3.2 Particle separation based on compressibility ...... 99 5.3.3 Cell separation ...... 101 5.4 Methods ...... 106 5.5 Conclusion ...... 106

Chapter 6 Conclusions and perspectives ...... 108

6.1 Achievements ...... 108 6.2 Perspectives ...... 110 References ...... 114 Appendix Highlights and medium reports ...... 124

LIST OF FIGURES

Figure 1-1 A uniform, metallic interdigital transducer (IDT) deposited on the piezoelectric substrate generates SAWs that propagate along the substrate surface in both directions...... 6

Figure 1-2 SAW-induced acoustic streaming. (a) Rayleigh SAW is excited by means of an IDT. Underneath a liquid, the SAW turns into a leaky SAW, radiating pressure waves at the Rayleigh angle ΘR into the fluid. The streaming map is from Koster. Ref. [30]. (b) Comparison of experimental results and numerical modelling for a hemispherical droplet positioned at the centre of the SAW propagation direction...... 8

Figure 1-3 (a) A cross-section schematic of a microfluidic channel and the two IDTs used to generate travelling SAWs that propagate in opposite directions, establishing a SSAW across the channel width. This SSAW radiates acoustic leakage waves into the liquid at the Rayleigh angle. (b) and (c) show the simulated interference pattern of a one-dimensional and two-dimensional SSAW field on the substrate surface, respectively...... 11

Figure 1-4 Time snapshots of displacement field of the longitudinal-mode leakage wave from a SAW propagating in (a) the positive x-direction and (b) the negative x- direction. The width of the channel is half the SAW wavelength and the height is one SAW wavelength. The pressure field resulting from the combined displacement fields (i.e. resulting from a SSAW on the substrate) is shown in the cross-section of the channel at (c) t = t0 and (d) t = t0 + . The dotted line denotes the pressure nodal plane in the middle of the channel...... 14

Figure 1-5 Fluid mixing by SAW-induced acoustic streaming. (a) Rapid mixing of glycerine (light) and water (dark) in a droplet. (b) Mixing of fluorescence dyes (light) with water (dark) in a rectangular microfluidic channel. Refs. [36, 38] ...... 16

Figure 1-6 TSAW-driven programmable bioprocessors. (a─d) Three droplets are moved individually and with precise control...... 17

Figure 1-7 (a) Schematic of experimental setup depicting the emergence of a thin film from a standing oil drop due to TSAW exposure. (b) Initial state of the oil droplet. (c) After excitation with TSAW, the bulk of the oil droplet is displaced along the SAW propagation direction while a thin oil film advances in the opposite direction. (d) Finger patterns form in the thin film. (e) Soliton-like wave pulses subsequently appear above the fingers and translate in the SAW propagation direction...... 19

vii

Figure 1-10 Jetting generation resulting from exposure to travelling SAW...... 24

Figure 1-11 (a) Sequential images of rotational acoustic streaming in a drop resulting from asymmetric exposure to TSAWs, with flow streamlines visualized using dye. (b) Concentration of particles in a 0.5 μl droplet via such rotational acoustic streaming...... 26

Figure 1-12 (a) Schematic of device setup for SAW-based carbon nanotube alignment. (b) AFM image of aligned multi-walled carbon nanotubes. Ref. [76]. (c) SEM image of aligned single-walled carbon nanotubes between pre-patterned gold electrodes on LiNbO3...... 27

Figure 1-13 (a) The device structure and working principle for the SAW-driven PDLC light shutter. The magnified part shows a reversible switching process between two different liquid crystal droplet configurations. (b) Transmission response of the PDLCs under different acoustic powers. The insets I and II show the imaging quality at the off and on states of the IDT, respectively...... 29

Figure 1-14 (a) Schematic of SAW device using a superstrate with embedded PC. TSAWs generated by IDT on the LiNbO3 substrate leak into a water-coupling layer, inducing Lamb waves in the silicon superstrate that are then spatially filtered by the phononic lattice to provide asymmetric exposure to the droplet. (b) Left panel shows the band structure of the designed phononic lattice; the shaded area depicts the absolute phononic band gap. Right panel shows simulations at two different frequencies. (c) Concentration of 10 μm polystyrene beads at center of the droplet due to circular acoustic streaming. (d) Concentration of human blood cells at the center of the droplet...... 31

Figure 1-15 Schematic and working mechanism of SSAW-based focusing of particles in a liquid flow stream...... 33

Figure 1-16 (a) Schematic of device for SSAW-based separation of particles in a liquid flow stream, based on particle size. (b) Schematic of device for SSAW-based separation of cell-encapsulating polymer beads in a liquid flow stream, based on bead density...... 35

Figure 1-17 (a) Schematic of protein separation mechanism. A LiTaO3 chip with IDTs (153 MHz/wavelength 26.6 μm) on opposite sides is used to generate a shear SSAW. The supported membrane (~4 nm thick) and protein segregation of membrane-bound streptavidin (red, Texas Red label) and avidin (green, Atto 488 label) can be observed by fluorescence microscopy through the optically transparent chip. (b) Fluorescence images of dye-labelled avidin and streptavidin. (c) Overlay images to demonstrate the segregation of avidin and streptavidin...... 37

viii

Figure 2-1 Device structure and working mechanism of the acoustic tweezers. (A) Schematic illustrating a microfluidic device with orthogonal pairs of chirp IDTs for generating standing SAW. An optical image of the device can be seen in Supporting Fig. S2. (B) A standing SAW field generated by driving chirp IDTs at frequency f1 th and f2. When particles are trapped at the n pressure node, they can be translated a distance of (Δλ/2)n by switching from f1 to f2. This relationship indicates that the particle displacement can be tuned by varying the pressure node where the particle is trapped...... 43

Figure 2-2 Device fabrication process. (AD) Fabrication of the chirp IDTs including a metal deposition process and a following lift-off process. (EH) Fabrication of PDMS-based microchannels using soft-lithography process. (I) The bonding between the substrate and the channel...... 48

Figure 2-3 Optical image of the acoustic device for single particle manipulation. The device consists of a single-layer PDMS microfluidic channel and two pairs of chirp IDTs deposited on a 128° Y-cut LiNbO3 wafer...... 49

Figure 2-4 Quantitative analysis of the acoustic tweezers. (A) Simulated pressure field between adjacent pressure anti-nodes. (B) One-dimensional particle motion induced by a constant frequency change at varying applied acoustic power (experimental results). (C) Velocity plots corresponding to the displacement curves in (B). The inset (smoothed with a moving-average filter of 5 data points) shows that a velocity of 30 µm/s is achieved at the power input of 11 dBm. (D) Experimentally measured acoustic radiation force (ARF) on the particles as a function of distance from the nearest pressure node (discrete points) at different input power levels. The fitted curves are shown in solid lines. (E) Demonstration of reproducible particle motion. Here x-direction particle motion is repetitively shown between two stationary frequencies to show reproducibility. (F) Demonstration of continuous particle translation along the x-direction in well defined steps, while holding stationary in the y-direction...... 52

Figure 2-5 Independent two-dimensional single particle and cell manipulation. (A) Stacked images used to demonstrate independent motion in x and y using a 10-µm fluorescent polystyrene bead to write the word “PNAS”. (B) Stacked images showing dynamic control of a bovine red blood cell to trace the letters “PSU”. The diameter of bovine red blood cell is about 6 µm ...... 54

Figure 2-6 Experimental results for cell viability and proliferation tests. HeLa cells were incubated for 20 h after being treated in SAW field for 6 s, 1 min, and 10 min respectively under the input power of 23 dBm, and then (A) metabolic activity was measured at 450 nm after 2-h BrdU labeling and following 2-h Reagent WST-1 reincubation, to verify the cell viability. Subsequently, (B) DNA synthesis was determined using Cell Proliferation ELISA to verify the cell viability. As control experiments, cells were examined without SAW treatment and at 65°C for 1 h. The culture medium with no cell was also measured as comparison. Each group was tested five times...... 55

Figure 2-7 The temperature of device was examined after applying input power of 23 dBm and 25 dBm for 10 mins. The results indicate a minimal heating effect...... 56

Figure 2-8 Single C. elegans manipulation. One single C. elegans was (A) trapped, (B) moved in y direction, (C) moved in x direction, and (D) moved in y direction again and released, with the average velocity of ~40 µm/s. An optical image of C. elegans (E) before and (F) after being fully stretched...... 57

Figure 2-9 This stacked image demonstrates parallel manipulation of groups of bovine red blood cells...... 58

Figure 2-10 This stacked image demonstrates parallel manipulation of hundreds of 7-μm polystyrene beads to create an array of “S”s...... 59

Figure 2-11 (a-f) show the process of massive manipulation of particles. All the particles were collected to one side of the channel. (g) and (h) show the distribution of large (10 µm) and small (1 µm) particles before and after we sweep the frequency...... 60

Figure 3-1 Diagram of FACS system. Cells are fluorescently tagged, analyzed by optical detection, encapsulated into droplets, and sorted by electric field...... 63

Figure 3-2 (a) Layout of the microfluidic sorter enabled by optical detection and optical switch. (b) Schematic of instrument of the optically switched microfluidic FACS system...... 64

Figure 3-3 Schematic illustrations of the SAW-based mechanism for droplet or cell sorting. With the absence of SAW propagation through the main channel, (a) the droplets or (c) cells in the main channel move into the upper outlet channel due to its lower hydrodynamic resistance. With SAW exposure, acoustic streaming moves the fluid downward, carrying with it the contained (b) droplets or (d) cells and moving them to the lower outlet channel...... 65

Figure 3-5 An optical picture of the SSAW-based cell-sorting device with 3 inlets and 5 outlets...... 68

Figure 3-6 A 15 μm fluorescence particle passing through the sorting region can be directed to three different outlet channels by modulating input frequency (14.5 MHz, off, and 13.9 MHz)...... 70

Figure 3-7 Sorting of human white blood cells (HL-60 Human promyelocytic leukemia cells) into 5 specific outlet channels at frequencies of (a) 9.8, (b) 10.0, (c) 10.2, (d) 10.6, and (e) 10.9 MHz...... 72

Figure 4-1 3D cells patterning process. (a) Prepolymer solution with cells is first introduced into the chamber (1) and cells are then patterned via dielectrophoretic forces in the 1.8-mm-thick dielectric layer (2), The hydrogel is polymerized by UV light exposure (3), embedding cells in a stable microorganization. The hydrogel then can be removed for consecutive culture (4), or incorporated into multilayer structure by repeating steps 1–3; each layer may include distinct cell organization, cell type and hydrogel formulation (5, 6). (b) Electric field is simulated by finite element modeling (CFD Research Corp.). (c–f) A microarray of fibroblast clusters are embedded in the DCP hydrogels. (g) Distinct fluorescently labeled fibroblasts in a bilayered are patterned above and in concentric rings, (6). Scale bars: c, 250 mm; d, g, 100 mm; f,50 mm...... 75

Figure 4-2 SSAW-based patterning of a group of particles in stagnant fluid. Schematic of (a) 1D patterning using two parallel IDTs, (b) 2D patterning using two orthogonal IDTs (the angle between the IDTs can be changed to achieve different patterns). (c) Distribution of fluorescent microbeads before and after the 1D (SAW wavelength was 100 um) patterning process, (d) and the 2D patterning process (SAW wavelength was 200 um)...... 76

Figure 4-3 (a) A schematic of the slanted finger IDT (SFIT); (b) The actuation of a droplet on the surface of the LiNbO3 sunstrate as well as on a cover slip coupled onto it at different input frequency and the theoretical calculation of the centre of the excited SAW. The inset is the S-parameter showing a stable response in the frequency range between 8 mHz and 14 MHz...... 77

Figure 4-4 (a) The left panel shows the configuration of a SFIT of which the period is changing continuously. The right panel describes the amplitude distribution of a SAW at different frequencies. (b) A schematic drawing of the acoustic-based, one- dimensional dynamic patterning device. (c) Relationship between frequency and spacing of one-dimensional patterning on different piezoelectric substrates...... 79

Figure 4-6 (a) The distribution of primary acoustic radiation force with dependence of frequency. (b) The motion tracking of particles in a SSAW field. (c) Distance dependence of particles’ velocity...... 86

Figure 4-7 Dynamic patterning of bovine red blood cells. I, II, and III show the 1D pattern at the frequencies of 31 MHz, 25 MHz, and 13 MHz, respectively...... 88

Figure 4-8 2D dynamic patterning of fluorescent polystyrene microbeads. (a) Schematic of the 2D dynamic patterning device: the particle pattern in the region corresponding to a smaller finger pitch (higher frequency) has a smaller period, while the pattern in the region corresponding to a larger finger pitch (lower frequency) has a larger period. (b) The simulated distribution of pressure nodes (PN) in SSAW field with the SAW excited from the low frequency region of the SFITs. (c) An optical image

of the 2D patterning device. Bottom figures I, II and III show the patterned arrays with periods of 150 μm, 195 μm, and 240 μm at the frequencies of 18.0 MHz, 13.3 MHz, and 10.9 MHz, respectively...... 89

Figure 5-1 Schematic and working mechanism of SSAW based separation. (a) Cross section showing that particles move towards the pressure node in SSAW field. (b) the SSAW applies greater force on larger particles and pushes them along the linear pressure node. (c) A photo of real separation device...... 95

Figure 5-2 Stacked image showing the trajectores of 15 µm polystyrene beads at inuput power of (a) 27dBm, (b) 23dBm, and (c) 15 dBm. The flow velocity is ~2 mm/s...... 97

Figure 5-3 Stacked images showing particles (2 and 10 µm) trajectories at the working region (a and c) and outlet region (b and d) when SAW is off (a and b) and on (c and d)...... 98

Figure 5-4 High resolution separation of fluorescent polystyrene beads with diameters of 9.9 µm (red) and 7.3 µm (green). The fluorescent profile represents the beads distribution after passing the working region...... 100

Figure 5-5 stacked images show the separation process of polystyrene beads (15 µm) and HL-60 cells. They have similar densities and sizes but different compressibilities...... 101

Figure 5-6 stacked images showing the separation process of HL-60 from human red blood cells...... 102

Figure 5-7 Flow cytomerey scattergrams of side scatter versus forward scatter, showing the cell population of sample before (a) and after (b and c) separation...... 102

Figure 5-8 (a) stacked image showed that a single CTC cell was pulled out from the rest WBC stream. EpCAM, CD45, and DAPI are used to identify MCF-7 from leukocytes. Green dots (positive to EpCAM) are recognized as MCF-7 cells while red dots (positive to CD45) are regarded as leukocytes. (b) erythrocyte-lysed human blood sample spiked with MCF-7 cells, the original cancer cells concentration is ~5- 10%. (c) Cells were collected from outlet after CTC isolation device. The results show CTC purification of as high as 98% is achieved...... 104

Figure 5-9 (a-b) experimental results for MCF-7 Cell viability and proliferation test: 1. positive control, 2. cells passing through device with SAW off, 3. cells passing through device with SAW on, 4 negative control. Cells apoptosis assay is carried out at conditions of 1, 2, and 3. (c-e) show the cell apoptosis test result under each condition...... 105

xii

LIST OF TABLES

Table 2-1 SAW properties of some piezoelectric material with relatively high transparency ...... 45

Table 2-2 Basic properties of LiNbO3 ...... 46

Table 2-3 Recipe used for Photoresist in DRIE ...... 47

xiii

ACKNOWLEDGEMENTS

First of all, I would like to express my deepest gratitude to my Ph.D. advisor, Prof. Tony

Jun Huang. I want to thank him for his constant supervision, support, tolerance and trust. His guidance makes my Ph.D. study such a pleasant journey. His passion for research and supporting student success and growth has truly inspired me to strive for excellence.

I would like to thank my thesis committee, Prof. Bernhard R. Tittmann, Prof. Corina

Drapaca, and Prof. Qiming Zhang for their valuable suggestions on my dissertation work in so many ways. I would also like to give special thanks to chair of Department of Engineering

Science and Mechanics, Prof. Judith A. Todd and his administrative assistants for their supports in the past several years. I gratefully acknowledge Prof. Stephen Benkovic, Dr. Poki Yuen, and

Dr. Ming Dao for their insightful suggestions and strong support to my research.

My Thesis work was done in collaborations with so many great people within and outside of BioNEMS research group. My sincere thanks to Jinjie Shi, Sz-Chin Steven Lin, Xiaole Mao,

Yuebing Zheng, Yanjun Liu, Bala Krishna Julury, Michael Ian Lapsley, Brian Kiraly, Hua

Huang, Qingzhen Hao, Bingxin Zhang, I-Kao Chiang, Zach Stratton, Shahrzad Yazdi, Hongjun

Yue, Danqi Chen, Zhangli Peng, Monica Diez, Sarah E Du, Daniel Ahmed, Mengqian Lu,

Ahmad Ahsan Nawaz, Yanhui Zhao, Sixing Li, Peng Li, Shikuan Yang, Chenglong Zhao,

Yuchao Chen, Feng Guo, Chung Yu Chan, Yuliang Xie, Zhangming Mao, Po-Hsun Huang,

Nitesh Nama, Xiang Guo, and Xiangjun Zhang for their help in different ways.

Lastly, I wish to thank my beloved family for their love, caring, and support that has made this dissertation possible. My sincerest thanks also go to my friends who have always been there for me.

Chapter 1

Introduction and Background

The past decade has witnessed rapid growth in the field of lab-on-a-chip. The continuous infusion of new physics into the on-chip world has enabled some fascinating innovations. For example, the introduction of electrokinetics gave rise to a plethora of effective cell manipulation methods, and the coupling of optics spurred the development of optofluidics. Recently, a new frontier of lab-on-a-chip has been opened, with surface acoustic wave (SAW) technologies being applied on-chip. The advantages of such SAW microfluidics are numerous: simple fabrication, high biocompatibility, fast fluid actuation, versatility, compact and inexpensive devices and accessories, contact-free particle manipulation, and compatibility with other microfluidic components. We believe that these advantages position SAW microfluidics to impact many applications in biology, chemistry, engineering, and medicine. In this chapter, we review the theory underpinning SAWs and their interactions with contacting fluids and contained particles.

We then review the SAW-enabled microfluidic devices demonstrated to date, including fluid control and particle manipulation.

1.1 Motivation

During the past two decades, microfluidics has emerged as an important platform in chemistry, biology, and medicine [1‒5]. The introduction of “lab-on-a-chip” (i.e., small-sized analytical devices) spurred steadily increasing interest and research efforts in the microfluidics discipline (upon which lab-on-a-chip devices are based). The promise of such lab-on-a-chip devices is grounded in the advantages of microfluidics relative to traditional laboratory techniques used in chemistry and biomedicine: system miniaturization, automation, low cost, reduced reagent and sample consumption, rapid turnaround time, and precise microenvironment control. With an aim to improve the functionality, versatility, and performance of lab-on-a-chip devices, researchers have continued to integrate new physics into the microfluidic platform. For example, incorporating electrowetting in microfluidics enabled effective on-chip manipulation of droplets, i.e., digital microfluidics [6]; incorporating magnetics in microfluidics led to the demonstration of highly efficient cell manipulation and separation [7]; and incorporating optics in microfluidics spurred the development of fluid-based optical components with unprecedented tunability, such as optofluidic lasers, lenses, and modulators [8‒10].

In recent years, surface acoustic wave (SAW) technologies have begun to receive significant attention in the microfluidics community. SAWs are acoustic waves that propagate along the surface of an elastic material. To date, SAW technologies have been used extensively in the telecommunication industry (e.g., cell phones) for signal processing and filtering [11]. Other established applications of SAW technologies include touch-sensitive screens and biological/chemical sensing [12‒14]. Recent research demonstrates that SAWs provide an effective means to control fluids and particles in lab-on-a-chip devices [15]. These SAW-based

3 microfluidic devices offer the following useful and inimitable combination of features:

Simple, compact, inexpensive devices and accessories: SAW devices have been used extensively in various compact commercial electronic systems, such as cell phones (each cell phone manufactured today contains multiple SAW devices which, along with their accessories, occupy only a small portion of the phone volume). This widespread commercial deployment demonstrates that SAW devices—and the accessories needed to drive them (such as driving circuits and power supplies)—are compact and inexpensive, and can achieve high reliability [11].

So SAW-based microfluidic devices would be simple and inexpensive to fabricate and integrate with other on-chip components in mass production.

High biocompatibility: The acoustic power intensity and frequency used in many SAW-based microfluidic devices are both in a range similar to those used in ultrasonic imaging, which has been used extensively for health monitoring during various stages of pregnancy and proven to be extremely safe [16]. Therefore, we expect that with proper design, SAW-based microfluidic devices will also be safe and biocompatible with cells, molecules, and other biological samples.

This expectation has been partially confirmed by cell viability and proliferation tests using existing acoustofluidic devices [17, 18].

Fast fluidic actuation and large forces: Current microfluidic technologies have difficulty generating fast fluidic actuation or large forces acting on particles. These drawbacks have limited their applications in medical diagnostics and biochemical studies. As indicated in a review article on microfluidics [4], “small feature sizes typically prevent flow velocities from being high enough to yield high (Reynolds) numbers. High-frequency acoustic waves, however, can circumvent such difficulties.” SAW-based devices can be used to selectively introduce chaotic advection [31] to a microfluidic system (in which laminar flow generally dominates), thereby enabling fast, effective manipulation of fluids and particles. Thus far, SAW-based microfluidic devices have proven to be able to pump fluids at 1−10 cm/s [19] and manipulate millimeter-scale

4 objects (e.g., C. elegans) [18]—neither of these two features can be readily achieved by other microfluidic techniques.

Versatility: SAW technologies enable biological/chemical detection, fluidic control (e.g., fluid mixing, translation, jetting, and atomization), and particle manipulation (e.g., focusing, patterning, separation, sorting, concentration, and re-orientation). SAW techniques are capable of manipulating most microparticles, regardless of their shape, electrical, magnetic, or optical properties; they are capable of manipulating objects with a variety of length scales, from nm to mm; and they are capable of manipulating a single particle or groups of particles (e.g., tens of thousands of particles).

Contact-free manipulation: SAWs manipulate particles and cells by means of the primary acoustic radiation force applied by the surrounding fluid; and SAWs manipulate fluid by means of acoustic waves leaked into the fluid. This represents contact-free manipulation, eliminating the potential for sample contamination.

Compared with bulk acoustic wave (BAW) microfluidics [20], another subset of acoustofluidics (i.e., the fusion of acoustics and microfluidics), SAW microfluidics has the following advantages:

1) It allows one to better control excitation frequencies in a wider range and utilize higher

excitation frequencies whenever needed. As a result, SAW microfluidics is more versatile

and can achieve more precise and controllable manipulation of fluids and particles.

2) It does not require that fluid channels be made of high acoustic reflection materials, making it

easier to employ in polymer-based devices, which have been widely used in microfluidic

platforms.

3) Because SAWs confine most of their energy to the surface of a microfluidic device substrate

(whereas BAWs expend energy traveling through the bulk of the device substrate), they

require less power than BAWs to achieve the same acoustic effects in the microfluidic device.

5 These recognized features and advantages of SAW microfluidics (relative to BAW microfluidics and other microfluidic approaches) have rendered it an attractive platform for many lab-on-a-chip applications [21, 22]. In this thesis, we introduce a SAW microfluidic platform called acoustic tweezers that can manipulate single particles, cells, and whole organism, then followed by some novel applications based on this platform.

1.2 Fundamentals of SAW

1.2.1 Generation of SAW

The best-known form of SAW, Raleigh SAW, is composed of a longitudinal and a vertically polarized shear component [23]. Rayleigh SAW strongly couples with media in contact with the wave propagation surface, enabling the sensing of mass perturbation and elastic properties of a medium introduced on the wave’s propagation path. Other SAW modes exist in elastic materials of different compositions. For example, Love SAW is a guided shear-horizontal wave that propagates in a thin layer on top of a substrate. However, as most SAW-based lab-on-a- chip technologies utilize Rayleigh SAWs, we use “SAW” to specifically indicate Rayleigh SAW in this review.

SAWs are generally created by applying an appropriate electric field to a piezoelectric material. The piezoelectric material, in turn, generates propagating mechanical stress. A typical

SAW device uses at least one set of metallic interdigital transducers (IDTs) fabricated on the surface of a piezoelectric substrate to apply this electric field, generating a SAW displacement amplitude on the order of 10 Å. An IDT consists of a set of connected metallic fingers interspaced with an opposite set of connected metallic fingers; an alternating current electrical signal in the radio frequency (RF) range is applied across the two sets of connected fingers. The structure of

6 the IDT determines the bandwidth and directivity of the generated SAW: by changing the number, spacing, and aperture (overlapping length) of the metallic fingers, one can change the characteristics of the resulting SAW. For example, a focused IDT consists of pairs of annular electrodes and can focus SAW energy to a spatially small focal point [24]; a chirped IDT has a gradient of the electrode finger width directed along the SAW propagation direction, allowing it to generate SAWs over a wide frequency range [18]; a slanted finger IDT has a gradient of electrode finger width directed perpendicular to SAW propagation direction, allowing it to generate narrow SAW beams of varying frequency along its finger length [25]. Each IDT variant has its own set of advantages and disadvantages relative to the other variants, and the choice of

IDT type for use in a SAW-based microfluidic device depends on the device requirements.

Additional IDT design types will be mentioned in this review article. A simple interdigital transducer is presented in Fig. 1-1 to show the generation and propagation of SAW.

Figure 1-1 A uniform, metallic interdigital transducer (IDT) deposited on the piezoelectric substrate generates SAWs that propagate along the substrate surface in both directions.

Generation of SAW on a piezoelectric material can be mathematically modeled, but is complicated by (1) the anisotropy exhibited in most piezoelectric materials and (2) the intrinsic electromechanical coupling in piezoelectric media. The anisotropy of a piezoelectric material

7 dictates whether the piezoelectric material generates shear-horizontal SAW [26], leaky SAW

[27], or psuedo-SAW [28]. As such, consideration must be given to anisotropy to ensure the generation of the desired type of wave. The full mathematical modeling of SAW propagation requires the analysis of the link between electrical signal and deformation in piezoelectric media, referred to as electromechanical coupling. The constitutive equations governing the linear theory of wave propagation in piezoelectric material are as follows [29]:

(1-1)

(1-2)

where are the components of strain tensor, is the electric field, are the

components of the piezoelectric stress tensor, are the dielectric constants at constant strain,

and are the elastic moduli.

1.2.2 SAW-induced streaming

When a travelling SAW contacts a liquid, which has a smaller speed of sound than that of the substrate, the liquid’s viscosity causes part of the SAW to refract into the liquid as a longitudinal wave. Because of this acoustic refraction, the mode of SAW changes to a form called

“leaky SAW” or “pseudo-SAW”. As shown in Fig. 1-2a, a SAW is excited by the IDT and propagates from left to right along the piezoelectric substrate surface within a wavelength depth

[30]. The refracted wave moves along the direction given by a refraction angle known as

Rayleigh angle:

, (1-3)

where cs and cl denote the acoustic wave velocities of the piezoelectric substrate and the fluid, respectively. For a SAW propagating on a 128° Y-cut lithium niobate (LiNbO3) substrate at

8 room temperature, the SAW velocity of about 3990 m/s [31] and the speed of sound in water of

1490 m/s result in a Rayleigh angle of about 22°. The refracted longitudinal waves generate a force in their propagation direction and induce flow within the confined liquid. The boundaries of the confined liquid reflect the actuated liquid and lead to internal streaming. Such a non-linear phenomenon that transforms the SAW attenuation into a steady fluid flow is called SAW-induced acoustic streaming [32‒40].

9 SAW-induced acoustic streaming can be visualized by introducing small particles or dyes into the effected liquid. The experimental and numerical results in Fig. 1-2b show a comprehensive view of SAW streaming patterns within a 30 µl hemispherical microdroplet positioned at the centre of the SAW propagation direction [34]. The SAW-induced streaming pattern varies dramatically with the shape of the confined liquid, as well as the incident position, angle, and operating frequency of the SAW [37]. Mathematical models, based on the compressible Navier-Stokes equations and numerical simulations of SAW-induced streaming, have been investigated by Vanneste et al [33].

The principles of mass and momentum conservation that govern the motion of continuous media, for a linear viscous compressible fluid, yield the following set of governing equations:

(1-4)

, (1-5)

where is the mass density, is the flow velocity, is the fluid pressure, and are the shear and bulk dynamic viscosities, respectively [40, 41]. These equations allow one to predict the motion of the fluid when complemented by adequate sets of boundary conditions and a constitutive relation linking the behavior of the pressure to that of the mass density . In many applications, and are assumed to be linearly related:

(1-6)

where is the speed of sound in the fluid at rest. Equations ((1-4)–(1-6)) form a nonlinear system of equations whose solution can be challenging. Since the fluid is a dissipative system, the response to harmonic forcing is not, in general, harmonic. It is convenient to think of the motion induced by harmonic forcing as having two components: (i) a harmonic, or more generally periodic, component with period equal to the forcing period; and (ii) remainder which can, in turn, be viewed as having a steady component and, possibly, a harmonic component with

10 frequency double that of the forcing. The first component of the motion is the fluid’s acoustic response. The second is the streaming motion. The acoustic component of the velocity is often referred to as the particle velocity, although this sort of vocabulary does not agree with the rigorous notion of particle velocity in continuum mechanics.

Further analysis and details are presented in Köster’s work [39]. From a computational viewpoint, the complexity inherent in the original nonlinear problem is overcome by the solution of two linear problems in succession. A discussion of how the equations for the determination of the acoustic motion are derived is offered by Friend and Yeo [19], who also include a discussion on the use of a nonlinear version of the constitutive relation in Eq. (1-6).

1.2.3 Formation of SSAWs

The basic principles that govern the interference of all waves give rise to the formation of

SSAWs (standing SAWs). If a pair of identical IDTs fabricated on a piezoelectric material generates two travelling SAWs propagating toward each other, the SAWs’ interference will result in a one-dimensional SSAW field as shown in Fig. 1-3a. Computational analysis of a one- dimensional SSAW on the surface of a substrate has led to the simulated interference pattern shown in Fig. 1-3b. The light and dark regions represent weak and strong amplitudes of the sound pressure field, respectively. The lightest and darkest lines show that SSAWs have a series of nodes (minimum field) and anti-nodes (maximum field) at fixed locations on the substrate surface. The distance between neighboring nodes—and also the distance between neighboring anti-nodes—is always half of the SAW wavelength on the substrate. Compared with a BAW approach, which typically uses acoustically reflective materials in channel walls to form standing waves within the channel (or cavity), SAW creates a standing wave through direct interference of

11 two or more identical SAWs (i.e. channel walls need not to be constructed of acoustically reflective materials).

Similar to the one-dimensional case, a two-dimensional SSAW field can be formed on the surface of a substrate by generating SAWs from two pairs of IDTs arranged orthogonally, as shown in Fig. 1-3c. Note that the two-dimensional interference pattern is tilted 45 degrees with respect to the propagation direction of each SAW, and the shortest distance between nodes—and also the shortest distance between anti-nodes—in this two-dimensional patterned array is times the SAW wavelength on the substrate.

12 1.2.4 Primary acoustic radiation force in SSAWs

Both experiments and theory have established that when particles are immersed in a fluid medium, and a standing acoustic field is established in that fluid medium, the particles experience primary acoustic radiation forces that move them toward pressure nodes or pressure anti-nodes, depending on the material properties of both the particles and the surrounding fluid. The primary acoustic radiation force Fax is expressed as

(1-7)

where p0 and Vp are the acoustic pressure and particle volume, respectively; and k are the wavelength and wavevector of the acoustic waves, respectively; β is compressibility; and ρ is density [42]. φ, the acoustic contrast factor, indicates whether particles aggregate at pressure nodes (positive φ) or pressure anti-nodes (negative φ) and is given by:

, (1-8)

where the subscripts and stand for particle and fluid, respectively. Most particles, such as polystyrene beads and blood cells, have a positive acoustic contrast factor φ, causing them to move toward the pressure nodes in a standing acoustic field.

SSAW-based microfluidic devices establish a standing acoustic field in fluid by means of

IDT pairs forming a SSAW on a piezoelectric substrate in contact with the fluid (such as the setup in Fig. 1-3a). As discussed in previous section, we have good understanding of the SSAW field present on the piezoelectric substrate. However, if we are to apply Eq. (1-7) in order to explain the working mechanism of SSAW-based microfluidic devices, we must understand the standing acoustic field present in the fluid (as induced by acoustic energy leakage from the contacting substrate).

13 A recent study by the authors [43] details a computational method used to model the

SSAW-induced pressure field in the microfluidic channel depicted in Fig. 1-3a. The pressure gradient above the surface of the piezoelectric substrate leads to the generation of a longitudinal sound wave in the liquid. The displacement fields of the leakage wave created by two opposing travelling SAWs are plotted in Figs. 1-4a and 1-4b. Figs. 1-4c and 1-4d show the pressure fields corresponding to the combined displacement fields at two time snapshots with half a period time difference. During the time between the two snapshots, the amplitude of the pressure in the vertical center of the channel remains zero (indicated by the dashed line in Figs. 1-4c and 1-4d), while pressure in the other channel regions simply changes sign. These results indicate the existence of a standing acoustic pressure field in the fluid of the channel. Furthermore, the pressure nodal plane of this standing acoustic field sits immediately above the pressure node of the substrate SSAW that induced it. Several groups have experimentally confirmed this correspondence between pressure nodes of the substrate SSAW and fluid acoustic field [44‒48].

The authors’ computational model of the SSAW also demonstrated that the resulting primary acoustic radiation forces consist of axial mode (acting along theh channel width) and transverse mode (acting along the channel height). The axial primary force is stronger than the transverse primary force. The partially reflected waves from the PDMS channel walls are included to calculate the transverse primary force.

Eq. (1-7) indicates that particles suspended in fluid in which a standing acoustic field exists will be subjected to primary acoustic radiation forces that drive them to either the field’s nodes or anti-nodes. And since both theoretical and experimental work has demonstrated that the nodes and anti-nodes of the standing acoustic field in the fluid lie immediately above those of the substrate SSAW, microfluidic devices can be designed according to knowledge of the substrate

SSAW. This is important when calculating device dimensions, because the behavior of substrate

SSAWs is simple and computationally straightforward, and no simulations of the much more

14 complex fluid behavior in acoustic fields are required. By using a glass channel (acoustically reflective) in a SAW device rather than a PDMS channel (acoustically absorbent), Johansson et al. found that the glass-channel-SAW system displayed features similar to BAW technology: the size and shape of the channel significantly affected the acoustic field, and the node-to-node distance was half the sound wavelength in liquid [49].

Figure 1-4 Time snapshots of displacement field of the longitudinal-mode leakage wave from a SAW propagating in (a) the positive x-direction and (b) the negative x-direction. The width of the channel is half the SAW wavelength and the height is one SAW wavelength. The pressure field resulting from the combined displacement fields (i.e. resulting from a SSAW on the substrate) is shown in the cross-section of the channel at (c) t = t0 and (d) t = t0 + . The dotted line denotes the pressure nodal plane in the middle of the channel. [43]

15 1.3 Literature Review of SAW microfluidics

1.3.1 Fluid mixing

Many lab-on-a-chip applications require the mixing of two or more fluids. However, the low

Reynolds number laminar flows that predominate at the micro-scale result in mixing that occurs via diffusion. Such diffusion-based mixing is too slow for most lab-on-a-chip applications, so researchers have looked to SAW-induced streaming for its ability to mix fluids quickly by generating chaotic advection.

A few groups have demonstrated methods using TSAWs to mix fluid in an unconstrained droplet. Shilton et al. [38] generated a TSAW using a single-phase unidirectional transducer

(SPUDT) to induce liquid recirculation inside a droplet. Their results, shown in Fig. 1-5a, demonstrate the fast mixing of dyed water and dyed glycerine solution. Frommelt et al. [50] demonstrated a more refined TSAW-based droplet mixing, using a pair of tapered IDTs (TIDTs) to generate a narrow SAW beam with a tunable launching point. By individually modulating the input signals of the two TIDTs, they demonstrated the ability to temporally modulate the flow patterns generated by each IDT to achieve efficient mixing. They also demonstrated that they could control mixing speed (within the droplet) by adjusting SAW amplitude and frequency.

Tseng et al. [36] also used IDTs to demonstrate TSAW-based mixing. Instead of mixing fluids in an unconstrained droplet, they mixed fluids inside a microchannel (Fig. 1-5b). They performed a comprehensive experimental study on the effects of various operational parameters, showing that mixing performance could be significantly improved by applying higher voltage signals to the IDTs.

Recently, Rezk et al. [51] incorporated SAW-induced mixing in a paper microfluidic device. They utilized a hue-based colorimetric technique to compare the mixing efficiency of

16 their device with that of capillary-based mixing: the SAW-based mixing showed greatly enhanced consistency and speed. By adjusting IDT design and input signal parameters, researchers have proven that TSAWs can effectively and controllably mix fluids in both open and confined fluid geometries. This versatility makes TSAW-based mixing techniques extremely attractive in microfluidics.

1.3.2 Fluid translation

1.3.2.1 Fluid translation in open space

When a liquid droplet is placed within the propagation path of TSAWs on a piezoelectric substrate, leaky SAW will be diffracted into the droplet at the Rayleigh angle. If the SAW has

17 low amplitude, it will induce acoustic streaming within the droplet. If the SAW has sufficiently intermediate amplitude, the leaky acoustic energy generates an acoustic force on the droplet along the SAW propagation path, causing the droplet to deform in an axisymmetrical conical shape and translate across the substrate [32]. Liquid droplet speeds of 1─10 cm/s can be achieved using this

TSAW-based actuation; this is more than an order of magnitude faster than other current micro- pumping actuation schemes [19].

Moreover, as electrical actuation of IDTs can be programmed, droplet movements can be automated. Through the automated control of multiple droplets, merging, mixing, splitting, and chemical or biochemical reactions can be performed in so-called “programmable bioprocessors”.

A simple example of a programmable bioprocessor is shown in Fig. 1-6, in which three droplets with different fluid content are moved independently in any desired direction and can be made to join together [52].

Figure 1-6 TSAW-driven programmable bioprocessors. (a─d) Three droplets are moved individually and with precise control. [52]

18 Many chemical and biological applications have been demonstrated with TSAW-driven droplets. Guttenberg et al. [53] precisely actuated oil-covered aqueous droplets between sinkers and heaters to perform a highly sensitive, fast, and specific DNA amplification reaction with droplet volumes as low as 200 nL.

Rezk et al. [54] subjected a droplet of silicone oil to TSAW excitation, observed behavior substantially different from that of water droplets, and developed theory for the behavior (Fig. 1-

7a). Silicone oil has a much smaller contact angle with a LiNbO3 piezoelectric substrate than does water. When a silicone oil droplet on such a substrate was exposed to TSAW, this small contact angle led to a majority of the droplet being propelled in the SAW propagation direction (Fig. 1-

7b), while a thin film spread in the opposite direction (Fig. 1-7c). As the thin film advanced, it formed finger-shaped patterns (Fig. 1-7d); eventually, soliton-like wave pulses appeared above the fingers and propagated along the TSAW direction (Fig. 1-7e). The authors suggest that this

TSAW-induced thin film spreading may have applications in film coating and microfluidic actuation.

Historically, most SAW microdevices have been fabricated on bulk LiNbO3 or quartz substrates; these bulk piezoelectric substrates may require some extra processing in order to integrate control electronics (e.g., for the IDTs). To facilitate further advancements in SAW microfluidics, Du et al. [35] presented a thin film piezoelectric material for translation of liquid droplets with volumes up to 10 µl. They first deposited a ZnO thin film on a plain silicon substrate. This hydrophilic ZnO thin film layer prevents effective droplet translation. To circumvent this, they treated the ZnO thin film with a self-assembled monolayer of octadecyltrichlorosilane (OTS), which made the substrate surface hydrophobic while producing no measurable acoustic damping. The new substrate is cheaper than bulk piezoelectric substrates, and the silicon base can easily be integrated with control electronics for the IDTs, enabling the potential for a fully automated microsystem.

1.3.2.2 Microfluidic pumping in enclosed channels

In addition to translating liquid droplets in open space, TSAWs can be employed to pump fluid through enclosed channels.83─89 Cecchini et al. [56] bonded a straight PDMS microchannel between two IDTs on a LiNbO3 substrate (Fig. 1-8a). After placing a water droplet between one

IDT and the channel inlet (this droplet was termed the water “reservoir”), two different operation modes were tested: direct drive mode and inverted drive mode. In direct drive mode, SAWs were excited by the IDT near the channel inlet and propagated from inlet to outlet (Fig. 1-8b), whereas inverted drive mode had SAWs excited by the IDT near the channel outlet and

Figure 1-8 (a) Schematic of the assembly of the microfluidic device and (b) activation of IDT1, called direct drive (DD) mode; activation of ID2, called inverted drive (ID) mode, results in TSAW propagation opposite that shown in this image. (c) Photographs of the (ineffective) water filling process under DD mode at different times. (d) Photographs of the water filling process under the ID mode at different times. [56] propagated from outlet to inlet. In direct drive mode, SAW excitation caused significant atomization in the water reservoir, resulting in rapid evaporation of the water and preventing the channel from being filled (Fig. 1-8c). In inverted drive mode, the water reservoir quickly translated into the microchannel, filling it with flow speed as high as 1.24 mm/s. The authors propose that this flow rate, occurring opposite the TSAW propagation direction, is attributable to

TSAW-induced atomization at the leading edge of the water. These droplets continuously form, coalesce, and rejoin the water-air meniscus, thereby dragging the water through the channel and

21 resulting in net motion opposite the direction of TSAW propagation (the water reservoir is not exposed to significant atomization under the inverted drive configuration because the TSAW power is sufficiently diminished by the time it reaches the reservoir location at the channel inlet).

Based on this TSAW inverted drive mechanism, Girardo et al. [57] developed a fully controlled low-voltage micro pump in a two-dimensional microchannel array. By combining a

5×5 orthogonal array of PDMS microchannels on a LiNbO3 substrate with 20 IDTs, the researchers could selectively generate either a single TSAW or multiple TSAWs. Thus, droplets at channel inlets could be directed through the microchannel grid to the desired outlets. This device achieved several SAW-driven fluid operations including extraction, deviation, splitting, and simultaneous multichannel filling.

TSAW-driven pumping has also been demonstrated in a closed microfluidic channel

(with no liquid-air interface) by Fallah et al. [58]. This device featured a microfluidic channel in a square-shaped loop, with an IDT to generate SAWs travelling along one of the four sides. When a

TSAW reaches the fluid-substrate interface, the acoustic energy leakage profile decays exponentially with continued wave travel along the fluid channel. With a sufficiently long channel, the location of initial TSAW contact with the fluid essentially acts like a point pressure source, driving the fluid through the channel loop according to conservation of mass. With this setup, fluid flow could be continuously actuated with very fast response. Fluid pumping in such a closed-loop chamber can be used to mimic the action of small blood vessels, making it useful for clinical applications and biomedical study [55].

1.3.2.3 Microfluidic rotational motor

Shilton et al. utilized SAW-induced acoustic streaming to drive a rotary micromotor in a microfluidic chip (Fig. 1-9a) [60, 61]. A pair of IDTs were patterned on a LiNbO3 substrate, and a

Figure 1-9 (a) Schematic of the SAW-induced rotary micromotor. Silicone gel was used to break the axisymmetry of the planar SAW and generate a centrosymmetric SAW, resulting in centrosymmetric acoustic streaming inside the drop. (b) The static and spinning states of the Mylar disk placed on top of the drop. [60, 61]. layer of hydrophobic Teflon film was coated in the space between the two IDTs in order to hold a fluid droplet in place. Silicone gel was placed on the SAW’s path to absorb half of the energy from each IDT, creating a centrosymmetric TSAW exposure in the droplet, resulting in centrosymmetric acoustic streaming inside the droplet. When a Mylar disc with 5 mm diameter and 100 µm thick was placed on top of the droplet, the droplet’s rotational inertia actuated rotational motion in the disc (shown in Fig. 1-9b). The disk’s angular velocity under a range of

SAW amplitudes was investigated. In the case in which a 50 µl water drop was used as the fluid- coupling layer, a maximum disk rotation speed of 2250 rpm was achieved with a maximum torque of around 69 nNm. The authors found that increasing SAW amplitude to surpass ~3 nm would reduce angular speed due to the unstable disc rotation caused by asymmetric flow in the drop.

23 As described in the above examples, SAW-induced acoustic streaming enables a simple and miniaturized on-chip rotational motor without the need for moving mechanical parts. In future studies, researchers will likely examine fluid-coupling layers other than water in order to circumvent the potential for evaporation.

1.3.3 Jetting and atomization

A fluid jet is commonly defined as a coherent stream of fluid projected into a surrounding medium. For a fluid to undergo jetting phenomena, it must possess sufficient inertia to overcome the restoring capillary forces acting on the interface of the fluid and surrounding media [26].

Jetting at micro-scales finds numerous applications in hand-held ink jet printers, ink jet highlighters, and ink jet brushes [62]. Such micro-scale jetting can be accomplished by exposing small fluid volumes to TSAW (with the SAW power higher than that used for droplet translation).

Recent research has demonstrated that SAW-based fluid jet production offers benefits over existing jetting techniques as it concentrates mechanical energy into a small drop, generating jets without the narrow confinement typically necessary to accelerate fluid.

Recently, Tan et al. [63] studied the nature of jet formation using TSAW devices, as shown in Fig. 8. They characterized jetting length as a function of the driving force and the jet

Weber number. They also predicted the jet’s velocity as a function of acoustic Reynolds number using the jet momentum equation of Eggers [64], and verified their results with the experimental data. These rigorous characterization efforts have helped to facilitate future TSAW-based jetting technologies.

Atomization, which is the making of an aerosol of small solid particles or liquid droplets, has long been applied in numerous areas, such as internal combustion engines, medicine, agriculture, and cosmetics [65]. Kurosawa et al. [66] first proposed and constructed a novel

24 ultrasonic atomizer using a TSAW device. Since then, SAW-based atomization has been used for numerous applications such as protein extraction and characterization for paper-based diagnostics

[67], portable pulmonary delivery of asthmatic steroids [68], and mass spectrometry interfacing with microfluidics [69].

Figure 1-10 Jetting generation resulting from exposure to travelling SAW. [63].

TSAW atomization has also been utilized for producing micro and nano particles.

Alvarez et al. [70] demonstrated the generation of protein aerosols and nanoparticles using SAW atomization. The small footprint of its generating mechanism, low power consumption, biocompatibility, and control over aerosol size make SAW an attractive method for atomization and a potential enabler for next-generation drug delivery devices.

TSAW-induced atomization is caused by the generation of capillary waves on the air- liquid interface [71]. The mathematical analysis of this phenomena is very challenging, owing to the presence of a free interface and non-linear, two-dimensional wave interaction coupled with vastly varying time scales. Despite the associated difficulty, numerous efforts have been made to understand this phenomenon from a theoretical perspective. Köster et al. [39] used a perturbation approach to investigate flow field inside a droplet, while Collins et al. [72] investigated the hydrodynamics associated with SAW atomization, and developed a model for thin film spreading behavior under SAW excitation. Despite recent progress in describing the physics governing

SAW-induced atomization, much remains to be understood.

25 1.3.4 Particle/cell concentration

Concentrating particle/cell suspensions is a basic but critical operation in many applications in chemistry and biomedicine. On the macro-scale, particles can be easily concentrated using centrifugation. On the micro-scale, however, concentrating cells or particles can be difficult. As volume decreases, surface forces acting on the particles/cells begin to dominate over body forces. Lack of significant body forces makes standard centrifugation impractical, so researchers have explored SAW-induced acoustic streaming to concentrate cells and particles for low-volume systems [38].

To initiate rotational fluid motion, Shilton et al. [38] placed a droplet within a portion of a SAW propagation pathway such that the droplet experienced TSAW exposure across a fraction of its width. This resulted in a circular pattern of SAW-induced acoustic streaming within the droplet (Fig. 1-11a). When microparticles in aqueous suspension were subjected to this rotational pattern of SAW-induced streaming, they were concentrated and deposited in the center of the droplet. The researchers attributed the concentration effects to a shear gradient between regions of high shear at the droplet’s periphery and regions of low shear at the droplet’s center. Particles tended to move towards the region of low shear, where the fluid linear velocities approached zero.

Figure 1-11b shows sequential images of the SAW-induced streaming for a water sample containing 0.5 µm white fluorescent beads. The progression from distributed individual dots at t=0 to the central bright spot at t=1 s demonstrates how quickly the rotational SAW-induced streaming moves the fluid and concentrates the particles. Thus TSAWs allow for quick and efficient particle concentration at the micro-scale with a simple fluid droplet device setup.

1.3.5 Reorientation of nano-objects

1.3.5.1 Carbon nanotube alignment

Carbon nanotubes (CNTs) have been a focus of nanotechnology research for over two decades [75]. Their mechanical strength and unique electrical properties make CNTs attractive for a wide variety of applications ranging from sports equipment to electronics. However, aligning large quantities of CNTs for use in these applications often proves to be challenging.

TSAWs present a practical solution, allowing for the simultaneous alignment of many CNTs.

Strobl et al. [76] demonstrated the alignment of multi-walled carbon nanotubes (MWNTs) on a LiNbO3 substrate using TSAWs. Figure 1-12a shows a typical setup for the SAW-induced nanotube alignment device. After activating the SAWs, the device aligned carbon nanotubes

25−45 degrees relative to the SAW field (shown in Fig. 1-12b). To discern between the effects of fluid movement and piezoelectric field in the alignment of these MWNTs, the authors covered the

LiNbO3 substrate with a 20 nm layer of NiCr; this thin metallic film then screened the

27 piezoelectric field of the TSAWs. Experiments conducted on this metal-coated substrate device yielded no observable alignment of the MWNTs. To confirm this observed alignment is dependent on the piezoelectric field, the authors then used a LiTaO3 substrate to propagate a shear-wave SAW (as opposed to the Rayleigh SAW propagated on a LiNbO3 substrate). The mechanical component of a shear-wave SAW exists in the plane of the crystal, yielding minimal fluidic coupling relative to that of Rayleigh SAWs. When using LiTaO3 substrate, the MWNTs aligned parallel to the SAW field, showing improved alignment relative to the LiNbO3 device.

Thus the authors concluded that the MWNT alignment is attributable to the piezoelectric field associated with SAW propagation, and the acoustic streaming generated by Rayleigh SAWs actually shifted the MWNTs from being directly aligned with the piezoelectric field.

Figure 1-12 (a) Schematic of device setup for SAW-based carbon nanotube alignment. (b) AFM image of aligned multi-walled carbon nanotubes. Ref. [76]. (c) SEM image of aligned single- walled carbon nanotubes between pre-patterned gold electrodes on LiNbO3. [77].

28 Smorodin et al. [77] improved upon the alignment performance of the SAW-only approach by first thiolating single-walled carbon nanotubes (SWNTs) and then placing them on a substrate with patterned gold electrodes (Fig. 1-12c). By applying SAWs across this setup, the authors found that a majority of SWNTs aligned at angles between 0 and 20 degrees relative to the SAW field. They attributed this improvement to the rapid attachment of thiolated carbon nanotubes to the gold electrodes, reducing the impact of acoustic streaming on the nanotube alignment.

Though the previous studies demonstrated successful alignment of CNTs, their use of piezoelectric substrates limits their usefulness in microelectronics applications, as most microelectronic devices use non-piezoelectric silicon substrates. Seemann et al. [78] used a “flip- chip” configuration to overcome this limitation, enabling the alignment of carbon nanotubes on silicon substrate. In the experiment, a silicon substrate was patterned with gold contacts and then brought into close proximity with a LiNbO3 substrate. A CNT solution was deposited to fill the space between the two substrates. Application of TSAWs to the CNT solution (by means of the

LiNbO3 substrate) resulted in a combination of piezoelectric field and acoustic streaming effects that successfully aligned CNTs between the pre-structured gold electrodes on the silicon substrate.

1.3.5.2 Liquid crystal reorientation

Polymer dispersed liquid crystals (PDLCs) consist of liquid crystal droplets randomly dispersed in a transparent polymer matrix. They are widely used in displays and optical elements, where the application of an electric field to the PDLCs readily adjusts the light transmittance through the material. In addition to PDLC realignment via an electric field, several reports have shown that liquid crystals can also be realigned using acoustics. Recently, Liu et al. [79] reported a light shutter effect induced by TSAWs applied to PDLCs. As shown in Fig. 1-13a, PDLCs were

29 placed between two IDTs on a LiNbO3 substrate. In this setup, one IDT generated SAWs and the other detected them. Results demonstrated that SAW-induced acoustic streaming successfully aligned PDLCs parallel to the SAW propagation direction, changing the PDLCs from opaque to transparent (Fig. 1-13b).

During experimentation, SAW propagation on the substrate heated the PDLCs to 40°C.

To verify that the PDLC realignment was attributable to SAW-induced acoustic streaming and not to thermal effects, the researchers set up an experiment using standing SAWs rather than

TSAWs. This standing SAW field minimized the acoustic streaming effect while still heating the substrate and PDLCs; no PDLC realignment was observed. Therefore, the authors concluded that

SAW-induced acoustic streaming is a viable actuation method for PDLC realignment.

30 1.3.6 Phononic Crystal-Assisted SAWs

In the past decade, there has been a tremendous growth of interest in two- and three- dimensional periodic structures due to their ability to manipulate the propagation of acoustic waves on the wavelength scale. These artificial periodic structures, coined phononic crystals

(PCs), are composed of arrays of elastic scatterers embedded in elastic host materials with properties different from those of the scatterers. PCs exhibit several unique behaviors as a result of their structures.

To accomplish particle concentration in a droplet on the surface of a superstrate, Wilson et al. [80] exploited the absolute phononic band gap of a PC. As shown in Fig. 1-14a, an IDT was fabricated on a LiNbO3 wafer to generate TSAWs. A PC consisting of circular holes in a square array was designed to exhibit a wide phononic band gap and fabricated on a silicon wafer (Fig. 1-

14b). When a SAW is generated at the proper frequency by the IDT, it travels along the surface of the LiNbO3 substrate, leaks into a water-coupling layer, and excites Lamb waves across the thickness of the silicon superstrate that propagate in the same direction as the SAWs. Half of the excited Lamb waves encounter and are reflected by the PC, and the other half travel through the superstrate undisturbed. Thus the water droplet experiences an asymmetric Lamb wave exposure, and an in-plane counter-clockwise acoustic streaming pattern is induced. Fig. 1-14c shows the induced circular acoustic streaming being used to focus small particles inside a droplet. The

Cooper group also used this method to concentrate human blood cells from a diluted blood sample, demonstrating its applicability in biology and medicine (Fig. 1-14d).

Despite the added complexity to the manufacturing processes, the use of PCs in acoustic streaming applications enables some notable advantages. For example, PCs with different design characteristics can be used to tune the frequency and amplitude of a single SAW, allowing users to program fluid manipulation without altering the input electrical signal. In addition, PCs and

31 IDTs can be fabricated independently on separate units (e.g., substrates and superstrates, coupled by an intermediate fluid layer), enabling device configurability and disposability. Several studies have applied these advantages to create useful technologies. Reboud et al. [81] developed a sophisticated SAW-based fluid manipulation tool for the detection of rodent malaria parasite

Plasmodium berghei in blood. Red blood cells and the parasitic Plasmodium berghei in a drop of blood were mechanically lysed via the strong acoustic rotation vortices generated by PC-aided acoustic streaming. The researchers then amplified the parasitic genomic sequence by heating the sample using different acoustic field and frequencies. Their results demonstrated that this PC- assisted device has the ability to detect around 0.07% parasite DNA in a microliter-size blood sample and, more broadly, that it has potential for disease diagnosis in developing world.

32 Recently Bourquin et al. [82] reported an interfacial jetting phenomenon by fabricating a conic PC on a superstrate. In this work, the authors forced a SAW in a piezoelectric substrate to couple with a superstrate and excite Lamb waves. The cone-shape design of the PC they used provided a means to effectively focus acoustic energy to different hot spots, depending on the frequency of the SAW. They used this spatial control of acoustic energy to dictate the location and direction of the interfacial jetting behavior; the results show that this focused acoustic energy can efficiently produce interfacial jetting phenomenon.

1.3.7 Microfluidic Technologies Enabled by SSAWs

In this section, we review the latest on-chip innovations enabled by SSAWs. Instead of harnessing the acoustic streaming, SSAW-based devices use the primary acoustic radiation forces acting on particles via the surrounding fluid. We will detail devices that use these primary acoustic radiation forces to accomplish (1) focusing a flow stream of particles into a single-file line, (2) separating a flow stream of particles based on particle properties, and (3) patterning a group of particles in stagnant fluid.

1.3.7.1 Focusing of particles in a flow stream

Microfluidic focusing of particles in a liquid flow stream has attracted attention mainly due to its direct applicability to on-chip flow cytometry [83]. To detect particles via flow cytometry, the particles need to be lined up single-file so they can each pass individually through a detection point. Shi et al. [44] accomplished particle focusing in a microfluidic channel by situating the channel between two IDTs such that a SSAW was established across the channel width with a single pressure node that was located at the channel center. When particles passed

33 through the region of SSAW exposure, the acoustic radiation force pushed them to the center of the microfluidic channel width (i.e., two-dimensional focusing), as shown in Fig. 1-15. The same group also showed that SSAW can effectively achieve 3D focusing [43]. Applying a SSAW to fluid in a microchannel, the researchers realized that a non-uniform acoustic field generates a primary acoustic radiation force transverse to the particles in the channel (i.e., in the z-direction, relative to the device plane). This force will direct all of the particles towards the point of maximum acoustic kinetic energy. However, the acoustic radiation force acting in the z-direction is weaker than that acting in the device plane. Therefore, particles focus first along the channel width and then migrate to a focal point along the channel height. The researchers used a prism placed adjacent to the microchannel to verify experimentally that particles did, in fact, focus in the z-direction; they then validated this with theoretical calculations.

Figure 1-15 Schematic and working mechanism of SSAW-based focusing of particles in a liquid flow stream. [44].

34 1.3.7.2 Continuous separation of particles in a flow stream

As implied by Eq. (7), a standing acoustic wave field exerts a primary acoustic radiation force whose magnitude and direction depend on particle size, density and compressibility. Therefore,

SSAW fields can differentiate particles or cells based on their physical properties. Shi et al. [84] first reported SSAW-based continuous separation of particles in a flow stream in a PDMS microfluidic device. The researchers situated a microfluidic channel between two IDTs such that a SSAW was established across the channel width with half a wavelength spanning the channel and a single pressure node at the channel center. The researchers introduced a particle mixture into the channel from side inlets while simultaneously injecting a sheath flow from the center inlet (Fig. 1-16a). Because larger particles tend to experience larger primary acoustic radiation force, they migrate to the pressure node faster than smaller particles. Therefore, with appropriate length of the SSAW exposure region, SSAW enabled size-based particle separation: larger particles move to the center outlet, while smaller particles remain in the side streams.

Experimental results yielded 80% separation efficiency when separating a mixture of 4.16 µm and 0.87 µm particles. This device demonstrated a simple, efficient, and cost-effective method for size-based particle separation. Following the work of Shi et al., Nam et al. [85] developed a similar device setup and applied it to sorting blood cells from platelets.

In addition to separating a particle flow stream based on particle size, SSAW devices can also separate based on particle density. Nam et al. [86] demonstrated that large-cell-quantity alginate beads (LQABs), small-cell-quantity alginate beads (SQABs), and empty beads could be collected from their respect outlets (Fig. 1-16b). Jo et al. [87] reported a sheathless method to separate equally-sized polystyrene and melamine beads based on their density differential. High- density melamine beads (1.71 g/cm3) were separated from low-density polystyrene beads (1.05 g/cm3) with a separation efficiency of 89.4% at a 2 µL/min flow rate.

Figure 1-16 (a) Schematic of device for SSAW-based separation of particles in a liquid flow stream, based on particle size. (b) Schematic of device for SSAW-based separation of cell- encapsulating polymer beads in a liquid flow stream, based on bead density. [84, 86].

36 1.3.7.3 Patterning a group of particles in stagnant fluid

Techniques that can noninvasively arrange particles or cells (contained in a stagnant fluid) into desired patterns are invaluable for many biomedical applications, including microarrays, tissue engineering, and regenerative medicine. The non-contact, non-invasive primary acoustic radiation forces generated by SSAWs serve as an excellent enabler of such particle/cell patterning techniques. Wood et al. [47] exploited SSAWs to demonstrate the linear alignment of particles.

Following this work, Shi et al. [46] developed “acoustic tweezers” to effectively and noninvasively arrange particles and cells. The acoustic tweezers device consists of a PDMS microfluidic channel and a pair of IDTs deposited on a piezoelectric substrate in a parallel or orthogonal arrangement. After applying a RF signal to both IDTs to generate a SSAW field, particles/cells were patterned in parallel lines or arrays.

1.3.7.4 Protein manipulation

Supported lipid bilayers (SLBs) are critically important in cellular biology, as they control the intracellular and intercellular movement of ions, proteins, and other molecules [88].

Recently, researchers have applied SSAWs to pattern SLBs, which has use in membrane biology study. Hennig et al. [89] showed that local concentrations of SLBs can be modulated by applying

SSAW to a substrate. In the experiment, standard IDTs generated shear SSAWs on a LiTaO3 substrate to induce lateral reorganization of a lipid bilayer. Higher membrane density was found in the antinodes of the in-plane shear SSAW, while lower membrane density was found in the nodes. The researchers carefully monitored pattern formation and decay by fluorescence microscopy in order to study diffusion times of supported bilayers during the process. Diffusion times matched those attained with conventional methods, confirming the accuracy of SSAW for

37 patterning of SLBs. Hennig et al. further demonstrated the modulation of DNA density on supported lipid bilayers by binding DNA to a catonic SLB [90].

Based on this lipid-patterning approach, Neumann et al. [91] further demonstrated that proteins bound to SLBs can be accumulated, transported, and segregated using shear SSAWs. As shown in Fig. 1-17, SSAW-induced membrane modulations can achieve accumulation of labeled lipids. When the lipids are labeled with biotin, a biotin-binding protein called neutravidin accumulates alongside the lipids at pressure antinodes. The authors finely controlled the movement of proteins on SLBs by adjusting the phase between the two SAWs that combine to form the SSAWs. Phase adjustments in turn changed the locations of pressure nodes and

38 antinodes, enabling protein transport in the 2D plane of the substrate. Finally, the group used shear SSAWs to separate avidin and streptavidin (two different biotin-binding proteins) based on their sizes (Fig. 1-17a). Since streptavidin is the smaller of the two, it migrated towards the regions with high lipid bilayer densities at the antinodes. In contrast, SSAWs pushed the larger avidin proteins towards the pressure nodes, resulting in size-based protein segregation (Fig. 1-

17b-c). Thus, SSAW-based microfluidic devices can be used to manipulate proteins on SLBs with operational simplicity and high success rate.

1.4 Goals and organization

This dissertation described our research effort on developing acoustofluidic platform that can manipulate particles and cells in microfluidic chip using standing surface acoustic waves. A series of lao-ob-chip applications are presented based on this acoustic tweezers platform.

Chatpers 2-5 are expended versions of published or to be published manuscript. Each chapter begins with a brief literature review and motivation.

In Chapter 2, we develop an on-chip single cell manipulation system using tunable surface acoustic waves, called acoustic tweezers. This technology enables the manipulation of single particle, cell, or even whole organism, in a microfluidic channel.

In Chapter 3, we present a multichannel cell sorter that can sort single cell into five separate outlet channel in on step, enabling the multicolor micro fluorescent activated cell sorter

(µFACS).

In Chapter4, we develop a tunable cell patterning system that can dynamically change the pattern of cells, which will find great potential in tissue engineering and cellular studies.

In Chapter 5, a high efficient cell separation technique is developed. Cells or particles with different sizes and compressibilities are successfully separated.

39 Chapter 6 outlines the summary of this dissertation and perspectives for the future research.

Chapter 2

Single particle, cell, and organism manipulation using acoustic tweezers

Techniques that can dexterously manipulate single particles, cells, and organisms are invaluable for many applications in biology, chemistry, engineering, and physics. Here, we demonstrate standing surface acoustic wave based “acoustic tweezers” that can trap and manipulate single microparticles, cells, and entire organisms (i.e., Caenorhabditis elegans) in a single-layer microfluidic chip. Our acoustic tweezers utilize the wide resonance band of chirped interdigital transducers to achieve real-time control of a standing surface acoustic wave field, which enables flexible manipulation of most known microparticles. The power density required by our acoustic device is significantly lower than its optical counterparts (10,000,000 times less than optical tweezers and 100 times less than optoelectronic tweezers), which renders the technique more biocompatible and amenable to miniaturization. Cell-viability tests were conducted to verify the tweezers’ compatibility with biological objects. With its advantages in biocompatibility, miniaturization, and versatility, the acoustic tweezers presented here will become a powerful tool for many disciplines of science and engineering.

2.1 Introduction

Cellular-scale manipulation is essential to many fundamental biomedical studies. For example, the ability to precisely control the physical location of a cell facilitates the investigation of cell-cell and cell-environment interactions [92]. Manipulation techniques could also provide tools to help researchers observe the behavior of entire organisms such as C. elegans

(Caenorhabditis elegans) [93, 94]. Additionally, these techniques might also aid in molecular

41 dynamics/mechanics studies by allowing researchers to precisely monitor and control the interactions between biomolecules.

Trapping and manipulating microparticles was first demonstrated via optical techniques in 1986, when Arthur Ashkin, Steven Chu, and colleagues first demonstrated trapping of single micrometer-sized dielectric particles with a laser beam, a method now commonly known as optical tweezers [95]. Optical tweezers have since been used to trap and manipulate many kinds of micro/nano objects, including dielectric spheres, metal particles, cells, bacteria, DNA, viruses, and molecular motors [96−98]. Although optical tweezers have demonstrated excellent precision and versatility for a number of functionalities, they have two potential shortcomings: first, they may cause physiological damage to cells and other biological objects from potential laser-induced heating, multiphoton absorption in biological materials, and the formation of singlet oxygen [99]; and second, they rely on complex, potentially expensive optical setups that are difficult to maintain and miniaturize. Many alternative bioparticle-manipulation techniques [100−113] have since been developed to overcome these shortcomings, however, each technique has its own potential drawbacks. For example, magnetic tweezers [118−110] require targets to be pre-labeled with magnetic materials, a procedure that affects cell viability; electrophoresis/dielectrophoresis based methods [100−102, 111−113] are strictly dependent on particle polarizibility and medium conductivity, and utilize electrical forces that may adversely affect cell physiology due to current- induced heating and/or direct electric-field interaction [114]. In this regard, acoustic-based particle manipulation methods present excellent alternatives [19]. Compared to their optical, electrical, or magnetic counterparts, acoustic-based methods are relatively non-invasive to biological objects and work for most microparticles regardless of their optical, electrical, or magnetic properties.

To date, many acoustic-based particle manipulation functions (e.g., focusing, separating, sorting, mixing, and patterning) have been realized [43, 44, 46, 84]. None of these approaches,

42 however, have achieved the dexterity of optical tweezers; in other words, none of the previous acoustic-based methods are capable of precisely manipulating single microparticles or cells along an arbitrary path in two dimensions. The standing surface acoustic wave (SAW)-based acoustic tweezers presented in this article represent the first acoustic manipulation method to precisely control a single microparticle/cell/organism along an arbitrary path within a single-layer microfluidic channel in two dimensions. In our system, SAWs are generated by interdigital transducers (IDTs) deposited on the surface of a piezoelectric substrate. The use of SAWs allows our device to utilize higher excitation frequencies which results in finer resolution in terms of particle manipulation compared to bulk acoustic waves (BAWs). Additionally, we demonstrate similar manipulation of biological objects including cells and entire organisms (C. elegans). C. elegans is an attractive model organism for many biological and medical studies, mainly because of its relatively small size (~1 mm long), optical transparency, well-mapped neuronal system, diverse repertoire of behavioral outputs, and genetic similarities to vertebrates [93]. However, trapping and manipulating C. elegans has proven to be difficult and generally involves anesthetics, vacuum, cooling, or direct-contact mechanical procedures [93, 94, 115]. To our knowledge, our acoustic tweezers are the first to achieve contact-free, non-invasive, precise manipulation of C. elegans.

2.2 Working mechanism

The working mechanism and device structure of the acoustic tweezers are illustrated in

Figure 2-1. A 2.5×2.5 mm2 polydimethylsiloxane (PDMS) channel was bonded to a lithium niobate (LiNbO3) piezoelectric substrate asymmetrically between two orthogonal pairs of chirped

IDTs.

Figure 2-1 Device structure and working mechanism of the acoustic tweezers. (A) Schematic illustrating a microfluidic device with orthogonal pairs of chirp IDTs for generating standing SAW. An optical image of the device can be seen in Supporting Fig. S2. (B) A standing SAW th field generated by driving chirp IDTs at frequency f1 and f2. When particles are trapped at the n pressure node, they can be translated a distance of (Δλ/2)n by switching from f1 to f2. This relationship indicates that the particle displacement can be tuned by varying the pressure node where the particle is trapped.

Chirped IDTs have a linear gradient in their finger period (Fig. 2-1A) which allows them to resonate at a wide range of frequencies [116]. The chirped IDTs in our experiment have 26 pairs of electrodes with the width of electrode and spacing gap increasing linearly from 25 µm to 50

µm by an increment of 1 µm. The aperture of the chirped IDTs is 3.5 mm, larger than the width of the 2.5×2.5 mm2 square channel to ensure full coverage of the standing field. Each pair of chirped IDTs was independently biased with a radio frequency (RF) signal to generate SAWs; the interference between them forms a standing SAW field on the substrate. The standing SAW leaks into the adjacent fluid medium and establishes a differential pressure field in the fluid; this field generates acoustic radiation forces which act on suspended particles. The acoustic radiation forces drive particles to nodes or anti-nodes in the acoustic pressure field, depending on their elastic properties [117−121]. Most objects, including polystyrene beads, cells, and C. elegans, are pushed to the pressure nodes because of density and/or compressibility variations relative to the background medium. The large bandwidth of the chirped IDTs translates into a wide spectrum of accessible standing SAW wavelengths, which defines the large manipulation range of the device.

Using chirped IDTs with varying input RF frequency, we can shift the location of the pressure nodes generated from standing SAW interference. As a result, a single particle/cell/C. elegans which is trapped in the pressure node can be freely manipulated in two dimensions.

Figure 2-1B shows a schematic of the standing SAW and related pressure field along one dimension (x-axis) of the device. We refer to the stationary pressure node in the center of the

IDTs as the 0 order node (shown as a long dash dot line in Fig. 2-1B), progressing to the 1st order, nd rd 2 order, 3 order, etc. outward from the center. Because absolute node location ( for nth order pressure node, is the SAW wavelength) is directly related to the SAW wavelength, which is dependent on the signal frequency ( , where is the SAW propagation velocity on the surface of substrate), all higher-order ( ) pressure nodes can be moved simply by altering the applied signal frequency. The node displacement ( ) can be described by:

, as shown in Fig. 2-1B for a frequency change from f1 to f2.

The equation indicates that the particle displacement is directly proportional to the node order.

2.3 Methods

2.3.1 Materials

To generate a high amplitude surface acoustic wave to manipulate particles within microfluidic channel under relatively low power consumption, a piezoelectric material with high electro-mechanical coupling coefficient is required. Secondly, the technology presented here provides a platform that can manipulate particles and cells, and will find a series of applications in biology and biomedicine, so the SAW device substrate is required to be compatible with the characterization system widely used in bio-application. As the most widely used characterization system, an inverted microscopy system required the device substrate to be transparent in visible light range (~400 nm to ~700 nm). The criteria for choosing a piezoelectric material as SAW device substrate in SAW microfluidic system include (1) transparent to visible light (~400 nm to

~700 nm); (2) high electro-mechanical coupling coefficient; (3) high mechanical robustness and yield stress; and (4) available in market at low price. A summary of the coupling coefficients and

SAW propagation velocity of some commonly used piezoelectric materials can be found in Table

2-1, they are all transparent in visible light.

Table 2-1 SAW properties of some piezoelectric material with relatively high transparency [47-49]

128˚ Y-cut LiNbO3 Y-cut LiNbO3 ST-X Quartz ZnO/Quartz

SAW propagation 3980 m/s 3488 m/s 3158 m/s 2900 m/s velocity

K2 (%) 5.5 4.9 0.14 1

Table 2-1 shows that 128˚ Y-cut LiNbO3 demonstrates a highest coupling coefficient and

SAW velocity. Since the acoustic radiation force is proportional to the working frequency of

SAW, the high SAW velocity will make it easier to generate higher frequency SAW and improve the performance of SAW device. Another reason to select 128˚ Y-cut LiNbO3 as substrate material in our study is its validation in the market at relatively low price (~60$ for 4” wafer).

The robustness render the compatibility with standard MEMS fabrication process such as photolithography, Plasmon etching, metal deposition and device bonding). Some other basic properties of 128˚ Y-cut LiNbO3 can be found in Table 2-2.

Table 2-2 Basic properties of LiNbO3 [50]

Density Heat Melting Hardness Transparency Optical Refractive

(Kg/m3) Capacity Point (˚C) (Mohs) range (nm) Anisotropy Indices

(J/K*mol)

4647 89 1260 5 350~5000 Uniaxial, n0=2.286

c-axic ne=2.203

2.3.2 Device fabrication

Two major steps were involved in the fabrication of the acoustic device (Fig. 2-2): (1) the fabrication of chirp interdigital transducers (IDTs), and (2) the fabrication of polydimethylsiloxane (PDMS) microchannel. Figure 2-2 (AD) shows the fabrication process of the chirp IDTs. A layer of photoresist (SPR3012, MicroChem, Newton, MA) was spin-coated on a Y+128° X-propagation lithium niobate (LiNbO3) wafer, patterned with a UV light source, and developed in a photoresist developer (MF CD-26, Microposit). A double-layer metal (Cr/Au,

50Å/500Å) was subsequently deposited on the wafer using an e-beam evaporator (Semicore

Corp), followed by a lift-off process to remove the photoresist and form the IDTs. The PDMS microchannels were fabricated using standard soft-lithography and mold-replica techniques (Fig.

2-2 (EH)). The silicon substrate for the microchannel mould was patterned by photoresist

(Shipley 1827, MicroChem, Newton, MA). The maximum etching depth of the silicon wafer is

determined by the thickness of photoresist, which serves as the protection layer during the etching

process. The thickness control of photoresist can be achieved by the spinning speed, spinning

time, and photoresist type. Table 2-3 listed some commonly used photoresists with combined

parameters.

Table 2-3 Recipe used for Photoresist in DRIE

Spin Thickness Prebake Exposure Post bake Develop Maximum

(rpm) (µm) time (s) (s) depth (µm)

Shipley 4000 2.6~2.9 105˚C, 60 s 10 NA 60~90 80

1827

4000 7~7.4 60˚C, 90 s; 10 s, for 60˚C, 90 s; 60~90 190

Z220 90˚C, 90 s; 9 cycles, 90˚C, 90 s; 120˚C, 90 s; 120˚C, 90 s;

3000 7.5~7.9 60˚C, 90 s; 8 s, for 9 60˚C, 90 s; 60~90 200

90˚C, 90 s; cycles 90˚C, 90 s; 120˚C, 90 s; 120˚C, 90 s;

2000 9.7~10 60˚C, 90 s; 7 s, for 9 60˚C, 90 s; 60~90 240

90˚C, 90 s; cycles 90˚C, 90 s; 120˚C, 90 s; 120˚C, 90 s;

The patterned silicon wafer was then etched by a Deep Reactive Ion Etching (DRIE,

Adixen, Hingham, MA). SylgardTM 184 Silicone Elastomer Curing Agent (Dow Corning,

Midland, MI) and SylgardTM 184 Silicone Elastomer Base were mixed at an 1:10 weight ratio,

cast onto the silicon mould, and cured at 65°C for 30 min. Finally, The IDT substrate and PDMS

were treated with oxygen plasma and bonded together. The width of electrode finger and spacing

48 gap of the chirped IDTs used in our setup decreases linearly from 50 to 25 µm, corresponding to

SAW frequency of 18.5 MHz to 37 MHz. The aperture of the chirped IDTs is 3.5 mm, larger than the width of the 2.5×2.5 mm2 square channel to ensure full coverage of the standing field over the whole channel. The particles of interests were injected into the square channel through the inlet.

And the whole device is as large as a dime, as shown in Fig. 2-3.

2.3.3 Experimental Setup

2 A 2.5×2.5 mm PDMS channel was bonded to a LiNbO3 piezoelectric substrate asymmetrically between two orthogonal pairs of chirped IDTs, to form the acoustic device. Two major steps were involved in the fabrication: (1) the fabrication of chirped IDTs, and (2) the fabrication of PDMS microchannel (see Supporting Information for more details). The manipulation device was mounted on the stage of an inverted microscope (Nikon TE2000U).

Two RF signals were generated from two function generators (Agilent E4422B) to drive the two pairs of chirped IDTs independently. Solutions of bovine red blood cells (~6 μm in diameter,

Innovative Research, Inc.), C. elegans or fluorescent polystyrene beads with diameters of 2 μm, 4

μm, 7 μm, 10 μm, and 15 μm were injected into the channel before the RF signals were applied.

A CCD camera (CoolSNAP HQ2, Photometrics, Tucson, AZ) and a fast camera (Casio EX-F1) were connected to the microscope to capture the manipulation process.

2.3.3 Cell Viability and Proliferation Assays

HeLa cell viability and proliferation tests were conducted through the measurements of metabolic activity and DNA synthesis, to further exam the non-invasiveness of our technique

(37). HeLa cells were incubated in Dulbecco’s modified Eagle medium, (DMEM)-F12 medium

(Gibco, CA), with 10% fetal bovine serum (Atlanta Biologicals, GA), penicillin (100 U/ml), and

100 μg/ml streptomycin (Mediatech, VA) to about 90% confluence before trypsinization (Trypsin

+ 0.05% EDTA, Gibco, CA) and dilution to 2 × 105 cells/ml in medium. We treated 100 µl HeLa cell suspensions in five different conditions respectively: 1) untreated cells (positive control); under SAW radiation for 2) 6 s; 3) 1 min; 4) 10 min; and 5) at 65°C for 1 h (negative control).

After treatment, HeLa cells were seeded into Costar 96-well black clear-bottom plate (Corning

Life Sciences) with seeding density of 2 × 104 cells/well in 100 µl culture medium. 100 µl fresh medium was added into separate well as blank control. Cells were then incubated for 20 h after which we added 10 µl /well BrdU labeling solution (Roche Applied Science, IN). After labeling for 2 h we added 10 µl /well WST-1 (Roche Applied Science, IN) and reincubated for another 2 h.

Then we measured absorbance of each well at 450 nm and 690 nm (reference) with an absorbance reader (BioTek, VT). Subsequently, BrdU incorporation was determined with Cell Proliferation

ELISA, BrdU (colorimetric) (Roche Applied Science, IN) and the absorbance of each well at 370 nm and 492 nm (reference) was measured with the absorbance reader. Five separate studies were conducted for each condition under input power of 23 dBm, and the averages were given.

2.4 Results and discussion

2.4.1 Characterization of the acoustic tweezers

Figure 2-4A shows the simulated two-dimensional pressure field surrounding each pressure node, with arrows denoting the acoustic radiation force vectors. The simulation results indicate that a particle between adjacent pressure anti-nodes will experience an attractive force toward the pressure node between them. Figure 2-4B examines one-dimensional particle motion under varying acoustic power in response to the same frequency shift (also see Supporting Video

1); Figure 2-4C plots the particle’s velocity during this process. At the lower end of the force spectrum (11 dBm, magenta curve in Figs. 2-4B and 2-4C), a 10-µm fluorescent polystyrene bead can be continuously moved with velocity of ~30 µm/s, while at the opposite end of the force spectrum (27 dBm, red curve in Figs. 2-4B and 2-4C), particle velocities as high as ~1600 µm/s are achieved.

We conducted a force analysis to quantitatively determine the magnitude of the acoustic radiation force exerted on the particles at various power inputs. The experimental radiation force is plotted in Fig. 2-4D, calculated from the difference between the drag force (found using the velocity data in Fig. 2-4C) and the time derivative of particle momentum (also calculated from the velocity data in Fig. 2-4C). The force exerted on a 10-µm particle (shown in Fig. 2-4D with discrete points) fits well with the theoretical sinusoidal dependence of the force on the distance to the pressure node (solid lines). Since the acoustic radiation force depends on particle size, compressibility, and standing SAW amplitude, the force exerted on particles (reaching as high as

150 pN for a 10-µm particle) can be predicted and tuned. To demonstrate the stability of our device, the displacement reproducibility is exhibited in Fig. 2-4E, where a 4-µm polystyrene bead is moved back and forth in the x-direction, while being held stationary in the y-direction. The

52 displacement can be reproduced over hundreds of cycles. Lastly, we repeatedly step a pressure node in the positive x-direction to demonstrate particle translation over a larger length scale (~100

µm in Fig. 2-4F), again holding the y-direction constant. The inset in Fig. 2-4F shows the stacked image of the linear movement of a 10-µm polystyrene bead.

Figure 2-4 Quantitative analysis of the acoustic tweezers. (A) Simulated pressure field between adjacent pressure anti-nodes. (B) One-dimensional particle motion induced by a constant frequency change at varying applied acoustic power (experimental results). (C) Velocity plots corresponding to the displacement curves in (B). The inset (smoothed with a moving-average filter of 5 data points) shows that a velocity of 30 µm/s is achieved at the power input of 11 dBm. (D) Experimentally measured acoustic radiation force (ARF) on the particles as a function of distance from the nearest pressure node (discrete points) at different input power levels. The fitted curves are shown in solid lines. (E) Demonstration of reproducible particle motion. Here x- direction particle motion is repetitively shown between two stationary frequencies to show reproducibility. (F) Demonstration of continuous particle translation along the x-direction in well defined steps, while holding stationary in the y-direction.

The power density required to manipulate 10-µm polystyrene beads in our setup is ~0.5

2 nW/µm for particle velocities of ~30 µm/s, which is much lower than its optical counterparts

(10,000,000 times less than optical tweezers and 100 times less than optoelectronic tweezers)

[111, 122]. The working frequency range of the chirped IDTs used in our setup was 18.5 MHz to

37 MHz, corresponding to SAW wavelengths of approximately 100 µm to 200 µm. The displacement resolution for this device is dependent on the node order and frequency, and is theoretically limited by the frequency resolution of the function generator. In our case, imperfect alignment of the PDMS channel, and slight frequency fluctuations from the function generator could be responsible for the minor dispersion in the step displacement in Fig. 2-4F. Finally, we note that with the current setup, the acoustic radiation force on objects smaller than 1 µm is theoretically equivalent to the drag force in solution, limiting our current device to micro-scale objects. Manipulating nano-objects might be possible by using high-frequency SAWs, since the acoustic force is proportional to SAW frequency [106, 120].

2.4.2 Two-Dimensional Manipulation of Single Particles, Cells, and Organisms

To demonstrate single particle/cell manipulation in two dimensions, we tuned the input frequency of both pairs of orthogonally arranged chirped IDTs (as shown in Fig. 2-1A). Each pair of chirped IDTs independently controls particle motion along a single direction, thus the orthogonal arrangement enables complete control in the device plane. The dexterity of this approach is shown in the layered image in Fig. 2-5A, where a 10-µm polystyrene particle is trapped and moved along a path to write “PNAS”. Figure 2-5B presents the capture and subsequent manipulation of single bovine red blood cell to trace the letters “PSU”, demonstrating the applicability of the acoustic technique to biological samples.

To further demonstrate the biocompatible nature of this technique, we conducted HeLa cells viability and proliferation assay after exposure to high-power (23 dBm) standing SAW fields for 6 s, 1 min, and 10 min. The results (see Fig. 2-6 and Methods) indicate that after 10 min in the standing SAW field, no significant physiological damage was found on the cell viability and proliferation.

Additional control experiments were conducted to examine the heating effects of our acoustic device on the channel. After 10 min of input power at 23 dBm, the temperature increased from 25°C to 27.9°C. For an exposure time of 1 min, the temperature increase was less than 2°C.

At a higher input power level (25 dBm), which was required for C. elegans manipulation, the temperature was stabilized at 31 °C even after 10 min, as shown in Fig. 2-7.

Figure 2-7 The temperature of device was examined after applying input power of 23 dBm and 25 dBm for 10 mins. The results indicate a minimal heating effect.

In addition to microparticles and cells, this acoustic device can also be used to manipulate entire multicellular organisms, such as C. elegans. This task is challenging for optical techniques because high power density must be applied over larger areas, leading to impractical total power requirements. Using the same experimental conditions as single cells and particles, we trapped and independently translated C. elegans in the x and y directions (Figs. 2-8A−D). The C. elegans was moved in either the x or y direction when we turned on the power and swept the frequency,

57 and returned to normal behavior after the power is shut off. As seen in Figs. 2-8E and 2-8F, the

C. elegans can also be immobilized along their entire length and be stretched in a standing SAW acoustic field. The immobilization and stretching can be maintained for extended periods of time without inducing physiological damage allowing long-term, full-body studies to be undertaken.

2.4.3 Massive manipulation of particles

Finally, our acoustic tweezers can also simultaneously manipulate large numbers of particles. Although at this stage the technique cannot select an individual particle from a group, parallel manipulation of multiple particles can be achieved with clusters of particles at a single pressure node, single particles at different pressure nodes, or clusters of particles at distinct pressure nodes. This manipulation can occur over a variety of length scales. These videos show that while the acoustic tweezers are capable of dynamically manipulating single particles/cells/organisms, they are also capable of simultaneously manipulating more than tens of thousands of particles/cells. Fig. 2-9 shows the parallel manipulation of four single/group bovine red blood cells. Fig. 2-10 is a stacked image of tens of beads cluster moving in parallel arrangement to write “S”. Parallel manipulation of thousands of particles can also be achieved using our device. As shown in Fig. 2-11, all the 10 µm particles were pushed and concentrated to one side of the channel after sweeping the frequency for ~1.5 minutes. By applying a optimized input power, we successfully collected all the large particles without moving the small particles within the same channel, achieving the on-chip particle separation based on size.

Figure 2-9 This stacked image demonstrates parallel manipulation of groups of bovine red blood cells.

Figure 2-10 This stacked image demonstrates parallel manipulation of hundreds of 7-μm polystyrene beads to create an array of “S”s.

2.5 Conclusion

In summary, we have demonstrated standing surface acoustic wave-based acoustic tweezers that can manipulate single particles/cells/organisms in a microfluidic chip. This acoustic device has significant advantages in biocompatibility and versatility. The lower power density requirement renders our technique extremely biocompatible. The simple structure/setup of these acoustic tweezers can be integrated with a small RF power supply and basic electronics to function as a fully integrated, portable, and inexpensive particle-manipulation system. The technique’s versatility has three aspects: 1) it is capable of manipulating most microparticles regardless of shape, electrical, magnetic, or optical properties; 2) it is capable of manipulating objects with a variety of length scales, from nm (if we use higher SAW frequency) to mm (as demonstrated in C. elegans); and 3) it is capable of manipulating a single particle or groups of particles (e.g., tens of thousands of particles). The acoustic tweezers’ versatility, biocompatibility, and dexterity render them an excellent platform for a wide range of applications in the biological, chemical, and physical sciences including the fundamental studies of mechanical properties of micro and nano scale particles such as cells, DNAs, proteins, and molecules. Additionally, the ability to massively move particles with great speed (up to 1600 µm/s) could facilitate this technique a key tool in many high-throughput assays such as cell sorting and separation. Finally, this device could be used to help researchers examine the behavioral and neuronal response of small organisms (such as C. elegans) to mechanical and chemical stimuli.

Chapter 3

Multichannel cell sorting using acoustic tweezers

3.1 Introduction

Cell sorting is essential in numerous biological studies and clinical applications, such as molecular biology, pathology, immunology, genetics, and medical diagnostics and therapeutics

[123, 124]. Conventionally cell sorting is performed by fluorescence-activated cell sorters

(FACS), in which cells are encapsulated into small liquid droplets, which are then selectively labeled with electric charges and are sorted by an external electric field [125, 126]. Flow cytometry is used to analyze the physical/chemical properties of particles, usually cells, in a continuous flow, through an optical detection. A conventional flow cytometry system normally consists of focusing, detection, and sorting. Cell focusing is typically achieved through a hydrodynamic process, in which a high-speed sheath flow is used to squeeze the sample flow.

Optical detection part requires light source, optical detectors assembly, and a computer system for the storage and analysis of detected signal. A schematic of the conventional flow cytometry system is shown in Figure 3-1, in which an electric field is applied to sort the droplet containing cells to different containers [127].

These droplet-based cell sorters, however, have several drawbacks: 1) they generate aerosols, thus requiring stringent protocols and precautions; 2) they require additional steps to enclose cells into droplets; 3) if one intends to post-process the cells after sorting (e.g., cell culture), removing cells from droplets is necessary and often inconvenient. In this regard, it is important to develop other techniques that can perform direct cell-sorting without enclosing cells

63 into droplets. In the past decade, researchers have developed several microfluidic-based methods for direct cell sorting, such as magnetic-activated cell sorting, optical tweezers, electrokinetic

Figure 3-1 Diagram of FACS system. Cells are fluorescently tagged, analyzed by optical detection, encapsulated into droplets, and sorted by electric field.

64 mobilization, and hydrodynamic flow [128131]. As an example, figure 3-2 shows a typical optical tweezers enabled cell sorter system. Cells are first hydrodynamically focused by two side sheath flow. A visible wavelength laser is used for detection and fluorescence measurement. A

20W CW Ytterbium fiber near-infrared laser is used as optical switch for actuating. An acousto- optic modulator (AOM) is used to control those laser beams.

Figure 3-2 (a) Layout of the microfluidic sorter enabled by optical detection and optical switch. (b) Schematic of instrument of the optically switched microfluidic FACS system.

These methods have pioneered many new avenues to on-chip cell sorting; however, they also suffer from drawbacks such as low controllability, low cell viability and proliferation, and/or requirements on bulky equipments. Acoustic-based approaches appear to be excellent alternatives for cell sorting. Compared to their optical, electrical, or magnetic counterparts, acoustic-based manipulations have been recognized as a more gentle approach to biological objects [17].

Recently, acoustic-streaming-based droplet/cell sorting has been developed by utilizing the acoustically induced fluid streaming effect [73, 74]. The device used to accomplish this sorting consists of a branched PDMS channel and an IDT positioned adjacent to the channel that generates TSAWs propagating across the channel width. Either droplets or cells are injected into

65 the main channel and hydrodynamically focused in its center before reaching the region of TSAW exposure. When the IDT is off, the droplets or cells travel through the upper outlet branch because of its larger cross-sectional area and correspondingly lower hydrodynamic resistance

(Figs. 3-3a and c). When the IDT is powered, the resulting TSAWs generate acoustic streaming in the channel fluid and push the fluid, which carries with it the contained droplets or cells, towards the lower outlet branch (Figs. 3-3b and 3-3d). In this way, the device harnesses SAW-induced acoustic streaming to sort droplets or cells into two outlets.

In this chapter, we present a fundamentally different cell sorting method that uses tunable standing surface acoustic waves (SSAWs) to manipulate suspended cells directly by the acoustic radiation force, rather than manipulating the fluid through the acoustic streaming effect as demonstrated in the acoustic-streaming-based cell sorting mechanism. Compared with the

66 acoustic-streaming-based cell sorting mechanism in terms of cell manipulation, our SSAW-based method is more controllable and stable than the SAW-based approaches in which chaotic streaming is employed. It also produces a higher range of translation (>100 µm), which renders the ability to sort cells into a greater number of outlet channels in a single step. This is a major advantage over most existing cells-sorting methods which typically only sort cells into two outlet channels. In this work, we demonstrate precise sorting of cells into five separate outlets in a single step without external labeling or droplet encapsulation. In addition, we conducted cell viability and proliferation tests to confirm the non-invasiveness of our approach.

3.2 Working mechanism

In previous chapter we have demonstrated that particles/cells in a microfluidic channel can be precisely manipulated in a tunable SAW field. They were driven by the acoustic radiation force to the pressure nodes (PNs) or anti-nodes (ANs) of the SSAW, depending on their material properties. Particles/cells can be acoustically manipulated by dynamically moving the position of

PNs/ANs. In this work we demonstrated a biomedical application, cell sorting, using tunable

SSAW. Our SSAW-based cell-sorting device consists of a single-layer polydimethylsiloxane

(PDMS) channel and a piezoelectric substrate with a pair of chirped IDTs, as shown in Fig. 3-4.

The gradient in IDT pitch allows the excitation of SAW over a wide range of frequencies.

Therefore by tuning the frequency of SSAW, we can precisely move the PN/AN in the lateral direction normal to the cell flow, which in turn drives the floating cells to designated outlets on the fly (Fig. 3-4a & 3-4b).

The SSAW field is formed by the two parallel chirped IDTs on the surface of the substrate (underneath the microfluidic channel). The SSAW couples into the liquid medium and generates an acoustic radiation force on objects suspended in the liquid. The density and

67 compressibility of the object and the surrounding liquid determine whether the objects will be driven to the PNs or ANs; however, most objects suspended in aqueous solutions, including

Figure 3-4 (a,b) Working mechanism of the SSAW-based cell-sorting device. When the SSAW is switched on at frequency of f1 (a) and f5 (b), all particles/cells are driven to the pressure node (solid lines) and directed in the upper (a) or bottom (b) outlet. (c) The stacked trajectories of two 10 μm polystyrene beads at two frequencies respectively in the SSAW working region. A lateral shift of 100 μm is achieved after passing through the sorting region. polystyrene beads and blood cells, will be pushed to the PNs. The PN location is determined by the SAW wavelength which is directly dependant on the signal frequency f ( , where c is the SAW velocity on the surface of the substrate). Note that the PN at the center of the two chirped IDTs is fixed (independent of frequency) and is therefore referred as 0 order. The subsequent PNs progressing outward from the 0 order are deemed the 1st order, 2nd order, 3rd order, etc. All other (n>0) PNs can be moved simply by modulating the applied signal frequency; however, the node movement at different order is different. The node movement ( ) can be described by: , where i, j>0. As shown in Fig. 3-4, we

68 arranged pressure nodes and anti-nodes parallel to the channel walls. The center of the microchannel is located a distance of around 1500 μm (5λ at 14MHz) from the 0 order mode.

Simply by varying the input frequency applied to the chirped IDTs, we can shift the location of the pressure nodes. As a result, particles/cells trapped in the PN can be translated laterally across the channel, and with the addition of a fluid flow, the particles/cells can be directed into any of the upper or lower output channels, which correspond to certain input frequency, as shown in Fig.

3-4 (a & b).

3.3 Methods

The details of device fabrication have been described in section 2.3.2 in chapter 2. Figure

3-5 shows a photographic image of our device.

Figure 3-5 An optical picture of the SSAW-based cell-sorting device with 3 inlets and 5 outlets.

We used 128° Y-cut LiNbO3 as the substrate for its optical transparency and excellent electrical- mechanical coupling efficiency. The electrodes of the chirped IDT are aligned parallel to the X- axis of the LiNbO3 substrate. The pitch of the chirped IDT electrodes ranges linearly from 140

μm to 200 μm, producing a working frequency range of ~9.5 MHz to ~14.5 MHz. The sorting device was mounted on the stage of an inverted microscope (Nikon TE2000U). A radio frequency

(RF) signal was generated by a function generator (Agilent E4422B) to drive the two chirped

IDTs simultaneously. Solutions of human white blood cells (HL-60, ~10 μm in diameter), or fluorescent polystyrene beads (dragon green) with diameters of 15 and 10 μm were injected into the microfluidic channel by syringe pumps (NEMESYS) and hydrodynamically focused to the center of the main channel by two side sheath flows before the RF signals were applied to the

IDTs. The flow rates of sample flow and the sheath flows were 0.2 μl/min and 1.0 μl/min, respectively. All the solution we used was phosphate buffered saline (PBS). A CCD camera

(CoolSNAP HQ2, Photometrics, Tucson, AZ) was connected to the microscope to monitor the sorting process.

3.4 Results and discussion

Our sorting device was first characterized by its sorting distance. Fig. 3-4c is the merged trajectories of 10 μm polystyrene particles moving through the sorting region at frequencies of

9.8 MHz and 10.1 MHz. A lateral shift of ~100 μm was achieved at the end of the sorting region.

Note that due to the wide working frequency range of the chirped IDT, precise alignment of

PDMS channel and IDTs are not required. With further optimization on the chirped IDT, a lateral shift greater than 100 μm can be obtained for applications that require larger sorting distances.

To demonstrate the functionality of our SSAW-based sorting method, we first sorted 15

μm fluorescent polystyrene beads (dragon green, Bangs Lab) with a density of 1.05 g/cm3 to 3

70 outlet channels. In this experiment, the width and height of the main channel were 200 μm and 70

μm, respectively. When SSAW was off, the particles did not experience any acoustic force and thus follow the flow to the center outlet channel (center image of Fig. 3-6b). When an input signal was applied with a frequency of 14.5 MHz, the closest PN to the particles was ~25 μm above the center of the channel and directs particles to the upper outlet channel (top image of Fig. 3-6a).

When the frequency was switched to 13.9 MHz, the PN was positioned below the center of the channel and particles were directed to the lower outlet channel (bottom image of Fig. 3-6c).

These images were produced by stacking a video which captured the continuous particle-sorting process. The particle velocity in the main channel in our experiment was ~2000 μm/s, and the power applied to the device was ~25 dBm.

Figure 3-6 A 15 μm fluorescence particle passing through the sorting region can be directed to three different outlet channels by modulating input frequency (14.5 MHz, off, and 13.9 MHz).

71 To further demonstrate the versatility of our method and its ability to sort cells into more outlets, we tested human white blood cells (HL-60 human promyelocytic leukemia cells) in our sorting device. The leukemia cells were first hydrodynamically focused to the center of the main channel by two side flows, and then sorted to five specific outlets at frequencies of 9.8, 10.0,

10.2, 10.6, and 10.9 MHz, respectively. The cell-sorting results are shown in Fig. 3 by stacking the images of experimental videos recorded under the various frequency conditions described above. In this way, we would be able to sort cells to five individual outlets in a single microfluidic chip. Our device could potentially extend from current 5 outlets to a higher number of outlet channels to increase the functionality of the device for applications such as whole blood analysis and antigen-based cell diagnostics where the separation of many cell types is necessary.

Finally, to prove the non-invasive nature of our acoustic based technique, we have conducted standard cell viability and proliferation assays using HeLa cells. HeLa cells were incubated for 20 h after being treated in SSAW field for 10 min under the input power of 23 dBm, and then metabolic activity was measured at 450 nm after 2-h BrdU (Roche Applied Science, IN) labeling and following 2-h Reagent WST-1 (Roche Applied Science, IN) reincubation.

Subsequently, DNA synthesis was determined using Cell Proliferation ELISA (Roche Applied

Science, IN) to verify the cell proliferation. For control experiments, cells were examined without

SSAW treatment and at 65°C for 1 h. The culture medium with no cell was also measured as comparison. Each group was tested five times. HeLa cells show no significant sign of decrease in both viability and proliferation after exposure to SSAW fields for 10 min, implying no significant physiological damage was induced by our technique.

Figure 3-7 Sorting of human white blood cells (HL-60 Human promyelocytic leukemia cells) into 5 specific outlet channels at frequencies of (a) 9.8, (b) 10.0, (c) 10.2, (d) 10.6, and (e) 10.9 MHz.

73 3.5 conclusion

In conclusion, we have demonstrated a tunable standing surface acoustic wave based cell sorting technique that can sort cells into five outlets in a continuous flow, rendering it particularly desirable for multi-type cell sorting. Our device is assembled on a low-cost and disposable microfluidic chip and requires small sample volumes (~100 μl), making it an ideal component for research analysis and diagnostics. Furthermore, this system is operated purely by electric control, and can be easily integrated into complex system such as fluorescent detection and high-speed electrical feedback system to develop a multicolor micro fluorescence activated cell sorting

(μFACS) system. Our next research goal is to integrate optical detection unit to develop such multicolor μFACS system.

Chapter 4

Tunable patterning of particles and cells using acoustic tweezers

We have developed an acoustic-based tunable patterning technique by which microparticles or cells can be arranged into reconfigurable patterns in microfluidic channels. In our approach, we use pairs of slanted-finger interdigital transducers (SFITs) to generate a tunable standing surface acoustic wave field, which in turn patterns microparticles or cells in one/two- dimensional arrays inside the microfluidic channels—all without the assistance of fluidic flow.

By tuning the frequency of the input signal applied to the SFITs, we have shown that the cell pattern can be controlled with the tunability of up to 72%. This acoustic-based tunable patterning technique has the advantages of wide tunability, non-invasiveness, and ease of integration to lab- on-a-chip systems, and shall be valuable in many biological and colloidal studies.

4.1 Introduction

The ordered arrangement of biological cells and microscale objects plays an important role in many biological and colloidal studies such as microarrays [132], regenerative medicine

[133], and tissue engineering [134]. Thus far, many cell/particle patterning and manipulation techniques have been demonstrated, including optical tweezers [97], magnetic tweezers [135], optoelectronic tweezers [111], hydrodynamic manipulation [83], electrokinetics [136], and bulk acoustic wave-based acoustophoresis [111]. Dielectrophoresis (DEP) applies DEP force when polarizable particles are subjected to a non-uniform electric field, and has been extensively used for cells or DNA manipulations. photopolymerizable polyethylene glycol (PEG) hydrogel has been widely used as inert biomataerials sue to its great biocompatibility, porosity, and

75 degradation. The combination of DEP and PEG has been demonstrated to successfully fabricate

3D tissue microstructure with varying patterns. The capabilities of this method are illustrated in

Fig. 4-1. Although this method offers the capability of achieving 3D structure patterning, since the electrode is fabricated in the substrate, once the device is manufactured, the electrodes can not be changed anymore, which means the pattern offered by such device is not tunable.

Among these approaches mentioned above, acoustic-based methods seem to be ideal for on-chip manipulation or patterning of microparticles/cells, because acoustic methods are inherently non-invasive and can be used to pattern virtually all kinds of microparticles [117−121].

Recently, we demonstrated an “acoustic tweezers” technique, in which micro-objects were effectively patterned using standing surface acoustic waves (SSAWs) generated by a pair of interdigital transducers (IDTs) [46]. The acoustic tweezers device consists of a PDMS

76 microfluidic channel and a pair of IDTs deposited on a piezoelectric substrate in a parallel (Fig.

4-2a) or orthogonal (Fig. 4-2b) arrangement. After applying a RF signal to both IDTs to generate a SSAW field, particles/cells were patterned in parallel lines or arrays. Figs. 4-2c and 4-2d show the distribution of microbeads before and after the 1D and 2D patterning processes. However, this acoustic tweezers technique shares the same limitation as many other on-chip patterning techniques: features (e.g., the period) of the pattern cannot be modified. Once the device is fabricated, the period of the pattern is established and can be adjusted only within the bandwidth of the IDT (a range of less than 10%) [47, 137]. In this sense, the previously demonstrated acoustic tweezers technique operates in a static manner.

In this article, we report a major advancement in the acoustic tweezers technique: achieving tunable particle/cell patterning by varying the SSAW field with slanted-finger interdigital transducers (SFITs). Since the period of a SFIT—a variation of the regular IDT—

77 changes continuously from one end to the other, SSAWs at different wavelengths can be generated in between a pair of identical SFITs [138]. The response of SFIT is characterized in

Fig. 4-3, in which the amplitude of SAW excited is not uniform. The control of the frequency of an input AC signal translates into a wavelength- and location-varying SSAW field that in turn patterns microparticles with different features. Such a tunable particle/cell patterning technique is expected to be valuable in many on-chip cell studies such as dynamic control of cell-cell interactions.

Figure 4-3 (a) A schematic of the slanted finger IDT (SFIT); (b) The actuation of a droplet on the surface of the LiNbO3 sunstrate as well as on a cover slip coupled onto it at different input frequency and the theoretical calculation of the centre of the excited SAW. The inset is the S- parameter showing a stable response in the frequency range between 8 mHz and 14 MHz. [138].

4.2 Working mechanism

The schematic of a SFIT is shown in the left panel of Fig. 4-4a. A SFIT can be conceptually considered as a cascade of thin, distinct IDTs with homogenous finger spacing separated by the dotted lines. If the number of sub-channels is large, each sub-channel can be

78 considered a conventional uniform IDT. For each sub-channel, the period of the slanted fingers,

Di, is given by

, (4-1)

where λi is the wavelength of the SAWs excited by the sub-channel, c is the SAW velocity on the piezoelectric substrate along the propagation direction (normal to SFIT), and fi is the resonance frequency of the sub-channel.

When a radio frequency (RF) signal is applied to the SFIT deposited on a piezoelectric substrate, SAWs are generated from the sub-channels where the period of the slanted fingers satisfies eqn (4-1) to support the resonance condition. Take one of our designs as an example:

The period of the SFIT decreases linearly to cover a frequency range of 12 MHz to 18 MHz.

When a RF signal is applied at a frequency near the center of the tunable range (15 MHz), the excited SAW has its maximum amplitude at the center of the SFIT with a narrow bandwidth, as shown by the solid black line in Fig. 4-4a. The bandwidth of the excited SAW is inversely proportional to the number of slanted fingers; the effective aperture Aeff of the excited SAW beam can be approximated as

, (4-2)

where n is the number of finger pairs, and f0, fh, fl, A0 are the center frequency, highest frequency, lowest frequency, and aperture length of the SFIT, respectively [139, 140]. If we tune the frequency of the input RF signal, the main SAW beam will be switched to corresponding high- frequency (red dash-dotted line in Fig. 4-4a) or low-frequency (blue dashed line in Fig. 4-4a) regions, respectively. Therefore, the SFITs allow dynamic control of the specific position of the maximum SAW amplitude and the acoustic wavelength. It is worth noting that the amplitude of the excited SAW generally varies with frequency when the input voltage is kept constant [141];

79 however, it is possible to calibrate the system to obtain a uniform output as shown in the right panel

Figure 4-4 (a) The left panel shows the configuration of a SFIT of which the period is changing continuously. The right panel describes the amplitude distribution of a SAW at different frequencies. (b) A schematic drawing of the acoustic-based, one-dimensional dynamic patterning device. (c) Relationship between frequency and spacing of one-dimensional patterning on different piezoelectric substrates.

80 of Fig. 4-4a.Since the propagation velocity of SAW varies depending on the material properties of the piezoelectric substrate [142, 143], as shown in Fig. 4-4c, we can also achieve large or narrow pattern spacing for different applications by choosing an appropriate substrate. In this experiment, we use a 128° Y-cut lithium niobate (LiNbO3) wafer for its optical transparency and high electro-mechanical coupling.

The acoustic-based tunable one-dimensional (1D) patterning devices consist of a polydimethylsiloxane (PDMS) microfluidic channel bonded in between an identical pair of SFITs deposited on a piezoelectric substrate. The pair of SFITs was deposited in a parallel geometry, as shown in Fig. 4-4b, and an RF signal was split and applied to each SFIT to generate two identical

SAWs propagating along the y-axis. These two SAWs propagate in opposite directions and interfere with each other to form a SSAW field in between the SFITs where the PDMS microchannel was bonded (Fig. 4-4b). Such a SSAW field leads to a periodic distribution of pressure nodes and antinodes on the device substrate. When the resonating SSAW encounters the liquid medium encapsulated in the microchannel, it generates periodic longitudinal pressure waves that cause pressure fluctuations in the medium. The acoustic radiation force, generated from the pressure fluctuations, pushes the suspended particles/cells toward pressure nodes or antinodes in the SSAW field, depending on the elastic properties of microparticles/cells. The primary acoustic radiation force that acts on microparticles can be expressed as

, (4-3) λ

, (4-4)

where p0 c c w c, and w are the acoustic pressure, wavelength, volume of the particle, density of the particle, density of the medium, compressibility of the particle, and compressibility of the medium, respectively [144]. Eqn (4-4) describes the acoustic contrast factor, φ, which

81 determines whether the particles move to pressure nodes or antinodes; the particles will aggregate at pressure nodes when φ is positive and pressure antinodes when φ is negative. So far as we know, most particles and cells have positive φ, and go to pressure nodes in the SSAW fields; bubbles and lipids usually have negative φ and move to pressure anti nodes. Since the acoustic radiation force is directly acting on the particles/cells, our tunable pattern is achieved without the assistance of fluidic flow. In a 1D SSAW field, the pressure nodes (or antinodes) are arranged in multiple lines, resulting in a 1D dynamic patterning of particles along these lines, as shown in

Fig. 4-4b. The period of the 1D pattern depends upon the frequency applied to the SFITs.

4.3 Methods

4.3.1 Device design and fabrication

Here we describe a general design rule for SAW-based 1D patterning. To obtain a tunable SAW frequency ranging from f1 to f2 on a piezoelectric wafer with a SAW velocity c, the period of each SFIT was designed to vary linearly from c/f1 (D1 in Fig. 1a) to c/f2 (D2 in Fig. 1a) per eqn (1). The width of each electrode is normally kept one quarter of the period. Two SFITs and a PDMS microchannel are aligned parallel to each other. The period of the particle pattern is equal to the period of the resultant SSAW field, which is half of the working wavelength of

SAW. As a result, the period of the tunable particle pattern ranges linearly from c/2f1 to c/2f2.

Device fabrication details have been described in section 2.3. 2 of chapter 2. To fabricate these acoustic-based tunable particle-patterning devices, a double layer of chrome and gold

(Cr/Au, 50 Å /800 Å) was deposited on a photoresist-patterned 128° Y-cut lithium niobate

(LiNbO3) wafer, followed by a lift-off technique to form the SFITs. A single-layer PDMS

82 channel was fabricated with standard soft-lithography and mold replica techniques, and was then aligned and bonded on the LiNbO3 substrate in between SFITS to obtain the desired devices.

4.3.2 System setup

The experiments were conducted on the stage of an inverted microscope (Nikon

TE2000U). Solutions of fluorescent polystyrene (Dragon Green) beads with diameter of 4.16 μm and 7.32 μm were injected into the microfluidic channel by a syringe pump (KDS210 Kd

Scientific) before the RF signal was applied. A signal generated by a RF signal function generator

(Agilent E4422B) was split into coherent signals after being amplified with a power amplifier

(Amplifier Research 100A250A). These coherent signals were then applied to the SFITs to generate SAWs at the designated frequency. A CCD camera (CoolSNAP HQ2, Photometrics,

Tucson, AZ) was connected to the microscope to record the patterning process. The applied RF power on devices was 26 dBm in patterning micro polystyrene beads and 23 dBm in patterning bovine red blood cells.

4.4 Results and discussion

4.4.1 Tunable 1D patterning

Dynamic patterning of micro polystyrene beads

The optical image of a tunable particle-patterning device is shown in the left panel of Fig.

4-4. The aperture and number of fingers of the SFITs are 5 mm and 10 pairs, respectively. The width of the SFIT fingers varies linearly from 50 μm to 75 μm along the aperture, corresponding

83 to a SAW wavelength range from 200 μm to 300 μm. The width of the PDMS microchannel is

200 μm. Inside the channel, the pattern period equals half the wavelength, which corresponds to the pressure-node separation. The propagation direction of the SFIT-generated SAW was set along the crystal x-axis of the 128˚ rotated, Y-cut single crystal LiNbO3 substrate, and thus the generated SAW frequency range was 12 MHz to 18 MHz. Here, we demonstrate tunable patterning using fluorescent (Dragon Green) polystyrene beads with a mean diameter of 7.32 μm suspended in water. The acoustic contrast factor φ of the beads is positive, causing them to aggregate at the pressure nodes upon SSAW generation. The microscope-captured images are displayed in the right panel of Fig. 4-5. When a RF signal was input into the SFITs at 12 MHz, polystyrene beads were patterned in two lines with a period of ~150 μm (Fig. 4-5-I) in the area where the period of the slanted fingers satisfies the resonance condition defined by eqn (1). Beads located outside of this area were not patterned. When the input RF signal was tuned to 15 MHz, the actively patterned area moved to the center of the device, with a corresponding period of ~120

μm (Fig. 4-5-II). When the RF signal was further increased to 18 MHz, the patterned area appeared only at the downstream end of the device with a period of ~100 μm (Fig. 4-5-III). By design, the measured results match well with the period of the SSAWs. Thus, we have shown that our device can dynamically pattern microparticles in 1D arrays with periods ranging from 100 μm to 150 μm. Note that due to the low diffusion rate of the microparticles, we can program the function generator to scan the RF signal frequency over the whole working range of the device to form a continuous pattern in the microfluidic channel, as illustrated in Fig. 4-4b. In our previous work [46], we have shown the stationary pattern, with which we can achieve only one specific pattern in one single device. In this work, we improved this stationary pattern into a dynamically tunable pattern, in which different particle’s/cell’s patterns can be dynamically achieved in a single device. This flexibility of particle patterning could be helpful in many fundamental

Figure 4-5 1D dynamic patterning of fluorescent polystyrene beads. Left: the optical images of the 1D tunable patterning device. The microchannel covers two lines of the pressure nodes of the SSAW. Right: I, II, and III show the 1D pattern at the frequencies of 12 MHz, 15 MHz, and 18 MHz, respectively. research and biomedicine studies. The ability to precisely control the physical distance between two particles/cells facilitates the investigation of cell behaviors such as cell-to-cell interactions and chemotaxis studies. For example, controlling the physical position of two cytochalasin D microsources would help study the migration behaviors of human neutrophil HL-60 cells [145].

Such tunable patterns can also be useful in tissue engineering. For example, we can first fill the channel with stromal cells, pattern and culture them on the substrate, and then refill with hepatocyte cells, and pattern them in a different array. By tuning the pattern, we can control the distance between these two types of cells, and thus facilitate investigation of their interactions

[146].

Quantitative force analysis

Particles in a microfluidic channel under the influence of a SSAW field are subjected to four forces: the primary radiation acoustic force, viscous drag force, gravity force, and buoyant force. Among these forces, the buoyant force and gravity force act along the channel height direction and typically balance one another due to their similar magnitudes and opposite directions. The viscous drag force is a passive force which is proportional to the particle velocity and acts along the direction opposite to the particle migration. As a result, particle patterning can be analyzed by examining the primary acoustic radiation force, which is given in eqn (3). Fig. 4-

6a shows the calculated distribution of the primary acoustic radiation force within a quarter wavelength range under different frequencies. The primary acoustic radiation force varies with distance from the pressure node in the form of a sinusoid function. The maximum force is located between a pressure node and a pressure antinode (0.25 λ), with its value proportional to the operating frequency. When a SSAW field is generated across the microfluidic channel, the primary acoustic radiation force pushes the suspended particles toward the closest pressure nodes.

The induced viscous drag force then counters the acoustic radiation force; therefore, examining particle velocity reveals information about the acoustic radiation force acting on the particles. We recorded the movement of the particles in a 1D SSAW field at the frequency of 12 MHz in Fig. 4-

6b. The white dots are the particles before SAWs were generated, and red lines are the trajectories of particles over the recorded time span. The applied RF power on devices was 26 dBm and the patterning time was around 10 s. The dependence of the particle velocity on the distance from the pressure node is given in Fig. 4-6c. The velocity distribution shows a maximum between the pressure node (x=0) and the pressure antinode (x=0.25λ), which agrees well with the theoretical analysis of the acoustic radiation force (Fig. 4-6a).

Dynamic patterning of human promyelocytic leukemia cells

To verify the applications of our tunable patterning technique to biological cells, we performed patterning of HL-60 Human promyelocytic leukemia cells. When the SFITs were excited with a frequency of 31 MHz, cells were patterned with a period of 60 μm in the region corresponding to a small finger pitch (as shown in Fig. 4-7-I). The RF power was 23 dBm and the patterning time was less than 10 seconds. When the frequency was tuned to 25 MHz and 13 MHz, cells were patterned in the regions corresponding to median and large finger pitch with a period of ~78 μm (Fig. 4-7-II) and ~150 μm (Fig. 4-7-III), respectively. The tunable bandwidth of the spatial period (150 μm – 60 μm = 90 μm) is 72% (90/125=72%) of the period at the center of the tunable range ((150 μm + 60 μm)/2=125 μm). Eqn (3) shows that the acoustic force is depending on the frequency and elastic properties. In this work, the acoustic force is larger on polystyrene beads than on cells; this is one of the reasons why the cell’s pattern is not as well confined as the bead’s pattern. The difference of the acoustic force exerting on cells and particles also results in a longer pattern time for cells than for particles. We have to note here that the pattern time at high frequency (as shown in Fig. 4-7-I) is less than that at low frequency (Fig. 4-7-III). This is because the acoustic force is larger at high frequency than that at low frequency.

4.4.2 Tunable 2D patterning

Our acoustic-based particle-patterning technique can be extended to obtain tunable 2D patterning. Fig. 4-8a shows the schematic of the 2D tunable patterning device, where two identical pairs of SFITs are arranged orthogonally. When a RF signal is applied to all SFITs, two

SAWs were generated at the corresponding frequency regions from all four SFITs and interfere

Figure 4-7 Dynamic patterning of bovine red blood cells. I, II, and III show the 1D pattern at the frequencies of 31 MHz, 25 MHz, and 13 MHz, respectively. with each other to form a 2D SSAW field. The simulated 2D SSAW field is portrayed in Fig. 4-

8b, where we see pressure nodes (PN) formed in orthogonal 2D arrays that are tilted at an angle of 45° with respect to the x-axis. The pattern of pressure nodes, and thus the patterned particles in the region corresponding to a smaller finger pitch (higher frequency) have a smaller period, while the pattern in the region corresponding to a larger finger pitch (lower frequency) has a larger period. The period of the 2D-patterned array is times the SAW wavelength. As a result, we can control the period of the 2D pattern simply by adjusting the frequency of the input RF signal.

The 2D dynamic patterning device consists of two identical pairs of SFITs deposited on a

LiNbO3 substrate in an orthogonal arrangement. The aperture of the SFITs is 2.5 mm— the same length as the square microfluidic channel. The period of the SFIT fingers varies linearly from 200

μm to 350 μm, corresponding to a range of SAW frequencies from 10.5 MHz to 18.5 MHz.

Therefore, in theory, the period of the generated 2D SSAW field and the 2D particle-pattern can be tuned from 141 to 250 μm. We used fluorescent polystyrene beads of 4.16 μm diameter to examine the 2D dynamic patterning effect. The two images at the bottom of Fig. 4-8 illustrate the

2D particle-patterning results at different frequencies. When the SFITs were excited with RF signals at 18 MHz, polystyrene beads in the upper corner of the channel were driven to the 2D

90 pressure nodes, as shown in Fig. 4-8-I. The measured period of the array was 150 μm, which is close to the theoretical prediction of 145 μm. When the frequency was tuned to 10.9 MHz, particles were patterned in the lower corner (Fig. 4-8-II) of the microfluidic channel. The measured periods of the patterned arrays are 240 μm; this is in absolute agreement with the theoretical prediction of 240 μm. Thus, we have demonstrated dynamic control of the period of a

2D pattern in the range from 150 μm to 240 μm, by tuning the RF signal frequency. The tunable bandwidth of the spatial period (240 μm – 150 μm = 90 μm) is 46% of the period at the center of the tunable range ((240 μm + 190 μm)/2=195 μm). This range can be further extended by increasing the gradient of the SFITs, so as to correspond with a wider range of working frequencies. However, one must consider that SAWs can be simultaneously excited at the fundamental and harmonic frequencies in large gradient SFITs to result in multiple patterning areas.

It is worth noting that the size of the bead aggregations at the pressure nodes expanded with decreasing frequency. This phenomenon is caused by two major factors. First, based on the assumption that the initial bead distribution was uniform in the solution, the period of the array at a higher frequency (shorter wavelength) is smaller than that of the array at a lower frequency.

Thus, fewer beads are collected at each pressure node at higher frequencies, resulting in smaller- size aggregations. Second, from eqn (3) we learned that smaller wavelengths (higher frequency) lead to larger acoustic radiation forces on particles, resulting in more compact aggregation.

4.5 conclusion

In conclusion, we have developed an effective tunable patterning technique using slanted finger interdigital transducers by which the wavelength and position of the excited surface acoustic waves can be dynamically adjusted. Different patterns of microscale particles and cells

91 in different regions can be dynamically achieved simply by controlling the frequency of the input signal without moving any on-chip or off-chip parts. Tunable periods with a range of 100 μm to

150 μm for 1D patterning and 150 μm to 240 μm for 2D patterning have been realized without the assistance of fluidic flow. This acoustic-based tunable particle/cell patterning technique is inherently non-invasive, increasing the potential for application in microarrays, cell studies, and tissue engineering.

Chapter 5

High efficiency cell separation using acoustic tweezers

5.1 Introduction

Efficient separation of suspended particles and cells is essential to many fundamental biomedical studies such as cancer cell detection and drug screening [147-150]. The most popular methods for cell separation in the life science laboratory so far are the centrifugal methods, which are capable of separating cells with differences in size and density [151, 152]. Another industrial and clinical standard for high quality cell separation is FACS (fluorescence activated cell sorter).

The FACS technology is performed in a sheath flow mode where cells are focused in the center of buffer and then pass through a laser beam for high speed and precise optical detection. The cells can be separated by the downstream electric field triggered by the optical signal [123, 124].

In the past years, fundamental advances in the lab-on-a-chip technologies have driven development for new approaches to cell separation. To date, many other microfluidic alternative of cell separator have been demonstrated, including magnetic [153, 7], hydrodynamic [154−156], optical lattice [157], electrophoresis/dielectrophoretic (DEP) [158−160], and acoustic [161] methods.

Magnetic method starts from labeling cells of interest with magnetic markers. Then an external magnetic field is applied to the sample, leading to the separation of labeled cells from the rest. The step of labeling in magnetic methods usually induces additional cost and processing time, but may also bring negative effect on the cells of interest. The hydrodynamic methods usually involve high flow speed (inertial force based method) or asymmetric obstacles inside the

93 channel (deterministic lateral displacement). These methods permit continuous operation without requiring additional labeling or external forces. However, the channel obstacles in the channel may exert high mechanical stress on cells and lead to low throughput. The optical lattice method provides a unique separation approach which can separate particles with different optical properties, however they have two potential shortcomings: 1) the potential laser-induced heating, the formation of singlet oxygen, and multiphoton absorption in biological materials may cause physiological damage to cells and other biological objects [99]; and second, they rely on complex, potentially expensive optical setups that are difficult to maintain and miniaturize.

Electrophoresis/dielectrophoresis based methods [158−160] are strictly dependent on particle polarizibility and medium conductivity, and utilize electrical forces that may adversely affect cell physiology due to current-induced heating and/or direct electric-field interaction [114].

In this regard, acoustic-based particle manipulation methods present excellent alternatives

[20−22]. Compared to their optical, electrical, or magnetic counterparts, acoustic-based methods are relatively non-invasive to biological objects and work for most microparticles regardless of their optical, electrical, or magnetic properties. The well developed bulk acoustic wave (BAW) acoustophoresis has demonstrated the separation of cells based on size and density in microfluidic chips without any labeling on the target particles or cells [161]. This BAW method, however, requires the channel material with excellent acoustic reflection properties (such as silicon and glass). The widely used soft polymer materials in microfluidic applications, such as PDMS, usually do not have those properties. Moreover, the transducer to generate BAW is bulky and hinders the system integration and miniaturization. We have previously reported a standing surface acoustic wave based continuous particle separation device, in which one linear pressure node was formed in the center of the microfluidic channel, parallel to the channel and flow direction [84]. Particles with different size were injected into the device through the side inlet channel. Large particles were pushed by the acoustic radiation force towards the pressure node in

94 the center of the channel, while small particles remained in the side flow due to insufficient lateral acoustic force. However, the separation efficiency and separation resolution in our previous studies are low, 85% for beads of 0.87 µm and 4.16 µm in diameter. Here in this section we proposed a unique design based on SSAW method, and demonstrated a high separation efficiency with separation resolution of 30%. Cell viability, proliferation, and apoptosis tests were carried out to confirm the excellent biocompatibility of our device.

5.2 Working mechanism

The standing SAW based separation device consists of a polydimethylsiloxane (PDMS) microfluidic channel bonded in between an identical pair of IDTs deposited on a piezoelectric substrate. The microfluidic channel consists of three inlets and two outlets, center inlet for sample and two side inlets for buffer. The pair of IDTs were deposited in a parallel arrangement, and aligned at a specific angle (15°, 30°, and 45° in this work, and can be variable) to the channel and flow direction. A RF signal was applied to each IDT to generate two identical SAWs. These two

SAWs propagate in opposite directions and interfere with each other to form a standing SAW in between the IDTs where the PDMS microchannel was bonde. Such a standing SAW generates a parallel distribution of pressure nodes and antinodes on the surface of substrate. The acoustic radiation force, generated from the pressure distribution, pushes the suspended particles towards pressure nodes or antinodes in the standing SAW field, depending on the elastic properties of microparticles. Figure 5-1a is a schematic of the cross section of the channel where particles are pushed towards the pressure node. Particles are injected through the center inlet channel and hydrodynamically focused by two side flows before entering the SSAW field. Particles in this SSAW field experience lateral acoustic radiation force, drag force, gravity force and buoyant force. Gravity force and buoyant force are similar in magnitude but opposite in

95 direction, and are almost balanced. The behavior of particles in the channel can be characterized by examining the drag force and acoustic radiation force.

The primary acoustic radiation force (Fr) and drag force (Fd) can be expressed as

, (5-1)

, (5-2)

, (5-3)

where p0, λ, Vp, ρp, ρm, βp, βm, η, r, and v are the acoustic pressure, wavelength, volume of the particle, density of the particle, density of the medium, compressibility of the particle, compressibility of the medium, medium viscosity, particle radius, and relative velocity, respectively [84]. Eqn (5-2) describes the acoustic contrast factor, φ, which determines whether the particles move to pressure nodes or antinodes: the particles will aggregate at pressure nodes when φ is positive and pressure antinodes when φ is negative. So far as we know, most particles and cells have positive φ, and go to pressure nodes in the SSAW fields, bubbles and lipids usually have negative φ and move to pressure anti nodes [46]. Eqns (5-1) and (5-3) indicate that the radiation acoustic force is proportional to the volume of the particle/cell while the drag force is proportional to the radius of particle. Large particles that experience larger acoustic force will be confined in the pressure node, and will be repositioned with large lateral displacements along the width of the channel, as shown in Fig. 5-1b. The green large particles will be collected in upper outlet channel. For the small particles, the force acting on them are not large sufficient to confine them in the pressure node, they remain in the center stream by the drag force and are collected in the bottom outlet channel, as shown the black particles in Fig. 5-1b. Device fabrication has been described in section 2.3.2 in chapter 2, and a photo of typical separation device is shown in Fig.

5-1c.

At a high input power, corresponding to large SAW amplitude, acoustic radiation force dominates and confine the particles trajectory along the tilted pressure node, achieving a large distance shift across the width of the channel. Low input power leads to small acoustic radiation force and drag force dominates on the particles, causing a small lateral distance shift. The trajectory of 15 µm particles at different SAW amplitude has been recorded at the flow velocity of ~2 mm/s, as shown in Fig. 5-2. Since the acoustic radiation force depends on the

97 mechanical properties such as volume, compressibility, and density, particles with differences in those properties can be differentiated and separated by our acoustic device.

Figure 5-2 Stacked image showing the trajectores of 15 µm polystyrene beads at inuput power of (a) 27dBm, (b) 23dBm, and (c) 15 dBm. The flow velocity is ~2 mm/s.

5.3 Results and discussion

5.3.1 Particle separation based on size

We first exam the SAW separator using polystyrene beads. Positions in the SAW working region and outlet of the channel were recorded to analyze the distribution of the particles, as shown in Fig. 5-3. A mixture of 10 µm and 2 µm particles were injected into the

Figure 5-3 Stacked images showing particles (2 and 10 µm) trajectories at the working region (a and c) and outlet region (b and d) when SAW is off (a and b) and on (c and d). channel and were hydrodynamically focused in the center of the channel by two side flow. When

SAW was off, small particles and big particles were flowing together along the stream and exited through the lower outlet channel, as shown in Fig. 5-3 a and b. when we turned on the SSAW device, particles entering the working region experienced acoustic radiation force, which pushed them towards the linear pressure nodes tilted with an angle of 30˚ along the flow direction. At the flow velocity of 6.5 mm/s and input power of 16-23 dBm, the acoustic radiation forces pushed large particles towards pressure node and confined them along the linear node with an angle along the channel until those particles exited the working region. The small particles, however, due to the insufficient acoustic radiation force acting on them, remained in the original flow

99 stream. Fig. 5-3 c and d indicate that large particles were pulled out from the mixture stream and were separated through upper outlet channel while small particles trajectory were not significantly affected and were collected in the lower outlet channel. The ratio of large and small particles collected from each outlet channel was analyzed to evaluate this method. The number of particles was counted through the recorded video. 98% of the large particles migrated to the upper outlet channel while 100% of the small particles remained in the lower outlet channel.

To further exam the resolution of our technique, we separated fluorescent polystyrene beads with diameters of 9.9 µm and 7.3 µm in aqueous buffer. A mixture of those beads were injected into this device and set to flow at ~1.5 mm/s. The small beads (green) and large beads

(red) were mixed before entering the SSAW working region (Fig. 5-4). The large beads were extracted from small beads stream during passing through the working region. The fluorescent intensity profile was scanned near the outlet channel to indicate the beads distribution. Two peaks for small beads were caused by the non-uniform flow velocity distribution in vertical direction.

This is attributable to the hydrodynamic effect within the laminar flow. This result shows that our method has achieved the separation resolution of 30% (=2(9.9-7.3)/(9.9+7.3)), which is better than most of other methods.

5.3.2 Particle separation based on compressibility

To further explore the versatility of our method, particle separation was carried out based on the difference of compressibility. HL-60 are human promyelocytic leukemia cell line, with diameter of ~15 µm. Hl-60 cells (with density of ~1.075 kg m-3, compressibility of ~4 *10-10 Pa-1) were mixed with 15 µm polystyrene beads (with density of 1.05 kg m-3, compressibility of ~2.16

*10-10 Pa-1) [42]. They have similar sizes and densities but different compressibilities. Fig. 5-5 shows the separation of particles with different compressibilities. Polystyrene beads (dark dots)

100 Were pulled out of HL-60 cells (light gray dots) in the SAW working region and eventually collected by upper outlet channel.

101

Figure 5-5 stacked images show the separation process of polystyrene beads (15 µm, dark dots) and HL-60 cells (grey dots). They have similar densities and sizes but different compressibilities.

5.3.3 Cell separation

Leukemia model

To demonstrate the ability of our device for biological applications, we carried out separation of human leukemia cancer cells from human blood. Human red blood cells (purchased from Zen-bio) was diluted with PBS (Phosphate buffered saline) buffer by 100 times and mixed with HL-60 (human promyelocytic leukemia cells). The ratio of blood cells and HL-60 was close to 1to 1. The stacked images show the cell separation process, in which HL-60 cells were selectively moved from the rest blood cells and collected from the upper outlet channel (Fig. 5-6).

To evaluate the separation efficiency, cells were collected by each outlet channel and then

102

Figure 5-6 stacked images showing the separation process of HL-60 from human red blood cells.

Figure 5-7 Flow cytomerey scattergrams of side scatter versus forward scatter, showing the cell population of sample before (a) and after (b and c) separation.

103 characterized using commercial standard flow cytometry (Coulter FC 500). As a comparison, the mixture sample was also counted through the flow cytometry, as shown in Fig. 5-7a. The results show that 82% of the cells from upper outlet channel were HL-60 and 81% were red blood cells from the lower outlet channel, as shown in Fig. 5-7 b and c.

Circulating tumor cell model

Circulating tumor cells (CTCs) have drawn increasing research attention in recent years due to their potential value in cancer prognosis, therapy monitoring, and metastasis research. Rare

CTCs in the blood of patients with metastatic cancer is a potentially accessible source for detection, characterization, and monitoring of non-hematological cancers. The isolation of CTCs is a tremendous technical challenge due to their low concentration, as few as one cell per 109 haematological cells in blood.

To demonstrate the applicability of our device to CTC, we studied isolation of cancer cells from human blood. In the study, 1 mL human whole blood was lysed using RBC Lysis

Buffer [eBioscience], and the white blood cells (WBC) concentration was measured to be 2-

4*106/mL. This erythrocyte-lysed blood sample was then mixed with 100uL cancer cell

(6*106/mL) to achieve a cancer cell concentration of 10%. Here MCF-7 cells (human breast cancer cell line) were used as cancer cell model. The mixed sample was then delivered into a

SAW-based CTC isolation device through a syringe pump. Since cancer cells are usually much larger than white blood cells (as shown in Fig. 5-8), when the cells entered the SSAW working region, cancer cells were isolated from WBCs. The process of one CTC cell isolated from leukocytes is shown in Fig. 5-8 a. CTC cells and leukocytes are eventually collected from different outlets for consecutive characterization. Here EpCAM, CD45 surface markers (green), and a nuclear stain (DAPI, blue) are used to investigate the purity of isolated CTC. Epithelia

104

Figure 5-8 (a) stacked image showed that a single CTC cell was pulled out from the rest WBC stream. EpCAM, CD45, and DAPI are used to identify MCF-7 from leukocytes. Green dots (positive to EpCAM) are recognized as MCF-7 cells while red dots (positive to CD45) are regarded as leukocytes. (b) erythrocyte-lysed human blood sample spiked with MCF-7 cells, the original cancer cells concentration is ~5-10%. (c) Cells were collected from outlet after CTC isolation device. The results show CTC purification of as high as 98% is achieved. cancer cells such as MCF-7 are positive to EpCAM (red), negative to CD45, and positive to

DAPI (blue), while leukocytes are negative to EpCAM, positive to CD45, and positive to DAPI

(blue). To evaluate the performance of cancer cell isolation using our device, the recovery rate and purity of cancer cell isolation were investigated. The recovery rate (%) and purity (%) of cell isolation are defined as the percentage of the isolated cancer cell number over the spiked cancer cell number and that of isolated cancer cell number over the total collected cell number, respectively. We use MCF-7 cell line as CTC model, and the preliminary results indicate a purity as high as 98%, much higher than that of the current commercial approach Cellsearch (0.1%)

[162, 163], and higher than that of other state-of-art label free CTC isolation methods (80%-90%)

[164, 165], as shown in Fig. 5-8 b and c.

105 Cell viability, apoptosis, and proliferation test

Biocompatibility of our CTC isolation device is very important since further CTC cell physiological studies will be conducted after CTCs are collected. Therefore, it is required for the isolation process with very little, if any, physiological impact on the cells. To demonstrate the biocompatibility of our device, we conducted cells viability, apoptosis, and proliferation assay after exposure to SAW field at working power level (25 dbm, or 2W/cm2). WST-1 cell viability test (Roche), BrdU Cell Proliferation ELISA (Roche), and Calcein AM and SYTOX Orange

(Invitrogen) were used to test cells viability, proliferation, and apoptosis, respectively. MCF-7 cells were delivered into the separation device at a flow rate of 2 uL/min under the input power of

25 dBm (2W/cm2). Cell tests were then conducted immediately after being collected from the outlet. The results indicate that no significant changes were found in cell viability, apoptosis and proliferation. The promising results in Fig. 5-9 show that our SAW device is ideal for CTCs isolation from blood for consecutive CTCs study without affecting cell physiological properties.

106 5.4 Methods

Blood Processing and Sample Preparation

Fresh human whole blood with Acid Citrate Dextrose (ACD) as anticoagulant was purchased from Zen-bio. To lyse the red blood cells, 1 ml of whole blood was incubated with 10 ml of 1X RBC Lysis Buffer (eBioscience) for 10-15 min at room temperature followed by centrifugation at 400 x g, resuspension in PBS, and cell counting with Hemacytometer to determine white blood cell (WBC) concentration. Then cultured MCF7 breast cancer cells were spiked into the prepared WBC suspension at a desired ratio. This prepared sample was injected into our SSAW device for MCF7 separation.

CTC Staining

After separation, cells from the CTC outlet were collected and fixed with 4% paraformaldehyde (Santa Cruz Biotechnology, Inc.) for 5 min and subsequently permeabilized with 0.2% Triton X-100 (Sigma-Aldrich) in PBS. These fixed cells were then stained with DAPI

(nuclei staining), FITC-conjugated anti-CD45 antibody (WBC staining) (Invitrogen), and

Phycoerythrin (PE)-conjugated anti-EpCAM antibody (MCF7 staining) (eBioscience). The stained cells were analyzed through epifluorescence imaging.

5.5 Conclusion

In this chapter, we have presented a unique cell separation microfluidic device using standing surface acoustic wave. Particles of varying size and compressibility can be effectively and continuously separated using our SAW device. We have successfully demonstrated on-chip

107 continuous separation of 1) polystyrene beads with different size, 2) beads and cells with same size but different compressibility, 3) Leukemia cancer cells from human red blood cells, and 4)

Human breast cancer cells from Human white blood cells as CTCs model. A series of cells viability, proliferation, and apoptosis tests were performed to prove excellent biocompatibility of our method. In addition, this SAW device is simple, low cost, miniaturized, and can be fabricated via standard microfabrication, allowing the easy integration into other lab-on-chip technologies.

108 Chapter 6

Conclusions and perspectives

6.1 Achievements

SAWs have demonstrated tremendous capability in microfluidic applications, including: rapid and localized fluid mixing, droplet translation, fluid pumping in microchannels, droplet- based rotational micromotors, droplet jetting and atomization, reorientation of nano-objects, as well as particle/cell focusing, sorting, manipulation, and patterning. Some of these functions have been achieved in microfluidics using techniques other than SAW as well; however, SAW has a combination of advantages that competing techniques lack: simple fabrication, compact and inexpensive devices and accessories, high biocompatibility, fast fluid actuation, versatility, contact-free particle manipulation, and compatibility with other microfluidic components. In this dissertation, we have demonstrated that our acoustic tweezers can achieve the following functions: 1) single cell/organism manipulation; 2) tunable cell patterning; 3) multichannel cell sorting; 4) high-efficiency cell separation.

Acoustic tweezers based single cell/organism manipulation The acoustic tweezers I developed was the first acoustic manipulation method which can trap and dexterously manipulate single microparticles, cells, and entire organisms (i.e., C. elegans) along a programmed route in two- dimensions within a microfluidic chip. We demonstrate that the acoustic tweezers can move a 10-

µm single polystyrene bead to write the word “PNAS” and a bovine red blood cell to trace the letters “PSU”. It was also the first technology capable of touchless trapping and manipulating C. elegans, a one-millimeter long roundworm that is one of the most important model systems for studying diseases and development in humans.

109 Acoustic tweezers based tunable cell patterning The ordered arrangement of biological cells and microscale objects plays an important role in many biological and colloidal studies such as microarrays, regenerative medicine, and tissue engineering. I have developed an acoustic-based tunable patterning technique that can control the cell pattern with the tunability of up to 72%.

This acoustic-based tunable patterning technique has the advantages of wide tunability, non- invasiveness, and ease of integration to lab-on-a-chip systems, and shall be valuable in many biological and tissue engineering studies such as investigation of cell-cell interactions.

Acoustic tweezers based multichannel cell sorting Cell sorting is essential for many fundamental cell studies, cancer research, clinical medicine, and transplantation immunology. I developed an acoustic-based method that can precisely sort cell into five separate outlets of cells, rendering it particularly desirable for multi-type cell sorting. Our device requires small sample volumes (~100 μl), making it an ideal tool for research labs and point-of-care diagnostics.

Furthermore, it can be conveniently integrated with a small power supply, a fluorescent detection module, and a high-speed electrical feedback module to function as a fully integrated, portable, inexpensive, multi-color, miniature fluorescence-activated cell sorting (μFACS) system.

Acoustic tweezers based high-efficiency cell separation Highly efficient, on-chip cell separation techniques are fundamentally important in biological and chemical analyses such as cancer cell detection, drug screening, and tissue engineering. In particular, the ability to separate cancer cells (such as leukaemia cells) from human blood can be invaluable for cancer biology, diagnostics, and therapeutics. I have developed an acoustic tweezers-based cell separation technique that can achieve high-efficiency separation of human leukemia cells (HL-60) and human breast cancer cells (MCF-7) from human blood cells. This method is simple and versatile, capable of separating virtually all kinds of cells (regardless of charge/polarization or optical properties) with high separation efficiency and low power consummation.

110 6.2 Perspectives

We believe that these unique advantages and functions position SAW to be a key enabler for microfluidics to fulfill its promise in medical diagnostics and biological/chemical studies. For this to happen—for SAW microfluidic technologies to progress from research labs to everyday use in real-world applications—a number of hurdles must be overcome. In this section, we identify some of the areas requiring progress from both theorists and technical innovators.

The physics of SAW microfluidics, encompasses a range of phenomena, including acoustic radiation forces, acoustic streaming, jetting, and atomization. Theoretical work over the last two decades has fleshed out our understanding of this wide-ranging physics; however, work remains if our theoretical understanding is to be complete. For instance, numerical simulations

(e.g., computational fluid dynamics modeling) of SAW-induced acoustic streaming have been hampered by the large discrepancy in the time and space domains between the driving SAW and the resulting liquid streaming: a very small time step and a very fine mesh are required to capture the SAW actuation of the liquid, but the resulting streaming occurs over a relatively large time scale and large spatial dimensions. As another example, quantitative estimates of primary acoustic radiation force acting on complexly-shaped biological particles are currently inadequate, impeding the understanding and optimization of SSAW-based microfluidic platforms.

Advancements in SAW-based microfluidic devices—and their adoption in real-world applications—will be strongly facilitated by SAW researchers improving the theoretical foundations of the discipline.

While the theorists do their part, there is much room for continued device innovations. As described throughout this article, much of SAW microfluidics to date has relied on either acoustic streaming or primary acoustic radiation forces to accomplish useful actions. For particle manipulation techniques that rely on SSAW, primary acoustic radiation forces are the enabler,

111 and acoustic streaming is considered an undesirable concurrence that disrupts otherwise predictable particle movements. To improve the precision of SSAW-based particle manipulation, it would be beneficial for researchers to find ways to minimize acoustic streaming in such devices.

It should be pointed out that little research has been done on the acoustic streaming induced by

SSAW. This is an area in need of attention; perhaps through some clever innovations, researchers may even be able to turn SSAW-induced acoustic streaming from a complicating factor to an enabling mechanism.

Further improvements to the exactness of SSAW-based particle manipulation are also needed. Current acoustic tweezers devices have micrometer-scale resolution. Their resolution could potentially be refined to the nanometer-scale by using high-frequency SAWs. However, increasing SAW frequency results in higher energy attenuation as the acoustic wave transfers from the substrate surface to the contacting fluid. High SAW frequencies also result in stronger acoustic streaming, which compounds the increasing effect of streaming on particle movement as the particle size decreases. Both the increased energy attenuation and augmented acoustic streaming will need to be addressed in order to improve the resolution—and thus the applicability—of SSAW-based particle manipulation. In addition to the issue of manipulation resolution, researchers have struggled to achieve precise control of particle manipulation in the device out-of-plane (z) direction. Applications such as tissue engineering and cell-cell communications rely on careful control of cells in all three dimensions. For SSAW-based methods to become an all-around particle/cell manipulation solution, researchers will need to gain precise control of the z direction.

Other device innovations may come from use of new piezoelectric substrates. Among hundreds of piezoelectric materials, LiNbO3 has found ubiquitous use in SAW microfluidics due to its transparency, which enables easy microscope observation of on-chip events. More effort could be applied to exploring the potential benefits of other substrate materials. For example, the

112 use of phononic crystals could help better manipulate SAW propagation and amplitude, and the use of piezoelectric thin films may lead to reduced device cost. In addition to considering new piezoelectric substrate materials, researchers should expand an existing trend: the use of disposable superstrates—to which are bonded the microfluidic components of the overall device—that are acoustically coupled to the substrate via a liquid layer. Such superstrates serve to separate the expensive piezoelectric substrate and its IDTs from the microfluidic components, making a single SAW-based device reusable by exchanging the superstrates. For example, if a cell patterning and culturing microfluidic device existed on a superstrate, multiple patterned superstrate cultures could be obtained with a single piezoelectric substrate. In combination with investigation of new substrate/superstrate materials, alternate acoustic modes, such as transverse

SAW or plate waves, may also offer opportunity for new innovations.

SAW-based microfluidics has come a long way in the past decade. It has been used to successfully perform a wide range of biomedicine-oriented lab functions, from cell focusing, to cell sorting, to cell manipulation, to cell enrichment, to cell lysis, to on-chip PCR, to biosensing, and many more. In fact, these functions span the full spectrum of laboratory functions, from sample preparation all the way to final analysis; in other words, SAWs have been used to, on the aggregate, create a lab-on-a-chip. Researchers should now focus on actually creating that lab-on- a-chip. This will require integrating multiple SAW techniques onto a single microfluidic device, with transitions from one stage to another.

In order to make this adoption of SAW-based microfluidics a reality, researchers must build integrated devices capable of performing a complete conventional laboratory process, from start to finish. In this regard, facilitating the application of SAW-based microfluidics to real- world laboratory functions will take more than integrating multiple SAW techniques onto a coordinated device. It will also require creative solutions to a classic problem of microfluidic devices: the mismatch between coin-sized microchips and the large peripheral control and

113 detection systems that still reside off-chip. Researchers will need to integrate function generators, amplifiers, SAW transducers, and signal processors onto a compact system. This integration and miniaturization is doable (SAW devices and accessories have already been integrated into compact, low-cost electronic devices such as cell phones), and it is necessary if SAW-based microfluidic devices are to achieve the adoption in biological studies and medical diagnostics that researchers have long predicted.

114 References

[1] G. M. Whitesides, “The origins and the future of microfluidics” Nature, 2006, 442, 368‒373. [2] P. Neuzil, S. Giselbrecht, K. Länge, T. J. Huang and A. Manz, “Revisiting lab-on-a-chip technology for drug discovery” Nat. Rev. Drug Discov., 2012, 11, 620‒632. [3] X. Mao and T. J. Huang, “Microfluidic diagnostics for the developing world” Lab Chip, 2012, 12, 1412‒1416. [4] T. M. Squires and S. R. Quake, “Microfluidics: Fluid physics at the nanoliter scale” Rev. Mod. Phys., 2005, 77, 977‒1026. [5] A. Arora, G. Simone, G. Salieb-Beugelaar, J. T. Kim and A. Manz, “Latest Developments in Micro Total Analysis Systems” Anal. Chem., 2010, 82, 4830–4847. [6] I. A. Eydelnant, U. Uddayasankar, B. Li, M. W. Liao and A. R. Wheeler, “Virtual microwells for digital microfluidic reagent dispensing and cell culture” Lab Chip, 2012, 12, 750‒757. [7] N. Pamme and A. Manz, “On-chip free-flow magnetophoresis: continuous flow separation of magnetic particles and agglomerates” Anal. Chem., 2004, 76, 7250‒7256. [8] X. Mao, J. R. Waldeisen, B. K. Juluri and T. J. Huang, “Hydrodynamically tunable optofluidic cylindrical microlens” Lab Chip, 2007, 7, 1303–1308. [9] H. Schmidt and A. R. Hawkins, “The photonic integration of non-solid media using optofluidics ” Nat. Photon., 2011, 5, 598–604. [10] D. Psaltis, S. R. Quake and C. Yang, “Developing optofluidic technology through the fusion of microfluidics and optics” Nature, 2006, 442, 381–386. [11] C. C. W. Ruppel, L. Reindl and R. Weigel, “SAW devices and their wireless communications applications” Microwave Magazine, 2002, 3, 65‒71. [12] A. Polh, “A review of wireless SAW sensors” IEEE Trans. Ultrason., Ferroelectr., Freq. Control, 2000, 47, 317‒32. [13] T. M. a Gronewold, “Surface acoustic wave sensors in the bioanalytical field: recent trends and challenges” Anal. Chim. acta, 2007, 603, 119‒128. [14] J. Lee, Y.-S. Choi, Y. Lee, H. J. Lee, J. N. Lee, S. K. Kim, K. Y. Han, E. C. Cho, J. C. Park and S. S. Lee, “Sensitive and simultaneous detection of cardiac markers in human serum using surface acoustic wave immunosensor” Anal. Chem., 2011, 83, 8629‒8635. [15] S.-C. S. Lin, X. Mao and T. J. Huang, “Surface acoustic wave (SAW) acoustophoresis: now and beyond” Lab Chip, 2012, 12, 2766‒2770. [16] D. L. Miller, N. B. Smith, M. R. Bailey, G. J. Czarnota, K. Hynynen and I. R. S. Makin. “Overview of therapeutic ultrasound applications and safety considerations” J. Ultrasound Med., 2012, 31, 623‒634. [17] M. Wiklund, “Acoustofluidics 12: biocompatibility and cell viability in microfluidic acoustic resonators” Lab Chip, 2012, 12, 2018‒2028.

115 [18] X. Ding, S.-C. S. Lin, B. Kiraly, H. Yue, S. Li, I. K. Chiang, J. Shi, S. J. Benkovic and T. J. Huang, “On-chip manipulation of single microparticles, cells, and organisms using surface acoustic waves” Proc. Natl. Acad. Sci. U.S.A., 2012, 109, 11105‒11109. [19] J. Friend and L. Y. Yeo, “Microscale acoustofluidics: Microfluidics driven via acoustics and ultrasonics” Rev. Mod. Phys., 2011, 83, 647‒704. [20] H. Bruus, J. Dual, J. Hawkes, M. Hill, T. Laurell, J. Nilsson, S. Radel, S. Sadhal and M. Wiklund, “Forthcoming Lab on a Chip tutorial series on acoustofluidics: Acoustofluidics—exploiting ultrasonic standing wave forces and acoustic streaming in microfluidic systems for cell and particle manipulation” Lab Chip, 2011, 11, 3579‒3580. [21] A. Wixforth, “Acoustically driven programmable microfluidics for biological and chemical applications” J. Lab. Automat., 2006, 11, 399‒405. [22] Z. Wang and J. Zhe, “Recent advances in particle and droplet manipulation for lab-on-a- chip devices based on surface acoustic waves” Lab Chip, 2011, 11, 1280‒1285. [23] M. Gedge and M. Hill, “Acoustofluidics 17: Theory and applications of surface acoustic wave devices for particle manipulation” Lab Chip, 2012, 12, 2998‒3007. [24] M. K. Tan, J. R. Friend and L. Y. Yeo, “Direct visualization of surface acoustic waves along substrates using smoke particles” Appl. Phys. Lett.,, 2007, 91, 224101. [25] X. Ding, J. Shi, S.-C. S. Lin, S. Yazdi, B. Kiraly and T. J. Huang, “Tunable patterning of microparticles and cells using standing surface acoustic waves” Lab Chip, 2012, 12, 2491– 2497. [26] K. Hashimoto and M. Yamaguchi, “Excitation and propagation of shear-horizontal-type surface and bulk acoustic waves” IEEE Trans. Ultrason. Ferroelectr. Freq. Control, 2001, 48, 1181‒1188. [27] C. T. Schröder and W. R. Scott Jr., “On the complex conjugate roots of the Rayleigh equation: The leaky surface wave” J. Acoust. Soc. Am., 2001, 110, 2867‒2877. [28] E. Adler, “SAW and pseudo-SAW properties using matrix methods” IEEE Trans. Ultrason., Ferroelectr., Freq. Control, 1994, 41, 876‒882. [29] B. Auld, Acoustic Fields and Waves in Solids, Wiley Press, New York, 1973. [30] D. Köster, “Numerical simulation of acoustic streaming on surface acoustic wave-driven biochips” SIAM J. Sci. Comput., 2007, 29, 2352‒2380. [31] L. Meng, F. Cai, Q. Jin, L. Niu, C. Jiang, Z. Wang, J. Wu and H. Zheng, “Acoustic aligning and trapping of microbubbles in an enclosed PDMS microfluidic device” Sens. Actuators, B Chem., 2011, 160, 1599‒1605. [32] A. Wixforth, C. Strobl, C. Gauer, A. Toegl, J. Scriba and Z. V. Guttenberg, “Acoustic manipulation of small droplets” Anal. Bioanal. Chem., 2004, 379, 982‒991. [33] J. Vanneste and O. Buhler, “Streaming by leaky surface acoustic waves” Proc. R. Soc. London, Ser. A Phys., 2011, 467, 1779‒1800. [34] M. Alghane, B. X. Chen, Y. Q. Fu, Y. Li, J. K. Luo and A. J. Walton, “Experimental and numerical investigation of acoustic streaming excited by using a surface acoustic wave device on a 128° YX-LiNbO3 substrate” J. Micromech. Microeng., 2011, 21, 015005.

116 [35] X. Y. Du, Y. Q. Fu, J. K. Luo, A. J. Flewitt and W. I. Milne, “Microfluidic pumps employing surface acoustic waves generated in ZnO thin films” J. Appl. Phys., 2009, 105, 024508. [36] W.-K. Tseng, J.-L. Lin, W.-C. Sung, S.-H. Chen and G.-B. Lee, “Active micro-mixers using surface acoustic waves on Y-cut 128 LiNbO3” J. Micromech. Microeng., 2006, 16, 539‒548. [37] T. Frommelt, D. Gogel, M. Kostur, P. Talkner, P. Hänggi and A. Wixforth, “Flow patterns and transport in Rayleigh surface acoustic wave streaming: combined finite element method and raytracing numerics versus experiments” IEEE Trans. Ultrason., Ferroelectr., Freq. Control, 2008, 55, 2298‒2305. [38] R. Shilton, M. K. Tan, L. Y. Yeo and J. R. Friend, “Particle concentration and mixing in microdrops driven by focused surface acoustic waves” J. Appl. Phys., 2008, 104, 014910. [39] D. Köster, PhD Thesis, Augsburg University, 2006. [40] W. L. Nyborg, in Acoustic Streaming, ed. W. P. Mason and R. N. Thurston , Academic Press, New York, 1965, 11, 265‒329. [41] S. S. Sadhal, “Acoustofluidics 13: Analysis of acoustic streaming by perturbation methods” Lab Chip, 2012, 12, 2292‒2300. [42] D. Hartono, Y. Liu, P. L. Tan, X. Y. S. Then, L.-Y. L. Yung and K.-M. Lim, “On-chip measurements of cell compressibility via acoustic radiation” Lab Chip, 2011, 11, 4072‒4080.. [43] J. Shi, S. Yazdi, S.-C. S. Lin, X. Ding, I.-K. Chiang, K. Sharp and T. J. Huang, “Three- dimensional continuous particle focusing in a microfluidic channel via standing surface acoustic waves (SSAW)” Lab Chip, 2011, 11, 2319‒2324. [44] J. Shi, X. Mao, D. Ahmed, A. Colletti and T. J. Huang, “Focusing microparticles in a microfluidic channel with standing surface acoustic waves (SSAW)” Lab Chip, 2008, 8, 221‒223. [45] M. Alvarez, J. R. Friend and L. Y. Yeo, “Surface vibration induced spatial ordering of periodic polymer patterns on a substrate” Langmuir, 2008, 24, 10629‒10632. [46] J. Shi, D. Ahmed, X. Mao, S.-C. S. Lin, A. Lawit and T. J. Huang, “Acoustic tweezers: patterning cells and microparticles using standing surface acoustic waves (SSAW)” Lab Chip, 2009, 9, 2890‒2895. [47] C. D. Wood, S. D. Evans, J. E. Cunningham, R. O’Rorke, C. Wälti and A. G. Davies, “Alignment of particles in microfluidic systems using standing surface acoustic waves” Appl. Phys. Lett., 2008, 92, 044104. [48] L. Johansson, J. Enlund, S. Johansson, I. Katardjiev and V. Yantchev, “Surface acoustic wave induced particle manipulation in a PDMS channel—principle concepts for continuous flow applications” Biomed. microdevices, 2012, 14, 1‒11. [49] L. Johansson, J. Enlund, S. Johansson, I. Katardjiev, M. Wiklund and V. Yantchev, “Surface acoustic wave-induced precise particle manipulation in a trapezoidal glass microfluidic channel” J. Micromech. Microeng., 2012, 22, 025018.

117 [50] T. Frommelt, M. Kostur, M. Wenzel-Schäfer, P. Talkner, P. Hänggi and A. Wixforth. “Microfluidic mixing via acoustically driven chaotic advection” Phys. Rev. Lett., 2008, 100, 034502. [51] A. R. Rezk, A. Qi, J. R. Friend, W. H. Li and L. Y. Yeo, “Uniform mixing in paper-based microfluidic systems using surface acoustic waves” Lab Chip, 2012, 12, 773‒779. [52] A. Wixforth, “Acoustically driven planar microfluidics” Superlattic Microst., 2003, 33, 389‒396. [53] Z. Guttenberg, H. Muller, H. Habermüller, A. Geisbauer, J. Pipper, J. Felbel, M. Kielpinski, J. Scriba and A. Wixforth, “Planar chip device for PCR and hybridization with surface acoustic wave pump” Lab Chip, 2005, 5, 308‒317. [54] A. R. Rezk, O. Manor, J. R. Friend and L. Y. Yeo, “Unique fingering instabilities and soliton-like wave propagation in thin acoustowetting films” Nat. Commum., 2012, 3, 1167. [55] S. W. Schneider, S. Nuschele, a Wixforth, C. Gorzelanny, a Alexander-Katz, R. R. Netz and M. F. Schneider, “Shear-induced unfolding triggers adhesion of von Willebrand factor fibers” Proc. Natl. Acad. Sci. U.S.A., 2007, 104, 7899‒7903. [56] M. Cecchini, S. Girardo, D. Pisignano, R. Cingolani and F. Beltram, “Acoustic- counterflow microfluidics by surface acoustic waves” Appl. Phys. Lett., 2008, 92, 104103. [57] S. Girardo, M. Cecchini, F. Beltram, R. Cingolani and D. Pisignano, “Polydimethylsiloxane–LiNbO3 surface acoustic wave micropump devices for fluid control into microchannels” Lab Chip, 2008, 8, 1557‒1563. [58] M. A. Fallah, V. M. Myles, T. Krüger, K. Sritharan, A. Wixforth, F. Varnik, S. W. Schneider and M. F. Schneider, “Acoustic driven flow and lattice Boltzmann simulations to study cell adhesion in biofunctionalized μ-fluidic channels with complex geometry” Biomicrofluidics, 2010, 4, 1‒10. [59] C. Fillafer, G. Ratzinger, J. Neumann, Z. Guttenberg, S. Dissauer, I. K. Lichtscheidl, M. Wirth, F. Gabor and M. F. Schneider, “An acoustically-driven biochip–impact of flow on the cell-association of targeted drug carriers” Lab Chip, 2009, 9, 2782‒2788. [60] R. J. Shilton, N. R. Glass, P. Chan, L. Y. Yeo and J. R. Friend, “Rotational microfluidic motor for on-chip microcentrifugation” Appl. Phys. Lett., 2011, 98, 254103. [61] R. J. Shilton, S. M. Langelier, J. R. Friend and L. Y. Yeo, “Surface acoustic wave solid- state rotational micromotor” Appl. Phys. Lett., 2012, 100, 033503. [62] B.-J. de Gans, P. C. Duineveld and U. S. Schubert, “Inkjet printing of polymers: state of the art and future developments” Adv. Mater., 2004, 16, 203‒213. [63] M. K. Tan, J. R. Friend and L. Y. Yeo, “Interfacial jetting phenomena induced by focused surface vibrations” Phys. Rev. Lett., 2009, 103, 024501. [64] J. Eggers, “Nonlinear dynamics and breakup of free-surface flows” Rev. Mod. Phys., 1997, 69, 865‒929. [65] A. Lefebvre, Atomization and Sprays, CRC Press, New York, 1989. [66] M. Kurosawa, T. Watanabe, A. Futami and T. Higuchi, “Surface acoustic wave atomizer” Sens. Actuators. A Phys., 1995, 50, 69‒74.

118 [67] A. Qi, L. Y. Yeo, J. R. Friend and J. Ho, “The extraction of liquid, protein molecules and yeast cells from paper through surface acoustic wave atomization” Lab Chip, 2010, 10, 470–476. [68] A. Qi, J. R. Friend, L. Y. Yeo, D. A. V. Morton, M. P. McIntosh and L. Spiccia, “Miniature inhalation therapy platform using surface acoustic wave microfluidic atomization” Lab Chip, 2009, 9, 2184–2193. [69] S. H. Yoon, Y. Huang, J. S. Edgar, Y. S. Ting, S. R. Heron, Y. Kao, Y. Li, C. D. Masselon, R. K. Ernst and D. R. Goodlett, “Surface acoustic wave nebulization facilitating lipid mass spectrometric analysis” Anal. Chem., 2012, 84, 6530–6537. [70] M. Alvarez, J. Friend and L. Y. Yeo, “Rapid generation of protein aerosols and nanoparticles via surface acoustic wave atomization” Nanotechnology, 2008, 19, 455103. [71] O. V. Abramov, High-Intensity Ultrasonics, Gordon and Breach Science Publishers, Amsterdam, 1998. [72] D. J. Collins, O. Manor, A.Winkler, H. Schmidt, J. R. Friend and L. Y. Yeo, “Atomization off thin water films generated by high-frequency substrate wave vibrations” Phys. Rev. E, 2012, 86, 056312. [73] T. Franke, A. R. Abate, D. A. Weitz and A. Wixforth, “Surface acoustic wave (SAW) directed droplet flow in microfluidics for PDMS devices” Lab Chip, 2009, 9, 2625‒2627. [74] T. Franke, S. Braunmüller, L. Schmid, A. Wixforth and D. A. Weitz, “Surface acoustic wave actuated cell sorting (SAWACS)” Lab Chip, 2010, 10, 789‒794. [75] R. H. Baughman, A. A. Zakhidov and W. A. de Heer. “Carbon nanotubes--the route toward applications” Science, 2002, 297, 787‒792. [76] C. J. Strobl, C. Schäflein, U. Beierlein, J. Ebbecke and A. Wixforth, “Carbon nanotube alignment by surface acoustic waves” Appl. Phys. Lett., 2004, 85, 1427‒1429. [77] T. Smorodin, U. Beierlein, J. Ebbecke and A. Wixforth, “Surface‐Acoustic‐Wave‐Enhanced Alignment of Thiolated Carbon Nanotubes on Gold Electrodes” Small, 2005, 1, 1188‒1190. [78] K. M. Seemann, J. Ebbecke and A. Wixforth, “Alignment of carbon nanotubes on pre- structured silicon by surface acoustic waves” Nanotechnology, 2006, 17, 4529‒4532. [79] Y. J. Liu, X. Ding, S.-C. S. Lin, J. Shi, I.-K. Chiang and T. J. Huang, “Surface Acoustic Wave Driven Light Shutters Using Polymer‐Dispersed Liquid Crystals” Adv. Mater., 2011, 23, 1656‒1659. [80] R. Wilson, J. Reboud, Y. Bourquin, S. L. Neale, Y. Zhang and J. Cooper, “Phononic crystal structures for acoustically driven microfluidic manipulations” Lab Chip, 2010, 11, 323‒328. [81] J. Reboud, Y. Bourquin, R. Wilson, G. S. Pall, M. Jiwaji, A. R. Pitt, A. Graham, A. P. Waters and J. M. Cooper, “Shaping acoustic fields as a toolset for microfluidic manipulations in diagnostic technologies” Proc. Natl. Acad. Sci. U.S.A., 2012, 109, 15162‒15167. [82] Y. Bourquin, R. Wilson, Y. Zhang, J. Reboud and J. M. Cooper, “Phononic crystals for shaping fluids” Adv. Mater., 2011, 23, 1458‒1462.

119 [83] X. Mao, S.-C. S. Lin, C. Dong and T. J. Huang, “Single-layer planar on-chip flow cytometer using microfluidic drifting based three-dimensional (3D) hydrodynamic focusing” Lab Chip, 2009, 9, 1583‒1589. [84] J. Shi, H. Huang , Z. Stratton, Y. Huang and T. J. Huang, “Continuous particle separation in a microfluidic channel via standing surface acoustic waves (SSAW)” Lab Chip, 2009, 9, 3354‒3359. [85] J. Nam, Y. Lee and S. Shin, “Size-dependent microparticles separation through standing surface acoustic waves” Microfluid. Nanofluid., 2011, 11, 317–326. [86] J. Nam, H. Lim, C. Kim, J. Y. Kang and S. Shin, “Density-dependent separation of encapsulated cells in a microfluidic channel by using a standing surface acoustic wave” Biomicrofluidics, 2012, 6, 024120. [87] M. C. Jo and R. Guldiken, “Active Density-based Separation using Standing Surface Acoustic Waves” Sens. Actuators A Phys., 2012, 187, 22–28. [88] R. P. Richter, R. Bérat and A. R. Brisson. “Formation of solid-supported lipid bilayers: an integrated view” Langmuir, 2006, 22, 3497‒3505. [89] M. Hennig, J. Neumann, A. Wixforth, J. O. Radler and M. F. Schneider, “Dynamic patterns in a supported lipid bilayer driven by standing surface acoustic waves” Lab Chip, 2009, 9, 3050–3053. [90] M. Hennig, M. Wolff, J. Neumann, A. Wixforth, M. F. Schneider and J. O. Radler, “DNA Concentration Modulation on Supported Lipid Bilayers Switched by Surface Acoustic Waves” Langmuir, 2011, 27, 14721–14725. [91] J. Neumann, M. Hennig, A. Wixforth, S. Manus, J. O. Radler and M. F. Schneider, “Transport, separation, and accumulation of proteins on supported lipid bilayers” Nano Lett., 2010, 10, 2903‒2908. [92] E. E. Hui, S. N. Bhatia, “Micromechanical control of cell-cell interactions” Proc. Natl. Acad. Sci. USA 2007, 104, 5722−6. [93] N. Chronis, M. Zimmer, C. I. Bargmann, “Microfluidics for in vivo imaging of neuronal and behavioral activity in Caenorhabditis elegans” Nature Methods 2007, 4, 727−731. [94] C. B. Rohde, F. Zeng, R. Gonzalez-Rubio, M. Angel, M. F. Yanik, “Microfluidic system for on-chip high-throughput whole-animal sorting and screening at subcellular resolution” Proc. Natl. Acad. Sci. USA. 2007, 104, 13891−5. [95] A. Ashkin, J. M. Dziedzic, J. E. Bjorkholm, S. Chu, “Observation of a single-beam gradient force optical trap for dielectric particles” Opt. Lett. 1986, 11, 288−290. [96] D. G. Grier “A revolution in optical manipulation” Nature, 2003, 424, 810−816. [97] X. Zhang, K. Halvorsen, C. Z. Zhang, W. P. Wong, T. A. Springer, “Mechanoenzymatic cleavage of the ultralarge vascular protein von Willebrand factor” Science, 2009, 324, 1330−1334. [98] A. H. J. Yang et al. “Optical manipulation of nanoparticles and biomolecules in sub- wavelength slot waveguides” Nature, 2009, 457, 71−75.

120 [99] M. B. Rasmussen, L.B. Oddershede, H. Siegumfeldt “Optical tweezers cause physiological damage to Escherichia coli and Listeria bacteria” Appl. Environ. Microbiol., 2008, 74, 2441−2446. [100] G. B. Salieb-Beugelaar, G. Simone, A. Arora, A. Philippi, A. Manz, “Latest developments in microfluidic cell biology and analysis systems” Anal. chem., 2010, 82, 4848−4864. [101] M. Krishnan, N. Mojarad, P. Kukura, V. Sandoghdar, “Geometry-induced electrostatic trapping of nanometric objects in a fluid” Nature, 2010, 467, 692−695. [102] S. Park, Y. Zhang, T. H. Wang, S. Yang, “Continuous dielectrophoretic bacterial separation and concentration from physiological media of high conductivity” Lab Chip, 2011, 11, 2893−2900. [103] S. Takayama et al. “Laminar flows: subcellular positioning of small molecules” Nature, 2001, 411, 1016. [104] Y. J. Liu, X. Ding, S. S. Lin, J. Shi, I.-K. Chiang, T. J. Huang, “Surface acoustic wave driven light shutters using polymer-dispersed liquid crystals” Adv. Mater., 2011, 23, 1656−1659. [105] J. Shi, D. Ahmed, X. Mao, S. S. Lin, T. J. Huang, “Acoustic tweezers: patterning cells and microparticles using standing surface acoustic waves (SSAW)” Lab Chip, 2009, 9, 2890−2895. [106] S. M. Woodside, B. D. Bowen, J. M. Piret, “Measurement of ultrasonic forces for particle- liquid separations” AIChE. Journal, 1997, 43, 1727−1736. [107] K. Wang, E. Schonbrun, P. Steinvurzel, K. B. Crozier, “Trapping and rotating nanoparticles using a plasmonic nano-tweezer with an integrated heat sink” Nature Commun., 2011, 2, 469. [108] A. Snezhko, I. S. Aranson, “Magnetic manipulation of self-assembled colloidal asters” Nature Mater., 2011, 10, 698–703. [109] M. Manosas, M. Spiering, Z. Zhuang, S. J. Benkovic, V. Croquette, “Coupling DNA unwinding activity with primer synthesis in the bacteriophage T4 primosome” Nature Chem. Biol., 2009, 5, 904−912. [110] A. Khademhosseini, M. H. May, M. V. Sefton, “Conformal coating of mammalian cells immobilized onto magnetically driven beads” Tissue Eng., 2005, 11, 1797−1806. [111] P.-Y. Chiou, A. T. Ohta, M. C. Wu, “Massively parallel manipulation of single cells and microparticles using optical images” Nature, 2005, 436, 370−372. [112] A. Jamshidi, P. J. Pauzauskie, P. J. Schuck, A. T. Ohta, P.-Y. Chiou, J. Chou, P. Yang, M. C. Wu, “Dynamic manipulation and separation of individual semiconducting and metallic nanowires” Nature Photon., 2008, 2, 86−89. [113] H.-Y. Hsu, A. T. Ohta, P.-Y. Chiou, A. Jamshidi, S. L. Neale, M. C. Wu, “Phototransistor- based optoelectronic tweezers for dynamic cell manipulation in cell culture media” Lab Chip, 2010, 10, 165−172. [114] J. T. J. Voldman, “Electrical forces for microscale cell manipulation” Annu. Rev. Biomed. Eng., 2006, 8, 425−454.

121 [115] K. Chung, M. M. Crane, H. Lu, “Automated on-chip rapid microscopy, phenotyping and sorting of C. elegans” Nature Methods, 2008, 5, 637−643. [116] C. K. Campbell, “Surface Acoustic Wave Devices for Mobile and Wireless Communications” (Academic Press, Orlando), 1998, pp 217−220. [117] H. Bruus “Acoustofluidics 1: Governing equations in microfluidics” Lab Chip, 2011, 11, 3742−51. [118] A. Lenshof, M. Evander, T. Laurell, J. Nilsson, “Acoustofluidics 5: Building microfluidic acoustic resonators” Lab chip, 2012, 12, 684−95. [119] J. Dual, P. Hahn, I. Leibacher, D. Möller, T. Schwarz, “Acoustofluidics 6: Experimental characterization of ultrasonic particle manipulation devices” Lab chip, 2012, 12, 852−862. [120] H. Bruus, “Acoustofluidics 7: The acoustic radiation force on small particles” Lab chip, 2012, 12, 1014−21. [121] P. Glynne-Jones, R. J. Boltryk, M. Hill, “Acoustofluidics 9: Modelling and applications of planar resonant devices for acoustic particle manipulation” Lab chip, 2012, 12, 1417−1426. [122] K. C. Neuman, A. Nagy, “Single-molecule force spectroscopy: optical tweezers, magnetic tweezers and atomic force microscopy” Nature Methods, 2008, 5, 491−505. [123] M. Eisenstein, “Cell sorting: divide and conquer” Nature 2006, 441, 1179–1185. [124] H. M. Shapiro, Practical flow cytometry, Wiley-Liss, New York, 2003. [125] J. F. Leary, “Ultra high-speed sorting” Cytometry 2005, 67A, 76–85. [126] F. A. Gomez, Biological Applications of Microfluidics, 1st ed. John Wiley & Sons, Hoboken, NJ, 2008. [127] http://www.bio.davidson.edu/courses/genomics/method/facs.html. [128] C. H. Chen, S. H. Cho, H.-I Chiang, F. Tsai, K. Zhang, and Y.-H. Lo. “Specific sorting of single bacterial cells with microfabricated fluorescence-activated cell sorting and tyramide signal amplification fluorescence in situ hybridization” Anal. Chem., 2011, 83, 7269–7275. [129] X. Wang, S. Chen, M. Kong, Z. Wang, K. D. Costa, R. a Li, and D. Sun, “Enhanced cell sorting and manipulation with combined optical tweezer and microfluidic chip technologies” Lab Chip, 2011, 11, 3656–3662. [130] T.-H. Wu, Y. Chen, S.-Y. Park, J. Hong, T. Teslaa, J. F. Zhong, D. D. Carlo, M. a Teitell, and P.-Y. Chiou, “Pulsed laser triggered high speed microfluidic fluorescence activated cell sorter” Lab Chip, 2012, 12, 1378–1383. [131] M. M. Wang, E. Tu, D. E. Raymond, J. M. Yang, H. Zhang, N. Hagen, B. Dees, E. M. Mercer, A. H. Forster, I. Kariv, P. J. Marchand, and W. F. Butler, “Microfluidic sorting of mammalian cells by optical force switching.” Nature biotechnology, 2005, 23, 83-7. [132] D. Gresham, M. J. Dunham and D. Botstein, “Comparing whole genomes using DNA microarrays.” Nat. Rev. Genet., 2008, 9, 291–302. [133] S. R. Khetani and S. N. Bhatia, “Microscale culture of human liver cells for drug development.” Nat. Biotechnol., 2007, 26, 120–126.

122 [134] A. I. Hochbaum and J. Aizenberg, “Bacteria Pattern Spontaneously on Periodic Nanostructure Arrays” Nano Lett., 2010, 10, 3717–3721. [135] M. Vilfan, A. Potocnik, B. Kavcic, N. Osterman, I. Poberaj, A. Vilfan and D Babic, “Self- assembled artificial cilia” Proc. Natl. Acad. Sci. U S A, 2010, 107, 1844–1847. [136] Dirk R Albrecht1,2, Gregory H Underhill1,2, Travis BWassermann1, Robert L Sah1 & Sangeeta N Bhatia, “Probing the role of multicellular organization in three-dimensional microenvironments” Nature Methods, 2006, 3, 369-375 [137] C. D. Wood, J. E. Cunningham, R. O’ Rorke, C. Walti, E. H. Linfield, A. G. Davies and S. D. Evans, “Formation and manipulation of two-dimensional arrays of micron-scale particles in microfluidic systems by surface acoustic waves” Appl. Phys. Lett., 2009, 94, 054101. [138] Y. Bourquin, J. Reboud, R. Wilson and J. M. Cooper, “Tuneable surface acoustic waves for fluid and particle manipulations on disposable chips” Lab Chip, 2010, 7, 221–223. [139] H. Yatsuda, K. Noguchi and K. Yamanouchi, “Surface Acoustic Wave Filter Using Floating Electrode Slanted Finger Unidirectional Transducers” Jpn. J. Appl. Phys., 2000, 39, 3041. [140] M. Goto, H. Yatsuda and T. Chiba, “Power Flow Angles for Slanted Finger Surface Acoustic Wave Filters on Langasite Substrate” Jpn. J. Appl. Phys., 2007, 46, 4744–4748. [141] T. Wu and I. Chang, “Actuating and detecting of microdroplet using slanted finger interdigital transducers” J. Appl. Phys., 2005, 98, 024903. [142] E. Papadakis, Ultrasonic Instruments and Devices, Academic Press, 1999, 580. [143] Bu G, Ciplys D, Shur M, Schowalter LJ, Schujman S, Gaska R., “Surface acoustic wave velocity in single-crystal AlN substrates” IEEE trans. ultrason. Ferroelectr. freq. control, 2006, 53, 251-254. [144] K. Yosioka and Y. Kawasima, “Acoustic radiation pressure on a complete sphere” Acustica, 1955, 5, 167–173. [145] H. Kress, J. Park, C. O Mejean, J. D Forster, J. Park, S. S Walse, Y. Zhang, D. Wu, O. D Weiner, T. M Fahmy & E. R Dufresne, “Cell stimulation with optically manipulated microsources” nature methods, 2009, 6, 905–909. [146] E. E. Hui & S. N. Bhatia, “Micromechanical control of cell–cell interactions” Proc Natl Acad Sci USA 2007, 102, 5722–5726. [147] C. W. Yung, J. Fiering, A. J. Mueller and D. E. Ingber, “Micromagnetic –microfluidic blood cleansin g device”, Lab Chip , 2009, 9, 1171– 1177. [148] O. Lara, X. Tong, M. Zborowski and J. J. Chalmers, “Enrichment of rare cancer cells through depletion of normal cells using density and flow-through, immunomagnetic cell separation”, Exp. Hematol., 2004, 32, 891–904. [149] N. Ye, J. Qin, W. Shi, X. Liu and B. Lin, “Cell-Based High Content. Screening Using an Integrated Microfluidic Device” Lab Chip, 2007, 7, 1696–1704. [150] P. O. Krutzik, J. M. Crane, M. R. Clutter and G. P. Nolan, “High-content single -cell drug screening with phosphospecific flow cytometry” Nature Chem. Biol., 2008, 4, 132– 142.

123 [151] J. Seo, M. H. Lean and A. Kole, “Membrane -free microfiltration by asymmetric inertial migration” Appl. Phys. Lett., 2007, 91, 033901. [152] A. J. Mach, J. H. Kim, A. Arshi, S. C. Hur, and D. Di Carlo, “Automated cellular sample preparation using a Centrifuge-on-a-Chip.” Lab on a chip, 2011, 11, 2827−34. [153] K. McCloskey, J. Chalmers and M. Zborowski, “Magnetic cell separation: characterization of magnetophoretic mobility” Anal. Chem., 2003, 75, 6868–6874. [154] D. Di Carlo, D. Irimia, R. Tompkins, and M. Toner, “Continuous inertial focusing, ordering, and separation of particles in microchannels,” Proc. Natl. Acad. Sci. U.S.A. 2007, 104, 18892. [155] L. R. Huang, E. C. Cox, R. H. Austin, J. C. Sturm, “Continuous particle separation through deterministic lateral displacement ” Science, 2004, 304, 987–990. [156] K. K. Zeming, S. Ranjan, and Y. Zhang, “Rotational separation of non-spherical bioparticles using I-shaped pillar arrays in a microfluidic device.” Nature communications, 2013, 4, 1625. [157] M. P. MacDonald, G. C. Spalding, and K. Dholakia, “Microfluidic sorting in an optical lattice” Nature, 2003, 426, 421−424. [158] N. Lewpiriyawong, K. Kandaswamy, C. Yang, V. Ivanov, and R. Stocker, “Microfluidic characterization and continuous separation of cells and particles using conducting poly(dimethyl siloxane) electrode induced alternating current-dielectrophoresis.” Anal. Chem., 2011, 83, 9579−85. [159] S.-B. Huang, M.-H. Wu, Y.-H. Lin, C.-H. Hsieh, C.-L. Yang, H.-C. Lin, C.-P. Tseng, and G.-B. Lee, “High-purity and label-free isolation of circulating tumor cells (CTCs) in a microfluidic platform by using optically-induced-dielectrophoretic (ODEP) force.” Lab chip, 2013, 13, 1371−83. [160] J. Sun, Y. Gao, R. J. Isaacs, K. C. Boelte, P. C. Lin, E. M. Boczko, and D. Li, “Simultaneous On-Chip DC Dielectrophoretic Cell Separation and Quantitative Separation Performance Characterization.” Anal. Chem., 2012, 84, 2017−24. [161] A. Lenshof and T. Laurell, “Continuous separation of cells and particles in microfluidic systems.” Chemical Society reviews, 2010, 39, 1203−17. [162] S. Maheswaran and D. A. Haber, “Circulating tumor cells: a window into cancer biology and metastasis” Current Opinion in Genetics & Development, 2010, 20, 96−99. [163] W. J. Allard, J. Matera, M. C. Miller, M. Repollet, M. C. Connelly, C. Rao, A. G. J. Tibbe, J. W. Uhr and L. W. M. M. Terstappen, “Tumor cells circulate in the peripheral blood of all major carcinomas but not in healthy subjects or patients with nonmalignant diseases” Clinical Cancer Research, 2004, 10, 6897−6904. [164] Igor Cima, ChayWen Yee, Florina S. Iliescu, Wai Min Phyo, Kiat Hon Lim, Ciprian Iliescu, and Min Han Tan, “Label-free isolation of circulating tumor cells in microfluidic devices: Current research and perspectives” Biomicrofluidics, 2013, 7, 011810. [165] S. K. Arya, , B. Lim, , & A. R. A. Rahman, “Enrichment, detection and clinical significance of circulating tumor cells.” Lab Chip, 2013, 13, 1995−2027.

124 Appendix

Highlights and medium reports

Acoustic tweezers based single cell/organism manipulation The acoustic tweezers we developed can trap and dexterously manipulate single microparticles, cells, and entire organisms

(i.e., C. elegans) along a programmed route in two-dimensions within a microfluidic chip. Due to its significance, this work was highlighted at National Science Foundation, National Institutes of

Health, Nature Methods, Lab on a Chip, and more than 300 public media such as MSNBC, Live

Science, Popular Science, Science Daily, Medical News Today, Nanowerk, SciTechDaily, Bio-

Medicine, and R&D Magazine.

Acoustic tweezers based tunable cell patterning

This acoustic-based tunable patterning technique we developed has the advantages of wide tunability, non- invasiveness, and ease of integration to lab-on-a-chip systems. Due to its significance, this work was selected as a cover article at the journal Lab on a Chip (Figure to the right).

125

Acoustic tweezers based multichannel cell sorting We developed an acoustic-based method that can precisely sort cell into five separate outlets of cells. It can be conveniently integrated with a small power supply, a fluorescent detection module, and a high-speed electrical feedback module to function as a fully integrated, portable, inexpensive, multi-color, miniature fluorescence-activated cell sorting (μFACS) system. Due to its significance, this work was selected as front cover article (Figure to the right) at the journal Lab on a Chip, and was highlighted at National Science Foundation and many public media such as Science Daily, The

Engineer, Device Space, UPI, Futurity, Human Health & Science, R&D Magazine, Quantum

Times, PhysOrg, and Laboratory Equipment.

VITA

Xiaoyun Ding

Education Ph.D., Engineering Science and Mechanics, The Pennsylvania State University, USA (2013) M.S., Microelectronics Engineering, Chinese Academy of Sciences, China (2009) B.S., Material Physics, Fudan University, China (2006) Honors and Awards Rustum and Della Roy Innovation in Materials Research Award, The Pennsylvania State University (2013) JALA Ten Honorees, Journal of Laboratory Automation (2012) Student Paper Competition Award, IEEE IUS2012, Dresden, Germany (2012) The Baxter Young Investigator Award, Chicago, USA (2012) Student Travel Award, IEEE International Ultrasonics Symposium, Dresden, Germany (2012) First place, Penn State ESM Today Graduate Research Symposium (2012) Innovation Award, Penn State ESM Today Graduate Research Symposium (2011) Selected Journal Publications 1. Xiaoyun Ding, Sz-Chin Steven Lin, Brian Kiraly, Hongjun Yue, Sixing Li, Jinjie Shi, Stephen J. Benkovic, and Tony Jun Huang, On-Chip Manipulation of Single Microparticles, Cells, and Organisms Using Surface Acoustic Waves, Proceedings of the National Academy of Sciences of the United States of America (PNAS), 2012, 109, 11105-11109. 2. Xiaoyun Ding, Sz-Chin Steven Lin, Michael Ian Lapsley, Sixing Li, Xiang Guo, Chung Yu Chan, I-Kao Chiang, J. Philip McCoy, and Tony Jun Huang, Standing surface acoustic wave (SSAW) based multichannel cell sorting, Lab on a Chip, 2012, 12, 4228–4231. (Cover image) 3. Xiaoyun Ding, Jinjie Shi, Sz-Chin Steven Lin, Shahrzad Yazdi, Brian Kiraly, and Tony Jun Huang, Tunable Patterning of Microparticles and Cells using Standing Surface Acoustic Waves, Lab on a Chip, 2012, 12, 2491-2497. (Cover image) 4. Xiaoyun Ding, Peng Li, Sz-Chin Steven Lin, Zackary S. Stratton, Nitesh Nama, Feng Guo, Daniel Slotcavage, Xiaole Mao, Francesco Costanzo, and Tony Jun Huang, Surface acoustic wave (SAW) microfluidics, Lab on a Chip, in press. 5. Sixing Li, Xiaoyun Ding, Feng Guo, Yuchao Chen, Michael Ian Lapsley, Sz-Chin Steven Lin, and Tony Jun Huang, “Microfluidic Droplet Sorting via Standing Surface Acoustic Wave”, Anal. Chem., in press. 6. Yuchao Chen, Xiaoyun Ding, Sz-Chin Steven Lin, and Tony Jun Huang, Tunable Nanowire Patterning Using Standing Surface Acoustic Waves (SSAW), ACS Nano, 2013, 7, 3306-3314. 7. Yan Jun Liu, Xiaoyun Ding, Sz-Chin Steven Lin, Jinjie Shi, I-Kao Chiang, and Tony Jun Huang, Surface Acoustic Wave Driven Light Shutters Using Polymer-Dispersed Liquid Crystals, Advanced Materials, 2011, 23, 1656-1659. (Front cover image) 8. Xiaoyun Ding, Fei Geng, and Le Luo, Process development in metal/BCB multilayer interconnections of MMCM with embedded chip in Si substrate, Microelectronic Engineering, 2009, 86 (3), 335-339. 9. Xiaoyun Ding, Fei Geng, and Le Luo, Wideband vertical transitions for multilayer BCB‐ based MCM, Microwave and Optical Technology Letters, 2009, 51 (6), 1584-1587.