Manipulations of micro/nanoparticles using gigahertz acoustic streaming tweezers

Efficient manipulation of bioparticles, cells, and biomolecules is of fundamental significance in many areas of modern biotechnology,¹ such as biosensing,² cell separation,³ therapeutics,⁴ and drug screening.⁵ In the past few decades, many manipulation techniques have been developed, including electrokinetic/dielectrophoresis,^6,7 optofluidics,^8–12 magnetophoresis,^13,14 hydrodynamic manipulation,^15–18 and acoustophoresis.^19,20 In particular, acoustophoresis has become increasingly popular owing to its lack of a need for labels, its biocompatibility, its independence of the medium, and its relatively easy integration with other microfluidic platforms. Theoretical and experimental studies of acoustic manipulations indicate that a wide range of particles can be separated based on their sizes, shapes, densities, and compressibilities.²¹ Therefore, acoustophoresis has been applied to many size-based manipulation techniques, such as separation,^22,23 concentration,²⁴ patterning,²⁵ and mixing of bioparticles.²⁶ 

Despite the success of acoustic systems in manipulating microscale objects, the acoustic manipulation of objects at the submicrometer and nanometer scales is still challenging. In particle manipulation by acoustic methods, two competing forces are involved: the acoustic radiation force (ARF)^27,28 and the acoustic streaming drag force (ASF).^29–31 The ARF pushes objects along the propagation direction of the bulk acoustic wave. For example, a standing wave helps to trap and enrich the particles to their force-equilibrium positions (e.g., pressure nodes). On the other hand, the ASF generated by acoustic streaming drags objects along the fluid flow, which may disturb the trapping stability.³⁰ The ARF is proportional to the cube of the particle radius, while the ASF is proportional to this radius. Therefore, as the particle radius decreases, the ARF decreases more rapidly than the ASF. At the submicrometer scale, the ASF starts to play the dominant role in particle motion. This is the main reason why the conventional acoustic manipulation method is limited to the micrometer scale (i.e., >500 nm).³² 

As many biosystems function at the submicrometer and nanometer scales (e.g., extracellular vesicles, exosomes, and proteins), there is an urgent need to develop acoustic tools at these scales. Recently, some reports have described a new concept of acoustic trapping in which the ARF and ASF are combined, with the two forces working together to realize manipulations of extremely small objects. Mao et al.³³ generated single-vortex acoustic streaming assisted by a glass capillary and realized the trapping of nanoparticles ranging in size from 80 to 500 nm. Collins et al.³⁴ produced a focused and high-frequency surface acoustic wave (SAW) that generated strong double vortices in a continuous flow system, enabling size-selective trapping and separation of nanoparticles down to the 100 nm scale.

In our group, we have been working on acoustic manipulation using microelectromechanical systems (MEMS)-fabricated bulk acoustic wave (BAW) devices that resonate above the gigahertz (GHz) frequency.³⁵ The use of a GHz frequency increased the strength of the body force generated in the fluid driving the particles in the system, while the short attenuation length of the acoustic wave dissipated the acoustic energy quickly, thus preventing the formation of acoustic standing waves. Through integration of this device into microfluidic channels, confined high-speed microvortices were induced by the GHz acoustic streaming, and this approach was successfully applied to ultrafast fluid mixing³⁶ and particle trapping in a microfluidic system.³⁷ Although different-sized particles (from several micrometers to less than 100 nm) can be trapped by these GHz acoustic streaming tweezers (AST), the particle behavior in such 3D microvortex systems remains largely unknown. In this work, we study GHz AST and the associated particle manipulation in an open solution chamber without any physical confinement. The formation of the vortices is thoroughly analyzed both theoretically and by 3D finite element numerical simulations, the results of which are subsequently confirmed by microparticle tracking experiments. Enrichment and separation of particles of different sizes are systematically studied and experimentally tested. The absence of any need for fluidic channels facilitates the integration of the technique developed here with other devices (e.g., biosensors), thereby providing a versatile particle manipulation tool for a variety of potential applications.

Both the ARF and ASF have been comprehensively investigated in the sub-GHz range, especially in the case of standing surface acoustic wave (SSAW) and acoustic streaming induced by microbubbles³⁸ or microstructures.²⁰ However, acoustic manipulation above the GHz range has been less well studied. Here, we carry out a theoretical analysis of the GHz acoustic streaming effect generated by traveling bulk acoustic waves (TBAWs). It should be noted that in this system, the fluid is treated as compressible when we model the acoustic part, but as incompressible when we model the steady flow in the liquid part.

The motion of the fluid follows the basic principles of mass conservation and Newton’s second law, expressed respectively by

and

where ρ is the fluid density and v is the fluid velocity, both of which are functions of time and space, ρf is the body force, and T is the stress tensor. When an acoustic wave propagates in a viscous fluid, the nonlinear attenuation of oscillating displacements in a viscous medium leads to a time-averaged momentum flux. Because the free acoustic field is nonuniform and the scale of the medium is so large in comparison with the GHz traveling BAW-based AST, this momentum flux is a typical Eckart streaming, which pushes the fluid along the propagation direction of the acoustic wave.³⁹ As a consequence, the constitutive equation of a Newtonian fluid giving the relationship between stress, shear strain, normal strain and pressure is

where ε is the strain rate tensor, I is the unit tensor, μ is the viscosity, η is the bulk viscosity, and p is the pressure. On substituting Eq. (3) into Eq. (2), the full Navier–Stokes equation is obtained as

In our case, owing to the single-frequency vibration excited by the resonator, a directional flow and a series of harmonic vibrations will be generated, and so v, p, and ρ can be expressed as complex series:

and

Here, subscript 0 indicates the steady-state components, subscript 1 indicates the fundamental frequency, subscripts 2, 3, … indicate its higher-harmonic components, and the overbar (־) indicates complex conjugation. ω, t, and i are the angular frequency, time, and imaginary unit. In the case of GHz fundamental-frequency motion, we adopt the following two reasonable assumptions: (i) the steady-state components are much smaller than the fundamental-frequency component; (ii) the higher-order components make a negligible contribution. Therefore, the fundamental-frequency components become the dominant terms in the equations. Substitution of the approximations from Eqs. (5a) and (5b) into Eq. (4) gives the fundamental-frequency field equation

and the steady-state equation

respectively. Here, κ is the isothermal compressibility of the fluid, and β and c₀ are respectively the attenuation coefficient and the acoustic velocity in the fluid.

In our system, Eq. (6) can be considered as a wave equation with a dissipation component, and Eq. (7) can be regarded as equivalent to the Navier–Stokes equation since they have the same form. ρ₀f_B describes the body force that is responsible for the steady flow in the fluid. The whole process can be simply described as follows: A high-frequency acoustic wave, vibrating at the fundamental frequency, experiences attenuation when it propagates in the fluid. From the point of view of energy conservation, the kinetic energy of the flow comes from the dissipation of the acoustic wave. As a result, acoustic streaming is generated by acoustic wave attenuation.

A schematic of the GHz AST induced by TBAWs is shown in Fig. 1(a). When a radio-frequency (RF) signal is applied, the device generates a series of TBAWs, which propagate along the direction perpendicular to the device surface. When the TBAWs interact with the liquid, longitudinal wave leakage leads to acoustic pressure and directional flow in the latter. The particles in such a TBAW field are subject to four forces: the gravity force, the buoyant force, the ARF and the ASF. Because the gravity and buoyancy are almost balanced, the motion of the particles is mainly determined by the ARF and the ASF and is governed by the following equation:

where m_p and v_p are the particle mass and the particle velocity, respectively. The ARF is determined by the negative spatial gradient in the acoustic pressure field. Previous works based on acoustic wave scattering theory have provided the following expression for the ARF when the particles are much smaller than the acoustic wavelength:^28,40

where a, 2p₁, 2v₁, and k are the particle radius, incoming pressure, incoming wave vibration velocity, and wave number, respectively. f₁ and f₂ depend on the relative compressibility and the densities of the particle and fluid. In our system, the wavelength at 2.5 GHz (in water at 23 °C and 1 atm) is λ = (1484 m/s)/2.5 GHz ≈ 0.6 μm, and the particles we use range from the nanoscale to the microscale. Therefore, we have to consider two different scenarios. In the first scenario, the particles are smaller than the wavelength, and we can use Eq. (9). In the second scenario, the particles are larger than the wavelength, which is a situation that has rarely been studied previously. In this case, we propose the following approximate formula, the derivation of which is included in the supplementary material:

where p₁ and ρ₀ are the amplitude of the acoustic wave pressure and the density of the medium, respectively, and B is an adjustable constant. The ASF is given by

where v is the fluid velocity and v_p is the particle velocity.

As can be seen from Eqs. (9) and (10), the ARF is proportional to the cube of the particle radius a when the particles are smaller than the wavelength, but to the square of a when they are larger than the wavelength. The ARF distribution is shown in Fig. 1(b). From Eq. (11), it can be seen that the ASF is proportional to a, and it drags particles along the directions of the streamlines. To be more specific, when a particle moves closer to the edge of the device, the ARF acts as a centripetal force on the particle, pushing it toward the inner side of the streamline, and it then moves along the inner part of the streamline. As the particle moves away from the device, the ASF becomes dominant. Thus, the particle keeps following the streamline after moving away from the edge of the device. According to Eqs. (9) and (10), the ARF is particle-size-dependent. Large particles experience a large ARF, which means that they will be pushed toward the inner streamlines. In this AST platform, the vortex center is close to the surface at the edge of the device. Therefore, the larger particles follow the inner streamlines. Conversely the ARF is smaller for smaller particles, and thus they follow the outer streamlines, as shown in Fig. 1(c).

The TBAW device consists of a piezoelectric sandwich structure and a Bragg reflector structure, fabricated by standard MEMS technology. The Bragg reflector layers are fabricated by alternately depositing SiO₂/AlN (650 nm/640 nm) on a Si wafer by plasma-enhanced chemical vapor deposition (PECVD) and RF reactive magnetron sputtering. The piezoelectric structure is deposited on the Bragg reflector structure and comprises an AlN film sandwiched between top and bottom molybdenum (Mo) electrodes. The 170 nm bottom Mo electrode is deposited by magnetron sputtering and then patterned by dry etching. A 1100 nm thick AlN film is then deposited as the piezoelectric layer by RF reactive magnetron sputtering. The 160 nm top Mo electrode is subsequently fabricated in the same as the bottom electrode and then patterned by reactive ion etching. Finally, 1000 nm Au pads are evaporated by physical vapor deposition (PVD) and then patterned by a lift-off process to serve as an electrical connection. The 2.5 GHz TBAW device thus fabricated is shown in Fig. 2. More detailed information about the fabrication process can be found in our previous paper.³⁹ 

The model is built using the multiphysics finite element simulation software COMSOL 5.2. The model building procedure can be split into three parts, as shown in Fig. S1 in the supplementary material. The first step is a frequency-domain study, in which the solid mechanics, electrostatics, and pressure acoustics modules are used. The TBAW device is modeled using the first two modules, which are coupled with each other through the piezoelectric stress equations. The pressure acoustics module is coupled with the solid mechanics module at the interface between the solid and fluid. Therefore, we can calculate the velocity v₁ at the fundamental frequency and further obtain the body force ρ₀f_B. As we have already mentioned, in this module, we treat the liquid as a compressible fluid. Second, in the laminar module, we take the body force from the results of the frequency-domain study to further analyze the acoustic streaming in a steady state. The dissipation of the acoustic wave is converted into kinetic energy of the fluid. Here, we treat the liquid as an incompressible fluid. Finally, the particle tracing module is added to the time-domain study. Under the influences of the ARF and the ASF, the trajectories of the particles are calculated dynamically.

The 3D model of the TBAW device in the fluid is shown in Fig. S2 in the supplementary material. A pentagonal TBAW device is immersed in the center of a cylindrical water domain that has an uncovered roof and two closed walls of height 50 μm. To simplify the model, we apply the RF power to the top and the bottom boundaries of the piezoelectric layer, which vibrates through the inverse piezoelectric effect to create acoustic waves. A schematic of the different layers of the model is shown in Fig. S2(a) in the supplementary material. The model is divided into four parts in the z-axis direction to facilitate the meshing step: the TBAW device, one water part (the overlap of the pressure acoustics and laminar modules), and the other two water parts (only the laminar module). A continuous boundary condition is imposed at the solid–liquid interface, which ensures the continuity of stress and acceleration at the interface. The walls and roof of the cylindrical water domain are set as no-slip boundaries and an open boundary, respectively. Figure S2(b) in the supplementary material shows the model in the x–y plane, where the smallest pentagon is the TBAW device and the surrounding pentagons are the substrate. The boundary of the smallest pentagon and the second smallest pentagon are set as low-reflecting boundaries to absorb waves, similarly to a perfectly matched layer. Meanwhile, the three concentric pentagons are also used to provide a balance between numerical accuracy and computational load. Details of a similar simulation can be found in our previous paper.⁴¹ 

A polydimethylsiloxane (PDMS) chamber filled with liquid was sealed on top of the device. The device was connected to an RF signal generator (Agilent EXG Vector Signal Generator N51728). The observation system was constructed by integrating cameras in two directions. Thus, we could track particle motion from both top and lateral views. The top-view observation system included a video camera (Olympus DP73, Japan) and an optical fluorescence microscope (Olympus BX53, Japan). The lateral observation system was a commercialized micro-particle image velocimetry (PIV) system (Lavision, Germany), which included a telephoto microscope (12×), a double-pulsed 532 nm Nd-YAG laser, a programmable timing unit (PTU), a CMOS camera, and Davis 8.3 software.

We used 10 μm polystyrene (PS) particles to characterize the profile of the microvortices. The particle concentration was dense enough to be observed clearly from both the top and lateral views. The flow profile was analyzed by the Davis 8.3 software. The RF signal generator connected with a power amplifier was used to excite the TBAW device to generate vortices at different powers.

All PS particles, of sizes 300 nm, 1 μm, 5 μm, and 10 μm, were dispensed in HEPES (pH 7.4) buffer. Typically, 50 μl of PS particle solution was added to the PDMS chamber. The device was connected to the RF signal generator with an output of 1 mW. The experimental phenomena were captured by the video camera with the optical microscope.

As an acoustic wave reaches the boundary between one medium and another, both reflected and transmitted waves are generated. The reflection coefficient is given by R = (Z₁ − Z₂)/(Z₁ + Z₂), where Z_i = ρ_ic_i(i = 1, 2) are the acoustic impedances, with ρ_i and c_i being the density of the medium and the wave speed in the medium, respectively. For the TBAW device working in air, which has a significantly different acoustic impedance from a solid, total reflection occurs. This helps to store the acoustic energy within the device, thereby ensuring a high quality factor (Q = 225), as shown in Fig. S3(a) in the supplementary material. However, when the GHz TBAW is working in a liquid, the vibrating device interacts with the latter. The acoustic wave from the device is largely transmitted into the liquid owing to the relatively small difference between their acoustic impedances. There is a substantial loss of acoustic energy into the liquid, as can be seen from the decrease in the Q value from 225 to 38 [Fig. S3(a) in the supplementary material].

The loss of energy to the environment can be illustrated by the change in the z component of the wave speed. The z component of the wave vibration velocity in an arbitrary position z_i can be expressed as

Thus, the wave vibration velocity decays exponentially with propagation distance. Figure S3(d) in the supplementary material shows the attenuation distribution in the fluid according to Eq. (12). This dissipation leads to a body force being applied to the liquid. Thus, a shorter energy dissipation distance will result in a higher efficiency of energy conversion. Owing to the high frequency of the TBAW device, the wave vibration velocity drops to 20% of its original value after the wave has traveled a distance of just 10 μm along the z axis. Under this condition, the fluid obtains most of its energy quickly.

The simulation results for the acoustic field [Fig. S3(b) in the supplementary material] demonstrate that the gradient of the acoustic pressure amplitude is in the vertical direction above the device and inclined to the horizontal direction above the edge of the device. The ARF magnitude can be obtained from Eqs. (9) and (10), and its direction is marked by arrows in Figs. S3(b) and S3(c).

A strong body force applied to the liquid will generate strong acoustic streaming. In this subsection, the acoustic streaming effects are carefully studied both with simulation and micro-PIV. According to Eq. (7), the body force can be calculated from the vibration velocity v₁. After the body force has been added to the water domain of the simulation, acoustic streaming is induced. Figure 3 shows the 3D simulation results for the flow field. To better demonstrate the acoustic streaming velocity distribution, we capture the cross-sections of the velocity field in planes −200, −100, 0, 100, and 200 μm away from the x–z plane, as shown in Figs. 3(a)–3(e), respectively. Figure 3(f) shows the detail of the velocity field in the x–z plane. It can clearly be seen that the streaming flows rise from the bottom of the fluid (the top of the device) and rotate clockwise or counterclockwise to form vortices around the center of the device. It should be noted that the maximum particle velocity occurs when the particle passes over the edge of the device.

For comparison with the simulation results, micro-PIV assisted with PS tracer particles was used to trace and analyze the streaming effects. The real-time particle motions induced by acoustic streaming in the x–y and x–z planes were obtained by high-speed cameras in both top and side views. The top view of the streaming is shown in Fig. 4(a). The dark areas are the microstreaming flows in which the particles are trapped. Ten petal-like streaming vortices are clearly observed above the device surface. The 10 generated petal-like 3D microvortices are related to the shape of the device. In our opinion, since the velocity of the vortex is proportional to the characteristic length of the device, the vortices are not uniformly distributed along the edge of the pentagon. It should also be noted that, after the device has been switched off, the petal-like vortices disappear, which indicates dynamic particle trapping (see Video S1 in the supplementary material).

Figure 4(b) shows a side view of the vortices. There are two nearly symmetrical vortices in a cross-section, where the bright areas indicate the concentrated particles. Figure 4(d) shows the particle velocity distribution in this cross-section. Note that the velocity has been homogenized in magnitude (and so the arrows indicate just the direction of the streaming flow) to provide a clear illustration of its direction, which would otherwise be obscured by the huge differences in velocity magnitude between particles at the edge of the device and those at other positions. It can be seen that the experimentally observed streaming directions are completely consistent with our simulation results.

The relationship between the applied power and the maximum particle velocity was also investigated experimentally. The maximum velocities at different powers (1, 3, 10, 32, and 100 mW) were measured at the edge of the device, with the motions of 1 μm PS tracer particles being recorded by a high-speed camera. We then calculated the velocity based on the particle displacement over a given time interval. Figure 4(c) shows a linear fit of the relationship between the maximum velocity and the applied power. This linearity can be explained by the velocity relationship in Eq. (7). Since the steady-state streaming velocity v₀ is much smaller than the velocity v₁ at the fundamental frequency, they are related by $v0∝(2v1)2$⁠. For an acoustic resonator, the velocity at the fundamental frequency is related to the applied power by v₁ ∝ P^1/2. Therefore, we can derive the relationship v₀ ∝ P, which agrees with the experimental results.

The ASF induced by these particular streaming flows and the ARF both play important roles in the vertical and horizontal manipulation of particles. According to the preceding theoretical analyses and simulations, the acoustic pressure amplitude decreases exponentially along the z direction, and the wave speed drops to 20% of its original speed after traveling a distance of just 10 μm along the z axis [Fig. S3(c) in the supplementary material]. Thus, the trapped particles will experience an effective ARF only when they are close to the device. Besides, the ARF is proportional to a power of the particle radius a (either a² or a³), while the ASF is simply proportional to a. Therefore, larger particles are subjected to greater ARFs, which push the particles toward the inner orbits of the trajectory. Therefore, large particles will be concentrated at the edge of the device (see also Video S2 and Fig. S4 in the supplementary material). Small particles, on the other hand, are subjected to a negligible ARF, and so they continue to move along the streamlines. We used simulations and experiments to analyze this phenomenon.

The particle tracing module was used for the transient state study in the simulation, with 200 particles of different sizes being added randomly to the water domain and subjected to the ARF, the ASF, and gravity and the buoyancy forces. Here, we defined the solution of the equation F_ARF/F_ASF ≈ 1 as the value of the particle radius. Figure 5 shows the simulated trajectories for particles of different sizes. The sizes of the trajectories increase as the particle radius decreases. For instance, the calculated ARFs on a 10 μm particle and a 5 μm particle are around 1 nN and 0.25 nN, respectively (Fig. S3 in the supplementary material), and so the 10 μm particle will move to an inner trajectory and finally become trapped at the edge of the device. The trajectories of 1 μm particles are relatively large and form streaming vortices, and 300 nm particles show similar behavior. These results indicate that particles smaller than 1 μm are influenced more by the ASF than by the ARF. As a consequence, we can estimate that the critical value is approximately 1 μm. This critical point can also be determined experimentally. It can be inferred that as the particle size decreases, the particles start to form petal-like streaming vortices, rather than concentrating at the edge of the device (on inner trajectories).

To validate our simulation results experimentally, different-sized PS particles (10 μm, 5 μm, 1 μm, and 300 nm) were introduced into the AST. We fixed the focal plane of the vertical microscope at the top of the device and applied a 1 mW, 2.5 GHz RF signal to the device. As shown in Figs. 6(a) and 6(b), particles of diameters of 10 and 5 μm were first trapped in the vortices and gradually concentrated to form a pattern at the edge of the device. However, when 1 μm particles were used, they did not form such a pattern, but instead were trapped inside petal-like streaming vortices [Fig. 6(c)]. The 300 nm particles showed similar but even more obvious behavior [Fig. 6(d)]. These results indicate that the critical value is approximately 1 μm, in good agreement with the simulation results. Particles larger than 1 μm are subject to a greater ARF and are steadily concentrated to form patterns at the edge of the device, while particles smaller than 1 μm are subject to a greater ASF and tend to follow the streamlines, becoming dynamically trapped and finally concentrated in the vortex zone.

In this work, we have demonstrated manipulation of micro/nanoscale particles by GHz AST. The behavior of a 2.5 GHz TBAW device in liquid was investigated by theoretical analyses, numerical simulations, and experiments. A 3D multiphysics finite element model was established to study the GHz AST and its application to particle manipulation. The TBAWs leak into the liquid, generating acoustic streaming and inducing multiple petal-like 3D microvortices above the device. Under the influence of the ASF and the ARF, particles in the liquid are trapped in the vortices and move along different trajectories depending on their sizes. Large particles (>1 μm) are influenced mainly by a strong ARF and are eventually concentrated to form a steady pattern at the edge of the device, while small particles (<1 μm) are only dynamically trapped and concentrated in the vortex zone. In summary, a simple and effective 3D GHz AST approach has been theoretically and experimentally demonstrated as a dedicated tool for noninvasive micro/nanoscale manipulation of various particles.

The supplementary material contains the derivation of Eq. (10), together with additional figures and videos.

The authors gratefully acknowledge financial support from the National Key R&D Program of China (2018YFE0118700), the Natural Science Foundation of China (NSFC No. 62174119), Tianjin Applied Basic Research and Advanced Technology (17JCJQJC43600), and the 111 Project (B07014).

The authors have no conflicts to disclose.

H.W. and Z.T. contributed equally to this work.

The data that support the findings of this study are available from the corresponding author upon reasonable request.

 Hang Wu received a B.S. degree in 2018 from Harbin University of Science and Technology, Harbin, China. And he is currently pursuing a M.A. Eng. Degree at Tianjin University. His research interests include manipulations of micro/nanoparticles by gigahertz acoustic streaming tweezers and manipulations of single particles by acoustic resonator array.

 Zifan Tang is a Ph.D. Candidate in Electrical Engineering at the Pennsylvania State University. She has a research background with microfluidics, nanofluidics, bioMEMS, and lab-on-a-chip devices. She is interested in sample-in answer-out molecular diagnostic system for sensitive, specific, and rapid detection of viral pathogens at the point of need. She is also studying the fundamental and applied challenges in solid-state nanopore for next-generation medical diagnosis.

 Rui You received a B.S. degree in 2018 from Tianjin University, Tianjin, China, where he is currently pursuing a Ph.D. degree. His research interests include film bulk acoustic resonator, microfluidics, acoustofluidic and drug delivery systems.

 Shuting Pan received the B.S. degree from Tianjin University, Tianjin, China, in 2014, and received her Ph.D. degree from Tianjin University in 2020. Her research focuses on the application of acoustic device in biomolecular interaction.

 Wenpeng Liu received his B.S. degree and Ph.D. degree in the Specialty of Instrument Science and Technology from Tianjin University in 2017. He is currently working as a Research Fellow at Harvard Medical School. His main research subjects focus on acoustofluidic, plasmonic nanomaterials, photonics, and integrated molecular diagnostic system.

 Hongxiang Zhang received a B.S. degree in 2013 and a Ph.D. degree in 2018 from Tianjin University, Tianjin, China. His research interests include modeling, design and simulation of MEMS devices and microfluidics.

 Tiechuan Li received a masters degree in 2016 from Liaoning University of Technology, Liaoning, China. And he is currently pursuing a Ph.D. degree in Tianjin University, Tianjin, China. His research interests include micro/nanodevices, microsystems, and acoustofluidic.

 Yang Yang received the B.S. degree from Chongqing University, Chongqing, China, in 2015, and he is a Ph.D candidate in Duan's group, Tianjin University. His research interest focuses on acoustofluidics for manipulation of biospecimens in micro/ nanoscale.

 Chongling Sun received a B.S. degree in Tianjin University, Tianjin, China. Currently, she is an Engineer at the School of Precision Instrument and Opto-electronics Engineering, Tianjin University. Her research focuses on the area of MEMS process.

 Wei Pang received a B.S. degree from Tsinghua University, Beijing, China in 2001, and a Ph.D. degree in electrical engineering from the University of Southern California, Los Angeles in 2006. He is currently a Professor at the College of Precision Instrument and Optoelectronics Engineering, Tianjin University, Tianjin, China. His research focuses on the areas of MEMS for RF wireless communications, biological detection, wireless sensor platforms, and medical ultrasound.

 Xuexin Duan received a Ph.D. degree from the University of Twente, Netherlands in 2010. After postdoctoral studies at Yale University, he moved to Tianjin University, Tianjin, China. Currently, he is a full professor at the State Key Laboratory of Precision Measuring Technology and Instruments, Department of Precision Instrument Engineering of Tianjin University. His research concerns MEMS/NEMS devices, microsystems, and microfluidics, and their interfaces with chemistry, biology, medicine, and environmental science.