Applied electric field analysis and numerical investigations of the continuous cell separation in a dielectrophoresis-based microfluidic channel

ri p Ng t Na o.16 Faculty of Materials Science and Engineering, Phenikaa University, Hanoi, 10000, Viet Nam idic-based devices TCs [5,6]. The sep- res for subsequent ]. c field sources is a anipulate a wide annel [8]. For the the movement of od, cancerous cells cell sample solu- rphology, deform- Since distinctive cell shape characteristics, cell size-based separa- tion techniques are also typically preferred to classify several types of cells [1,8,9,15,16]. From the practical point of view, recently, cell size and DEP-based separation methods have arisen in micro- fluidics. For instance, a DEP-based microfluidic device with two inlets and two outlets for continuous-flow separation of platelets from other blood cells due to their size difference was demon- strated (platelets: 2e3 mm and RBCs: 7e8 mm) [17,18]. The non- * Corresponding author. Faculty of Electrical and Electronic Engineering, Pheni- kaa University, Yen Nghia, Ha Dong, Hanoi, 12116, Viet Nam. ** Corresponding author. Faculty of Electrical and Electronic Engineering, Pheni- kaa University, Yen Nghia, Ha Dong, Hanoi, 12116, Viet Nam. E-mail addresses: viet.nguyenngoc@phenikaa-uni.edu.vn (N.-V. Nguyen), hieu. nguyenvan@phenikaa-uni.edu.vn (N. Van Hieu). Contents lists availab Journal of Science: Advanc journal homepage: www.el Journal of Science: Advanced Materials and Devices 6 (2021) 11e18Peer review under responsibility of Vietnam National University, Hanoi.advantages, such as simplicity, cheapness, good compatibility, and ability, mechanical, electrical, or magnetic properties [1,13,14].1. Introduction Separation, isolation, capture, identification, and enumeration of rare cell types, including circulating tumor cells (CTCs) are important for a large number of medical and clinical studies [1e3]. The type of CTCs which may appear with a low concentration in the bloodstream is considered a good biomarker at the early stages of the primary and even metastatic tumors [4]. The detection and quantification of CTCs from a blood sample can serve as a key means for real-time cancer prognosis. Microfluidic technology has been demonstrated useful to operate bioparticles due to lots of effectiveness. Consequently, numerous microflu have been developed for the manipulation of C aration of CTCs is one of themost crucial procedu detection and aggressive treatment of cancer [7 Dielectrophoresis (DEP) with external electri rapid, simple, and well-known technique to m range of biological particles within the microch said purposes, DEP is also utilized to separate different cancer cells [9e12]. Using the DEPmeth could be isolated from normal blood cells or the tion by various cell properties, such as size, moNumerical simulation open access article under the CC BY license ( r t i c l e i n f o Article history: Received 17 August 2020 Received in revised form 27 October 2020 Accepted 4 November 2020 Available online 7 November 2020 Keywords: Continuous cell separation Circulating tumor cells (CTCs) Dielectrophoresis (DEP) Microfluidicshttps://doi.org/10.1016/j.jsamd.2020.11.002 2468-2179/© 2020 Publishing services by Elsevier B creativecommons.org/licenses/by/4.0/).a b s t r a c t Dielectrophoresis (DEP) is known as an attractive and frugal technique to manipulate biological particles in microfluidics. This study presents the advanced solution strategy of a DEP-based microfluidic channel for focusing and separating cancerous cells in continuous flow. Theoretical calculations were carried out to define the favorable parameters in the electric field operation of the microchip. A simulation model was also used to explore the performance of the design in the isolation of circulating tumor cells (CTCs). It revealed that the optimal conditions of the device are suitable to effectively separate CTCs from red blood cells (RBCs) within the channel structure, with a high flow rate of 1.5 mL/min, and an electric amplitude as low as 10 Vpp, at the frequency of 1 kHz. The proposed method has shown potential as a simple, easy-to-operate, and low-cost approach enable to enhance the diagnosis systems for cancer detection at early stages. © 2020 Publishing services by Elsevier B.V. on behalf of Vietnam National University, Hanoi. This is anPhenikaa Institute for Advanced Study (PIAS),d Faculty of Biotechnology, Chemistry and Environmental Engineering, Phenikaa University, Hanoi, 10000, Viet Nam e Phenikaa University, Hanoi, 10000, Viet NamOriginal Article Applied electric field analysis and nume continuous cell separation in a dielectro channel Ngoc-Viet Nguyen a, b, **, Tu Le Manh b, c, Tang Son Nguyen Van Hieu a, b, e, * a Faculty of Electrical and Electronic Engineering, Phenikaa University, Hanoi, 12116, Vie b Phenikaa Research and Technology Institute (PRATI), A&A Green Phoenix Group JSC, N c.V. on behalf of Vietnam Nationalcal investigations of the horesis-based microfluidic uyen b, d, Viet Thong Le a, b, m 7 Hoang Ngan, Trung Hoa, Cau Giay, Hanoi, 11313, Viet Nam le at ScienceDirect ed Materials and Devices sevier .com/locate/ jsamdUniversity, Hanoi. This is an open access article under the CC BY license (http:// drag force ðFHD
!Þ [21]. Several forces are generally ignored due to their minor effects, such as gravitational force. The total force on each cell is the sum of the vector forces of DEP and hydrodynamics. The description of the cell movement by Newton's second law is frequently expressed by: FDEP

!þ FHD
!¼mpdvp ! dt (1) where vp! and mp are the velocity and mass of the cell, respectively. The FDEP

! is determined based on DEP properties. It is worthy of mention that DEP is known as the interaction of polarizable par- ticles in a non-uniform electric field. It means that when a particle approaches the gap between the electrodes, the particle will experience a DEP force. The DEP phenomenon is dominated by the difference of electrical factor between the particle and the sur- rounding medium solution. The DEP force inducing on a particle with the spherical shape of Rp radius, in a suspension mediumwith N.-V. Nguyen, T. Le Manh, T.S. Nguyen et al. Journal of Science: Advanced Materials and Devices 6 (2021) 11e18uniform electric field was generated by the arrangement of alter- nating polarity electrodes. The device obtained high separation efficiency with the low voltage operation. Cancer cells and eryth- rocytes could be distinguished using DEP within open-top micro- chamber devices [19]. A microchip, which can continuously separate and selectively concentrate bacterial cells from the com- plex physiological sample, was proposed using specific DEP tech- niques [20]. Based on the substantial size difference of cells, this chip was used to separate bacteria (z1 mm) and RBCs (z6e10 mm). Another polymeric microfluidic device with 100 mm-thick elec- trodes was built to perform DEP trapping for the separation of CTCs (MDA-MB-231) from the RBCs sample [21]. However, the chip only reached high efficiency at flow rates below 0.5 mL/min. Using a DEP- based microfluidic chip with optically transparent electrodes, MCF- 7 CTCs were successfully separated from HCT116 CTCs at the optimal conditions of 9 Vpp, 3.2 MHz, 0.1 mL/min [22]. Therefore, one of the critical demands is to improve the cell-throughput of these continuous microfluidic channels. On the other hand, experimental investigations of the cell sep- aration process are merely difficult, time-consuming, and expen- sive in many circumstances. Meanwhile, modeling and simulation methods have been shown as a versatile tool for many cell manipulation applications in microfluidic chips [23e26]. Computer simulations can generate important and useful information for understanding phenomena that occurred in these devices. The electromechanical behavior of various cell types, as well as, the motion of cells can be analyzed and predicted before performing laboratory experiments. Computational methods are thus possible to significantly contribute to the design improvement of the microfluidic devices [27]. Several DEP-based chip designs for the separation of platelets from RBCs have been improved by simula- tion tools [17,18]. Microfluidic devices and numerical simulations for the manipulation of CTCs in different stream-flow conditions have also been performed [28,29]. Although understanding of these types of studies is not completed, the development of DEP-based CTCs separation techniques could be rapidly achieved by combining with multiphysics simulation. In the present work, an advanced design of the DEP-based microfluidic platform is pro- posed based on the fabricated one [22] to monitor the size-based cell separation process and the isolation of CTCs. Using theoret- ical calculations and numerical simulations, the effects of medium solution, applied electric field, and fluid flow speed on the sepa- ration of binary cell mixture is investigated to define the optimal parameters. The performance of the proposed design is validated by comparing it with other theoretical and experimental results reported in the literature. 2. Chip design and numerical simulation 2.1. Design and working principle of the microfluidic chip The proposed design of the microfluidic platform for continuous cell focusing and DEP-based cell separation is depicted in Fig. 1. The chip is composed of three basic parts, such as a glass substrate, polydimethylsiloxane (PDMS) channel configuration, and an array of coplanar metal microelectrodes. It can be fabricated using a similar soft lithography process in the laboratory [33e35]. The microfluidic structure contains one inlet and two outlet ports. The channel is divided into two main sections: a concentration region (main channel) for continuous inertial focusing of randomly distributed cells within the fluid flow to a single streamline, and a separation region (right straight and lateral channels) for isolation and collection of the cell types. A cell sample mixed with a buffer12solution is injected at the inlet towards the main channel. The inlet connected to the main channel is the injection port of the cell sample into the concentration part, and the two other outlets are the cell enrichment ports of the separation part. The main channel with a curving symmetric geometry has been demonstrated to continuously order different cells at high rates, and without externally applied forces in the channel [36]. Another microfluidic design utilizing the DEP technique can also be studied and applied to cell focusing [37,38]. The second part of the channel, which in- cludes the straight channel connected to the main channel on the right side, and the lateral channel deviating at an angle from the straight channel, is the important area to separate cell types by the effects of DEP and drag forces. The DEP response is generated by the electrodes located at the intersection area between the right straight channel and lateral channel. Under suitable conditions, the bigger cells are repelled to go out at the side channel, while the smaller cells are still collected at the right main channel. Thus, the method can be used to achieve continuous separation of cells based on their size. This ability of the microfluidic chip is applied for separating CTCs and RBCs. 2.2. Computational methodology In a microfluidic channel, cell particles follow the streamlines of fluid flow. In the case of using the DEP technique, cell motion is mainly impacted by the DEP force ðFDEP

!Þ, and the hydrodynamic Fig. 1. Sketch of the microfluidic chip design for continuous cell focusing and separation. 2.3. Simulation flow To study the separation process, the movement trajectories of cell particles in the microchannel are simulated using COMSOL Multiphysics software (version 5.0). Three-dimensional (3D) geometrical parameters of the proposed design are given in Table S1. The AC/DC electric module, the laminar flow module, and the particle trajectory tracking module were applied. Finite element analyses returned the distribution of the electric potential and the flow speed inside the microfluidic channel geometry. Laplace equations and NaviereStokes equations were used as governing equations for electric and flow computations, respec- tively. Then, Newton's second law was employed to solve the par- ticle trajectory. Human normal erythrocytes and MDA-MB-231 breast cancer cells [39] are chosen as simulated objects of RBCs and CTCs, respectively. Cells were assumed as spherical particles with the sizes and main electrical properties of the two types of cells found in the previous literature, and listed in Table S2. To indicate the operation of the separation chip, cell concentrations were set up about 25 cells/mL and 103 cells/mL for CTCs and RBCs, respectively. Fig. 2 shows a flowchart of the solution algorithm used in this work. First, the geometrical parameters and physical properties of the two cell types are filled in the parameter table of the software. Then, building the 3D geometry of the chip, setting the boundary conditions of the applied modules, and creating a mesh for the model are performed, respectively. An AC voltage excitation with defined amplitude and frequency is set on the surface of the elec- trodes. The inlet of the fluid flow channel is set to a flow rate, the Journal of Science: Advanced Materials and Devices 6 (2021) 11e18the relative permittivity (or dielectric constant) εm is described by [30]: FDEP

!¼2pR3pε0εmReðfCMÞVjErms

!j2 (2) where Erms is the root mean square of the electric field, ε0x 8:85 1012 F  m1 is the vacuum permittivity, the subscripts p and m are the represented characters of the particle and the medium, respectively. Re(fCM) is the real part of the Clausius-Mossotti factor (fCM), given by: fCM ¼ ε * p  ε*m ε * p þ 2ε*m (3) with ε* is the complex electrical permittivity. It is expressed as: ε *¼ ε j s u (4) where s is the electrical conductivity, ε is the permittivity, j is the imaginary unit and u is the angular frequency of the electric field. The DEP force alters with the magnitude and the sign of the Re(fCM) factor. It is easy to verify that the value of Re(fCM) varies in a range from 0.5 to 1. According to the sign of fCM, DEP is frequently classified into positive DEP (pDEP) if Re(fCM) > 0 and negative DEP (nDEP) if Re(fCM) < 0. As a result, the particles forced by pDEP response are guided toward the location of the high electric field strength (around the edges of the electrodes), whereas the particles forced by nDEP response are moved to the low electric gradients (away from the electrodes). A biological cell could be assumed as a heterogeneous spherical structure in the evaluation of its electric properties [31]. The model was simplified to a single-shell dielectric model with a cytoplasm surrounded by a thin lipid membrane layer. The Clausius-Mossotti factor for viable cells can be rewritten as: fCMðuÞ¼  u2  tmt*p  tpt*m  þ ju  t*m  tm  t*p   1 u2  2tmt*p  tpt*m   ju  t*m þ 2tm  t*p   2 (5) where t*m ¼ cmrp=sm; tm ¼ ε0εm=sm; t*c ¼ cmrp=sp; and tp ¼ ε0εp=sp are time constants of the suspension medium and the cell, separately. And cm is the cell membrane capacitance. The FHD
! for a particle can be defined by Stokes's law in micro- channel [32]: FHD
!¼6phRp vm
! vp! (6) with h is the dynamic viscosity coefficient of the medium and vm
! is the flow velocity. In our study, the combined action of DEP force and fluid force was employed to achieve the purpose of cell separation in the microfluidic device. The applied electrical parameters were chosen to induce nDEP force against the hydrodynamic drag force on each cell. Both CTCs and RBCs responded to nDEP at an alternating cur- rent signal with a low frequency. Moreover, because of the distinctive size (the diameter of CTCs is commonly larger 2e5 times than that of blood cells) [7], there were high differences in DEP force between cancerous and normal cells. Therefore, the DEP chip enables an efficient method for the cell size-based separation, N.-V. Nguyen, T. Le Manh, T.S. Nguyen et al.typically CTCs and RBCs. 13Fig. 2. Solution algorithm of the numerical method. electric impulses. For the higher media conductivities, the Re(fCM) factor of CTCs is bigger than that of RBCs in the frequency range channel A first simulation was performed to test the applied fields of the design. Tetrahedral elements were used in the 3D simulationmodel mesh. More grid points were placed on the area around the elec- trodes due to the necessary impacts of the force fields in this region. Simulated results of the velocity field in the channel and electric field around the electrodes of the microfluidic chip are shown in Fig. 4(a) and 4(b), separately. As mentioned in the theoretical model, the separation of cells depends on the balance between DEP and hydrodynamic drag forces. The distribution of the DEP force closely relates to the applied electric field, while the drag force is mainly affected by the fluid flow field in the channel. It is assumed that the microfluidic channel was filled up with a medium solution (εm ¼ 78; sm ¼ 0:055 S=m). Cells suspended in this carrier fluid solution were injected into the channel with a flow rate of 1.0 mL/ min. The velocity of the fluid flow could reach up to 5.0 mm/s at the center of the cross-section on the left side of the straight channel (see in Fig. 4(a)). A sinusoidal peak-to-peak voltage of 10 V at the frequency of 1 kHz was applied. As can be seen, the electric field peaks at the edges of the electrodes, and the DEP effect decreases as increasing the z-distance of the cell from the electrode top surface channel (see in Fig. 4(b)). Thus, each cell can obtain a stronger DEP force if it is close to the electrode surfaces and edges. N.-V. Nguyen, T. Le Manh, T.S. Nguyen et al. Journal of Science: Advanced Materials and Devices 6 (2021) 11e18boundary condition on the channel walls is set to no-slip mode, and the pressure is set to zero at the channel outlets, which is the at- mospheric pressure. Subsequently, the electric and flow fields are solved to obtain the key parameters for the calculation of DEP and hydrodynamic drag forces in the particle tracing module. The governing equations are solved herein using the finite difference method. The position and velocity of the cell particles are calculated since the influence of coupling the electric field and flow field. The trajectory of the cells in the channel can be tracked. After running the simulation process, the important results are exported to the graphs for analysis. In these numerical simulations, we only focus on analyzing the separation region, wherein the non-uniform electric field is created by the microelectrodes. As previously dis- cussed, all cells are concentrated into a single streamline at the center of the flow before they approach the electrodes [36]. In the area around the electrodes, DEP and drag forces are the dominant components acting on a cell. Their net force significantly affects the movement of each cell. The important parameters, which impact the separation process, are investigated in simulations, including applied electric frequency and voltage amplitude, and flow rate. These boundary properties can be put on the setting by a para- metric sweep mode, to produce many different results, and to repeat cell trajectory calculations. This allows us to quickly find out the optimal conditions for the proposed chip design. 3. Results and discussion 3.1. Applied electric frequency for DEP From theoretical models, the DEP force acting on a cell is dependent upon the cell size, the electric factors of cell and media, and the applied electric field. The DEP-based cell mani