1

Book of Abstracts for the 4th International Conference on Dielectrophoresis

July 26-28, 2021 (originally planned for July 2020)

Flagstaff, Arizona, USA and ONLINE www.dep2020.org

2

Dear colleagues,

Building on the success of our 3 previous Dielectrophoresis meetings and with more than 70 abstract submissions from researchers from 15 different countries and interdisciplinary backgrounds, the organizing committee of DEP 2020.1 is excited to welcome you in Flagstaff and Online!

In this document you will find the abstracts of the works to be presented in the conference. This is an open access document so please share freely. If you have any questions about the contents of an abstract, please contact the authors directly. The Organizing Committee is not responsible for any claims listed in the abstracts.

With all the best wishes,

Conference Co-chairs, on behalf of the organizing committee

Mark A. Hayes Professor Arizona State University USA mhayes@asu.edu

Lisa A. Flanagan Associate Professor and Vice Chair for Research Department of Neurology Associate Professor Biomedical Engineering and Anatomy & Neurobiology Sue & Bill Gross Stem Cell Research Center University of California Irvine USA lisa.flanagan@uci.edu

Rodrigo Martinez-Duarte Associate Professor Department of Mechanical Engineering Clemson University USA rodrigm@clemson.edu

3

Organizing Committee

Jonathan Cottet, Massachusetts Institute of Technology, USA

Zach Gagnon, Texas A&M University, USA

Pablo Garcia-Sanchez, Universidad de Sevilla, Spain

Nic Green, University of Southampton, England

Carlotta Guiducci, École Polytechnique Féderale de Lausanne, Switzerland

Ralph Hölzel, Freie Universitat Berlin, Germany

Michael P. Hughes, University of Surrey, England

Fatima Labeed, University of Surrey, England

Blanca Lapizco-Encinas, Rochester Institute of Technology, USA

Georg Pesch, University of Bremen, Germany

Antonio Ramos, Universidad de Sevilla, Spain

Alexandra Ros, Arizona State University, USA

Soumya Srivastava, University of Idaho, USA

Nathan Swami, University of Virginia, USA

Gilad Yossifon, Technion-Israel Institute of Technology, Israel

4

5

6

Contents PLENARY ................................................................................................................................................................ 9

PROTEIN DIELECTROPHORESIS: STATUS OF EXPERIMENTS AND THEORY ..................................................................................... 9

SESSION 1 ............................................................................................................................................................ 10

AC ELECTROKINETICS FOR THE IMMOBILISATION OF NANOPARTICLES AND MOLECULES ............................................................... 10 DIELECTROPHORETIC FILTRATION FOR SELECTIVE PARTICLE RECOVERY AT HIGH THROUGHPUT ....................................................... 11 DIELECTRIC PROPERTIES OF INFECTED PORCINE KIDNEY CELLS FOLLOWING GLYCINE TREATMENT ................................................. 12 NEURAL NETWORKS MEET IMPEDANCE CYTOMETRY ............................................................................................................. 13

POSTER SESSION 1 ............................................................................................................................................... 14

MICROFLUIDIC ARRAYS COMBINING DIELECTROPHORETIC AND HYDRODYNAMIC FORCES FOR TRAPPING AND RETRIEVAL OF SELECTED

CELLS ......................................................................................................................................................................... 14 DEP-ON-CMOS: A MICROFLUIDIC PLATFORM FOR DIELECTROPHORETIC MANIPULATION, TRAPPING, AND DIFFERENTIAL SEPARATION OF

VIABLE AND NON-VIABLE YEAST CELLS ............................................................................................................................... 15 AC FIELD ASSISTED DEPOSITION OF INFLUENZA VIRUSES ON NANOELECTRODES .......................................................................... 16 AC ELECTRIC FIELD MEDIATED PREPARATION OF REGULAR ENZYME ARRAYS AND THEIR FUNCTIONAL CHARACTERIZATION ................... 17 AC ELECTROKINETIC IMMOBILIZATION OF K562 EXOSOMES ON NANOELECTRODE ARRAYS .......................................................... 18 ELECTROROTATION OF SINGLE CELLS FOR THE ANALYSIS OF MEMBRANE DAMAGE INDUCED BY TOXINS MIMICKING THE

NEURODEGENERATIVE EFFECT OF AMYLOID BETA IN THE ALZHEIMER'S DISEASE ......................................................................... 19 LABEL-FREE CELL SORTING IN FLOW WITH REAL-TIME MODIFICATION OF SEPARATION PARAMETERS ............................................... 20 A STUDY OF OPTIMAL ELECTRODE DESIGN FOR ELECTRIC FIELD DRIVEN TARGET CAPTURE IN SPR BIOSENSORS .................................. 21 MODIFIED RED BLOOD CELLS AS MULTIMODAL STANDARDIZED PARTICLES FOR BENCHMARKING CELL ELECTROPHYSIOLOGY ................ 22 RAPID DETECTION OF CANCER RELATED CELLULAR LYSIS EVENTS IN PATIENT PLASMA USING HIGH CONDUCTANCE DIELECTROPHORESIS

................................................................................................................................................................................. 23 DEP SEPARATION OF DIFFERENT CANDIDA STRAINS USING 3D CARBON ELECTRODES ............................................................... 24 THE EFFECT OF OXIDATIVE STRESS ON DIELECTROPHORETIC PARAMETERS OF HUMAN RED BLOOD CELLS ......................................... 25

Session 2 .............................................................................................................................................................. 26

SIMULTANEOUS USE OF METAL COATED THREE-DIMENSIONAL SU-8 PILLARS AS PASSIVE POSTS AND ELECTRODES ............................ 26 USING DEP TO DETECT CELL PHENOTYPE AND CELL SURFACE COMPOSITION ............................................................................. 27 DIELECTROPHORESIS QUANTIFIES THE DYNAMIC HETEROGENEITY OF MESENCHYMAL STEM CELLS ............................................... 28 3D PRINTING MICROFLUIDIC DEVICES USING LIQUID DIELECTROPHORESIS ................................................................................. 29

Session 3 .............................................................................................................................................................. 30

PANCREATIC CANCER DETECTION THROUGH MULTIOMIC ANALYSIS OF BIOMARKERS COLLECTED FROM PLASMA USING HIGH

CONDUCTANCE DIELECTROPHORESIS ................................................................................................................................ 30 A SEQUENTIALLY SELECTIVE DEP PLATFORM FOR STUDYING CYTOKINE AND CYTOLYTIC RESPONSES FROM THE INTERACTION OF A

CONTROLLED NUMBER OF TUMOR AND IMMUNE CELLS IN CONFINED VOLUMES ......................................................................... 31 INTEGRATION OF SELECTIVE CAPTURE OF TUMOR CELLS WITH ELECTROCHEMICAL BIOSENSING AT A WIRELESS ELECTRODE ARRAY ......... 32

7

CHARACTERIZATION OF PARTICLE TRAJECTORIES DURING DIELECTROPHORESIS COLLECTION AND THE PHYSICAL PHENOMENA OF

INSULATOR COVERAGE OVER THE ELECTRODES ................................................................................................................... 33

Keynote ............................................................................................................................................................... 34

MODELING THE ELECTROKINETIC BEHAVIOR OF METALLODIELECTRIC PARTICLES ......................................................................... 34

Session 4 .............................................................................................................................................................. 35

SIMULTANEOUS ELECTROROTATION SYSTEMS TO DETERMINE THE MEMBRANE CAPACITANCE AND CYTOPLASM CONDUCTIVITY OF CELLS35 RAPID CELL VIABILITY AND ANTIMICROBIAL SUSCEPTIBILITY TESTING USING IMPEDANCE SPECTROSCOPY .......................................... 36 COMPUTATIONAL MODELING OF THE ELECTRIC FIELD DISTRIBUTION TOWARDS EXOSOME CHARACTERIZATION .............................. 37 NOVEL RECOGNITION AND TARGETING OF TEMOZOLOMIDE RESISTANT CELLS IN GLIOBLASTOMA ................................................... 38

Session 5 .............................................................................................................................................................. 39

MULTIDIMENSIONAL SORTING OF MIXED MICROPARTICLES VIA INSULATOR-BASED ELECTROKINETICS AND DIELECTROPHORESIS ........... 39 STATIONARY ELECTRO-OSMOTIC FLOW VORTICES ON INSULATING SURFACES INDUCED BY AC ELECTRIC FIELDS .................................. 40 COMBINING HYDRODYNAMIC AND DIELECTROPHORETIC TRAPPING TO CONTROL THE INTERACTION BETWEEN BEADS AND CELLS .......... 41 LATTICE BOLTZMANN SIMULATIONS OF DIELECTROWETTING ................................................................................................. 42 DIELECTROPHORESIS OF AIR ........................................................................................................................................... 43

Keynote ............................................................................................................................................................... 44

POTENTIALS OF DIELECTROPHORESIS FOR AN ANALYTICAL CHEMIST ....................................................................................... 44

Session 6 .............................................................................................................................................................. 45

OPTIMIZED ELECTROMANIPULATION BUFFER FOR ENHANCED CELL VIABILITY AND DIELECTROPHORETIC CONSISTENCY ..................... 45 DETECTION OF AUTOLOGOUS BLOOD TRANSFUSIONS USING DIELECTROPHORESIS .................................................................... 46 INTEGRATION OF ELECTRODES ON GLASS SUSPENDED MICROCHANNEL RESONATORS FOR DEP TRAPPING OF PARTICLES ..................... 47 ISOLATING AND CONCENTRATING UNLABELED SARS COV-2 FROM SALIVA WITH IDEP.............................................................. 48

Poster Session 2 ................................................................................................................................................... 49

SELECTIVE TRAPPING AND RETRIEVAL OF SINGLE CELLS USING MICROWELL ARRAY DEVICES COMBINED WITH POSITIVE- AND NEGATIVE-DIELECTROPHORESIS ..................................................................................................................................................... 49 TRANSITION TO HIGH-FREQUENCY (MHZ) AC ELECTROOSMOSIS OBSERVED ON THE NANOSCALE .................................................. 50 HIGH THROUGHPUT ASSAY TO MEASURE CELLULAR DIELECTROPHORETIC MOBILITY FROM INDIVIDUAL-BASED CELL MOVEMENT...... 51 ADHESION MEASUREMENT OF WATER DROPLETS ON SUPERHYDROPHOBIC SURFACES VIA ELECTRIC FIELDS ...................................... 52 DIGITAL DROPLET MICROFLUIDICS WITH PROGRAMMABLE LIQUID HANDLING BASED ON CONTACT CHARGE ELECTROPHORESIS ........ 53 NUMERICAL STUDY OF JANUS DROPLET SPONTANEOUSLY MIGRATION BY OPENFOAM .............................................................. 54 DIELECTROPHORESIS BASED SENSOR USING SURFACE CONDUCTIVITY DIFFERENCE OF PARTICLES .................................................... 55 SIDEWALL ELECTRODES IN A MICROCHANNEL FOR HIGH-THROUGHPUT DIELECTROPHORETIC SEPARATIONS ...................................... 56 AUTOMATED FLUORESCENCE QUANTIFICATION OF EXTRACELLULAR VESICLES COLLECTED FROM HUMAN PLASMA VIA

DIELECTROPHORESIS ..................................................................................................................................................... 57 A STUDY ON ALTERATIONS IN MEMBRANE PHYSIOLOGY OF ULTRASONICALLY IRRADIATED HUMAN ERYTHROCYTES THROUGH

DIELECTROPHORESIS ..................................................................................................................................................... 58 STUDYING THE DEP BEHAVIOR OF THE TRYPANOSOMA BRUCEI PARASITES ............................................................................... 59 MONITORING ERYPTOSIS USING DIELECTROPHORETIC CHARACTERIZATION .............................................................................. 60

Keynote ............................................................................................................................................................... 61

AC ELECTROKINETICS: APPLICATIONS FOR CELL BIOMECHANICS AND FATIGUE FAILURE .............................................................. 61

Session 7 .............................................................................................................................................................. 62

8

EFFECT OF DIELECTROPHORESIS ON COLLOIDAL CYLINDERS NEAR A PLANAR BOUNDARY ............................................................... 62 DIELECTROPHORETIC-DRIVEN DEFORMATIONS OF A LIQUID-FLUID INTERFACE ........................................................................... 63 FREQUENCY MODULATED DIELECTROPHORETIC PARTICLE CHROMATOGRAPHY ........................................................................... 64 TUNABLE NANOCHANNELS FOR DYNAMIC CONTROL OF THE CONCENTRATION-POLARIZATION-BASED PRECONCENTRATED ANALYTE PLUG

................................................................................................................................................................................. 65

Session 8 .............................................................................................................................................................. 66

DIFFERENT BLOOD CELLS EXHIBIT RHYTHMIC ELECTROPHYSIOLOGICAL BEHAVIOUR ..................................................................... 66 THE ELECTROME AND THE ROSETTA STONE ........................................................................................................................ 67 CHARACTERIZING THE HETEROGENEITY OF NEURAL STEM CELL POPULATIONS USEFUL FOR TRANSPLANTATION .............................. 68 IMPLICATIONS OF DIELECTROPHORETIC ISOLATION ON TRANSMEMBRANE POTENTIAL-INDUCED CELL DAMAGE .............................. 69 ELECTROPHYSIOLOGY-BASED STRATIFICATION OF TUMORIGENICITY AND DRUG SENSITIVITY OF PANCREATIC CANCER SUBPOPULATIONS

USING MACHINE LEARNING APPROACHES ........................................................................................................................... 70

Keynote ............................................................................................................................................................... 71

MICROPIPETTE DIELECTROPHORETIC DEVICE FOR RAPID PURIFICATION OF CIRCULATING SMALL EXTRACELLULAR VESICLES ............... 71

Session 9 .............................................................................................................................................................. 72

CONTROLING THE LOCATION OF AN ELECTROKINETICALLY PRECONCENTRATED PLUG OF BIOMOLECULES.......................................... 72 FROM LAB-ON-A-CHIP TO LAB-ON-A-PARTICLE ELECTRICALLY-POWERED PLATFORMS ................................................................. 73 WHY 3D ELECTRODES FOR DIELECTROPHORESIS? A CRITICAL REVIEW ON THE FABRICATION TECHNIQUES THAT CAN ENABLE HIGHER

THROUGHPUT IN DEP DEVICES ........................................................................................................................................ 74 NONLINEAR ELECTROKINETICS FOR ASSESSMENT OF BIOPARTICLES IN MICRODEVICES .................................................................. 75 ELUCIDATING CELL PHENOTYPE BASED ON BIOPHYSICAL ANALYSIS OF SECRETED SUBCELLULAR BODIES ............................................ 76

Poster Session 3 ................................................................................................................................................... 77

HIGH-RESOLUTION 3D-PRINTED INSULATOR-BASED DIELECTROPHORESIS DEVICES TOWARDS MANIPULATION OF BIOANALYTES ....... 77 A TUNABLE INSULATOR-BASED DIELECTROPHORESIS SYSTEM FOR THE SEPARATION OF BIOMOLECULES ........................................ 78 CYLINDRICAL AND TEARDROP SHAPED ACTIVE POSTS IN DETERMINISTIC LATERAL DISPLACEMENT DEVICES ....................................... 79 DETERMINISTIC IDEP RATCHET DEVICES FOR HIGH-THROUGHPUT ORGANELLE SEPARATION ....................................................... 80 SEPARATION OF DIPLOID AND TETRAPLOID CANCER CELL POPULATIONS USING HIGH-FREQUENCY DIELECTROPHORESIS ...................... 81 A MICROFLUIDIC DEVICE FOR CHARACTERIZING VARIABILITY OF BETA-GALACTOSIDASE IN SINGLE MDA-MB-231 CELLS .................... 82 DC G-IDEP TRAPPING OF GOLD NANOPARTICLES ............................................................................................................... 83 TOWARDS SEPARATING MICROPLASTIC PARTICLES WITH INSULATOR-BASED DIELECTROPHORESIS ................................................ 84 A NUMERICAL INVESTIGATION TO EXTEND QUANTITATION OF GRADIENT-INDUCED FORCES WITHIN AN INSULATOR-BASED SAWTOOTH

DESIGN ...................................................................................................................................................................... 85

9

PLENARY Protein Dielectrophoresis: Status of Experiments and Theory

Ronald Pethig ron.pethig@ed.ac.uk

Institute for Integrated Micro and nano Systems

The University of Edinburgh United Kingdom

Dielectrophoresis (DEP) studies have progressed from the microscopic scale of cells and bacteria, through the mesoscale of virions to the molecular scale of DNA and proteins [1]. Beginning in 1994 with the pioneering work of Washizu et al [2] at least 20 different globular proteins have now been studied for their DEP responses [3]. Analyses of these responses have employed the so-called Clausius-Mossotti (CM) function, even though its theoretical derivation arguably fails to describe the situation for nanoparticles such as proteins that possess an intrinsic dipole moment and other subtle physico-chemical attributes [4, 5]. The CM function of DEP is in fact an analogue of (but not the same as) the CM-relation that formed the bedrock of classical dielectric theory [e.g., 6] used to describe the electrical polarization of proteins [e.g., 7]. In this presentation the derivations of the molecular CM-relation of dielectrics and the macroscopic CM function of DEP will be summarized and discussed in terms of a critical assessment of the work reported to date for proteins [8]. The current approaches [9, 10] for an evolving theory of protein DEP will also be presented. References: 1. Kim, D., Sonker, M. and Ros, A., Dielectrophoresis: From molecular to micrometer-scale analytes, Anal. Chem., 91: 277-295 (2019). 2. Washizu, M., Suzuki, S., Kurosawa, O., Nishizaka, T. and Shinohara, T., Molecular dielectrophoresis of biopolymers, IEEE Trans. Ind. Appl. 30(4): 835-843 (1994). 3. Hayes, M. A., Dielectrophoresis of proteins: Experimental data and evolving theory, (2020) - in press. 4. Pethig R., Limitations of the Clausius-Mossotti function used in dielectrophoresis and electrical impedance studies of biomacromolecules, Electrophoresis, 40: 2575-2583 (2019). 5. Matyushov D.V., Electrostatic solvation and mobility in uniform and non-uniform electric fields: From simple ions to proteins, Biomicrofluidics, 13: 064106 (2019). 6. Debye, P., Polar Molecules: The Chemical Catalog Co., New York (1929). 7. Oncley, J. L., The investigation of proteins by dielectric measurements, Chem. Rev. 30: 433-450 (1942). 8. Holzel, R. and Pethig, R., Protein Dielectrophoresis: Status of experiments and an empirical theory, Micromachines 11, 533 (2020). 9. Holzel, R. and Pethig, R., Protein Dielectrophoresis: Key dielectric parameters and evolving theory, Electrophoresis, DOI: 10.1002/elps.202000255 (2020). 10. Heyden, M. and Matyushov, D.V., Dielectrophoresis of proteins in solution, J. Phys. Chem. B., 51, 11634-11647 (2020).

10

SESSION 1

AC electrokinetics for the immobilisation of nanoparticles and molecules Ralph Hölzel, Xenia Knigge, Eva-Maria Laux, Mareike Noffke, Sandra Stanke, Christian Wenger, and

Frank F. Bier ralph.hoelzel@izi-bb.fraunhofer.de

Biosystems Integration and Process Automation / Biomolecular Nanostructures and Measurement

Technology Fraunhofer Institute for Cell Therapy and Immunology, Branch Bioanalytics and Bioprocesses, Potsdam (IZI-

BB), Am Mühlenberg 13, 14476 Potsdam, Germany Germany

A universal prerequisite for the preparation of biosensors is the functionalisation of the sensor surface. This can be conveniently accomplished by the combination of AC electrokinetics with micro- and nanoelectrodes leading to the attraction and immobilisation of bioparticles and molecules at a conductive surface. Classical dielectrophoretic theory calls for typical electrode geometries in the order of the particle size, i.e. for proteins well below a micrometer, preferentially a few nanometers. For this purpose, we have developed quite different electrode structures: Interdigitated electrodes with gap widths below 1 µm, planar triangular electrodes with mutual distances of around 100 nm, and regular arrays comprising up to 1 million pin-like electrodes with tip diameters ranging from 500 nm down to the size of a single protein molecule, i.e. 10 nm or less. Successful immobilisation has been achieved for quite different nanoparticles like polystyrene nanospheres, viruses and exosomes, as well as for dissolved molecules: single molecules of the autofluorescent protein R-phycoerythrin [1], the enzyme horseradish peroxidase [2], antibodies [3] and small organic dye molecules [4]. Nanospheres have been immobilised deterministically as individual singles in regular arrays [5]. Observation and proof of a successful immobilisation on the nanoscale are challenging. Therefore, quite different microscopical techniques are employed: optical, scanning force as well as scanning electron microscopy. Fluorescence microscopy shows that protein function is preserved in the course of DEP immobilisation. Fluorescence polarisation microscopy reveals the immobilisation of eGFP in a properly aligned manner and allows to determine the orientation of the protein’s fluorescing subunit in relation to the whole molecule [6]. Metal containing dyes can be localised by X-ray spectroscopy in the electron beam [4]. Results are discussed in the light of recent hypotheses about the actual interaction mechanisms between dissolved molecules and polarising fields. [7] [1] R. Hölzel, N. Calander, Z. Chiragwandi, M. Willander, F. F. Bier, Phys. Rev. Lett., 2005, 95, 128102. [2] E.-M. Laux, U. Kaletta, F.F. Bier, Ch. Wenger, R. Hölzel, Electrophoresis, 2014, 35, 459. [3] S. Otto, U. Kaletta, F.F. Bier, Ch. Wenger, R. Hölzel, Lab Chip, 2014, 14, 998-1004. [4] E.-M. Laux, C. Wenger, F. F. Bier, R. Hölzel, 2020, 412, 3859. [5] X. Knigge, Ch. Wenger, F. F. Bier, R. Hölzel, J. Phys. D: Appl. Phys., 2018, 51, 065308-065318. [6] E.-M. Laux, X. Knigge, F.F. Bier, Ch. Wenger, R. Hölzel, Small, 2016, 12, 1514-1520. [7] R. Hölzel, R. Pethig, Electrophoresis 2020, (https://doi.org/10.1002/elps.202000255).

11

Dielectrophoretic filtration for selective particle recovery at high throughput Georg R. Pesch, Malte Lorenz, Michael Baune, Jorg Thöming

grp@uni-bremen.de grp@uni-bremen.de

Chemical Process Engineering (CVT), Faculty of Production Engineering University of Bremen

Germany

Separation of small particles according to properties such as material, form, or size is a difficult task with many applications. For example, the recovery of precious metals from waste streams requires a material-selective separation of particles. Traditional methods (e.g., separation by density) fail for sub-micron and micron-sized particles. State-of-the-art waste recovery produces large amounts of dust that contains highly valuable materials that are currently lost. In this talk, we utilize dielectrophoretic (DEP) filtration, a versatile particle separation technique based on polarization effects which is capable of solving such difficult separation tasks. In general, the dielectrophoretic force, i.e., its direction and strength, depends not only on particle’s volume but also on its relative polarizability in the suspension medium (mainly defined by the particle material) and dielectrophoresis can thus be used to move, separate, and fractionate particles. DEP is inherently limited to small length scales as the DEP force depends on the gradient of the electric field. Consequently, small electrode designs are required to produce large enough gradients to act on small particles limiting overall channel dimensions. Thus, so far, dielectrophoresis has mostly been applied in microfluidic systems for (bio-)analytical chemistry at low flow rates and at very high discriminatory ability. Application of DEP at preparative or even industrial scale is rare. In this talk, recent advances in DEP filtration are presented. In DEP filtration, the field gradient is decoupled from the channel dimensions thus allowing for much higher throughputs by scattering an originally homogeneous field at the solid-liquid interface of a macroscopic and highly porous dielectric. Then, particles can be selectively trapped from a feed suspension at the electric field maxima in the filter. This allows for (switchable) particle separation at several orders of magnitude larger flow rates compared to conventional DEP designs. In this talk, we address both selectivity as well as recovery in selective separation of metal particles from mixtures containing plastic and metal oxides. We will further discuss future directions for DEP filtration.

12

Dielectric Properties of Infected Porcine Kidney Cells following Glycine Treatment Sanaz Habibi, Pratik U. Joshi, Xue Mi, Caryn L. Heldt, and Adrienne R. Minerick

habibi@mtu.edu minerick@mtu.edu

Department of Chemical Engineering University of Michigan, Michigan Technological University

USA

Evaluating and monitoring viral infections is of critical importance to develop and screen new pharmaceutical drugs. Osmolyte glycine, a naturally occurring organic compound, has been shown to significantly reduce porcine parvovirus (PPV) infection in porcine kidney (PK-13) cells. It is postulated that glycine stops the PPV infection cycle by disturbing capsid assembly into viable virus particles. In this study, dielectrophoresis (DEP) was utilized to further investigate host/virus/osmolyte (i.e. PK-13/PPV/glycine) mechanisms. Dielectrophoretic frequency spectra were compiled from cell velocities to determine DEP force and crossover frequency (fco) for non-infected (control), PPV-infected, glycine-treated/non-infected (glycine control), and glycine-treated/PPV-infected PK-13 cells at 0, 1.5, 4, 8, and 10 hours. Sinusoidal voltages with 5 Vpp were applied to a star-shaped electrode configuration to create a non-uniform electric field. The distance between adjacent, alternating electrodes was 50 µm, resulting in the electric field density of 0.1 Vpp/µm. Frequencies between 0.1-0.9 MHz were examined, within which Maxwell–Wagner polarization is known to occur. Non-infected PK-13 cells had a fco between 0.48-0.46 MHz reproducibly from 0-10 hours. Upon PPV infection, the DEP spectra were comparable to those for non-infected PK-13 cells at 0 and 1.5 hours. However, the fco gradually shifted to 0.4-0.38 MHz at 4 hours post-infection. The fco further decreased to 0.27-0.29 MHz 10 hours post-infection. During the first 4 hours of glycine treatment, the DEP spectrum of PPV-infected cells was similar to that obtained in the absence of glycine. This suggests that glycine did not stop PPV penetration into the PK-13 cell. However, after 8 and 10 hours, the glycine-treated/PPV-infected cells exhibited attenuated fco shifts down to 0.31-0.35 MHz at both 8 and 10 hours. Two models, the single-shell spherical DEP polarization and a modified theoretical model were utilized to extract the PK-13 cell dielectric properties from fco measurements. With infection, the single-shell spherical model showed a 70% increase in cell membrane capacitance over the 10-hour experiment. Glycine lowered this elevated cell membrane capacitance by 20.5%. These results were in good agreement with those derived by the modified theoretical model. The modified theoretical model extracted cells’ dielectric properties through optimizing an expression for the medium conductivity in terms of fco. This model revealed that cytoplasm conductivity of PPV-infected PK-13 cells slightly increased from 1.81 S/m to 1.85 S/m after 10 hours. This trend mirrors the rise in the cytoplasm following the glycine treatment of the PPV-infected PK-13 cells. The cytoplasm conductivity was deduced to increase only from 1.81 S/m to 1.83 S/m after 10 hours of glycine treatment. These variations are not as significantly different when compared to the dramatic change in cell membrane capacitance, but do provide insights into cytoplasm changes in addition to membrane capacitance. The results from this work are remarkably insightful since they indicate DEP potential to be used as a mechanistic viral diagnostic as well as high throughput screening of new anti-viral pharmaceuticals.

13

Neural networks meet impedance cytometry Paolo Bisegna, Riccardo Reale, Carlos Honrado, Nathan Swami, and Federica Caselli

caselli@ing.uniroma2.it caselli@ing.uniroma2.it

Department of Civil Engineering and Computer Science University of Rome Tor Vergata, Rome, Italy

Italy

We present our recent activity dealing with the use of neural networks for fast processing of impedance cytometry data streams. We believe that this combination of microfluidic impedance cytometry and machine learning can serve as a stepping stone to real-time single-cell recognition and sorting. Microfluidic impedance cytometry is a label-free electrical sensing technique for the analysis and characterization of single particles and cells, such as mammalian cells, yeast cells, bacteria, droplets, algae and pollen grains. In last years, the technique has been evolving from an analysis essentially based on signal amplitude at limited bandwidth, towards novel multiparametric characterization approaches. This calls for tailored strategies of signal processing to extract the information which is embedded in the electrical fingerprints, potentially in real-time to drive online sorting and enrichment. Towards this vision, we explore inline recognition of phenotypes using a combination of microfluidics and machine learning. In particular, we have considered three case studies as detailed below. The first application uses an impedance chip design for multiparametric data extraction. The measured signal waveforms are fed to a 4-layer recurrent neural network (RNN), which is suited to process data that have a temporal dimension. The network is trained to predict cell size, velocity, and position. The trained network is able to characterize with good accuracy beads, red blood cells, and yeasts, with a unitary prediction time of 0.4 ms. The second application deals with the problem of particle coincidences, i.e., two or more particles visiting the sensing zone in close proximity, which limits the throughput and accuracy of impedance cytometers and Coulter-type devices. In order to solve this problem, we propose an approach based on an impedance chip with two electrical sensing zones, whose signals are processed by a classification RNN that predicts the number of coinciding particles, followed by a regression RNN that extract the features of each particle. Finally, we present a multimodal (electro-optical) approach to solve cell classification problems via microfluidics and machine learning. Electrical signals and synchronized optical images are processed by two independent machine learning-based classifiers, whose predictions are then combined to provide the final classification outcome. We demonstrate a proof-of-concept experiment dealing with the classification of eight classes from different pollen species.

14

POSTER SESSION 1

Microfluidic arrays combining dielectrophoretic and hydrodynamic forces for trapping and retrieval of selected cells

Pierre-Emmanuel Thiriet, Joern Pezoldt, Gabriele Gambardella, Kevin Keim, Bart Deplancke, and Carlotta Guiducci

pierre-emmanuel.thiriet@epfl.ch carlotta.guiducci@epfl.ch

CLSE EPFL

Switzerland

Personalized medicine appears as one of the next breakthrough in the healthcare field. As each patient requires unique profiling, a microfluidic platform that permits analysis, retrieval and re-expansion of single cells could open new avenues for diagnostics and single cell manipulation. Even though hydrodynamic trapping has proven to be a reliable and efficient tool for single cell arraying and over-time analysis, the off-chip recovery of individual cells released from such traps is still a perilous task. To overcome this current limitation, we integrated three-dimensional electrodes within hydrodynamic traps so that electric fields created by those electrodes can gently and selectively push a single cell out of the trap through insulator-based dielectrophoresis (iDEP). iDEP is an approach for cell manipulation commonly based upon application of a high electric voltage at both inlet and outlet of the chip. Curvature of electric field lines around microfluidic structures within the chip induces a gradient distribution of charges causing the appearance of a DEP force. This technique offers benefits in terms of cost and versatility for structure design, but it suffers from electrolysis and fouling issues when using high electric potentials. To overcome this challenge we miniaturized electrodes to bring them in close vicinity to the cell trapping area. While it is common for DEP to place cells in low conductivity medium, we can effectively release cells using minimal electric potential and keep them in their culture medium. The electrodes integrated close to the trap are vertical standing electrodes based on SU-8 photoresist. SU-8 pillars are patterned on the wafer and then covered with metal that is etched everywhere but on the pillars walls. This approach enables the integration of 3-D electrodes within the microfluidic with micron-range accuracy. After selection and retrieval from their trapping site, individual cells can be guided towards a recovery well using novel hybrid SU-8/PDMS pneumatic valves. To ensure that cellular viability is maintained subsequent to DEP, we carried out through mRNA sequencing a comparison of cells phenotypes before and after injection in the microfluidic platform. Phenotype changes induced by handling the cell in the microfluidic device and by an exposure to DEP forces could be for the first time studied independently, revealing minor changes of the molecular profile upon DEP forces application, underlined by equivocal minor transcriptional modifications induced by the injection in the microfluidic device.

15

DEP-on-CMOS: a microfluidic platform for dielectrophoretic manipulation, trapping, and

differential separation of viable and non-viable yeast cells Honeyeh Matbaechi Ettehad, and Christian Wenger

matbaechi@ihp-microelectronics.com wenger@ihp-microelectronics.com

Material, Adaptive Material IHP–Leibniz-Institut für innovative Mikroelektronik, Im Technologiepark 25, 15236 Frankfurt/Oder, Germany

Germany

Dielectrophoresis (DEP) is a widely used technique for the detection of early-stage diseases in microfluidic lab-on-chip (LOC) applications. This technique has the advantage of using label-free samples and, as a result, less complex systems for the experiments, which makes them more favourable for point-of-care (POC) diagnostics. The integration of microfluidic channels with complementary metal-oxide-semiconductor (CMOS) can scale down multiple-stage laboratory procedures in a single chip and process micro and nano-liters samples within a fully isolated manner. In this work, we reported a CMOS-based microfluidic platform integrated with interdigitated electrodes (IDEs) for manipulating, trapping, and separating viable and non-viable yeast cells using DEP technique. Herein, the DEP spectrum analysis of various cell suspensions with different medium conductivities (ranging from 10-4 to 1) was investigated thoroughly by finite element simulation and experimentally. Trapping experiments with viable and non-viable yeast cells suspended in aqueous solutions were then conducted, showing that cells can be trapped at the electrodes by positive DEP force. Depending on the relative permittivity of the cells and medium flown over electrodes as well as amplitude and frequency of the AC, yeast cells exhibit translational movement in two opposite directions (positive and negative). This method allowed us to switch the propulsion direction of cells by varying the frequency of the applied electric field. DEP force-frequency profile indicated that changing frequency influences the cell trajectory and impacts the trapping yield (cell accumulation at the electrode) over the pDEP band. It was also seen that the evolution of the pDEP disappears in high conductivity suspending mediums. Beyond the achievements of cell manipulation and trapping, this system permits effective control of different cells simultaneously. Upon distinct DEP responses of viable and non-viable cells, high-throughput cell isolation and separation could also be achieved using the same structure in a noninvasive format. In this content, the desired cells absorbed to high electric field intensity regions and isolated at the electrodes. In contrast, undesired cells moved to low-intensity electric fields by nDEP and repelled from the electrodes. Thus, the undesirable cells were carried in the stream of the fluid flow and released from the microfluidic channel and collected at the outlet. This resulted in continuous-flow separation of a cell type from another type in the mixture. The proof-of-concept CMOS integrated microfluidic device with interdigitated electrodes ensured the cells' characterization and manipulation for detection and separation applications. The differential separation approach was found to work sufficiently well to be suited for implementing the ultimate design. These validations showed promise towards using this microfluidic device in a more complex LOC platform for biological and clinical applications. Monolithic integration of device and its potentiality for sample preparation, trapping, separation and analysis on a single platform as well as operational simplicity offer excellent benefits from a cost and commercialization perspective.

16

AC field assisted deposition of influenza viruses on nanoelectrodes Sandra Stanke, Christian Wenger, Frank F. Bier, Ralph Hölzel

sandra.stanke@izi-bb.fraunhofer.de Ralph.Hoelzel@izi-bb.fraunhofer.de

Biomolecular Nanostructures and Measurement Technology Fraunhofer Institute for Cell Therapy and Immunology, Branch Bioanalytics and Bioprocesses

Germany

A rapid characterization of viruses and virus subtypes is of great biomedical interest. Here we present the use of AC electrokinetic forces, like dielectrophoresis and AC electroosmosis, as a simple and fast method to functionalize nanoelectrode arrays as a potential biosensor. The permanent immobilization of polystyrene nanoparticles, antibodies and other proteins on electrodes has already been demonstrated [1, 2, 3]. The sensor itself consists of four individual arrays, each built up of 6256 tungsten nanoelectrodes with a diameter of 500 nm each. The immobilization, detection and characterization of influenza material is done without any prior chemical modification of the electrode surface. The accumulation of virus material over time has been observed, showing that the largest amount has already been drawn to the electrodes within 60 seconds and reached a saturation after 180 seconds of applied AC electric field. Due to side effects such as fluid streaming, a concentration gradient is created decreasing from the outer to the inner electrodes. It has been demonstrated, that the virus material is permanently immobilized even after switching off the electric field. Furthermore, each functionalized electrode can be considered as a single event. Comparing these single events it seems like the virus material is distributed randomly across the nanoelectrodes. But after deconvolving the fluorescence image and merging the images of around 100 electrodes it reveals that the major part of virus material is collected at the electrode edge. This is in line with theory, as this is the region of the highest field gradient, and thus here the AC electrokinetic forces have the greatest impact on the sample. The universal chip design does not limit the application to influenza viruses but also works for different viruses, bacteria, parasites or any other object that can be manipulated by AC electrokinetic forces. Additionally, each electrode can be used as part of an on-chip resonant circuit, whose frequency changes with surface coverage of the electrode [4] and, hence, serves as a measure of the amount of viruses attached to the electrodes. So in future, the evaluation by fluorescence microscopy can be changed to an electrical evaluation. Thus, combined with microfluidics this chip has the potential for a small and rapid Point-of-care system. References [1] X. Knigge, Ch. Wenger, F. F. Bier, and R. Hölzel, J. Phys. D: Appl. Phys., 2018, 51, 065308. [2] S. Otto, U. Kaletta, F. F. Bier, Ch. Wenger, and R. Hölzel, Lab Chip, 2014, 14, 998. [3] E.-M. Laux, X. Knigge, F. F. Bier, Ch. Wenger, and R. Hölzel, Small, 2016, 12, 1514-1520. [4] S. Guha, K. Schmalz, Ch. Wenger, and F. Herzel, Analyst, 2015, 140, 3262-3272.

17

AC electric field mediated preparation of regular enzyme arrays and their functional

characterization Mareike Noffke, Christian Wenger, Frank F. Bier, Ralph Hölzel

Mareike.Noffke@izi-bb.fraunhofer.de Ralph.Hoelzel@izi-bb.fraunhofer.de

Biomolecular Nanostructures and Measurement Technology Fraunhofer IZI-BB

Germany

For the observation of single enzyme molecules, the molecules are immobilized or entrapped as isolated singles. In most cases this is achieved by the dilution of the enzymes before entrapment or immobilization or by the creation of a low surface concentration of acceptors for enzyme binding. Singling then follows a Poisson distribution and only a few molecules can be studied. In this work, a new platform for a parallel, label free, deterministic singling of active enzyme molecules is developed. Enzyme molecules are immobilized by dielectrophoresis on pin-like electrodes arranged in planar arrays. The electrodes are available in different materials, shapes and tip diameters ranging from 500 nm down to about 1 nm. Thousands of electrode tips are arranged in each array, allowing to perform many experiments in parallel. So far, we successfully immobilized nanoparticles deterministically as singles (Knigge et al, 2018: J. Phys. D: Appl. Phys. 51, 065308). Immobilization of clusters of the enzyme horseradish peroxidase in its active form was also successful (Laux et al, 2014: Electrophoresis 35, 459-466) and the activity of immobilized enzymes was estimated. These arrays are being characterized and optimized to reach the singling of active enzyme molecules. For real-time monitoring of enzyme activity on individual electrodes, laser scanning microscopy and fluorescence correlation spectroscopy are used. For a quantitative interpretation of the results, experimental side effects like photooxidation and adsorption of fluorescent product molecules to surfaces have to be accounted for. In order to reduce side effects, a microfluidic setup is being developed.

18

AC electrokinetic immobilization of K562 exosomes on nanoelectrode arrays

Eva-Maria Laux (1), Christian Wenger (2, 3), and Ralph Hölzel (1) eva-maria.laux@izi-bb.fraunhofer.de eva-maria.laux@izi-bb.fraunhofer.de

Department: Biosystems Integration and Process Automation, Group: Biomolecular Nanostructures and Measurement Technology

(1) Fraunhofer Institute for Cell Therapy and Immunology, Branch Bioanalytics and Bioprocesses, Potsdam (IZI-BB), Am Mühlenberg 13, 14476 Potsdam, Germany. (2) IHP - Leibniz Institut fuer innovative

Mikroelektronik, Im Technologiepark 25, 15236 Frankfurt (Oder), Germany. (3) Brandenburgische Technische Universität Cottbus-Senftenberg, Platz der Deutschen Einheit 1, 03046 Cottbus, Germany.

Germany

The forces occurring in non-uniform AC electric fields have been exploited extensively for the directed movement of various particles, and for particle immobilization on electrode surfaces. Here, fluorescently labelled exosomes isolated from K562, a myelogenous leukemia cell line, are immobilized by AC electric fields on silicon nanoelectrode arrays. These arrays consist of more than 700 000 free-standing, truncated cone electrodes with 30 nm to 50 nm tip diameters. The counter electrode is integrated as a top layer of titanium nitride leaving concentric holes around each tip. Exosomes are vesicles that are secreted from most cell types, which usually contain specific mixtures of, e. g., proteins, lipids and nucleic acids. Since their typical sizes range from 40 nm to 120 nm, immobilization of few or even single exosomes on the tips with diameters of 30 nm to 50 nm appears feasible, as has previously been investigated by Knigge et al. [1]. In the results presented here, permanent immobilization of exosomes is accomplished, and the number of immobilized exosomes is evaluated from fluorescence intensities located at the electrodes. The high number of electrodes within an array enables parallel analysis of about eight thousand electrodes comprised in a single fluorescence image. The composition of an exosome’s content can be characterized using Raman scattering. During the development of a disease, the composition of the exosome content can change significantly, which will produce changes in the Raman spectra. Thus, exosomes can be used as markers for early detection, diagnosis and monitoring of diseases. With arrays of immobilized exosomes, characterization on a single-exosome basis is possible. [1] X. Knigge, C. Wenger, C., F. F. Bier, R. Hölzel, Dielectrophoretic immobilisation of nanoparticles as isolated singles in regular arrays. Journal of Physics D: Applied Physics 2018, 51, 065308.

19

Electrorotation of single cells for the analysis of membrane damage induced by toxins mimicking the neurodegenerative effect of amyloid beta in the Alzheimer's disease

Till Ryser, Kevin Keim, Guillaume Pillet, Widad Lahbichi, Anne-Laure Mahul-Mellier, Hilal Lashuel and Carlotta Guiducci till.ryser@epfl.ch

carlotta.guiducci@epfl.ch CLSE

École polytechnique fédérale de Lausanne (EPFL) Switzerland

Toxic aggregates of the protein amyloid beta (Aβ) are hypothesized to be the primary cause for Alzheimer's disease (AD). At present, diagnosis is largely based on symptoms, however, the early detection of toxic Aβ species could facilitate the study and treatment of the disease. In our study, we use electrorotation to monitor changes in the membrane capacitance of neuronal cells to assess the neurotoxicity of toxins related to AD and aim for a diagnostic tool to evaluate membrane damage induced by Aβ. The principle of electrorotation makes use of the fact that cells polarize when an AC electric field is applied. Upon rotation of the field, the polarized cell tries to align with the electric field and starts spinning around its z-axis. The rotation speed, dependent on the applied electric field frequency, can be fitted with the Clausius Mossotti factor (Gascoyne et al., Bioelectrochemistry and bioenergetics, 1995), uncovering parameters describing the integrity of the cell membrane, such as the membrane capacitance. A possible approach to control individual single cells in a microfluidic channel is to capture cells in cages by means of electrokinetic forces and analyze their rotational behavior (Rohani et al., Electrophoresis, 2014). With the help of three-dimensional electrodes developed in our lab (Kilchenmann et al., JMEMS, 2016), we designed and fabricated a device able to analyze multiple single cells in parallel. In previous works, our lab showed the capability of the system to distinguish different cell lines with similar diameters by comparison of their membrane capacitance (Keim et al., Electrophoresis, 2019). Furthermore, we investigated the impact of the changes in osmolarity of the medium on the membrane capacitance to further optimize our system (Keim et al., µTAS, 2019). In the next step, we tried to quantify the membrane toxicity of various neurotoxins replicating the damage induced by the protein Aβ in vitro. Aβ has been shown to interrupt the internal ion homeostasis of neuronal cells by forming pores into the peptide bilayer (Lashuel et al., Nature, 2002) and having a detergent-like effect by integrating into the membrane (Bode et al., JBC, 2019). Hence, the membrane integrity is compromised, which affects the membrane's capability to separate charges, possibly resulting in a change in membrane capacitance. Small molecules, such as Miltefosine and Lovastatin, can be used to mimic the lipid bilayer perturbation induced by amyloid beta. Miltefosine is a synthetic phospholipid analogous to phosphatidylcholine, the major constituent of the cell membrane, and it inhibits the de novo membrane synthesis. Lovastatin is an inhibitor of the HMG CoA synthase, an enzyme responsible for the synthesis of cholesterol. After incubating the neuroblastomal cell line M17 with the toxins off-chip, we introduce the cells in our device to characterize the membrane capacitance. Overall, our system aims at being a label-free approach for rapid evaluation of neurotoxicity of proteins found in the cerebrospinal fluid enabling early detection of AD.

20

Label-free cell sorting in flow with real-time modification of separation parameters Gloria Porro, Rita Sarkis, Kevin Keim, Olaia Naveiras and Carlotta Guiducci

gloria.porro@epfl.ch carlotta.guiducci@epfl.ch

Laboratory of Life Sciences Electronics - CLSE École polytechnique fédérale de Lausanne - EPFL

Switzerland

Deterministic lateral displacement (DLD) is an established technology to continuously sort particles in flow by mechanical properties. Rows of pillars are embedded in a microchannel, arranged at an angle with respect to the flow direction. Particles under the critical size of the system, which is intrinsically determined by its geometry (post size, post gap, array displacement angle), are transiting the device following the microfluidic flow lines. Larger particles, instead, are displaced from their original flow line and follow the array displacement angle, resulting in a change of trajectory which enables their separation. DLD is a powerful tool for particle separation, exhibiting large throughput and purity. Nonetheless, one main drawback is the lack of versatility, being the sorting dynamics purely mechanical and pre-determined by the array structure. Combining dielectrophoretic effects with passive DLD devices enables to tune the critical size of the system. Previous studies have exploited planar electrodes, either placed at the inlet and outlet of the microfluidic channel to reduce the critical size through positive DEP (Beech et al., Lab Chip 2009) or at the side boundaries by means of negative DEP (Calero et al., Lab Chip 2019). The CLSE laboratory recently introduced a microfabrication process to embed three-dimensional electrodes in microfluidics for dielectrophoretic applications (Kilchenmann et al., JMEMS 2016). These active pillars can be employed to replace the passive DLD posts obtaining a DEP-assisted DLD device, where the particles’ trajectories can be electrokinetically tuned in real-time. This approach - DLD-DEP with active 3D posts - has been validated on polystyrene beads in low conductivity media (Beech et al., Adv. Mater. Technol. 2019). Applying an AC signal to the three-dimensional electrodes, a local dielectrophoretic force is exerted on the particles. If compared to the use of planar electrodes, a homogeneous DEP force throughout the entire channel thickness enables for devices with high volumetric capacity, therefore enhancing the system throughput. Moreover, being the electric field delivered locally, the DEP force to deviate the particles can be set with a minimal applied voltage (few tens of volts). This makes our technology particularly suitable to perform bioparticle sorting directly in the original sample media since the application of low voltages allows to minimize bubble generation by electrolysis, often observed in high conductivity solutions. We recently introduced new designs for the sorting platform based on DLD-DEP with active 3D posts. We fabricated devices with active posts up to 100 µm in height and 4 in aspect ratio, to sort multiple cell species directly in culture media. We are developing systems in which different signals can be delivered to different regions of the active pillars array, thus enabling for the simultaneous and tunable sorting of more than two distinct populations by variation of the applied voltage. Our technology can be eventually employed as a powerful and versatile tool to sort cells in high conductivity media, with a method that can be easily implemented as a module in more complex microfluidic devices for bioanalytics.

21

A study of optimal electrode design for electric field driven target capture in SPR biosensors

Oleh Andreiev, Marion Costella, Marie Frénéa-Robin, Julien Marchalot, Antonio Ramos, Paul G Charette, Michael Canva and Jean-Pierre Cloarec

oleh.andreiev@ec-lyon.fr oleh.andreiev@ec-lyon.fr

Laboratoire Ampère; Laboratoire Nanotechnologies Nanosystèmes; Institut des Nanotechnologies de Lyon École Centrale de Lyon; Université de Lyon; Université de Sherbrooke

France; Canada

SPR (Surface Plasmon Resonance) has become a widely adopted technology for the detection of biomolecular species and their interactions, because of its label free and real time nature. Moreover, SPR imaging (SPRI) enables the visualization of hundreds of active spots simultaneously, thereby enabling high-throughput screening of drugs and biomarkers. Yet, SPR biosensors need improvement for the detection of analytes present at very low concentration in the sample. Since SPR detects changes in refractive index in the immediate vicinity of the sensor surface (~200 nm), an efficient mass transfer must be established between the bulk and the surface in the case of very low concentration targets to increase the sensor response that is otherwise limited by passive diffusion. A promising approach to increase the concentration of target species on the detection surface and improve sensor response is to exploit electrokinetic active mass transport, specifically dielectrophoresis (DEP) and AC electroosmosis (ACEO). To this end, the metallic film used for SPR sensing can be structured and used as electrodes to generate the electric field driving electrokinetic effects. Two main electrode designs are explored by our team, namely the interdigitated (IDE) and face-to-face (F2F) configurations. While there are numerous reports on particle trapping and fluid mixing using IDE, there are few studies relating to ACEO effects produced with the F2F configuration. In this work, we study the trapping efficiency of both setups with further investigation of the F2F configuration using numerical simulation compared to experimental results. Based on the literature on electrokinetic effects, we developed a numerical model for predicting induced flow behavior. Results from bright field microscopy experiments and simulations tend to show the superiority of the F2F setup regarding target concentration ability. Optimal parameters for this configuration were also determined. Then, the proposed optimal setup was tested on a SPR bench and showed a significant reduction in the detection limit as well as a faster sensor response time. All in all, the proposed system may lead to improved SPR biosensing with more sensitive and rapid detection

22

Modified red blood cells as multimodal standardized particles for benchmarking cell electrophysiology

Armita Salahi, Carlos Honrado, Aditya Rane, Nathan Swami as7kf@virginia.edu

nswami@virginia.edu Electrical and Computer Engineering

University Of Virginia USA

Cellular systems often display a degree of phenotypic heterogeneity that has important implications on biological function and disease response. Thus, there is increasing interest in methods for label-free biophysical analysis and separation of single-cells, i.e., based on their inherent biophysical properties. AC electrokinetics represents a group of methods capable of performing such label-free analysis and separation based on the electrophysiology of single cells. Cell electrophysiology represents an aggregate of biophysical properties that are influenced by genomic and micro-environmental factors, being not only sensitive to whole-cell characteristics, such as size or shape, but also subcellular features, such as plasma membrane structure or cytoplasmic organization. However, when developing novel approaches for AC electrokinetics methods (e.g., dielectrophoresis or impedance cytometry), there is a lack of reliable, reproducible model particles with well-tuned electrophysiological phenotypes that may be compared versus other relevant cell types. There are two common model particles. One of them is plastic beads, which besides good approximations to cell volume, lack any other similarities to a live cell. The other model cells frequently used are yeast cells, which have a wide range of size and shape distributions. Hence, a central need for the advancement of cytometry is the availability of model cells with modulated properties for application as “standards” to compare against each cell type for enabling accurate and rapid phenotypic recognition. In this work, we systematically alter the subcellular electrophysiology of RBCs by using glutaraldehyde fixation to vary membrane capacitance and by electrolyte penetration from the buffer to vary cytoplasmic conductivity. Glutaraldehyde is an agent that crosslinks proteins in the cell membrane and cytoplasm. Hence, the increase in electrical opacity (the ratio of high to low-frequency impedance magnitude) with higher glutaraldehyde fixation levels reflects the successively lower membrane capacitance of the respective modified RBCs. A single-shell model was fitted to the impedance data to estimate the membrane capacitance of fixed cells. The results show that by increasing the ratio of fixation agents, the membrane capacitance of altered cells is lowered. Furthermore, for altering the cytoplasmic conductivity of RBCs, ghost RBCs (RBCs without internal content) were resealed in different conductivity buffers. Impedance measurements of ghost RBCs showed a higher phase contrast (the difference of high to low frequency impedance phase) among the cells that were resealed in higher conductivity buffers. We also confirmed higher cytoplasmic conductivity in these cells by using a single shell model. To assess resealing protocol effectiveness, ghost RBCs were resealed in different conductivity buffers that include FITC-dextran to offer identifying multimodal information on electrophysiology and fluorescence to the model particle. The flow cytometry results confirmed a similar level of FITC in ghost cells, regardless of the buffer conductivity. We envision the application of these standard particles with identifying multimodal information for rapid label-free benchmarking of the electrophysiology of cells with unknown phenotypes.

23

Rapid Detection of Cancer Related Cellular Lysis Events in Patient Plasma using High Conductance Dielectrophoresis

Sean Hamilton, Kyle Gustafson, Stuart Ibsen hamiltse@ohsu.edu

ibsen@ohsu.edu Cancer Early Detection Advanced Research - Biomedical Engineering

Oregon Health and Science University USA

The ability to differentiate between benign non-expanding masses and aggressively expanding malignant tumors is essential to reduce false positive rates in early stage cancer screening. A physical characteristic that could be used to help make this differentiation is the necrotic tissue specific to aggressive tumors. These necrotic regions release cellular organelles into circulation making the organelles themselves potential biomarkers in a liquid biopsy test. However, the small size and low buoyant density of the particles and their fragments makes it difficult to recover them from undiluted patient plasma using gold standard ultracentrifugation based techniques, which are time consuming and labor intensive. Here we demonstrate for the first time that high conductance dielectrophoresis (DEP) can be used to rapidly recover organelle-based particles from undiluted cancer patient plasma samples providing an alternative technique that shows promise for clinical translation. Regions of necrosis are inherent to rapidly expanding aggressive tumors due to hypoxic areas that develop from poor vascularization and increased interstitial pressure created by unspecified directional growth. These necrotic regions are symptomatically represented by local inflammation and cellular lysis. In healthy tissue, the programmed cell death pathway of apoptosis slowly breaks down a dying cell which blebs off fractions of itself in the form of large apoptotic bodies, which are in turn phagocytosed by macrophages. This prevents the release of free organelles into circulation. In contrast, the cellular lysis occurring in necrotic tissue results in the cells falling apart releasing cellular organelles into circulation. Freely circulating organelles could be an indication of abnormal cell death related to lysis. The microfluidic chip used here for high conductance DEP has been previously demonstrated for exosome isolation (particle size 80-120nm), making it a promising technology to recover organelle based particles of a similar size range. Specific organelles were selected based on particle size and the ability to amplify the signal of the lysis event. Mitochondria are of particular interest based on these criteria, as they are present in large numbers within cells. Depending on the cell type, each lysis event releases hundreds or thousands of mitochondria effectively amplifying the signal from a single lysis event. Signal amplification is important for early stage cancer detection because the overall signal produced is proportional to tumor size. We confirmed successful isolation and collection of whole intact mitochondria from buffer and plasma in addition to fragment collection from undiluted plasma using immunofluorescent staining for the mitochondria specific TOM20 biomarker. Mitochondria in the cell are approximately 1µm in diameter, but can fragment into smaller particles due to shear forces in circulation creating particles in the size range where an effective positive DEP force can be created. The ability to detect mitochondria with no manipulation to the plasma sample and the simplicity of the high conductance dielectrophoresis technique makes it a promising technology for in plasma detection of cellular lysis events and future clinical translation.

24

DEP Separation of Different Candida Strains Using 3D Carbon Electrodes Mohammad Amin Davarzani, Rodrigo Martinez-Duarte

mdavarz@clemson.edu rodrigm@clemson.edu

Multiscale Manufacturing Laboratory, Department of Mechanical Engineering Clemson University

USA

Candida infections are reported as the fourth most common hospital-acquired invasive bloodstream infections in the United States, with increasing prevalence in nosocomial infections. Recently, they have also been recognized as the 2nd most common cause of death because of such infections. Among more than 15 Candida strains, C. albicans, C. glabrata, C. tropicalis, C. parapsilosis are dominant causes of candidiasis, subsequently, which are involved in 95% of diseases. The traditional identification of candida species and anti-fungal therapy will take up to 1-2 days. Hence, a rapid assay that offers the potential for their separation or sorting can reduce the treatment time significantly is inevitable. Dielectrophoresis (DEP), a fast, accurate and label-free technique, presents the capability of separation of different Candida strains inside a microfluidic chip, including carbon electrodes. A previous work shows that cells will show a specific response to a non-linear electric field varying in frequency in the range 10–750 kHz and impedance (Martinez-Duarte et. al). In this research, we study the extensive and promising DEP response of Candida albicans and Candida glabrata species to the polarized 3D carbon electrodes. Three-step photolithography of negative photoresist on a Silicon Oxide wafer was used in the fabrication procedure. To obtain carbon microstructures, the SU-8-based microarrays were carbonized at 900°C for 75 minutes in a nitrogen atmosphere and cooled down with a 5 C/min rate. The chip consists of more than 3000 carbon microelectrodes with a diameter of 50 µm diameter and a height of 100 µm. A thin layer of SU-8 covered the electrodes to planarize the chip and insulate the connecting planar interdigitated leads. Utilizing our DEP platform, we separate Candida albicans and Candida glabrata from the mixed cell suspension and validate our results using a fluorescence-activated cell sorter accordingly. The results will show promise towards using 3D carbon electrode DEP as a tool to separate up to 70% of Candida albicans from Candida glabrata. Controlling the origin of candidiasis and early commencement of effective antifungal medication have been repeatedly reported as key factors in survival, and a growing body of data now supports the superiority of rapid assays for the primary treatment of invasive candidiasis. Within the next 5-10 years, the rapid development of technologies will likely generate tools to select the highest-risk patients for Candida species. Such methods will likely lead to lower mortality among patients with invasive candidiasis.

25

The effect of oxidative stress on Dielectrophoretic parameters of human red blood cells Timothy Swimmer, Sydney Joseph, Ronald Blum, and Erin A Henslee

swimtp19@wfu.edu hensleea@wfu.edu

Department of Engineering Wake Forest Engineering

USA

Oxidative stress (OS) in cells occurs due to imbalance of reactive oxidative species (ROS) and the cell’s antioxidant defense system. Red blood cell (RBC) OS is particularly important as it irreversibly damages the cells, impairs oxygen delivery and induces red blood cell aging and hemolysis. Not only are RBCs particularly susceptible to OS during normal processes occurring in circulation, oxidative damage and OS are hallmarks of many RBC pathologies such as malaria and sickle-cell disease; has been problematic in blood storage and transfusion; and associated with other diseases such as Diabetes and Parkinson’s disease. Previous work has examined changes in RBCs due to OS including deformability, mechanical stability, and elastic properties. Some attention has been given to RBC electrophysiology as a marker for OS and in examining anti-oxidant agents. Changes Zeta potential, for example has shown correlation to oxidative stress. Membrane dipole potential demonstrated how electrophysiology can reflect mechanistic pathways in oxidative stress. While these studies demonstrate the utility of RBC electrophysiology, there remains a need for more thorough investigation, cost-effective, and high-throughput assays. In this work, we present Dielectrophoresis (DEP) as a rapid tool to characterize OS in RBCs. We conducted a dose response study to determine the dose and exposure time of hydrogen peroxide, a ROS which induces OS. Isolated RBCs were suspended in 0, 0.5, 1, 2, 5, and 10mM concentrations of hydrogen peroxide (H2O2), incubated for 2 h and 4 h. DEP analysis was conducted on the 3DEP system for three biological repeats. Our results showed a significant increase in membrane conductance after 2 h incubation in 1 mM H2O2 when comparing our controls against treatment groups for dose effect. These results are in agreement with previous electrophysiological measures of H2O2 exposure demonstrating an increase in electrical properties of the RBC membrane. No changes in membrane capacitance were observed across all treatments. This is also consistent with previous RBCs studies that along with no morphological change, peak membrane conductance during peak PRX-SO2/3 oscillations (which counteract OS) were observed. Though previous work has indicated no significant affects to RBC shape or size due to H2O2 exposure, we had expected some change due to the stiffening of RBCs due to OS. Decreased Zeta potential has previously been shown as a result of H2O2-induced OS. Future work will correlate DEP measures with those of RBC deformability and Zeta potential analysis, for a more robust study of this process. Since ROS-mediated OS activates Ca2+ permeable pathways and triggers Eryptosis, we will also compare Phospholipid phosphatidylserine (PS) expression via Annexin-V to determine if changes on a shorter time scale in cytoplasmic conductivity may provide an early indicator of Eryptosis. This works presents preliminary evidence that DEP may provide a novel approach to RBC oxidative stress analysis. Future work to correlate DEP response to other RBC biophysical parameters will confirm its potential in monitoring oxidative stress progression, therapeutic targets, and the potential to monitor RBCs in vitro and in vivo.

26

Session 2

Simultaneous use of metal coated three-dimensional SU-8 pillars as passive posts and electrodes

Kevin Keim, Mohammed Z. Rashed, Jason P. Beech, Widad Lahbichi, Till Ryser, Gloria Porro, Jonas O. Tegenfeldt, Carlotta Guiducci

kevin.keim@epfl.ch carlotta.guiducci@epfl.ch

Institute of Bioengineering / Laboratory of Life Sciences Electronics École Polytechnique Fédérale de Lausanne

Switzerland

Three-dimensional electrodes, which span the complete height of a microfluidic channel, create homogeneous electric fields at any vertical position of the device. Correspondingly the dielectrophoretic force is of equal strength over the complete channel height, which can be an advantage over planar electrodes. Our group has developed a process coating the sidewalls of insulating SU-8 structures with metal and connecting them to an underlying metal layer, enabling three-dimensional electrodes of different shapes with photolithographic precision (Kilchenmann et al., JMEMS 2016). This enables exact shapes, not suffering from shrinking in comparison to carbon 3D electrodes out of SU-8 (Wang et al., JMEMS 2005) and fast production in comparison to for example electroplating (Voldman et al., Anal. Chem. 2002; Wang et al., JMEMS 2007). These metal coated three-dimensional electrodes of different shapes can be positioned free standing or integrated in an insulating SU-8 structure on a microfluidic chip and supplied with different electric signals. Using these electrodes, we could construct different systems, in which the electrodes are either used purely as electrodes, supplying the electric signal or additionally as restrictive structures for the flow and particles. Using the electrodes in a double array, dielecrophoretic micro cages of the size of a single cell can be created, which enable parallel trapping of multiple single cells in suspension. Individually controlling the applied signal to each electrode by an off-chip printed circuit board, allows to trap and release cells selectively. Alternating the DEP trapping signal with a signal creating a rotating electric field (Rohani et al., Electrophoresis 2014) allows to analyze the cells in the individual cages by electrorotation. With this method four different cell lines could be distinguished (Keim et al., Electrophoresis 2019) and specific membrane capacitance changes in one cell population of HEK 293 cells due to different osmolarities (Keim et al., µTAS 2019) could be observed with a throughput of up to 600 cells per hour. The same electrodes can be used as posts in deterministic lateral displacement (DLD) devices. DLD uses lines of restrictive posts under an angle to the microfluidic flow in order to separate different sized particles. Depending on the angle, the post size and the post distance every structure has a specific sorting size. Applying an AC electric field at these electrode posts, local dielectric forces can be implemented and the sorting size can be tuned and reduced. The critical size of device without an electric field of 6 µm is reduced to 250 nm with an electric field applied (Beech et al., Adv. Mater. Technol. 2019). Further one could possibly use this device to sort particles based on their dielectric properties. In the future devices will be further investigated and, based on the three-dimensional electrodes, new devices will be developed, which can for example release cells from hydrodynamic traps or to merge multiple cells in a controlled way in droplets.

27

Using DEP to detect cell phenotype and cell surface composition Alan Y.L. Jiang, Andrew R. Yale, Shubha Tiwari, Estelle Kim, Tayloria N.G. Adams, Lisa A. Flanagan

lisa.flanagan@uci.edu lisa.flanagan@uci.edu

Departments of Neurology, Biomedical Engineering, Anatomy & Neurobiology, and Chemical and Biomolecular Engineering

University of California, Irvine USA

Cells are dynamic and can shift phenotype and function in response to external cues. The means by which we detect the phenotypes of living cells are limited and often rely on minimal numbers of cell surface markers that are not sufficiently specific. DEP is a unique method for the analysis and separation of living cells that does not require labels or markers, but instead measures inherent composite cellular properties. Different cellular compartments, such as the plasma membrane or cytoplasm, can be probed in AC DEP by changing the frequency of the applied electric field. Notably, inherent cellular properties detected by DEP are sufficient to identify subtle differences in cell phenotype. Neural stem/progenitor cells that are similar in size and marker expression but vary in ability to form differentiated cell types can be separated by DEP because they differ in whole cell membrane capacitance. We found recently that other phenotypically important brain cells can be enriched by DEP-based sorting. Glioblastoma is the deadliest brain cancer and a significant problem for treatment is the presence of cells resistant to the most commonly used glioblastoma chemotherapeutic. Resistant cells are difficult to identify using markers, making development of novel therapeutics challenging. Our new studies show that chemotherapeutic resistant glioblastoma cells can be enriched by DEP, paving the way for characterization of these critical cells and identification of effective treatments. Phenotypically distinct brain cells can be enriched by DEP because they differ in whole cell membrane capacitance. We found that membrane capacitance is sensitive to the composition of the plasma membrane, specifically cell surface glycosylation. This realization enabled the identification of specific glycosylation pathways that regulate neural stem/progenitor cell fate, impacting the formation of differentiated cells in culture and in the brain. In summary, DEP provides a novel way to investigate cell identity and discover new determinants of important cellular functions such as differentiation and chemotherapeutic resistance.

28

Dielectrophoresis Quantifies the Dynamic Heterogeneity of Mesenchymal Stem Cells Tunglin Tsai, Lexi L. Crowell, Mary Tran, Angie M. Castro, Destiney Z. Ward, Prema D. Vyas, and

Tayloria N. G. Adams tunglit@uci.edu tayloria@uci.edu

Department of Chemical and Biomolecular Engineering, Sue & Bill Gross Stem Cell Research Center University of California – Irvine

USA

Human mesenchymal stem cells (hMSCs) can be derived from a variety of vascularized tissue and have gained traction in transplantation therapy due to their immunomodulatory, paracrine, immune-evasive, and multipotent differentiation potential. Given the heterogeneous nature of hMSCs, there is a demand to find additional biomarkers to ensure reproducibility when using these stem cells both in vivo and in vitro. In this work, we utilized dielectrophoresis (DEP), a label-free electrokinetic phenomenon, to investigate and quantify the heterogeneity of hMSCs derived from the bone marrow (BM) and adipose tissue (AT). Our hypothesis is that hMSCs have unique dielectric properties, detectable using DEP, which are linked to their inherent heterogeneity. Through computer simulation, we identified that the transient slope of the DEP response spectra can be used as a metric to quantify the relative dielectric heterogeneity that exists within the hMSC populations. While the membrane capacitance and cytoplasm conductivity can be used as biomarkers for identifying different cell types. To experimentally validate our hypothesis, we measured the DEP spectrum of BM-hMSCs and AT-hMSCs along with terminally differentiated mouse fibroblasts (NIH-3T3), human fibroblasts (WS1), and human embryonic kidney cells (HEK-293). We compared the membrane capacitance and cytoplasm conductivity of BM-hMSC against NIH-3T3, WS1, and HEK-293 and found that BM-hMSC exhibits distinct dielectric signatures. In addition, when we compared BM-hMSC to AT-hMSC, we found that while the two hMSC populations present similar membrane capacitances, their cytoplasm conductivities and the transient slopes differ. Inducing both populations to differentiate into adipocyte and osteocyte cells revealed that the two populations behave differently in response to differentiation-inducing cytokines. These similarities and differences were observed using histology and RT-qPCR analysis of the differentiation-related genes. Our results further strengthen the ability to use membrane capacitance, cytoplasm conductivity, and transient slopes, and DEP in general, to characterize hMSCs.

29

3D printing microfluidic devices using liquid dielectrophoresis

Josie L. Duncan, Jeff Schultz, Zeke Barlow, Rafael V. Davalos josied@vt.edu

davalos@vt.edu Bioelectromechanical Systems Laboratory, Department of Mechanical Engineering

Virginia Tech USA

3D printing has been an appealing yet challenging proposition for microfluidic researchers. Microfluidics has relied on stereolithography (SLA) methods for additive manufacturing of new devices, but SLA is limited in resolution, material selection, post-processing requirements, and electrode integration. Here, we present a novel 3D printing technology that uses liquid dielectrophoresis (L-DEP) to manipulate dielectric fluids such as PDMS to create microfluidic devices. Electric Field Fabrication (EFF), a technology developed by Phase, Inc., uses uniquely designed interdigitated electrode arrays to transform a droplet of dielectric fluid into a thin-film corresponding to the shape of a custom interdigitated electrode array. The build plate then comes down to control layer thickness and the resin is cured via thermal or UV energy sources, producing a smooth surface finish. Once the layer is cured, the build plate removes the cured layer from the electrodes and creates room for the next layer. This process is repeated until the desired 3D device is achieved. EFF differs from SLA, not only by the technology used to selectively manipulate the fluid, but also that the part is not fabricated in a vat of resin allowing for decreased post-processing time and the ability to integrate electrodes mid-print. This has advantages for both prototyping as well as factory-scale production of microfluidics. To date, UV or thermally cured resin has been used to fabricate microfluidic devices with simple channel geometries and the ability to fabricate more complex geometries is being investigated. Other current efforts are focused on optimizing parameters to directly 3D print PDMS using dielectrophoresis. This technology has many other applications in the biomedical field using an array of build materials.

30

Session 3

Pancreatic Cancer Detection through Multiomic Analysis of Biomarkers Collected from Plasma Using High Conductance Dielectrophoresis

Augusta Modestino, Jesus Bueno Alvarez, Michael Heller, Stuart Ibsen ibsen@ohsu.edu ibsen@ohsu.edu

Biomedical Engineering Department and Cancer Early Detection Advanced Research (CEDAR) Center, Knight Cancer Institute

Oregon Health and Science University USA

Tumors shed different types of nanoparticles containing valuable cancer biomarkers into circulation. One major class of shed particles are extracellular vesicles, such as exosomes, that contain protein biomarkers. A second class are nanoparticles comprised of cell-free DNA (cfDNA) containing cancer related mutations. Despite their importance, these two particle types are a challenge to use in clinical cancer screening applications because they are difficult to recover from plasma. To address this challenge we have developed a method using high conductance dielectrophoresis (DEP) chip technology that can simultaneously collect both types of particles from undiluted plasma. A buffer wash removes the bulk plasma purifying the desired nanoparticles. This enables quantifiable immunostaining of protein biomarkers on the vesicles and fluorescent dye staining of the cfDNA. We show the cfDNA can also be recovered and analyzed using digital PCR to detect cancer related mutations. We applied this technique in a blinded study that analyzed plasma samples from a 36 patient cohort consisting of patients with pancreatic cancer and benign pancreatic disease. Based on the relative levels of the protein biomarker Glypican 1 and the overall level of cfDNA we were able to use a bivariate analysis to successfully differentiate patients with pancreatic cancer from patients with benign pancreatic diseases such as pancreatitis, precancerous intraductal papillary mucinous neoplasm (IPMN), and healthy individuals. Separating the nanoparticles from the bulk plasma removed interfering proteins and enabled PCR based analysis of the collected cfDNA. KRAS mutations are common in pancreatic cancer and we show that the KRAS mutation status for each patient can be determined from the DEP recovered cfDNA using digital PCR and was confirmed using Sanger Sequencing. Once optimized, this technique takes less than an hour to complete and can be highly automated making this a promising technology for clinical translation compared to the current gold standard methods for recovering both extracellular vesicles and cfDNA. Traditionally, the two different particle types require entirely different isolation methods which are challenging to translate to a clinical laboratory environment because they are labor intensive, time consuming, and difficult to automate. Exosomes require overnight ultracentrifugation steps for purification and the cfDNA requires the use of labor-intensive Qiagen kits that have many involved steps. Each method requires separate plasma samples of several milliliters. The method developed here using high conductance dielectrophoresis chip technology, purchased from Biological Dynamics, can collect both particle types simultaneously from a single 50µl plasma sample. The chip uses a platinum electrode array at the bottom of a microfluidic chamber to create a DEP force on both the cfDNA particles and the extracellular vesicles. Both are drawn to the high field region around the electrode edge and are held in place through a purification buffer wash. This locally concentrates the biomarkers increasing the signal-to-noise ratio for fluorescent staining based detection. It also puts the biomarkers in a known location that allows for quantification and background noise subtraction. These properties make high conductance dielectrophoresis an ideal technology for future clinical translation and use in cancer screening applications.

31

A sequentially selective DEP platform for studying cytokine and cytolytic responses from the interaction of a controlled number of tumor and immune cells in confined volumes

Darshna Pagariya, Quinlan Pollak, and Robbyn K. Anand darshna@iastate.edu rkanand@iastate.edu

Department of Chemistry Iowa State University

USA

Environmental stimuli can induce the secretion of regulatory peptides, known as cytokines, from immune cells. Among their many functions, cytokines mediate the responses of cells participating in host immune and inflammatory defense against pathogens or tumor cells. Insight into cell-to-cell differences in cytokine secretion is a piece of vital information for understanding immune response and for tailoring therapeutic interventions such as cytokine therapy. In response to stimuli, individual cells secrete cytokines, which reach local concentrations in the pico- to nanomolar range; therefore, it is challenging to monitor secretion at single-cell resolution. Traditionally, cytokine secretion for single cells has been measured by intracellular cytokine staining (ICS) flow cytometry and by enzyme-linked immunospot (ELISpot) assay. Although both methods provide reliable and sensitive cytokine measurements, they do not provide real-time analysis and are not amenable to subsequent downstream assays due to sample degradation. Several microfluidic chips have partially addressed these shortcomings by co-encapsulating immune and tumor cells with reagent-coated beads in wells or traps. While these systems yield excellent sequential loading, the mechanism for loading is frequently size-based. This size-based selection is not ideal for heterogeneous target cell populations and lacks capture specificity. Further, the handling of the sample requires multiple manual interventions leading to sample loss and contamination. On the other hand, droplet-based co-encapsulation is limited by Poisson statistical loading and therefore, the success rate drops significantly. Here, we report a dielectrophoretic technique to sequentially and selectively load a controlled number of targeted tumor cells, immune cells, and bioconjugated beads into nanoliter-scale chambers. To prevent cross-contamination from the adjacent chambers, we isolate the chambers using an immiscible phase during measurement. Our selective phenotype approach requires minimum sample handling, low sample volume (~2.0 microliters), and sample retrievable for downstream analysis. We detect cytokine secretion on a bioconjugated bead for each tumor cell-immune cell pair in many isolated chambers in parallel.

32

Integration of selective capture of tumor cells with electrochemical biosensing at a wireless electrode array

Janis S. Borchers, Kira L. Rahn, Savanah Van Scoy, Morgan J. Clark, Robbyn K. Anand rkanand@iastate.edu rkanand@iastate.edu

Department of Chemistry Iowa State University

USA

Cancer evolution and the resurgence of refractory disease following therapeutic intervention is potentiated by the diversity of distinct subpopulations of tumor cells. A minority of cells with unique phenotypic and genotypic characteristics can drive disease progression and response to therapeutic agents. Therefore, single-cell analysis is critical to the accurate characterization of disease states. A major challenge for the development of next-generation devices for the evaluation of individual cells is the integration of all steps of analysis into a single platform. Such integration would greatly reduce the time and cost of diagnosis while increasing accuracy. Major challenges to the realization of this goal include a need for high volumetric throughput, selectivity for the cells of interest, and parallel and sensitive analysis. In this presentation, we discuss methods that leverage electrokinetic processes at arrays of bipolar electrodes (BPEs) to address each of these challenges. Specifically, we will discuss selective capture of cancer cells by dielectrophoresis (DEP) at BPEs aligned to nanoliter-scale reaction chambers and integration of electrochemical sensing of cell-surface antigens. In this platform, antigen expression is reported by electrochemiluminescence (ECL), which is amenable to implementation in resource-limited and point-of-care settings. We anticipate that these methods can be extended to a wide range of cell types and biomarkers.

33

Characterization of Particle Trajectories During Dielectrophoresis Collection and the Physical Phenomena of Insulator Coverage over the Electrodes

Ramona Luna, Daniel Heineck, Stuart Ibsen lunara@ohsu.edu ibsen@ohsu.edu

Biomedical Engineering - Cancer Early Detection Advanced Research Center Oregon Health and Science University

USA

It is known that tumors shed various nanoparticles into circulation even at the early stages. Due to their small size and low abundance, it is challenging to recover and detect these particles from plasma, especially at early cancer stages where biomarker levels are low. Early cancer detection is critical to improve treatment outcomes. High conductance dielectrophoresis (DEP) is a promising technology to recover these particles based on their dielectric differences with the surrounding plasma. Better understanding the interaction between the particles and the electric field will aid in designing new methods to improve collection of rare cancer derived particles and increase device sensitivity when biomarker abundance is low in early stage cancer samples. Consequently, we studied the behavior of particles and physical phenomena that occur inside these microarray chips when DEP is applied. The chips were purchased from Biological Dynamics and have a unique hydrogel layer to protect the sample and electrodes from electrochemical destruction. It has been observed that when DEP is applied, the hydrogel partially detaches from the electrode surface forming domes. When domes are present, the collection of particles is enhanced at the edges of the circular electrodes. We used COMSOL Multiphysics Software to model the effect of the dome on the electric field, and various insulating materials that could be within the dome. Our results suggest that DEP high field regions intensify and localize at the edges of the electrodes when domes are formed leading to enhanced collection of particles. Moreover, we analyzed the trajectories of individual particles under the influence of negative DEP as they moved towards collection in low regions of the electric field to understand how well the theoretical field gradients matched particle motion. For these experiments, we used 10mm particles and recorded videos when DEP was applied for 40s. We used FIJI’s TrackMate plug-in to analyze the videos and obtain experimental track data. From COMSOL, we obtained theoretical path values created by the electric field gradient. Particle trajectory was found to be highly dependent on the starting location of the particle relative to the electrode itself and the particles were found to closely follow the theoretical gradient of the electric field. The results from these experiments help validate the theoretical field gradients obtained from COMSOL simulations. The analysis of the relationship between the enhanced collection and the dome formation in addition to the study of particle trajectories, enable future optimization of protocol development to enhance collection when there is low abundance of biomarkers. Further studies will involve particle tracking experiments for different sizes of particles and analyzing how the formed domes affect the particle trajectories in DEP collection experiments.

34

Keynote

Modeling the electrokinetic behavior of metallodielectric particles Pablo Garcia-Sanchez

pablogarcia@us.es pablogarcia@us.es

Departamento de Electronica y Electromagnetismo, Facultad de Fisica Universidad de Sevilla

Spain

The talk deals with the theoretical and numerical modeling of recent experiments with metallodielectric microparticles. We focus on the following problems: A) We discuss novel experimental results with Lollipop shaped microparticles consisting of a dielectric tail and a gold disk like head. These particles move towards stable equilibrium positions within a quadrupolar electrode array. The location of these positions can be controlled by tuning the frequency of the ac signals on the electrodes. We make a detailed comparison between numerical results and experimental data of equilibrium positions. We also discuss the requirements on the electrode geometries and complex particle shape that lead to a stable equilibrium. B) Janus metallodielectric spheres near an electrode show a complex electrokinetic response. At low frequencies, induced-charge electroosmosis leads to net motion of the Janus sphere traveling with their dielectric hemisphere forward. At higher frequencies, the sphere reverses its motion. The latter phenomenon has been termed self-dielectrophoresis and we show that this motion is caused by the interaction of the Janus sphere with the polarizable electrode.

35

Session 4

Simultaneous electrorotation systems to determine the membrane capacitance and cytoplasm conductivity of cells

Shikiho Kawai, Suzuki Masato, and Tomoyuki Yasukawa kawai.shikiho@jp.panasonic.com

yasu@sci.u-hyogo.ac.jp Graduate school of material science

University of Hyogo Japan

We developed a simultaneous electrorotation systems to determine membrane capacitances and cytoplasm conductivities of hematopoietic cells by using a device with three-dimensional interdigitated array (3D-IDA) electrodes at the single operation. The device consists of two IDA electrodes that were arranged orthogonally to form grids surrounded by four microbands in a fluidic channel. Constant rotation of electric fields was generated in grids by applying AC voltages with a 90° phase difference with each microband electrode. The cells experienced the rotational electric fields were rotated at the center of grids. The rotation rates of four different types of cells with similar shape as a function of the applied frequency were investigated to obtain rotation spectra. The spectra were utilized to determine the membrane capacitances and cytoplasm conductivities of cells by fitting with theoretical curves of electrorotation rates. K562 cells were simultaneously electrorotated by 3D-IDA devices. When AC signals (700 kHz, 5 Vpp) were applied, K562 cells dispersed randomly in the 3D-IDA device started to rotate and moved to the center of each grid. AC signals with the frequencies from 100 kHz to 1 MHz were applied to the electrodes to obtain the rotation spectrum. The rotation rate of K562 cells increased with increment of applied frequency up to 600 kHz, and then decreased. The spectrum experimentally obtained well corresponded with the curve obtained theoretically. The membrane capacitance and cytoplasm conductivity of K562 cells were determined as 9.16±0.07 mF/m2 and 0.32±0.02 S/m, respectively, which are almost the same values or orders compared to those reported previously. Likewise, other hematopoetic cells (THP-1 and Jurkat) were subjected to the simultaneous electrorotation to obtain membrane capacitances and cytoplasm conductivities. The determined membrane capacitances and cytoplasm conductivities for Jurkat and THP-1 cells coincides with those reported. The present system is advantageous to the unnecessity of repeated operations by exchanging cells to the area for the electrorotation. The time required to obtain the data for rotation rates of 50 cells with 10 different frequencies is only 5 min. Therefore, the simultaneous electrorotation of multiple single cells by using the 3D-IDA device allows to the determination of electric properties of cells simply and rapidly with a single operation. The membrane capacitance and the cytoplasm conductivity of WEHI-231 were also estimated to be 8.89±0.25 mF/m2 and 0.28±0.03 S/m, respectively. This is the first report of the dielectric properties of WEHI-231 which is a mouse immature B-cell and widely used as model cells for research of B-cell development and activation. These four types of cells were clearly separated by using three parameters, which are membrane capacitance, cytoplasm conductivity, and cell radii. In particular, membrane capacitance of THP-1 was the highest compared to those of other types of cells as similar to previous reports. The complex surface morphology of THP-1 would lead to the increase of membrane capacitance. The results supported the correlation between simultaneous electrorotation by using the 3D-IDA device and the conventional electrorotation.

36

Rapid cell viability and antimicrobial susceptibility testing using impedance spectroscopy Pragya Swami, Ayush Sharma, Satyam Anand, Gajanand Verma, and Shalini Gupta

pragyaswami92@gmail.com shalinig@chemical.iitd.ac.in

Department of Chemical Engineering, IITD Indian Institute of Technology, Delhi

India

The current diagnostic procedures for bacterial detection are time-consuming as they rely on gene amplification and cell culture both of which are intrinsically slow processes, limited by the doubling rate of the involved species. Further, biochemical methods for species identification and antibiotic susceptibility testing (AST) for drug/dose effectiveness take several days and are non-scalable. Early diagnosis and timely therapeutic intervention can greatly reduce empirical treatment and thus curb antimicrobial resistance. The use of impedance spectroscopy as a simple, label-free, cost-effective, high-throughput and real-time tool for AST is a rapidly emerging area, however, the current trends in impedance biosensors focus on electrode design and cell enrichment strategies to improve signal sensitivity and limits of detection, making the technique less viable. On the other hand, the role of buffers to enhance impedance sensitivity is relatively unexplored. We have found that Good’s zwitterionic buffers (e.g. HEPES) show improved performance over standard electrolyte buffers (e.g. PBS) that are currently widely used in impedance spectroscopy. This is due to the significantly higher ion-ion interactions in zwitterionic buffers as compared to classical electrolytes at the concentrations at which they are used. This and the fact that zwitterionic molecules have larger sizes leads to the lowering of their conductivity which significantly improves their impedance sensing ability. We have exploited this fact to measure highly sensitive conductivity changes in clinical isolates of S. typhi, E. coli and S. aureus for their rapid AST profiling and MIC/MBC determination. When exposed to different classes of surface-acting antibiotics (colistin, methicillin, ampicillin, carbapenem etc.) in 15 mM HEPES buffer, the resistant and susceptible bacterial strains could be differentiated as early as 20 min, using ionic release from dying cells as a proxy indicator of their viability status. Further, using novel modified growth buffers, the detection process was also extended to non-surface acting drugs (ciprofloxacin, rifampicin etc.) in which cell growth was measured within just 1 to 2 doubling cycles (~ 60 min). As a result, MIC values could be obtained in an hour, greatly reducing the overall time for bacterial diagnosis and management.

37

Computational Modeling of the Electric Field Distribution towards Exosome Characterization

Gladys Guadalupe Diaz-Armas, Ana Paola Cervantes-Gonzalez, Rodrigo Martinez-Duarte, Victor Hugo Perez-Gonzalez

gladysg.1998@gmail.com vhpg@tec.mx

School of Engineering and Sciences Tecnológico de Monterrey

Mexico

Exosomes are crucial for multiple cellular processes, including cell-cell communication, inflammation, and homeostasis. Their content and surface markers relate to the state of the origin cell, making them valuable biomarkers for degenerative diseases. From all the electroactive phenomena, DEP has shown to be the most effective for exosome manipulation. The DEP force relies on the difference between the dielectric properties of the particles of interest and their suspending medium, as well as the frequency and non-uniformity in the spatial distribution of the electric field. In particular, the electric field must be properly designed to generate a viable field gradient, but also to ensure the viability of the exosome is not compromised. Since electroactive strategies like DEP do not require any kind of biomarker or moving parts to operate, such approaches have been proposed as a feasible solution for exosome manipulation. However, the electrical forces that exosomes are submitted to while being manipulated can still be strong enough to deform their membrane. Hence, a methodology to design electrode arrays tailored for exosome characterization is necessary. In this work, a series of electrode designs based on six different geometries were modeled to characterize the magnitude and spatial distribution of the electric field and its gradient, depending on the polarizing voltage between different intercalated electrodes featuring an increasing gap. For each design, a two-dimensional model was created in COMSOL Multiphysics. In a stationary study, the Laplace equation was solved in order to obtain the distribution of electric potential. Boundary conditions were selected based on the electrode polarity. Preliminary findings show that, by applying a potential difference of 5 V, linear electrodes can reach up to 3.6x10^5 V/m, while a sawtooth-based geometry can reach up to 5x10^5 V/m, although the second geometry has a more heterogeneous field distribution among the length of the channel. This study becomes relevant since the literature remarks that an electric field above a threshold of 10^5 V/m in magnitude is enough to manipulate particles such as viruses, proteins, and DNA, which are within the size range of interest (30-100 nm). Ongoing work is on modeling the behavior of a lipid bilayer sphere, with the physicochemical characteristics of an exosome, under the electric fields studied so far, to test whether an interface with such characteristics would be able to maintain its stability under electrical polarization. The long-term goal of our work is the manipulation of nanoscale biological structures without compromising their viability, which represents a critical step in their analysis towards applications in healthcare, diagnostics, and therapeutics.

38

Novel recognition and targeting of temozolomide resistant cells in glioblastoma Alan Y.L. Jiang, Andrew R. Yale, Jaclyn N. Hanamoto, Christopher R. Douglas, Clarissa C. Ro, Shubha

Tiwari, Kaijun Di, Naomi Lomeli, Daniela A. Bota, Lisa A. Flanagan yljiang@uci.edu

lisa.flanagan@uci.edu Departments of Neurology, Biomedical Engineering, and Anatomy & Neurobiology

University of California, Irvine USA

Glioblastoma (GBM) is the most common and deadly primary brain tumor and is characterized by a heterogeneous population of cells that are genetically unstable, highly infiltrative, angiogenic, and resistant to chemotherapy. Affected patients have very poor prognosis with median survival of 15 months and extremely low survival rates – 27% at 2 years and 9% at 5 years post-diagnosis. Temozolomide (TMZ) is an oral alkylating agent that is the first-line chemotherapeutic treatment for GBM. In many cases, patients do well with an initial TMZ treatment but have tumor regrowth that is refractory to TMZ treatment. Resistance has been attributed to intra-tumor heterogeneity, which can occur through selection of TMZ-resistant cells in the tumor as well as conversion of cells to a resistant state. Rapid identification and enrichment of TMZ-resistant tumor cells could lead to new and innovative treatments targeting resistant cells. We discovered that TMZ-resistant GBM cells significantly differed from TMZ-sensitive cells in whole cell electrophysiological properties measured by dielectrophoresis (DEP). More specifically, the whole cell membrane capacitance of TMZ-resistant cells was significantly lower than that of control cells while the midpoint membrane frequency of TMZ-resistant cells was significantly higher. There was no significant difference in cell size of resistant and sensitive populations. We found TMZ-resistant cells differed from controls in cell surface glycosylation , providing a molecular basis for capacitance differences between the cells. Differences in midpoint membrane frequency between the resistant and sensitive cells suggested TMZ-resistant cells could be separated using DEP. To test this, we developed a novel DEP buffer containing ROCK inhibitor (Y-27632) to improve cell recovery after DEP separation and used a high-throughput continuous DEP cell sorter to separate the cells. We separated GBM cells into two fractions that differed in membrane capacitance. The cell fraction with lower membrane capacitance was more resistant to TMZ, having a ~41% higher TMZ IC50 value than that of unsorted controls. In contrast, the cell fraction with greater membrane capacitance was more sensitive to TMZ, with a ~36% lower TMZ IC50 value than that of unsorted controls. To our knowledge this is the first study showing (a) TMZ resistance is correlated with whole cell membrane capacitance and cell surface glycosylation, and (b) TMZ-resistant GBM cells can be enriched using DEP. These findings could lead to a better understanding of the molecular profiles of resistant cells, identification of mechanisms of resistance, and discovery of novel treatment options for GBM including informed combinations of molecularly targeted classes of chemotherapeutics.

39

Session 5

Multidimensional sorting of mixed microparticles via insulator-based electrokinetics and dielectrophoresis

Laura Weirauch, Malte Lorenz, Blanca H. Lapizco-Encinas (*), Georg R. Pesch, Jorg Thöming lweirauch@uni-bremen.de thoeming@uni-bremen.de

Faculty of Production Engineering, Chemical Process Engineering (CVT), (*) Microscale Bioseparations Laboratory and Biomedical Engineering Department

University of Bremen, (*) Rochester Institute of Technology Germany

Insulator-based electrokinetics (iEK) is a powerful tool commonly used in fields of biomedical/bioanalytical chemistry. However, it has rarely been used to solve challenging separation problems involving non-biological particles or facing multidimensional separation problems. Sorting of suspended nano and submicron particles is very relevant for the production of highly specific particle systems that are required in many fields, such as the production of ceramics, powder-metallurgical components, and semiconductors. Here we present the capability of iEK and of dielectrophoresis devices to sort non-biological particles by multiple properties, specifically by size, surface functionalization or shape, while passing only one microfluidic unit. For this, we use a reservoir-based and a pressure-driven approach with selective trapping and remobilization. We found that, when conventional DC-iEK routines are applied, challenging conditions such as initially low absolute particle zeta potentials and high particle concentrations make sorting impossible in case of polystyrene-based microparticles. This is due to particle agglomeration and retention of particles even under the minimum trapping voltage during long experiments and when high trapping voltages are needed. We have shown the capability of material- and size-selective separation in a reservoir-based, DC-biased AC-iEK approach. A conventional iEK microchannel made out of PDMS with cylindrical insulating posts (200 μm diameter, 20 µm spacing) and pipette tips as liquid reservoirs were used. In order to prolong the experiment duration, which is necessary for multi-step experiments, the classic iEK design was extended by a reservoir connection, which balances out the reservoir level difference build up due to electroosmotic flow. Particle agglomeration could be significantly reduced by adjusting the pH and conductivity of the suspending medium [1]. A multidimensional sorting was achieved for mixtures of differently sized polystyrene (PS) particles and gold-coated polystyrene particles. In a first step, a high voltage was applied until all particles injected into the system were retained within the first columns of the post array. Then, the voltage was reduced in time steps of at least 30 s (to give particles enough time to penetrate the channel) so that first, all small PS particles, then all large PS particles and at the lowest voltage also gold-coated particles were released. This led to three fractions and demonstrates that iEK is an option for multidimensional sorting of micron-sized particles. In order to achieve technically relevant throughputs, we transfer the concept to a scalable design. In this, we use AC-DEP with pressure-driven flow for a consecutive selective trapping and remobiliziation. To verify the functionality, we first use microchannels containing insulating pillars as filter media. Contrasting to the reservoir-based approach we can selectively trap and release particles by adjusting the field frequency, allowing for a multidimensional separation. [1] L. Weirauch et al., Biomicrofluidics 13, 064112 (2019)

40

Stationary electro-osmotic flow vortices on insulating surfaces induced by ac electric fields

Raul Fernandez-Mateo, Victor Calero, Pablo Garcia-Sanchez, Hywel Morgan, Antonio Ramos r.fernandez-mateo@soton.ac.uk

pablogarcia@us.es Departamento de Electrónica y Electromagnetismo

Universidad de Sevilla Spain

We report stationary flow vortices of electrolytes on dielectric micropillars and corners induced by ac electric fields. Microfluidic channels are made of PDMS and the electrolytes are aqueous KCl solutions with conductivities around 1-10 mS/m. Electric fields with magnitude 10-100kV/m and frequency 10-1000 Hz are applied to the solutions seeded with fluorescent colloids. The flow characteristics are theoretically predicted based on the well-known phenomena of surface conductance and concentration polarization around a charged dielectric object. Hence, we refer to this fluid flow as concentration-polarization electro-osmosis (CPEO). Experimental data of the electro-osmotic velocity magnitude and frequency dependence agree with the theoretical estimates and are significantly different from predictions based on the standard theory for induced-charge electro-osmosis, which has previously been postulated as the origin of the stationary flow around dielectric objects. Our electrokinetic model reveals two distinct contributions working together: an oscillating nonuniform zeta potential induced on the insulating surface and a rectified electric field induced by the ion concentration gradients. We also discuss the role of these phenomena in insulating-DEP experiments.

41

Combining hydrodynamic and dielectrophoretic trapping to control the interaction between beads and cells

Clémentine Lipp, Jonathan Cottet, Philippe Renaud clementine.lipp@epfl.ch clementine.lipp@epfl.ch

LMIS4 EPFL

Switzerland

Cell-cell interaction studies are of great interest to understand physiological functions and are used for both fundamental and applied biomedical research. Most devices use bulk assays and measure a signal averaged over the cell population. However, measurements at the single-cell level are crucial to study heterogeneity in a population and rare events, which are not accessible in bulk assays. Different methods exist to study the affinity of a bond between two single-cells, measuring either ions or molecules resulting from the cell’s signaling pathway, or the mechanical force that the bond can withstand. Methods measuring the strength of a bond such as dual pipette[1], AFM[2] or optical tweezers[3] are difficult to operate and lack throughput. Microfluidics simplifies the manipulation of single-cells in larger quantities compared to the previously mentioned methods, and different devices were proposed to study cell-cell interactions based on mechanical[4] or ions and molecules[5] measurements. However, none of these strategies combine both parallelization possibility and interaction time control. Because it combines two different trapping mechanisms, the presented novel microfluidic device is able to independently manipulate and immobilize single beads or cells (here named “particles”) coming from two populations. Particles from population A are immobilized in flow conditions in a microchannel using single-particle hydrodynamic traps while population B is brought and maintained in close contact with the former using dielectrophoresis. A novel fabrication process was developed to fabricate two superimposed levels of microchannels. This process was used to fabricate a drain composed of a hole of diameter smaller than the particle of interest placed at the bottom of the upper microfluidic network and connected to the lower level of microchannel. The fraction of the upper channel’s flow passing through the hole can be controlled so that the particles following the streamlines in the upper channel will obstruct the hole thus stopping the fluid flow inside the lower channel. The trapped particles can be selectively released by reversing the direction of the flow inside the lower channel. This type of hydrodynamic traps has several advantages, comprising the possibility to trap and release, but also a minimum cluttering of the upper channel since nothing else than the trapped particle occupies space, contrary to the widely used in-plane traps[6]. This advantage is exploited by patterning around the drain coplanar electrodes used to generate an electric field to stop and position another particle by dielectrophoresis, in flow conditions, and induce contact with the first hydrodynamically trapped particle. Turning the voltage off releases the dielectrophoretically trapped particle which is dragged by the flow, unpairing the particles unless a specific interaction occurred between them. As proof of concept, we demonstrated the controlled pairing of hydrodynamically trapped polystyrene beads coated with antigens (anti-CD3 and anti-CD28) together with Jurkat cells. This combination of trapping means allows to control the contact time between particles. This will be used in further interaction studies to measure the bond strength and its dislocation constant. [1]Husson et al., PLoS ONE, vol. 6, no. 5, 2011 [2]Lim et al. J. Immunol., vol. 187, no. 1, 2011 [3]Gou et al., BioMed. Eng., vol. 14, no. 1, 2015 [4]Duchamp et al., RSC Adv., vol. 9, no. 70, 2019 [5]Dura et al., Nat. Comm., vol. 6, no. 1, 2015 [6]Faley et al., LOC, vol. 9, no. 18, 2009

42

Lattice Boltzmann Simulations of Dielectrowetting Élfego Ruiz-Gutiérrez, Andrew M. J. Edwards, Glen McHale, Michael I. Newton, Gary G. Wells, Carl V.

Brown, Rodrigo Ledesma-Aguilar elfego.ruiz@ed.ac.uk

glen.mchale@ed.ac.uk Institute for Multiscale Thermofluids, School of Engineering

University of Edinburgh United Kingdom

When a dielectric fluid interacts with an electric field, its molecules will respond to the field by polarising the medium. As a consequence, this produces forces that act on the fluid, a phenomenon called liquid dielectrophoresis (L-DEP) [1]. The ability to accurately control this forces has realised and revolutionised technologies such as inkjet printing, optical devices, micro-assembly and electrospinning. One important application of L-DEP is dielectrowetting [2], similar to electrowetting, it is a novel technique used to control the spreading of a droplet on solid surfaces. However, dielectrowetting differs from electrowetting in that dielectrowetting occurs due to the electromechanical force that arises from the polarisation the molecules of the droplet due to a non-homogeneous electric field, as opposed to the free charges of a conducting liquid in electrowetting setups. Also, dielectrowetting does not suffer from the contact angle saturation observed in electrowetting. In this contribution, we present a lattice-Boltzmann method (LBM) [3] capable of solving the hydrodynamic equations for dielectric fluids in the presence of electrostatic fields. The ability to perform simulations on the dynamics of multiphase fluids coupled with electric fields opens the door for a wide range of applications, for example, they can aid the design of complex electrode arrays for micrifluidic devices with the intention of manipulating of small amounts of liquid with the fine control that electronic devices allow or to induce specific film and droplet morphologies dielectrowetting. The LBM belongs to the family of computational fluid dynamic numerical simulations, and has gained reputation for accurately modelling flows of a wide range of situations, from rarefied supersonic gases to capillary phenomena. By adopting a modern formulation to the method inspired by spectral methods, we endow the lattice-Boltzmann with the capacity to include dielectrics. In this way, we improve the efficiency and stability of the method. The result is the coupling of hydrodynamic and electrostatic equations solved within a single method, which alleviates the need of running a separate algorithm concurrently, thus eliminating the need for alternative methods to tackle the electrostatic part separately. We have analysed the numerical method in terms of its accuracy and convergence for a wide range of electric permittivities surpassing most of real liquids. We first validate these findings by comparing the numerical method against exact solutions of simple systems. Then we validate the method against representative experiments with good agreement. This numerical method is not restricted to dielectrowetting, and it opens the possibility of modelling and analysing electrocapillary systems that are still challenging to tackle. In this way, the broad phenomena of electrocapillarity can be modelled efficiently within a single framework [4]. [1] Jones, T.~B. Liquid dielectrophoresis on the microscale. J Electrostat, 2001, 51-52:290--299. [2] McHale, G.; Brown, C.~V.; Newton, M.~I.; Wells, G.~G.; Sampara, N. Dielectrowetting driven spreading of droplets. Phys Rev Lett, 2011, 107:186101, 10. [3] Krüger, T.; Kusumaatmaja, H.; Kuzmin, A.; Shardt, O.; Silva, G.; Viggen, E.~M. The Lattice Boltzmann Method: Principles and Practice. Springer, 2016. [4] Ruiz-Gutiérrez, É.; Ledesma-Aguilar, R. Lattice-Boltzmann simulations of electrowetting phenomena. Langmuir, 2019, 35(14):4849--4859.

43

Dielectrophoresis of Air Abigail Rendos, Lucas Soffer, Aleksandrs L. Zosuls, Brian M. Walsh, and Keith A. Brown

alrendos@bu.edu brownka@bu.edu

Materials Science and Engineering, KABLab Boston University

USA

Dielectrophoresis (DEP) describes neutral particles moving in non-uniform electric fields. Typically, the force generated by DEP is proportional to the volume of the object, which makes it difficult to observe the effect of DEP on molecular-scale objects. In this work, we experimentally observe the DEP of gas in non-uniform electric fields generated by macroscopic electrodes and show that this effect can be large enough to generate audible sound. In order to prove that this sound originates from DEP, we perform a series of experiments with mixtures of carbon dioxide and nitrogen and find the magnitude and frequency of emitted sound to be in agreement with analytical predictions. Additionally, we find that ethanol or isopropanol vapors in air produce predictable and discernable acoustic signatures when driven using DEP. Overall, the observed sound agrees with a multiscale model of DEP of gas. Since the active elements that generate this sound can be substantially sub-wavelength, produced as part of standard printed circuit board fabrication, and feature no moving parts, they are highly amenable for developing active metamaterials or phase-controlled antenna array applications. Reference: Soffer, L.; Rendos, A.; Zosuls, A. L.; Walsh, B. M.; Brown, K. A., Dielectrophoresis of air. Appl. Phys. Lett. 2020, 116 (8), 084101.

44

Keynote

Potentials of Dielectrophoresis for an Analytical Chemist Alexandra Ros

alexandra.ros@asu.edu alexandra.ros@asu.edu

School of Molecular Sciences, Center for Applied Structural Discovery, The Biodesign Institute Arizona State University

USA

Dielectrophoresis has been successfully employed in the past for the manipulation of a variety of particles ranging from inorganic nanomaterials to biological entities such as cells or organelles. While theoretical models describing particle dielectrophoresis are excellent to predict and exploit dielectrophoresis for a variety of analytical applications, dielectrophoresis on the molecular level remains challenging both in describing the underlying polarization and in suitable experimental settings. A better understanding of (bio)molecular dielectrophoresis could however contribute to the development of novel separation, purification and fractionation techniques needed in the complex field of biomolecular characterization. Due to the comparatively small dielectrophoretic forces expected on the molecular level, the experimental challenge remains to design platforms with which dielectrophoresis can be evoked, but also studied without interference from other forces, especially such induced in electric fields and in microenvironments. Here, novel microfluidic realization platforms are presented that allow to study the DEP response of bioparticles and biomolecules, based on innovative microfabrication approaches including 3D printing. In addition, the potential of dielectrophoresis for separation and fractionation of biomolecules such as DNA and proteins as well as organelles will be presented.

45

Session 6

Optimized Electromanipulation Buffer for Enhanced Cell Viability and Dielectrophoretic Consistency

Alexandra R. Hyler, Daly Hong, Rafael V. Davalos, Nathan S. Swami, Eva M. Schmelz, ahyler@cytorecovery.com

eschmelz@vt.edu CytoRecovery, Inc. Biomedical Engineering and Mechanics (VT), Electrical and Computer Engineering (UVA),

Human Nutrition, Foods and Exercise (VT) CytoRecovery, Inc., Virginia Tech, University of Virginia

USA

Cell separation has become a critical diagnostic, research, and treatment tool for personalized medicine. By isolating subpopulations from heterogeneous cell mixtures, researchers have started identifying genetic markers, protein expression or activation levels, cell-cell interactions, and cell-environment communications that initiate, propagate, and alter disease progression. Alternative approaches through the use of microdevices that do not rely on protein markers or antibodies could provide opportunities to enrich cell populations using their biophysical and phenotypic properties inherent to each cell type. Dielectrophoresis (DEP) provides one such biophysical separation technique that can target cell subpopulations without labels and return native cells back to the researcher for downstream analysis. One challenge in employing any DEP device is the sample being separated must be transferred into an ultra-low conductivity medium, which can be detrimental in retaining cells’ native phenotypes for separation. Determining cell viability before, during, and after DEP is often neglected, and cell health greatly affects properties like membrane capacitance which can alter DEP frequency spectra. These changes can, therefore, alter the crossover point and slope, aspects crucial for determining DEP separations from heterogenous samples. Here, we investigate the impact of DEP buffer formulations on cell viability and DEP properties. We optimize a formulation, Cyto Buffer TM, that significantly improves cell viability by about 20% for long-term (up to 3h) exposure for cell lines regardless of origin, malignancy, or sensitivity. We also find that the newly formulated buffer retained DEP properties and diameters more similar to that of their native media when compared to the traditional DEP buffer formulation. Finally, we vetted Cyto Buffer’s usability on a cell separation platform (Cyto R1 TM), employing contactless DEP, to determine combined efficacy for cell separations. In this work, cells were determined to be >70% viable following exposure to Cyto Buffer, flow stimulation, electromanipulation, downstream collection, and subsequent 48h growth. The developed buffer demonstrated improved shelf life and cell viability for use in electrical cell manipulation, enrichment, and recovery for next generation cell separations.

46

Detection of Autologous Blood Transfusions Using Dielectrophoresis Edwin D. Lavi, Zachary Gagnon

elavi1@tamu.edu zgagnon@tamu.edu

Artie McFerrin Department of Chemical Engineering/Gagnon Group Texas A&M University,

USA

Dielectrophoresis (DEP) is a powerful electrokinetic technique for quantifying cell electrical properties, including the cell cytoplasm conductivity, membrane conductance, and membrane capacitance. In this talk, we present our efforts to quantify storage-induced human red blood cell (RBC) electrical property changes for use in detecting athlete cheating by an autologous blood transfusion (ABT) in endurance sports. In an ABT, athletes transfuse their own previously stored blood in order to boost oxygen capacity and athletic performance. It remains difficult to develop a method for directly detecting ABTs. In this talk, we present a novel approach for detecting ABTs using the DEPtech 3DEPreader. In particular, we demonstrate the ability to measure electrical property changes of RBCs stored in a bloodbank refrigerator for up to 20 days. Our experimental approach is enabled using our novel DEP glutaraldehyde crosslinking chemistry and high permittivity electrolyte buffer technique. We first show the ability to repeatedly measure the DEP spectra of crosslinked human red blood cells using the 3DEP reader.We next show the optimal DEP buffer conductivity/permittivity and glutaraldehyde crosslinking conditions to measure both cellular crossover frequencies (COFs) and the complete DEP spectrum of human RBCs in the 3DEP and compare these COF measurements with those obtained using traditional DEP quadrupole electrode arrays. Finally, we use this assay to quantify the storage induced changes of human red blood cells over a period of 20 days. We conclude our talk discussing the role dielectrophoresis can play in detecting ABTs.

47

Integration of electrodes on glass suspended microchannel resonators for DEP trapping of particles

Roberta Calmo, Jonathan Cottet, Clémentine Lipp, Stefano Stassi, Philippe Renaud and Carlo Ricciardi

cottet@mit.edu carlo.ricciardi@polito.it

EPFL-STI-IMT-LMIS4 École Polytechnique Fédérale de Lausanne

Switzerland

Suspended microchannel resonators (SMRs) have recently attracted great interest as highly sensitive biosensors to study cell populations, proteins and other micron-sized analytes [1]. Unfortunately, time-consuming, challenging and expensive fabrication techniques are required [2], limiting the widespread diffusion of this sensing technology as point-of-care devices for single cell or particle analysis. In this context, one of the main challenges remains the possibility to manipulate directly the physical location of cells or particles inside the SMR. So far the only strategy of dynamic trapping was obtained creating a physical constriction inside the microchannel. The possibility to trap particles inside a SMR by dielectrophoresis was always considered challenging taking into account the multilayered geometry and the complex design of the mechanical resonator. In fact, this technique would require the further deposition of electrodes either in direct contact with liquid [3] or through an insulating layer [4]. In this work, we improved the femtosecond laser fabrication approach, which allowed to develop these platforms entirely in glass by a 3D direct writing technique, followed by a mask-free wet etching step, optimized in a previous work by Calmo et al. [5]. The femtosecond-based process flow has been modified in order to easily deposit two coplanar electrodes on top of the suspended beam that can be exploited to integrate the dielectrophoresis system for cells and particles trapping. Considering the possibility to separate the two characteristic phases of the femtosecond based fabrication, laser writing and wet etching, we introduced between them different steps of photolithography and dry etching, defining the electrodes on top of a flat surface and avoiding the complex procedures of deposition on top of non-homogeneous substrate like a suspended beam. The electrodes material was optimized to withstand highly concentrated Potassium Hydroxide etching. The resonance properties of the fabricated SMR were accurately characterized in terms of frequency, quality factor and Allan deviation. The glass SMR with electrodes provided similar performances to the monolithic glass SMR. Preliminary tests with beads of different materials showed that it was possible to trap in flow conditions beads for various combinations of voltage and flow rate. This open the possibility to perform trapping and concentration of cells inside the SMR for an increased sensitivity, as alternative to the mechanical constriction approach. References [1] Cetin, A. E., Stevens, M. M., Calistri, N. L., Fulciniti, M., Olcum, S., Kimmerling, R. J., Munshi, N. C., Manalis, S. R., Nature Communications 2017, 8, 1613. [2] Burg, T. P., Manalis, S. R., Applied Physics Letters 2003, 83, 2698-2700. [3] Cottet, J., Kehren, A., Lasli, S., van Lintel, H., Buret, F., Frénéa-Robin, M., Renaud, P., Electrophoresis 2019, 40, 1498-1509. [4] Gascoyne, P. R. C., Vykoukal, J. V., Schwartz, J. A., Anderson, T. J., Vykoukal, D. M., Current, K. W., McConaghy, C., Becker, F. F., Andrews, C., Lab on a Chip 2004, 4, 299-309. [5] Calmo, R., Lovera, A., Stassi, S., Chiado, A., Scaiola, D., Bosco, F., Ricciardi, C., Sensors and Actuators B-Chemical 2019, 283, 298-303.

48

Isolating and Concentrating Unlabeled SARS CoV-2 from Saliva with iDEP Alex Ramirez, Matthew McFadden, Hoai Nguyen, AKM FK Rasel, Kallen Ruddle, Jared Smithers, Jerry Sheu, Bereket Estifanos, Michael Sauer, David Charlot, Anna Vu, Mikayla Carlson, Sean Seyler, Peter

Dawson, Zhiyun Li, Houpu Li, Alex McLaren, Brenda Hogue, Brian Richardson, Mark A. Hayes*

mhayes@asu.edu Arizona State University

School of Molecular Sciences Tempe, Arizona USA

Dielectrophoresis generates a force on particles which varies with subtle changes in the particle-solvent-electric field system. In certain configurations, this force can be used to uniquely separate biological particles with very small differences between them. We use this capability to isolate and concentrate virus from biological fluids, in this case saliva. Our target is SARS CoV-2, the virus responsible for the current pandemic. Our goal is to provide such high resolution separations that SARS CoV-2 and only this bioparticle is present at a predetermined location on a microfluidic system. This capability would allow for a simple readout: if no particles are present, no SARS CoV-2 is present in the sample—allowing for a new paradigm in diagnostic capabilities. However, this is an intensely difficult task and one that has not been previously envisioned. Using DEP as part of the final separation step, our system uses photolithographically fabricated filter-enrich-recovery system to select only particles of a similar size to the targeted virion. This removes background materials, increases the concentration of the virus and reduces the volume to be processed. For the dielectrophoresis step, we are using a conventional gradient insulator based DEP (g-iDEP) to quantify the specific electrokinetic/dielectrokinetic/higher-order electric field effects of the bioparticles. With this system we have shown capture of coronavirus (mouse hepatitis virus, MHV; very similar particle to SARS CoV-2, non-BSL3 space) and are targeting wild-type and variant-variant pairs to shown proof of principle. This entire project was enabled by the recent discussions on the Classius-Mossotti factor and its application to small particles, including proteins. This system contrasts dramatically with traditional diagnostic approaches where the selectivity is held in molecular recognition, sequence specificity and/or spectroscopy, whereas the selectivity is generated by separations science and a reagentless non-specific detection system of light scattering is used. This presentation will focus on our recent results with the unaltered, unlabeled and viable MHV particles in g-iDEP with light scattering detection.

49

Poster Session 2

Selective Trapping and Retrieval of Single Cells Using Microwell Array Devices Combined with Positive- and Negative-Dielectrophoresis Masato Suzuki, Misaki Hata, Tomoyuki Yasukawa

suzuki@sci.u-hyogo.ac.jp yasu@sci.u-hyogo.ac.jp

Graduate School of Material Science/ Analytical Chemistry University of Hyogo

Japan

This study demonstrates proposed manipulation techniques for retaining and retrieving target cells arrayed in microwells by microwell array devices that employ dielectrophoresis (DEP). Currently, the recovery of the target cells after determining the cellular function of individual cells trapped in a cell array employs the suction of the target cell in microwell by a microdispenser. It is necessary to arrange a micropipette above microwells with the target cells by a mechanical micromanipulator and carefully regulate the collection of each cell. Unfortunately, this method is very time consuming and labor intensive. This study proposes a simple device for flexible dielectrophoretic manipulation of cells based on the combination of positive DEP (p-DEP) and negative DEP (n-DEP). The upper substrate with microband electrodes was mounted on the lower substrate with microwells on the same design of microband electrodes by 90 degree relative to the bottom substrate. A cell array was prepared by directing cells in each microwell with an attractive force of p-DEP. Thereafter, a repulsive force of n-DEP was employed to retrieve the target cells from the microwell array. We demonstrated the retrieval of a single target cell from the cell array formed by p-DEP. Mixtures cells stained in green or red were prepared in a ratio of 10:1 to demonstrate the retrieval of red cells designated as the target cell. The application of the AC signal (1 MHz, 3 Vpp) to both the upper and the lower microband electrodes with opposite phasing resulted in the formation of the cell array. A single red cell was trapped in the 3–I well (positioned at intersection band-electrode 3 and I) and eight green cells were trapped in the others. Subsequently, the frequency applied to band electrode 3 on the upper substrate and the band electrode I on the lower substrate was switched from 1 MHz for p-DEP to 300 kHz for n-DEP, while the frequency applied to the other band electrodes was maintained at 1 MHz for p-DEP. The target red cell in 3–I well was gradually removed over a few seconds after switching the frequency, and it was then transferred downward in the image by slight fluidic flow. In contrast, the other green cells remained in the original position. The results indicated that the repulsive force of n-DEP from the strong electric field region acts on the cell in the 3–I well that comprises both band electrodes switched in the n-DEP frequency region. It is noted that p-DEP still acted on cells in wells comprising band electrodes that applied an AC signal in the p-DEP and n-DEP frequency regions, respectively. Thus, this system would make it possible to retrieve target cells selectively from the array of cells and recover them in an outlet without the microdispensers after removing the upper substrate.

50

Transition to high-frequency (MHz) AC electroosmosis observed on the nanoscale Gerhard Blankenburg, Leonardo Lesser-Rojas, Chia-Fu Chou

Gerhard.Blankenburg@gmail.com cfchou@phys.sinica.edu.tw

Nanobioscience Lab, Institute of Physics, Academia Sinica Department of Physics, National Taiwan University (NTU); Taiwan International Graduate Program, Nano

Science and Technology Program (TIGP-NANO) Taiwan

AC electroosmosis (ACEO) has been known for about 20 years after it was reported for electrodes in the size range of hundreds of micrometers down to a few micrometers. At the time, scaling laws were derived implying that the frequency range of ACEO can be tuned to the MHz range if the experimental assay could be downscaled to the nanometer range, but this transition has never been tested experimentally. To explore whether this transition exists, we fabricated a device comprised of 15 pairs of elongated microelectrodes with a gap between their tips. Devices with 9 μm gaps and 50 μm gaps were used in the experiments. Additionally, a configuration with another pair of nanoelectrodes connected to each pair of the microelectrodes, with a 5-100 nm gap in between the tips of nanoelectrodes, was employed to obtain results for the nanoscale behavior. Experimental results were obtained by fluorescence microscopy of polystyrene nanospheres in PBS buffers with low conductivity (<1 mS/cm). From the results, evidence is presented for ACEO on nanoscale electrodes in the high-frequency range. Videos obtained from the microelectrode-only system clearly demonstrate that ACEO is observed, creating a flow pattern that can also be identified on the much smaller nanoelectrodes of similar geometry. In addition to this, the transition of ACEO flow patterns from the kHz to MHz range has been observed on the device with nanoelectrodes. The competition of DEP force and ACEO flow over a range of AC frequencies and amplitudes is elucidated. These results indicate a new advance into the unexplored regime of ACEO and prove the existence of an effect in the MHz range mostly ignored in existing literature. Our finding has important implications for nanoscale DEP experiments and show potential for sample delivery or mass transport by frequency-tuning based pumping mechanisms on devices that combine electrode features of different size scales.

51

High Throughput Assay to Measure Cellular Dielectrophoretic Mobility from Individual-Based Cell Movement

Hyunwoo Lee, Sena Lee, Sejung Yang, Sang Woo Lee* sychoi0091@gmail.com yusuklee@yonsei.ac.kr

Department of Biomedical Engineering Yonsei University Republic of Korea

We demonstrated a novel automatic method of simultaneous segmentation and tracking of multiple individual cell trajectories, which were induced by a dielectrophoretic (DEP) force acting on the cells inside a microfluidic device. In the demonstration, our algorithm enabled segmentation and tracking of numerous MCF-7 cell DEP trajectories while the DEP force was oscillated between positive and negative. In addition, the cross-over frequency was measured by analyzing the segmented and tracked trajectory data of the cellular movements caused by the positive and negative DEP force. The measured cross-over frequency was compared with previous results. The multi-cellular movements investigation based on the measured cross-over frequency was repeated until the viability of cells was unchanged in the same environment as in a microfluidic device. The results were statistically consistent, indicating that the developed algorithm was reliable for the investigation of DEP cellular mobility. This study can be used as a powerful platform to measure simultaneously the DEP-induced trajectories of a multitude of cells, and to investigate in a robust, efficient and accurate manner the DEP properties at the single cell level as well as at an ensemble cell level

52

Adhesion measurement of water droplets on superhydrophobic surfaces via electric fields

Bin Li, Longquan Chen, and Sangwoo Joo libin@ynu.ac.kr swjoo@yu.ac.kr

Mechanical Engineering Yeungnam University

Republic of Korea

Adhering droplets on superhydrophobic surfaces may deform and lift off when the electric field is employed to reaches a certain critical value, due to unbalancing of the electrostatic force and the summation of the adhesion force and the gravitational force. Therefore, utilization of an electric field on superhydrophobic surface is a potential method which in turn evaluates the adhesion force. In this work, we aims to measure the adhesion force between the water droplets and the nano-coated superhydrophobic surfaces via electric field. The droplet deformation and motion are recorded by a high speed camera and analyzed in sequential frames. The influences of the contact angle hysteresis and the gap width of two parallel electrodes on the critical voltage are investigated. Finally, the theoretical models are proposed to predict and support our observations. This work can offer a platform for improving the performance of self-cleaning surfaces, anti-icing surfaces, and condensation heat transfer.

53

Digital Droplet Microfluidics with Programmable Liquid Handling Based on Contact Charge Electrophoresis

Huai Zheng, Fan Bai, and Sang W. Joo huai_zheng@whu.edu.cn huai_zheng@whu.edu.cn

Department of mechanical engineering Yeungnam University

Republic of Korea

Microfluidics is a system to control micrometer-scale liquid flow, has vast applications in biomedical and chemical analysis and microreaction. In microfluidics system, droplet-based microfluidics is an important subcategory thanks to its many advantages which include a small volume of reagents consumed, massive production of monodisperse droplets, high surface-area-tovolume ratio that facilitates fast reaction, and independent control of each droplet. Traditional droplet-based microfluidics is based on generated droplets in closed microchannels. But it brings some drawbacks, such as chip with one-time use, large flow resistance, and complexity of fabrication. Droplet microfluidics with open microchannels or without microchannels has emerged as a development, and it is named open microfluidics. The open microfluidics has been studied widely. The liquid flow drive mechanism is various and droplets are driven by surface tension, gravity, magnetic force, electrical force and acoustic force. In current open microfluidics, droplets move with sliding on surfaces. It leads to liquid fouling problems on surfaces and can not realize complex droplet manipulation. In this study, we proposal a new open microfluidics technique based on contact charge electrophoresis. Droplets are immersed into silicone oil, and deposited on electrode surfaces due to their higher density. With applying the voltage between a needle and plate electrodes, droplets obtain charges and are levitated at the interface between silicone and air. Moving the needle electrode can digitally manipulate droplets. The charge transfer and flow behavior of droplets are studied in detail. Through this droplet handling method, we realize the functions of droplets transportation, coalesce and mixing in microfluidics.

54

Numerical study of Janus droplet spontaneously migration by OpenFOAM

Fan Bai, Hongna Zhang, and Sang W. Joo baifan@ynu.ac.kr baifan@ynu.ac.kr

Department of mechanical engineering Yeungnam University

Republic of Korea

Numerical simulation of droplet manipulation on a flat, solid substrate has long been of interest to the digital lab-on-a-chip system, fog-harvesting, DNA micro-array, and biological engineering. So far, various researches have been performed to develop efficient and stable simulation methods for general single-phase droplet motions. The study of Janus droplet (multiphase droplet) motion on a solid surface has not been reported up to now, while Janus droplet can also be frequently met in the aforementioned applications. Janus droplets have attracted a wide range of interest over the past decades, which have unique architecture properties of at least two different kinds. They thus provide a wider range of applications than single-attribute structures due to their centrally asymmetrical structure. The method of droplet manipulation can be cataloged into two methods according to the droplet generation approach: passive method and the active method. For the passive method, the droplet is manipulated by the special design of the substrate that creates a velocity gradient locally. For the active method, the droplet is manipulated by localized energy gradients generated by external forces such as electric force and thermal energy. Unlike the previous two methods, however, it is worth investigating whether Janus droplets can move autonomously on smooth horizontal surfaces without any treatment owing to their unique properties. It is foreseen that a wider range of imaginable applications of liquid combinations based on two different chemical properties will be opened. Moreover, the numerical simulation can reasonably investigate the transient motion of Janus droplet on the horizontal substrate in theory. In this work, we present numerical simulations based on a multiphase-field model of the open-source CFD-software OpenFOAM, on account of the collocated finite-volume method (FVM). It is based on appropriately designed relations between surface tension and contact angle in order to ensure the right contact angles at multiphase junctions. A new solver for multiphase fluid flows called multiphaseInterDyMFoam was used to simulate the Janus droplet migration. As known, the interface tracking method of VOF depends on the mesh, and more accurate results can be obtained by choosing an appropriate mesh. Therefore, a multiphase dynamic refine mesh method was compiled to solve this problem. The unsteady laminar Navier-Stokes equation is solved using an unfixed Eulerian unstructured grid in each phase. The spontaneous migration of Janus droplet on a substrate was first achieved through numerical simulation method. The Janus droplets spontaneously move in a direction without any external force due to the difference in surface tension caused by the difference in physical properties of the Janus droplets. The Janus droplets are set with different rheological properties to investigate the effect of the interaction between two phases. It has been found that the manipulation of Janus droplet is independent of the droplet diameter; however, it largely depends on surface tension, rheological properties and volume ratio without any external force. Moreover, the variations of the velocity and acceleration of Janus droplets change with the change of aforementioned properties.

55

Dielectrophoresis based sensor using surface conductivity difference of particles Kang In Yeo, Yewon Kim, Seungyeop Choi

rkddls1905@gmail.com yusuklee@yonsei.ac.kr Biomedical Engineering

Yonsei university Republic of Korea

Dielectrophoresis (DEP) based measuring system is a promising multi-probe sensor installed in microfluidic devices, which specializes in measuring electrical properties of a given sample. In the DEP system, the characteristic of frequency-dependent DEP force exerted on the microparticles is generated by conductivity differences between the microparticles and medium. Here, we developed DEP sensor using negative DEP force exerted on the polystyrene microparticles. Different chemical components on the homogenous particles creates different surface charges on the microparticles, thereby allowing for efficient detection. Surface conductivity changes of the microparticles in different surface conductivity conditions are measured by particle movement, which occurs owing to the negative DEP force. As a result, our method which selectively detect different surface charge particles can be practically applied as a simple and efficient device for various detection purpose.

56

Sidewall electrodes in a microchannel for high-throughput dielectrophoretic separations Karina Torres-Castro, XuHai Huang, Carlos Honrado, Walter B. Varhue, Nathan S. Swami

kt2fe@virginia.edu ns5h@virginia.edu

ECE Department/SwamiLab University of Virginia

USA

Sidewall electrodes in a microchannel for high-throughput dielectrophoretic separations Karina Torres-Castro*, XuHai Huang, Carlos Honrado, Walter B. Varhue, & Nathan Swami * Presenting Author Phenotypic heterogeneity of biosystems, wherein fractional cellular subpopulations can determine the net biological function, challenges the ability to associate specific markers to disease outcomes. Hence, there is much interest in single-cell methods for separation and downstream validation of phenotypes from heterogeneous sample. We present a microfluidic device methodology for enabling high-throughput dielectrophoretic (DEP) deflection of cells (~10 mL/min flow rate and deflection of ~106 cells/min) by creating sequential field non-uniformities that extend over the microchannel depth (50 mm) for downstream validation by single-cell impedance cytometry. In this manner, flow focused single-cell streamlines can undergo progressive dielectrophoretic deflection with minimal dependence on the cell starting position, its orientation versus the field and intercellular interactions, thereby enabling separations at enhanced throughput (i.e., greater sample flow rates and cell concentration levels). Using a sample of red blood cells at concentration levels of ~109 cells/mL from 48% hematocrit human blood, the dielectrophoretic deflection of single cells is optimized for field conditions (Vpp/cm and frequency), flow rates (1-10 mL/min) and throughputs (~106 cells/min) to ensure sufficient spatial separation (>20 mm over the microchannel length) for cells of varying membrane capacitance to enable their collection at different outlets. Cell separation on this device is validated by downstream impedance cytometry using heterogeneous samples containing, 50% each of healthy human red blood cells (h-RBCs) and 1% glutaraldehyde-fixed RBCs (f-RBCs). Phenotype-specific separation is confirmed based on electrical opacity determined by impedance cytometry data of the respective collected fractions after DEP separation that reflects the expected membrane capacitance differences (capacitance of h-RBCs > f-RBCs), while invariance of electrical opacity of DEP collected versus input h-RBCS confirms the maintenance of cell functionality after DEP separation. Based on this, we envision application of this device configuration in future work for selective cell isolation towards quantifying phenotypic heterogeneity of cellular systems. References: K. Torres-Castro, C. Honrado, W.B. Varhue, V. Farmehini, N. S. Swami*, “High-throughput dynamical analysis of dielectrophoretic frequency dispersion of single-cells based on deflected flow streamlines”, Anal. Bioanal. Chem. (2020). DOI: 10.1007/s00216-020-02467-1 XuHai Huang, Karina Torres-Castro, Walter Varhue+, Armita Salahi, Ahmed Rasin+, Carlos Honrado+, Audrey Brown, Jennifer Guler, and Nathan S. Swami*, “Self-aligned sequential lateral field non-uniformities over channel depth for high throughput dielectrophoretic cell deflection”, Lab on a Chip (2021), DOI: 10.1039/D0LC01211D.

57

Automated Fluorescence Quantification of Extracellular Vesicles Collected from Human Plasma via Dielectrophoresis

Kyle Gustafson, Katherine Huynh, Jesus Bueno, Augusta Modestino, Daniel Heineck, Michael J. Heller, Stuart Ibsen gustafky@ohsu.edu

ibsen@ohsu.edu Department of Biomedical Engineering, School of Medicine, Oregon Health & Science University

Cancer Early Detection Advanced Research (CEDAR) Center, Knight Cancer Institute, Oregon Health & Science University

USA

Proteins on the surface and interior of extracellular vesicles (EVs) are an essential class of target biomolecules for the early detection of cancer via liquid biopsy. High conductance dielectrophoretic (DEP) chip technology enables isolation of EVs directly from biological fluids, such as blood plasma, enabling subsequent on-chip detection and quantification of EV proteins through immunofluorescent labeling. Low physiological concentrations of circulating EVs in early disease present a challenge in separating the immunofluorescent signal of labeled biomarkers from background noise. Here we demonstrate two new approaches to improve the fluorescence analysis of EV proteins collected and labeled on DEP chips. The first approach aims to control for variability in biomarker collection efficiency across individual chips so that differences in output indicate differences in physiological concentrations of circulating EVs across patients. Collection efficiency depends on multiple factors including sample conductivity and consistency in chip performance. We measure the collection efficiency of each chip by spiking a known concentration of fluorescent nanoparticles (referred to as the “internal standard”) into plasma samples prior to EV isolation. DEP forces act concurrently on the spiked particles and plasma EVs, allowing collection and purification of both. After immunolabeling the EV proteins, fluorescence images of the spiked nanoparticles and the immunofluorescent biomarker labels are acquired using appropriate optical filters. A custom algorithm quantifies fluorescence levels. EV protein fluorescence is then normalized to the fluorescence level of the spiked particles, improving the accuracy of direct comparisons between patient samples by accounting for differences in collection efficiency across chips. The ability to make direct comparisons between patients is essential for establishing patterns in EV protein concentrations that distinguish healthy patients from those with cancer. The second approach aims to standardize the analytical method for quantifying fluorescence levels of spiked particles, labeled EV proteins, and other labeled biomarkers collected at high-field (positive DEP) regions of DEP chips. An important component for standardizing analysis is minimizing user bias, which is achieved by full automation of the analytical process. We developed a custom, automated image processing algorithm in MatLab to identify high-field regions at each chip electrode and to subtract local background noise levels in each fluorescent region. Localized noise reduction and fluorescence quantification treats each region as a technical replicate (unique sampling) of biomarkers. With sample sizes between 100 and 200 electrodes per image, the population mean can be estimated within a narrow range with high confidence. The average value determined by the algorithm is a robust measurement of the amount of fluorescence signal emitted from immunolabeled EVs or spiked particles isolated via DEP. Experiments varying exposure time and particle concentration established a dynamic range for analytical output. The described techniques reduce the unwanted effects of collection efficiency differences across chips and user bias in fluorescence signal quantification on measured physiological concentrations of circulating EV proteins across patients. Researchers can take advantage of these methods to develop low-volume DEP assays that detect early disease states directly from high-conductance biofluids.

58

A study on alterations in Membrane Physiology of ultrasonically irradiated Human Erythrocytes through Dielectrophoresis

Kaleem Ahmed Jaleeli and Mohd Abdul Saleem kaleemjaleeli@gmail.com kaleemjaleeli@gmail.com

Physics Nizam College, Osmania University

India

The cells are basic building blocks of any biological system. To get an insight into the physiological reality of biological cells it is necessary to obtain the knowledge regarding their electrical makeup. In the present investigation, the changes in the membrane physiology of ultrasonically irradiated Human Erythrocytes were observed using a novel technique – Dielectrophoresis. Blood samples of volume 5ml were collected form the donors with their consent and were stored in heparin to prevent it from coagulation. The samples were irradiated to ultrasonic standing waves using ultrasonic interferometer at a frequency of 3 MHz up to 1 hour at an interval of 15 minutes. Dielectrophoretic Collection Rate (DCR) of Normal and ultrasonically irradiated Human erythrocytes were determined as a function of time at constant frequency, applied voltage, and cell concentration. Further a parameter known as Excess permittivity has been calculated form the knowledge of DCR. The present study reveals that the membrane physiology of ultrasonically irradiated Human erythrocytes is perturbed and this perturbations are observed at cellular level using the technique of Dielectrophoresis. The values of DCR and Excess permittivity of ultrasonically irradiated erythrocytes are found to be high as compared with that of the normal Human Erythrocytes.

59

Studying the DEP behavior of the Trypanosoma brucei parasites Megan Giacobbi and Rodrigo Martinez-Duarte

megan1023g@gmail.com rodrigm@clemson.edu

Multiscale Manufacturing Laboratory, Department of Mechanical Engineering Clemson University

USA

Human African Trypanosomiasis (HAT) is a disease caused by the Trypanosoma brucei parasite in sub-Saharan Africa. There is a strong need for a diagnostic tool able to identify this disease in its first stage to prevent neuropsychiatric effects that develop in the second chronic stage. The use of dielectrophoresis (DEP) to directly observe the Trypanosoma brucei parasite could help enable a reliable diagnostic method for HAT. To this end, our current goal is a protocol to specifically enrich the T. brucei parasites in specific locations in order to facilitate their direct observation. A previous study shows promise of high enrichment, up to 780% in 50 seconds of the parasite on a titanium device at 750 kHz and 5 Vpp (Keck et. al). Here we present a preliminary study of how different parasite properties can impact the DEP behavior of T. brucei when exposed to a given electric field, and thus the efficiency of its enrichment. Our experimental device features an array of indium tin oxide (ITO) interdigitated electrodes (semicircles, triangles, and lines). Exposing the parasites to a positive DEP force is expected to result in parasite enrichment at the tip of the electrodes. We will measure such enrichment and correlate it to parasite properties such as age, fluorescence staining and changes in their dimensions. This is important towards a diagnosis assay because it will allow characterization of the disease stage through the subtle differences in the parasite DEP response. This has the potential to lead to a more accurate and robust diagnostic method and more focused drug treatments for HAT in the future.

60

Monitoring Eryptosis using Dielectrophoretic Characterization Sydney Joseph, Timothy Swimmer, Ronald Blum, and Erin A Henslee

josesc20@wfu.edu hensleea@wfu.edu

Department of Engineering Wake Forest University

USA

Apoptosis (programmed cell death) is a biological process that controls cell growth. Eryptosis is a form of programmed cell death akin to apoptosis. Eryptosis is characterized by cell shrinkage, membrane blebbing, activation of proteases, and phosphatidylserine exposure at the outer membrane leaflet. Exposed phosphatidylserine is recognized by macrophages that engulf and degrade the affected cells. Phosphatidylserine (PS) is a phospholipid that is flipped from the inside to the outside of dying cells and signal macrophages to degrade the dying cell. Failure or an unbalance in the process can lead to severe consequences such a liver failure or anemia. Other pathological conditions that alter Erypotic process include malaria, sickle-cell disease, and thalassemia. Though Eryptosis is a well-known process, and effects such a malarial-anemia due to Eryptosis have been established, mechanisms of these processes are still unknown. Current RBC analysis such as blood smears, flow cytometry, and deformability relies on manual observations and often-costly equipment. Moreover, these techniques depend on observable changes in cellular morphology or behavior. This work demonstrates DEP’s ability to monitor the Eryptotic process. We followed a recently published step-by-step procedure using the calcium ionophore, ionomycin, to enhance intracellular Ca2+ concentration and induce Eryptosis. We treated RBCs with 1 μM and 10 μM ionomycin for 2 h and monitored RBC DEP properties for up to 4 days using the 3DEP system. We used microscopy to monitor RBC count and morphology and Annexin-V to monitor PS flipping. We did not observe any significant changes to bulk DEP properties for the 1 μM 2 h treatment. We observed a significant spike in membrane conductance and cytoplasmic conductivity in the 10 μM 2 h treatment group at the 24 h time point. We also saw a slightly significant (p=0.048) decrease in membrane capacitance. From 48-96 h these properties showed no significant trends though they all decreased. Annexin-V staining revealed that 1 μM 2 h treatment had less than 20% Eryptosis, which could account for our lack of detection via bulk DEP analysis (compared to >50% for the 10 μM treatment). Future work will investigate multi-populations to determine if this small percentage of Eryptosis and possibly hemolysis can be detected. The marked spike of parameters at the 24 h timepoint of the 10 μM treatment presents a promising hypothesis for future analysis. An increasing Ca2+ concentration which stimulates ion pathways such as the Gardos Channels permits the efflux of K+ and Cl- and also water due to osmosis. Therefore, it can be hypothesized that the increasing activity of such channels and shrinkage of the cells may be localized over even shorter periods between 0 and 24 h. The steady decrease in parameters from 24 to 96 h likely reflects the degradation of the cell and ion pathways. Taken together, this work demonstrates the potential of DEP to monitor Eryptosis, however further investigation is needed to examine the significance of the characterization.

61

Keynote

AC Electrokinetics: Applications for Cell Biomechanics and Fatigue Failure E (Sarah) Du du@fau.edu du@fau.edu

Living Devices & Biosensors Lab, College of Engineering and Computer Science Florida Atlantic University

USA

Mechanisms of biophysical property degradation, damage accumulation under cyclic loads as well as their contributions to the onset and progression of pathological states in biological cells are not fully understood. This knowledge gap is largely due to a lack of experimental methods that can apply cyclic loads for characterization of fatigue behavior of biological cells. Classical experimental methods such as optical tweezers, micropipette aspiration and atomic force microscope cannot be readily adapted for generation of such cyclic loading profiles. Here we present a general method that employs unique AC electrokinetics phenomena, namely dielectrophoresis and electrodeformation, for characterization of cell biomechanics and mechanical fatigue in biological cells. We demonstrate the capability of this method by characterizing human red blood cells for their dynamic and fatigue behavior upon monotonic and cyclic loads.

62

Session 7

Effect of dielectrophoresis on colloidal cylinders near a planar boundary Atakan Atay, Barbaros Cetin barbaros.cetin@bilkent.edu.tr barbaros.cetin@bilkent.edu.tr

Mechanical Engineering/Microfluidics & Lab-on-a-chip Research Group Bilkent University

Turkey

The electrokinetic behavior of the particles in a microchannel is widely investigated due to their potential use in microfluidic applications such as separation or characterization of bio molecules, bacteria, self-assembly processes etc. When nearby particles and/or channel walls present in a microchannel, the electrokinetic behavior of the particles under strong particle-particle and particle-wall interactions which may also include strong induced dielectrophoretic forces. These boundary effects can be well captured with a computational model which solves the field variables with the presence of the particles and channel wall. However, algorithmic and computational challenges in resolving complex particle shapes and inefficient implementation of particle motion come into view when volume based numerical models are implemented. Alternatively, Boundary Element method can be implemented to eliminate the aforementioned challenges through discretization of the particle and channel surfaces, which effectively reduces the dimension of the problem of interest by one. In this study, the electrokinetic motion of colloidal cylinders near a planar boundary is presented, and the effect of the dielectrophoresis on the motion of colloidal cylinders are investigated.

63

Dielectrophoretic-driven deformations of a liquid-fluid interface Israel Gabay, Federico Paratore, Evgeniy Boyko, Antonio Ramos, Amir Gat, Moran Bercovici

Israelgabay1@campus.technion.ac.il mberco@technion.ac.il

Mechanical engineering, Microfluidic Technologies Laboratory Technion - Israel Institute of Technology

Israel

DEP has been primarily investigated in the context of manipulating discrete objects, particularly particles and cells. However, DEP can be used to apply forces to any system exhibiting permittivity gradients. A particularly interesting application, first demonstrated by Brown et al., is the deformation of liquid-fluid interfaces. To date, demonstrations of such actuation have been limited to periodic structures. In this work, we investigate theoretically, numerically, and experimentally the ability to use DEP as a method for shaping a liquid-fluid interface into desired topographies. By choosing a polymer as the liquid and curing it, this approach can also serve as a fabrication method for smooth microstructures. We will present a theoretical model for the deformations of a thin liquid layer due to an electric field established by a pair of closely-spaced parallel electrodes. We modeled the spatial electric field created by the electrodes and use it to evaluate the force distribution on the interface through Maxwell’s stresses. By coupling this force with the Young-Laplace equation, we obtained solutions for the deformation of the interface. To validate our theory and demonstrate the feasibility of this mechanism, we designed an experimental setup that allows spatial dielectrophoretic actuation while providing measurements of microscale deformations. The system is based on microfabricated thin metal electrodes deposited on a glass substrate, which we cover with a thin film of dielectric liquid prior to the activation of the electric field. We used digital holographic microscopy to measure the induced deformations of the liquid-air interface, showing localized deformations with a dynamic range in amplitude ranging from hundreds of nanometers to hundreds of microns. We characterized the deformation as a function of the electrode-pair geometry and film thickness, showing very good agreement with the model. Based on the insights from the two-electrode system, we will present the ability to create desired two-dimensional deformation patterns. Furthermore, using a photopolymer as the dielectric liquid and curing it, this method enables the fabrication of solid objects with smooth (nanometer roughness) topographies. Lastly, we will discuss the response time of the system and the stability of the deformations (while still in liquid form) to external disturbances.

64

Frequency modulated dielectrophoretic particle chromatography Jasper Giesler, Laura Weirauch, Marc Peter Schmidt*, Michael Baune, Georg Pesch

j.giesler@uni-bremen.de gpesch@uni-bremen.de

Chemical Process Engineering University of Bremen, *TH Brandenburg

Germany

Frequency modulated dielectrophoretic particle chromatography utilizes periodic cycles of positive and negative dielectrophoresis to achieve a retardation of particles according to their polarizability. To achieve a chromatographic separation, particles are injected into a pressure driven laminar carrier flow, which transports the particle mixtures via viscous drag through a microfluidic channel. When an inhomogeneous electric field is generated in the device, polarizable particles will show dielectrophoresis, which either is pointed towards field maxima (pDEP) or away from them (nDEP). In our device the maxima are located at the bottom of the microfluidic channel where interdigitated electrodes induce the electric field. Since the direction of the dielectrophoretic movement for many (bio-)particles depends on the frequency of the electric field, a change in frequency can lead to a change in the direction of dielectrophoretic movement. Here, the frequency is modulated, by changing between a maximum and a minimum frequency periodically. When the crossover frequency of the suspended particle is within this frequency range, particles will show pDEP and nDEP (depending on their polarizability) for different time windows. Consequently, three scenarios can be distinguished: I) A particle shows substantial more pDEP than nDEP during the modulation spectrum and therefore predominantly moves towards the bottom of the channel, where the electrodes induce a high electric field strength. Since the fluids’ velocity close to the bottom is low, particles are slowed down by the lower fluid velocity or by getting reversibly trapped at the electric field maxima. The particles are then pushed away from the electrodes by nDEP when the frequency changes. This scenario effectively increases the particle’s residence time. II) When a particle exhibits a balanced pDEP and nDEP movement, the retardation is less pronounced. These microparticles travel towards high field regions when the CM factor is positive and away from them when it is negative. Due to their constant movement orthogonal to the fluid flow direction, they spend less time in regions with low fluid velocities and therefore are eluted fast. III) If nDEP outweighs pDEP, particles are predominantly pushed towards low field regions, which here are present at the channel’s ceiling. Like in scenario I), only low fluid flow is present at the field minima (ceiling) and the particle’s residence time is going to be enlarged. Although the polarizability of particles from scenario I) and III) is different, retention times can be the same. Nevertheless, since the extent of retardation depends on the chosen process parameters (e.g., frequency, voltage), a separation can be realised with a different set of parameters. According to the upper three different scenarios three different binary particle mixtures of polystyrene particles are separated experimentally, supported by a numerical approach to predict retention times, to demonstrate the feasibility of this technique.

65

Tunable nanochannels for dynamic control of the concentration-polarization-based preconcentrated analyte plug

Barak Sabbagh, Sinwook Park, Gilad Yossifon Baraksabbagh@campus.technion.ac.il

gilad.yossifon@gmail.com Faculty of Mechanical Engineering, Micro- and Nanofluidics Laboratory

Technion – Israel Institute of Technology Israel

Passage of an electric current through a permselective medium (e.g., membrane, nanochannel) results in regions of depleted and enriched ionic concentrations at the opposite sides of the membrane interfacing an electrolyte solution, due to an imbalanced ion transport. This phenomenon termed concentration polarization (CP). CP has been the focus of intensive research in the past decade in relation to lab-on-a-chip microfluidic applications, and mainly has been used for highly enhanced immunoassay bioanalysis and sensing by the preconcentration of analytes. Upon the application of electric field, a preconcentrated plug of the charged analyte forms at the outer edge of the depletion layer due to counteracting advection and electromigrative ion fluxes. CP-based preconcentration can continuously accumulate charged molecules of a sample solution and operate under a wide range of electrolyte conditions. However, one of the main drawbacks of current CP based preconcentration systems is the inability to precisely and dynamically control the location of the preconcentrated biomolecule plug to ensure its overlap with the surface immobilized molecular probes. Recently, approaches to control the diffusion layer, and the corresponding preconcentrated molecule plug, by inducing electro-thermal (ET) and alternating-current-electro-osmosis (ACEO) have been demonstrated. Herein, we showed for the first time the ability to gain a precise control over the biomolecule plug by using a pair of tunable nanochannels in series. The ability to dynamically tune the nanochannel dimensions was achieved by using an elastomeric microvalve, consisting of a deformable membrane that is sandwiched between a fluidic microchannel and a control channel. Upon closure of the microvalve the average height of the fluidic channel crosssection reduces to a sub-micron dimension where electric-double-layers (EDLs) overlap between the bottom microchannel surface and that of the membrane wall becomes important and results in ionic permselectivity. An extension to an array of such tunable nanochannels connected in series was also investigated and demonstrated the ability to form multiple biomolecule plugs as well as to enhance the parallelization of diagnosis.

66

Session 8

Different blood cells exhibit rhythmic electrophysiological behaviour Emily Kruchek, Rebecca Lewis, Jonathan Gibbins, Krista Clarke, Matthew Johnson, Rita Jabr, Debra

Skene, Andrew Beale, John O'Neill, Fatima Labeed F.Labeed@surrey.ac.uk F.Labeed@surrey.ac.uk

Centre for Biomedical Engineering University of Surrey

United Kingdom

Endogenous circadian rhythms are present in almost all physiological and behavioural processes. In normal physiology, the central circadian clock (located in the hypothalamic suprachiasmatic nuclei-SCN) synchronises the timing of the peripheral clocks of many organs via multiple neural and hormonal pathways . Desynchronisation of this timing system, e.g. by misalignment of internal timing with behavioural schedules (such as light/dark exposure; sleep/wakefulness; rest/activity and feeding/fasting) increases the risk of metabolic disorders, type 2 diabetes and is associated with cardiovascular events. Dielectrophoresis offers a novel method of assessing cell electrophysiology. For example, we have over a number of years used DEP to identify, track, and understand the presence of circadian rhythms in the electrical properties of red blood cells (RBCs). These cells lack a nucleus, and hence the exhibition of this circadian variation in both cytoplasm conductivity and membrane conductance cannot be attributed to gene expression, as found in the majority of cells; instead, we showed that the mechanism is primarily ionic, with intracellular potassium being used as part of the timekeeping mechanism. However, the study did not explain why RBCs exhibit such rhythms. In this work, we study the phenomenon and its impact on functional RBC behaviour by examining susceptibility to haemolysis; and expand the application of DEP analysis to examine other types of blood cell, both nucleated and non-nucleated. Our results suggest that the rhythms are conserved across a gamut of blood cell types, and that the presence of this circadian mechanism acts to protect against negative impacts of systemic circadian biology in the body, as demonstrated by variations in cell behaviour. Taken across the different cell types, this work help build a wider appreciation of blood cell electrophysiology as a response to wider aspects of the human circadian rhythm. It has potential implications for health and blood borne diseases.

67

The electrome and the Rosetta stone

Michael P. Hughes m.hughes@surrey.ac.uk m.hughes@surrey.ac.uk

University of Surrey

United Kingdon

When we use dielectrophoresis for cell analysis, we are accustomed to using the multi-shell adaptation of the Clausius-Mossotti factor to determine quantities such as “membrane conductance”, “cytoplasm conductivity” and so on. These parameters, whilst easy to grasp and relate to cellular function, are simplifications of the actual electrical behaviour of different cellular components. The reality is that these phenomena are in fact manifestations of other cellular electrical phenomena, combined in different ways. To understand these interrelationships it is necessary to simultaneously determine as many of these electrical measurands as possible, ideally under multiple conditions, in order to establish the cellular electrome and enable the determination of the relationships of the parameters therein. To this end, we have taken a multi-analytical approach (combining measurement of both membrane and zeta potentials with dielectrophoresis) to study cells in order to better understand the relationships between different electrical characteristics of the cell. We have found that these different electrical characteristics are in fact tightly connected; and that understanding these connections can help us better interpret parameters such as cytoplasm conductivity, membrane conductance and membrane potential as a connected model of ion concentrations and potentials. Furthermore, this interconnection of different electrical parameters (some more commonly understood by the wider electrophysiology community than others) helps to establish a common language between different electrophysiological communities by providing a “Rosetta stone” of translation between, for example, cytoplasm conductivity membrane potential. Understanding how these phenomena interconnect can also give insights into previously unexplored aspects of physical biology in areas as diverse as cancer, developmental biology, and cardiovascular disease.

68

Characterizing the Heterogeneity of Neural Stem Cell Populations Useful for Transplantation

Tayloria N.G. Adams, Shubha Tiwari, Clarissa C. Ro, Andrew R. Yale, Brian J. Cummings, Aileen J. Anderson, and Lisa A. Flanagan

tayloria@uci.edu lisa.flanagan@uci.edu

Department of Chemical and Biomolecular Engineering, Department of Neurology, Department of Anatomy and Neurobiology

University of California Irvine USA

Human embryonic stem (ES) cells provide great opportunities in stem cell therapeutics because they can differentiate into the three germ layers. Human ES cells can be directed toward human neural stem/progenitor cell (HuNSPC) differentiation and used to treat neurological diseases and injuries. Sufficiently characterizing HuNSPC’s functional behavior before using them in transplant therapy is essential to the development of reliable therapeutic treatment options. Currently, the process for cell characterization is challenging due to a lack of biomarkers that provide an adequate picture of HuNSPCs functional profile. Therefore, we have implemented dielectrophoresis (DEP), a label-free cell analysis technique, and flow cytometry to characterize HuNSPCs dielectric properties and surface protein expression, respectively. Two sets of HuNSPCs derived from GMP-grade human ES cells, Shef4 and Shef6. With these cells, we generated three batches of Shef4 (4-1, 4-2, 4-3), a batch of Shef6 sorted (6S), and unsorted (6U) cells. Each batch of cells was screened using a multimodal profile consisting of cell size, cell proliferation, DEP spectra, membrane capacitance, and surface protein (alpha6 integrin, alphaV integrin, and CD24) assessments. Our results show that the 4-1, 4-2, and 4-3 cells have significant changes in size, proliferation, membrane capacitance, and protein expression with increased in vitro age (passage number). Cell size and proliferative capacity were inversely correlated to passage number for Shef4 cells. The membrane capacitance depends on the passage number for the 4-1 and 4-2 cells (gradually decreasing) and fluctuates minimally for the 4-3 cells. For alpha6 integrin, there was a strong correlation to the membrane capacitance of 4-2 cells (R2=0.882), and essentially no correlation to 4-3 cells (R2=0.081). For alphaV integrin, there was a weak correlation to the capacitance of 4-2 cells (R2=0.244), and no correlation for the 4-3 cells (R2=0.082). The 4-1 cells were not included because alpha6 and alphaV integrin expression were low (less than 14%). Additionally, the proliferative capacity and integrin expression of 4-2 and 4-3 cells were directly correlated across passage numbers. We assessed Shef6 cells because they were previously effective in a rodent spinal cord injury model. The Shef6 cells were FACS sorted for CD133+/CD34- markers producing the 6S batch of cells. The 6U batch of cells was not FACS sorted. Our results show that the 6S and 6U cells have significant changes in size, proliferation, membrane capacitance, and protein expression. The 6S cells were smaller and more proliferative than 6U cells. From the DEP screening, 6S cells have higher membrane capacitance than 6U cells. The expression of alpha6 integrin and CD24 was higher for 6S cells than 6U cells. Also, there was no difference in the expression of alphaV integrin for 6S and 6U cells. Significant batch-to-batch variability exists among the Shef4 and Shef6 cells. DEP plus flow cytometry provides a good quantitative assessment of cell phenotype variability. Careful screening to assess cellular heterogeneity will be critical for the development of robust stem cell therapies.

69

Implications of Dielectrophoretic Isolation on Transmembrane Potential-Induced Cell Damage

Walter Varhue, Karina Torres, and Nathan Swami wbv8vt@virginia.edu wbv8vt@virginia.edu

Department of Biomedical Engineering; Department of Electrical Engineering University of Virginia

USA

Label-free dielectrophoretic isolation of biological cells, based on their characteristic subcellular electrophysiological phenotypes, such as membrane capacitance, cytoplasmic conductivity and nucleus to cytoplasmic volume has emerged as a method for enabling subtle distinctions between a variety of cell types. However, in order to ensure minimal damage to the isolated cells, there is a critical need to understand how particular dielectrophoretic conditions can be optimized in microfluidic device geometries to maximize enrichment, while reducing cell damage due to excessive transmembrane potential. Herein, we utilize the juxtaposition of the Schwan equation for induced transmembrane potential and the Clausius-Mossotti factor calculated from a single-shell spherical cell model as a framework to describe the transmembrane potential across the cell during dielectrophoresis. This framework allows us to examine a range of factors that influence electric field-induced viability loss, including the frequency range and polarity of trapping, field profile at the cell focusing position within the microfluidic geometry and the dielectric properties of the cell and suspending media used to create the dielectric contrast for enabling trapping. Through this examination we interpret various dielectrophoretic cell isolation strategies, to assess the role of trapping conditions, device platforms, and cell type on cell survival and viability under dielectrophoresis. Cell damage is most likely in situations of trapping within the frequency region wherein the force response shows inflection towards positive dielectrophoresis, especially within media of high conductivity or low membrane capacitance, wherein this force inflection region is spread over a wider frequency range. On the other hand, trapping under negative dielectrophoresis in the frequency region exhibiting inflection from steady state force levels to the crossover point enhances dielectric distinction between different cell types, while reducing risks from cell damage. However, this is only the case as long as the trapping force balance does not occur at locations within the microfluidic device geometry wherein the field reaches substantially high levels. Finally, trapping in the frequency range wherein positive dielectrophoresis reaches its steady-state level can enable distinctions between various cell-types based on their cytoplasmic conductivity, while reducing risks from cell damage, since utilization of high trapping frequency ensures that the transmembrane potential is minimal. We envision that these considerations can be applied to optimize protocols for dielectrophoretic isolation of large numbers of viable cells.

70

Electrophysiology-based stratification of tumorigenicity and drug sensitivity of pancreatic cancer subpopulations using machine learning approaches

Carlos Honrado, Sara Adair, John Moore, Armita Salahi, John McGrath, Walter Varhue, Vahid Farmehini, Bernadette Goudreau, Sarbajeet Nagdas, Edik Blais, Todd Bauer and Nathan Swami

cmh5nt@virginia.edu nswami@virginia.edu

Electrical & Computer Engineering University of Virginia

USA

Pancreatic ductal adenocarcinoma (PDAC) is an aggressive cancer with a 5-year survival rate of just 8%. Most fatalities are caused by metastases, but there is a lack of specific biomarkers that can be correlated to disease progression and metastasis potential. While a significant majority of PDAC patients (~95%) exhibit KRAS mutations that activate oncogenic proteins, tumorigenicity cannot be assessed solely based on specific mutations. Chemotherapy is often the only treatment option, with ~80% of patients having inoperable disease at presentation. However, chemotherapy drugs exhibit a high degree of patient-to-patient variability in drug sensitivity and cytotoxicity. Since these drugs broadly interfere with cell replication or DNA repair pathways rather than being targeted towards particular cell receptors or proteins, their action cannot be predicted by genetic and transcriptional markers. Given the short timeframe available to PDAC patients (median survival duration of just 3-7 months), there is an urgent need for exploring subcellular phenotypic alterations to assess tumor aggressiveness and tools for in vitro screening of pre-clinical drug targets. Due to the heterogeneity of tumor cells, lack of surface markers to stratify tumorigenicity and variability in drug cytotoxicity, there is much interest in biophysical characterization of single tumor cells to yield phenotypic markers that correlate with cancer onset, progression and drug sensitivity. Cell electrophysiology represents an aggregate of biophysical properties that are influenced by genomic and micro-environmental factors, being affected by whole-cell characteristics (e.g., size or shape) as well as subcellular features (e.g., cell membrane structure or cytoplasmic composition). Herein, using patient-derived PDAC expanded in mice as xenografts, cells with varying tumorigenicity and drug sensitivity are studied by single-cell impedance cytometry for high-throughput electrophysiology measurement (>5000 cells per population) at sub-cellular sensitivity and validated by dielectrophoretic measurements. We seek to distinguish PDAC cells from metastatic versus primary tumor sites (of both KRAS mutant and wild-type genotypes), and their sensitivity to gemcitabine, a common apoptosis-inducing chemotherapy agent. Specifically, we find that highly tumorigenic cell lines obtained from metastatic sites exhibit higher impedance phase versus those obtained from primary sites or those lacking KRAS mutations. Since impedance phase varies as the ratio of permittivity to conductivity, these observations of lowered conductivity of the cell interior were consistent with lowered positive dielectrophoresis levels and upshifted dielectrophoretic crossover frequencies for metastatic versus primary PDAC cell lines and for those with versus without KRAS mutations. Moreover, PDAC cells were treated with gemcitabine at 0.01, 0.1 and 1 µg/mL for 24 or 48 h. Using electrophysiology biometrics for size, membrane capacitance and internal conductivity, unsupervised machine learning clustering methods are used to quantify subpopulations, stratified as drug-resistant, apoptotic and necrotic, based on hallmarks in the electrophysiology data for each cell state. Alterations in the data clusters versus drug treatment conditions track with the drug sensitivity trends, as assessed by proliferation studies for each cell line studied. We envision the implementation of these methods to rapidly inform clinicians about metastatic potential and drug sensitivity of an individual’s PDAC, enabling personalized therapies.

71

Keynote

Micropipette Dielectrophoretic Device for Rapid Purification of Circulating Small Extracellular Vesicles

Leyla Esfandiari leyla.esfandiari@uc.edu leyla.esfandiari@uc.edu

Integrative Biosensing Laboratory, Department of Electrical Engineering and Computer Science and Department of Biomedical Engineering

University of Cincinnati

USA

Early stage detection of cancer is essential for improved long-term survival of patients. Currently, costly, extensively complex and invasive procedures, such as surgical tissue biopsies have been used for cancer screening. Thus, over the past few decades, advancements in microfluidics and lab-on-a-chip approaches have been made to develop minimally invasive and miniaturized platforms to identify and segregate the circulating cancer biomarkers such as small cell secreted extracellular vesicles (exosomes), circulating tumor cells (CTCs) and cell-free DNA (cfDNA) from body fluids. Exosomes, 30-120 nm in diameter, have drawn a great deal of attention due to their high abundance in all body fluids and their enriched and highly stable gene regulatory content. Tumor-derived exosomes have potential use as circulating biomarkers in liquid biopsy for early stage diagnostics and routine clinical monitoring of cancer progression in difficult to access tumor sites. However, because of the complex nature of samples and the heterogeneous physicochemical properties of exosomes, their accurate isolation and characterization from body fluids raises significant challenges in clinical settings. We have recently developed a simple, yet powerful micropipette dielectrophoretic chip capable of rapid and label-free purification of exosomes from body fluids by applying a significantly low electric field. The device also tailored with a sensing module to further characterize exosomes based on their cellular origins. Thus, it has a potential to be evolved as a liquid biopsy technology for early detection of circulating cancer biomarkers.

72

Session 9

Controling the location of an electrokinetically preconcentrated plug of biomolecules Sinwook Park, Barak Sabbagh, Ramadan Abu-Rjal, and Gilad Yossifon

psinwook@technion.ac.il yossifon@technion.ac.il

Faculty of Mechanical Engineering Technion I.I.T

Israel

We present a generic microfluidic-nanochannel platform enabling to perform dynamic and precise control of the locations of multiple preconcentrated plugs of biosamples. The analyte preconcentration is based on the ionic concentration-polarization (CP) phenomenon which is a robust and powerful electrokinetic technique that is applicable to a wide range of biosample size from biomolecules to cells. [1 ,2] However, the main deficiency of CP based preconcentration is the inability to precisely and dynamically control the location of the preconcentrated biosamples to overlap a desired detection area as the stable equilibrium position at which the biosample preconcentrate, resulting from exact cancellation of the electromigrative and counteracting convective forces, is highly sensitive to the system operating parameters (e.g. voltage, flow rate, geometry, surface charge etc.). One approach to achieve direct control over the length of the depletion layer, which in turn controls the location of the preconcentrated plug, uses either embedded electrodes or heaters for local stirring of the fluid, via either alternating-current-electro-osmosis (ACEO)[3] or electro-thermal (ET) flow[4], correspondingly. Here we focused on an alternative approach to control of the location of the preconcentrated plug using array of tunable ion permselective medium [5]. By activating individually addressable ion-permselective medium on demand, we can control the length of the diffusion layer, which in turn controls the location of the preconcentrated plug. This technology opens future opportunities for such smart and sophisticated manipulations of preconcentrated biological molecules plugs for various on-chip biological applications. [1] J.H. Lee, J. Han, Microfluid. Nanofluidics. 9 (2010) 973–979. [2] S.H. Ko et al. Lab Chip. 12 (2012) 4472–4482. [3] S. Park, G. Yossifon, I Phys. Rev. E. 93 (2016) 062614. [4] S. Park, G. Yossifon, Anal. Chem. 92 (2020) 2476–2482. [5] B. Sabbagh, E. Stolovicki, S. Park, D. A. Weitz, and G. Yossifon, Nano. Lett. 20, 8524-8533 (2020).

73

From Lab-on-a-chip to Lab-on-a-particle electrically-powered platforms Gilad Yossifon

yossifon@technion.ac.il yossifon@technion.ac.il

Micro and Nano-Fluidics Laboratory Technion - Israel Institute of Technology

Israel

Towards lab-on-a-particle platforms we suggest using mobile engineered active (“self-propelling”) carriers to revolutionize diagnostic testing and sample analysis; with advantages of the traditional lab-on-a-chip (e.g. portability, efficiency) but overcoming current challenges (e.g. complexity, predetermined design). Our novel generic active carrier, acting as a mobile floating microelectrode, uses a single externally applied electric/optical field to selectively trap, transport and deliver user-specified payload(s). Our unified solution is simpler and more robust than current systems where carrier propulsion and cargo manipulation are controlled by separate mechanisms. Moreover, current cargo loading requires specific and predefined targets and release of the cargo (if possible) is complicated. Our recent work [1] demonstrated that using dielectrophoresis (DEP), a frequency-dependent mechanism can selectively load and release the transported cargo. This offers a label-free method to generically, selectively and dynamically manipulate (load and release) a broad range of organic/inorganic matter [2]. Adding directed motion via magnetic stirring enables to develop these active particles into in-vitro assays with single cell precision and building blocks for bottom-up fabrication. Besides the local electric field gradient intensification essential for DEP, an important novelty of our mobile microelectrodes is also the strong local electric field intensification induced by the inherent small gap between the metallic patches of the active particle and the conductive substrate underneath. This property was recently exploited by us [3] to demonstrate a novel method of local and targeted (i.e. only those cells that are in contact with the active particle) electroporation of bacteria as well as micromotor based biosensing. [1] Boymelgreen, T. Balli, T. Miloh and G. Yossifon, Mobile Microelectrodes: Unified Label-Free Selective Cargo Transport by Active Colloids, Nature Communications 9:760 (2018). [2] X. Huo, Y. Wu, A. Boymelgreen and G. Yossifon, Analysis of cargo loading modes and capacity of an electrically-powered active carrier, Langmuir. [3] Y. Wu, A. Fu, and G. Yossifon, Active Particles as Mobile Microelectrodes for Selective Bacteria Electroporation and Transport, Science Advances (2020). [4] S. Park and G. Yossifon, Micromotor-Based Biosensing Using Label-Free and Directed Transport of Functionalized Beads, ACS Sensors (accepted).

74

Why 3D Electrodes for Dielectrophoresis? A critical review on the fabrication techniques that can enable higher throughput in DEP devices

Rodrigo Martinez-Duarte rodrigm@clemson.edu rodrigm@clemson.edu

Multiscale Manufacturing Laboratory, Department of Mechanical Engineering Clemson University

USA

The identification and sorting of targeted cells and microorganisms in a sample is a cornerstone of healthcare diagnostics and therapeutics. This presentation will focus on the use of Dielectrophoresis (DEP) for the selective sorting of targeted bioparticles in a sample and how the lack of throughput has been one important practical challenge to the widespread practical implementation of DEP to this end. Increasing the cross-sectional area A of a channel can lead to higher flow rates Q and thus the capability to process a larger sample volume per unit of time. However, the electric field gradient that is generated by polarized electrodes, and that is required for DEP, drastically decreases as one moves away from the electrodes. Hence, the scaling up of the channel cross section must be done asymmetrically; with one dimension preferably below 100 um and the other as large as possible to maximize A. One desires a channel aspect ratio AR=height/width that is much smaller or much larger than 1. Due to the constraints of current techniques that allow for fabrication of critical dimensions in the order of tens of micrometers, the real estate or surface area of a substrate is important in determining the price of the device. Thus, reducing footprint of the DEP device is important to ensure affordability, and the use of channels with AR>>1 is desired to increase the cross-section area of the channel while maintaining an optimized footprint. Furthermore, parallelization is needed to meet the time constraints, few minutes, of processing large sample volumes, > 1 ml, that are common for diagnostics and therapeutics. This creates the challenge to fabricate electrodes on the sidewalls of multiple channels with AR>>1, or a channel embedding an array of electrodes with a gap in between them with AR >>1. This proposed presentation will first detail the motivation for using three-dimensional (3D) DEP devices to improve throughput and then provide a critical review of selected techniques that have been used to fabricate them. The goal is to provide the attendee with a comparison of the advantages and limitations of selected fabrication techniques that would let them fabricate 3D DEP devices that best fit their envisioned application. Techniques include electrodeposition, doped Silicon, carbonization, coating of microstructures, and co-fabrication. Electrode materials to be addressed include metals, silicon, carbon, PDMS-based composites as well as conductive polymers and fluids.

75

Nonlinear electrokinetics for assessment of bioparticles in microdevices Blanca H. Lapizco-Encinas

bhlbme@rit.edu bhlbme@rit.edu

Microscale Bioseparations Laboratory and Biomedical Engineering Department Rochester Institute of Technology

USA

Electrokinetics (EK) is the movement of particles and fluids under the action of electrical fields. In microfluidic devices, an applied electric potential can be used for both, manipulating the liquid and the bioparticles within a microdevice. Since all bioparticles possess electrical charge to a distinct degree, electrophoresis (EP) effect are always present when assessing bioparticles. The use of electroosmotic (EO) flow can allow for “on the fly” dynamic re-direction of the liquid and particles to specific reservoirs within a microdevice for further collection and analysis. The super-position of EP and EO is often denoted as linear EK. Nonlinear EK phenomena such Dielectrophoresis (DEP) and electrophoresis (EP) of the second kind have been successfully employed for bio-particle manipulation. One key characteristic of these effects is their ability to work across scales, since both nano and micron-sized bioparticles can be manipulated with non-linear EK forces. A common approach employed to generate non-uniform electric fields and produce non-linear EK effects is the use of simple microchannel with embedded insulating structures, commonly known as “insulating posts.” Our group has reported a series of studies focused on the geometry and arrangement of insulating posts to enhance DEP effects on particles. We have also investigated extensively the use of DC potentials and low-frequency DC-biased AC potentials, as well as cyclical potentials. This presentation is focused on the design of microdevices that combine both linear and nonlinear EK effects for effective bioparticle manipulation: sorting, enrichment and separation. Special attention will be given to the new parameter of electrokinetic equilibrium condition, which includes the effects on electrophoresis of the second kind on particle migration. This work includes extensive experimentation with microdevices made from PDMS and mathematical modeling with COMSOL Multiphysics. The results obtained from several projects in our laboratory for the manipulation of bioparticles ranging from protein nanoparticles to eukaryotic cells will be presented, underlining the diverse schemes utilized in each type of application. Acknowledgements: The authors would like to acknowledge the financial support provided by the National Science Foundation (Award CBET-1705895).

76

Elucidating cell phenotype based on biophysical analysis of secreted subcellular bodies Carlos Honrado, John Moore, Walter Varhue, and Nathan Swami

nswami@virginia.edu nswami@virginia.edu

Electrical Engineering / Swami Group University of Virginia

USA

Extracellular secretions of subcellular bodies, such as vesicles, exosomes, apoptotic bodies or germinating bacterial spores that are obtained from cell cultures can offer selective information on the phenotypic transformations of parent cell, without the need to disturb the cell culture. However, due to the compositional diversity of subcellular bodies, their stratification for enabling correlations to the function, health and disassembly of cells under culture is challenging. Currently, phenotypic analysis requires cells to be lifted from culture for staining and flow cytometry quantification, while analogous methods with subcellular bodies requires the development of tailored labeling strategies for identification and downstream quantification. This is especially difficult to conduct for in vitro disease modeling using patient-derived tumor cells or ex vivo microbiota samples, which have fewer cell numbers and phenotypic transformations with unknown markers for staining. We present device and data analytics strategies for quantifying alterations to the inherent biophysical properties of subcellular bodies by impedance cytometry and dielectrophoretic isolation for the purpose of elucidating functional alterations in the cultured cells. In the first example, we compare exosomes obtained from cultures of pancreatic tumor cell lines of varying invasiveness to study the correlation of electrical conductance of the respective exosomes to the metastatic potential of the parent cell. Specifically, the lipid composition and fluidity of the exosomes that prime the tumor microenvironment for metastaticity is correlated to its electrical conductance and is modulated for specificity to exosome type by metal nanoparticle binding to surface antigen sites for cytometry and selective dielectrophoretic isolation. In the second example, we study the susceptibility of gut microbiota to Clostridium difficile (C. difficile) infection by exploring how the loss in diversity of gut microbiota affects the balance of secreted metabolites that regulate germination of C. difficile spores. Specifically, the high frequency impedance phase spectra can distinguish spore aggregates from germinated bacterial cells during ex vivo culture, thereby allowing us to monitor antibiotic-induced microbiota disruption in a mouse model for monitoring susceptibility to C. difficile infection. In the third example, subcellular apoptotic bodies secreted in the media of pancreatic tumor cultures under drug treatment are used to indicate drug sensitivity, without the need to lift cells under culture. Based on the phenotypic similarity of apoptotic bodies in the media of the drug-treated culture to lifted apoptotic cells, dielectric properties determined by high-throughput single-particle impedance cytometry can detect the temporal drug response of the culture by distinguishing the hallmarks of progressive apoptosis; i.e. apoptotic microvesicles secreted by intact cells, beaded apoptopodia caused by membrane transformations and larger subcellular bodies due to progressive cell disassembly. We envision application of this tool in conjunction with the appropriate in vitro model to assess phenotypic transformations of cultures. Acknowledgements: This work was supported by NIH Grants: 1R21AI130902-01, R01 CA200755 and U.S. National Center for Advancing Translational Sciences (NIH UL1TR003015)

77

Poster Session 3 High-Resolution 3D-Printed Insulator-Based Dielectrophoresis Devices towards

Manipulation of Bioanalytes Mohammad Towshif Rabbani, Mukul Sonker, Jorvani Cruz Villarreal, and Alexandra Ros

mrabban2@asu.edu Alexandra.Ros@asu.edu

School of Molecular Sciences/Ros Lab Arizona State University

USA

Microfluidics has enabled a wide range of biological and biochemical applications such as high-throughput drug testing or point of care diagnostics, to name a few. The introduction of insulator-based dielectrophoresis (iDEP) has provided a new dimension for the precise manipulation of bioparticles and biomolecules. DEP describes the phenomenon of a force experienced by a dielectric particle when it is subjected to a non-uniform electric field. The DEP force depends on the size and dielectric properties of the particles and the suspending medium. Despite the advancements in iDEP, it has been hampered due to often cumbersome and expensive fabrication methods. Recently, 3D-printing has been successfully implemented in microfluidics due to several advantages over conventional fabrication processes that require sophisticated instruments with cleanroom facilities. 3D-printing offers rapid prototyping, a high level of reproducibility, and truly 3-dimensional geometries that cannot be realized with conventional fabrication techniques. There are several approaches to 3D-printing, and among those, the two-photon polymerization (2-pp) process offers unique capabilities with unprecedented resolution compared to a standard polymer 3D-printing technology such as stereolithography. For various iDEP applications, higher electric field gradients are required, and 2pp 3D-printing can be employed to create nanometer-resolution constrictions and gaps to achieve orders of magnitudes higher electric fields. Here, we report the first iDEP-based manipulation of biomolecules, namely λ-DNA and phycocyanin, within high resolution 3D-printed microfluidic devices fabricated using 2-pp technology. iDEP microfluidic devices with different post geometries were successfully printed and developed with a gap resolution between posts down to 1 µm. Experimentally, 0.87 µm polystyrene beads were successfully trapped according to negative DEP (nDEP) in the insulating post array at 10 kHz, and 200 V applied over 0.5 cm. For λ-DNA and phycocyanin, positive DEP (pDEP)-based trapping was observed at 500 Hz with 350 V applied over the same channel length and at 100 Hz with 300 V, respectively. The nDEP trapping positions were also confirmed with a numerical model based on Comsol’s tracking module with the real part of the Clausius-Mossotti of -0.5 for polystyrene beads previously reported. In contrast, a Multiphysics module was developed, allowing to employ the polarizability, α, of biomolecules to account for their DEP force. For λ-DNA, an α value of 〖3.3*10〗^(-29) 〖Fm〗^2 was used according to previous reports,1 resulting in pDEP trapping. Both numerical models were in excellent agreement with experimentally observed trapping conditions for the two biomolecules and beads in the post array. Furthermore, sub-micron resolution was achieved using IP-Dip photoresist, where a resolution down to 800 nm was confirmed with scanning electron microscopy. Additionally, the 2-pp device material was also optimized with pentaerythritol triacrylate polymer and a photo-initiator (phenylbis(2,4,6-trimethylbenzoyl) phosphine oxide), reducing background fluorescence for improved detection. Our study provides insight on a novel approach of high-resolution 3D-printed microfluidic devices which offers the potential for prototyping novel iDEP microdevices and with a scope for further improvement in resolution down to few hundred nanometers opening the opportunity to explore iDEP of many nm-scale biomolecules in the future. 1. Regtmeier et al., Anal. Chem. 2007, 79 (10), 3925-3932.

78

A Tunable Insulator-Based Dielectrophoresis System for the Separation of Biomolecules Ricardo Ortiz, Mukul Sonker, Alexandra Ros

rortiz15@asu.edu Alexandra.Ros@asu.edu

School of Molecular Sciences Center for Applied Structural Discovery

USA

Insulator-based dielectrophoresis (iDEP) has been exploited to manipulate particles, organelles, and nucleic acids among other biomolecules. Typical iDEP devices integrate insulating post arrays with different geometries to induce a nonuniform electric field. The dielectrophoretic (DEP) force scales with the magnitude of the squared electric field gradient and the cubed radius of a particle. The DEP force can be maximized by increasing the electric potentials. To maximize the DEP force without increasing the electric potentials, we developed a microfluidic device that integrates a normally-open valve that serves as a dynamic insulating constriction on application of electric potential while the valve is deflected into the fluidic channel by a positive pressure.1 In this device, a two-layer platform that integrates a thin membrane made out of polydimethylsiloxane was integrated by soft lithography. During the application of a pneumatic pressure, the thin membrane was pushed down into the channel generating a constriction of varying distances and a nonuniform electric field. Here, we further optimize the valve by designing a normally-closed valve to fine-tune the electric field and valve actuation. A testing device was composed of a series of stem valves ranging from 10 to 100 µm in width that act as pneumatic valves and dynamic insulating constrictions. Experimental results showed that the pressure required for a valve stem to actuate into the control chamber while a vacuum is applied increases with the aspect ratio. The aspect ratio is defined as the length of the vacuum patch (lp) times the width of the vacuum patch (wp) over the length of the stem valve patch (lp) times the width of the stem valve (wv). Furthermore, a numerical study was developed to understand the relationship between the fine-tuning of the electric field gradient and the valve stem to substrate distance, d. A 2D cross-section of an actuated stem valve was designed (wv=10 µm width and stem valve length of 20 µm) with d ranging from 10 µm to 100 nm. The electric field gradient values ranged from 7.68 x 1016 to 1.52 x 1019 V2/m3 when adjusting d from 10 µm to 100 nm for 1000 V/cm applied across the channel embedding the valve. The gradient values for 1µm (1.06 x 1018 V2/m3) reached the same order of magnitude (6.9 x 1018 V2/m3) as previously reported for constrictions in post arrays (d=1 µm), where iDEP devices were used for trapping DNA.2 As a proof of principle, polystyrene beads were used as a model analyte to demonstrate the application of the dynamic constriction. We further suggest that serially aligning multiple dynamic constrictions will allow the design of a separation system for heterogeneous analyte mixtures based on iDEP. 1. Ros, A.; Kim, D.; Luo, Jinghui, Yang, M. Tunable Insulator-Based Dielectrophoresis (iDEP) with Membrane Valves. Patent Pending 2019 / 0224689 A1. July 25, 2017. 2. Camacho-Alanis, F.; Gan, L.; Ros, A. Transitioning Streaming to Trapping in DC Insulator-Based Dielectrophoresis for Biomolecules. Sensors Actuators, B Chem. 2012, 173, 668–675.

79

Cylindrical and teardrop shaped active posts in deterministic lateral displacement devices Widad Lahbichi, Gloria Porro, Kevin Keim, Jason P. Beech, Jonas O. Tegenfeldt, Carlotta Guiducci

widad.lahbichi@epfl.ch carlotta.guiducci@epfl.ch

Laboratory of Life Sciences Electronics (CLSE) Ecole Polytechnique Federale de Lausanne (EPFL)

Switzerland

Deterministic lateral displacement (DLD) is a microfluidic technology to sort particles by size in a continuous flow (L.R.Huang et al., Science 2004). Using connected metal coated posts in a DLD device as electrodes and applying electric fields can decrease the critical size for separation allowing sorting of nanoscale particles in micro-scaled posts arrays (Beech et al., Adv. Mater. Technol. 2019). The device consists out of 3D electrodes. These free-standing metal covered SU-8 electrodes are connected by a metal layer from underneath (Kilchenmann et al., JMEMS 2016). This allows the application of an AC signal, thus generating an electrical field gradient between the 3D electrodes and exerting a dielectrophoretic (DEP) force on the polarizable particles. In our case, the negative DEP force pushes the particles from one trajectory to another causing them to displace (travelling along the posts direction) inside the device. This force depends on the particle size, the polarizability of the particle compared to the medium, on the frequency of the signal applied, as well as on the electric field gradient. In order to maximize the electric field gradient, our group introduced a new geometry of the electrodes (teardrop shape) (K.Keim, PhD Thesis 2020). This new geometry creates a localized high electric field gradient reducing the sorting size while keeping comparable device dimensions. With our new design, we are able to generate similar field gradients as with cylindrical electrodes using lower applied voltages due to the teardrop shaped geometry. A comparison between the cylindrical and the teardrop shaped 3D electrodes is performed by flowing latex fluorescent beads of different sizes through the devices and observing the traveling mode at different electric field amplitudes. The critical diameter of both layouts is 6 µm and decreases with the applied electric field. When applying a 20 V (peak to peak) electric field, the particles of a 100 nm diameter get displaced in the devices with teardrop shaped 3D electrodes, while they remain in the zig-zag mode (straight through the device) in the devices with cylindrical shaped 3D electrodes. Our experiments show that teardrop shaped 3D electrodes are more effective than their round equivalent. The usage of the teardrop shaped 3D electrodes enables to sort smaller particles without reducing the device dimensions. Additionally, this will enable to sort nano-sized particles using micro-scaled DLD devices preventing the clogging and improving the throughput.

80

Deterministic iDEP Ratchet Devices for High-throughput Organelle Separation Domin Koh, Mukul Sonker, Ricardo Ortiz, Daihyun Kim, Alexandra Ros

dkoh9@asu.edu Alexandra.Ros@asu.edu

School of Molecular Sciences, Center for Applied Structural Discovery, The Biodesign Institute Arizona State University

USA

Organelle size and morphological characteristics are vital for their cellular functions and changes to these traits lead to cellular dysfunctions that are often associated with diseases like Alzheimer’s disease, obesity, diabetes, cancer etc. Fractionation of such subcellular organelle population can be of great importance to investigate these cellular dysfunctions. Thus, a size-selective separation tool may offer high potential for studying these diseases at the biomolecular level and development of innovative therapeutic interventions. While many separation methods have been developed for fractionation of organelles, it often requires cumbersome sample preparation and extraction steps leading to sample loss. Thus, a simple, predictable, and continuous high-throughput (HT) fractionation tool is required. Recently, dielectrophoresis (DEP) has gained immense interest as a bioparticle manipulation tool due to its predictive and non-destructive nature. Previously, we demonstrated a novel DEP-based deterministic ratchet migration phenomenon for size-based separation of polystyrene beads, mouse-liver mitochondria, and liposomes with baseline resolution.1 This was realized in an insulator-based dielectrophoretic (iDEP) microfluidic device containing an array of insulating posts to successfully fractionate bioparticles, exploiting size-heterogeneity. Here, we further develop these devices for large-scale HT organelle fractionation. Novel device designs and geometries were explored to provide better control over sample manipulation and capability for continuous as well as batch injections. These large-scale device designs have been designed to process large sample volumes >5 µL and up to 106 organelles per run so that a sufficient amount of purified organelles can be recovered for further characterization and biomolecular assessment. We are also optimizing the previously characterized electrical driving parameters in-silico using numerical modeling and experimentally, as required for large-scale devices. The numerical model predicted a baseline resolution for different size of mitochondria in ~100 s in a continuous HT iDEP device design. For better demonstration of size-based organelle fractionation, we are using mitochondria with knockdown of either Mfn-1 or Drp-1 genes that cause size differentiation of mitochondria. Additionally, to achieve the faster and optimized data analysis, we installed Optosplit that splits the wavelength of light and enables recording of florescent particles of different wavelength simultaneously, consequently, we could record the mitochondria of different size simultaneously that were stained with different florescent dyes. Also, for faster analysis, we utilized a ImageJ automatic particle tracking plugin, Mosaic particle tracker, to collect mitochondria tracing data automatically, which then, analyzed via MATLAB code. The MATLAB code contains various filters to enhance the quality of the analyzed data and the result shows good overlap with the manually analyzed data. Furthermore, various device designs are being explored to enable contactless-DEP, DEP in contact with liquid metal electrodes, and efficient sample recovery after HT organelle separation. In the future, we plan to further optimize the parameters required for ratchet-based separation and to gain knowledge for future fractionation of diseased mitochondria samples and other organelles. 1. Kim, D.; Luo, J.; Arriaga, E. A.; Ros, A., Deterministic Ratchet for Sub-micrometer (Bio)particle Separation. Analytical Chemistry 2018, 90 (7), 4370-4379.

81

Separation of diploid and tetraploid cancer cell populations using high-frequency dielectrophoresis

Josie L. Duncan, Mathew Bloomfield, Daniela Cimini, Rafael V. Davalos josied@vt.edu

davalos@vt.edu Bioelectromechanical Systems Laboratory, Department of Mechanical Engineering

Virginia Tech USA

Aneuploidy, or the presence of incorrect chromosome number, is ubiquitous in most cancers. However, aneuploidy has detrimental effects on cell fitness and can promote or inhibit tumor formation, making its role in cancer unclear. Recent evidence indicates that tetraploidy, a result of whole genome duplication without cell division, may lead to cancer by buffering against the fitness costs of aneuploidy. Therefore, isolating diploid and tetraploid cells from mixed cancer cell populations is useful for understanding how these cancer cell subpopulations contribute to tumorigenesis and respond to therapeutic treatments or other perturbations. Here we used high-frequency dielectrophoresis (DEP) to detect and isolate diploid and tetraploid human cancer cells based on their concentration of nuclear charge. We conducted experiments using DLD1 colorectal cancer cells (DLD1-2N) with a pseudodiploid karyotype and two tetraploid clones, DLD1-C1 and DLD1-C2 with similar nuclear to cytoplasm (N/C) ratios as DLD1, but substantially different cell and nuclear sizes. Specifically, DLD1-C2 cells were only slightly larger, whereas DLD1-C1 cells were twice as large compared to the DLD1-2N. The mixed sample was suspended in optimized low-conductivity buffer. The device consists of gold, planar, interdigitated electrodes powered by a function generator (10-100MHz and 6Vpp) to induce a gradient to manipulate a mixed population of diploid and tetraploid cells under flow conditions in order to balance dielectrophoretic forces and drag forces, optimizing separation. Preliminary results indicate there is significant separation between DLD1-2N and DLD1-C2 at 50MHz, 60MHz, and 80MHz, with no significant separation between DLD1-2N and DLD1-C1. The reversible trapping efficiency, or the percentage of cells released after turning off the electric field, of the two populations differs by upwards of 13% at these frequencies. Given that DLD1-C2 contained the same amount of DNA as DLD1-C1, but in a nucleus that is about half the size, our results suggest that high-frequency DEP has the resolution to separate populations based largely on concentrations of DNA content within the nucleus. The lack of significant separation of DLD1-2N and DLD1-C1 indicates an enlarged nuclear and cytoplasmic membrane holding more DNA content is negligible and rather separation is detected when there is a higher concentration of charge within the nucleus. The separation between DLD1-C2 and DLD1-C1, at the same frequencies, similarly suggests that overall chromosome number is not being detected; rather, chromosome concentration is altering the bioelectric signature of these cells. These results also indicate low trapping efficiencies at 90MHz and 100MHz (<5%) which indicate that the drag force is beginning to overcome that of the DEP force, limiting trapping capabilities. This study introduces a novel application of dielectrophoresis for manipulating and isolating viable cell populations of differing DNA concentration for subsequent study of tumor progression. This study also suggests that DNA concentration alters the bioelectric properties of the nucleus and thus, can be detected using dielectrophoresis.

82

A microfluidic device for characterizing variability of beta-galactosidase in single MDA-MB-231 cells

Joseph T. Banovetz, Min Li, Darshna Pagariya, Robbyn K. Anand joebano@iastate.edu rkanand@iastate.edu

Chemistry Iowa State University

USA

Expression of the enzyme beta-galactosidase can be used as a marker for cancer cell senescence and thus drug effectiveness. MDA-MB-231 is an invasive breast cancer cell line which exhibits a high degree of phenotypic heterogeneity. Most cellular assays focus on aggregate statistics from large numbers of cells. Because cancer cells are heterogeneous with respect to enzyme activity, the development of treatments is difficult because bulk statistics mask the variability in the enzymatic activity of individual cells. In order to obtain a more accurate picture of enzyme activity, a method to measure the distribution of single-cell activities is needed. We have developed a microfluidic device for analyzing beta-galactosidase expression in a relatively large number of MDA-MB-231 cells on a single-cell basis. The device uses an array of bipolar (floating) electrodes to capture cells by dielectrophoresis (DEP). DEP offers a label free method of manipulating cells with minimal impact on cellular function. Importantly, the fluorogenic beta-galactosidase assay composition was optimized for high activity in a low conductivity environment suitable for DEP. The captured cells are fluidically transferred and recaptured in reaction chambers and the fluorogenic assay buffer is added. An ionic liquid is used to isolate the cells and prevent cross-talk between chambers. Finally, a high AC voltage is used to lyse the cells and the enzyme activity is measured by fluorescence microscopy. Small scale implementation shows successful beta-galactosidase assay results from captured and isolated MDA cells. The device developed can be used to characterize the variability of beta-galactosidase activity in MDA-MB-231 cells. Characterization of this enzymatic activity will facilitate studies of MDA cellular senescence following drug challenges and other stressors. Furthermore, the device is adaptable to other types of fluorometric assay such as other enzymes or nucleic acid amplification. The authors gratefully acknowledge funding from the NIH Career Award.

83

DC g-iDEP Trapping of Gold Nanoparticles Alex Ramirez, Mark Hayes

ajrami11@asu.edu mhayes@asu.edu Mark Hayes Group

Arizona State University USA

Dielectrophoretic trapping is a separatory/analytical method that is capable of achieving high levels of analyte differentiation using a combination of electroosmotic flow, electrophoresis, and dielectrophoresis. The form of dielectrophoretic device used in these trials was of a gradient insulator-based design that induced the non-uniform electric fields necessary for dielectrophoretic trapping to occur. Development of such microfluidic devices began in the early 2000s and has produced several successful trials and refinements since then. Improvements have led to the ability of these devices to separate analytes to extremely high degrees of resolution as was demonstrated by the simultaneous separation of antibiotic resistant and antibiotic susceptible strains of bacteria in other experiments. The majority of analytes examined with these microfluidic devices have been biological in nature and on the scale of micrometers in size. The objective of this experiment was to test the lower limit of the device’s resolution by attempting to use dielectrophoresis to trap gold nanoparticles via the balancing point between electrophoretic and dielectrophoretic mobilities. Trials successfully captured 10 nm fluorophore-tagged gold nanoparticles at a mobility ratio of 6.16 x 1011 V2/m3, 60 nm citrate-capped gold nanoparticles at approximately 3.61 x 1010 V2/m3, and bare 10 nm gold nanoparticle aggregates at both 1.63 x 1010 V2/m3 and 1.68 x 1010 V2/m3. The corresponding voltages that were applied to achieve trapping were -1500 V, -2000 V, and -1500 V respectively. These findings were promising but reproducibility of the results was very low, largely due to matters of contaminants entering the devices and preventing the even, continuous flow of the analyte solution. Refinement of the analytical process should be pursued.

84

Towards Separating Microplastic Particles with Insulator-based Dielectrophoresis Shulin Bu, Mohammad Towshif Rabbani, Alexandra Ros

shulinbu@asu.edu Alexandra.Ros@asu.edu

School of Molecular Science, Center for Applied Structural Discovery, The Biodesign Institute Arizona State University

USA

The presence of microplastics has become an emerging threat to terrestrial and marine systems as they contaminate oceans, lakes and soils which impacts biodiversity and the ecosystem. According to the literature, from 1950 to 2004, the global release of microplastics has significantly increased from 1.7 million tons to 299 million tons [1]. Although there is no agreed definition, microplastics are generally referred to small ubiquitous pieces of plastic particles less than 5-millimeter in dimension. They mainly originate from two sources: those categorized as primary microplastics are manufactured in small size purposely for particular applications such as facial scrubs, and those categorized as secondary microplastics are formed from large plastic pieces broken down into smaller fragments by mechanical abrasion or UV radiation. Several types of microplastics are more likely to be found in the environment including polyethylene (PE), polypropylene (PP), polystyrene (PS) and polycarbonate (PC) [1]. Recently, researchers have revealed that microplastics are found in animals and potentially humans. Here, we propose to study the dielectrophoretic properties of microplastics in body fluids to develop detection and analysis tools that will help in studying microplastics found in living systems. Dielectrophoresis (DEP) is the movement of polarizable particles in a non-uniform electric field resultant from spatial electric field gradients and the interaction between dipoles. DEP possesses several advantages such as low cost and simple operating processes, it is time-saving, and can easily be integrated into devices [2]. Our project aims to fractionate microplastic particles from blood serum and other body fluids. As microplastic particles tend to attract proteins, cells or other entities that float in blood, bovine serum albumin (BSA) is introduced and attached onto the particles in order to mimic coated particles. BSA is known to easily adsorb to both positively charged and negatively charged particles as it consists of amino acids that can be both positively charged (lysine, histidine) and negatively charged (glutamic acid, aspartic acid) [3]. Microplastic samples with and without BSA coating were first characterized using dynamic light scattering (DLS). A decrease in the Zeta potential was observed upon protein coating. While polystyrene beads have recently shown negative DEP behavior [4], the differences in Zeta potential also suggest a differing DEP response [5]. We expect that future DEP studies on coated and non-coated microplastics will shed light onto the variations in DEP behavior upon environmental conditions as well as microplastics in body fluids. The insight into the DEP behavior of microplastics has the potential to serve as a highly valuable analytical tool related to pre-concentration, analysis and separation of microplastics. [1] Auta, H et al. (2017). Environment International, 102, 165-176. [2] Wang, Y et al. (2018). Electrophoresis, 40, 969-978. [3] Phan, H et al. (2015). PLOS ONE, 10, 0141282. [4] Luo et al., Anal. Chem. 2016, 88 (11), 5920-5927. [5] Rabbani, M et al. (2017). Analytical Chemistry, 89(24), pp.13235-13244.

85

A Numerical Investigation to Extend Quantitation of Gradient-induced Forces within an Insulator-based Sawtooth Design

A K M Fazlul Karim Rasel, Sean L Seyler, and Mark A Hayes arasel@asu.edu

MHayes@asu.edu Hayes Lab, School of Molecular Science

Arizona State University USA

Gradient insulator dielectrophoresis using a sawtooth design is an approach that has been used to quantitate the gradient-related forces, which includes dielectrophoresis (DEP), using known counter forces and zero-velocity (i.e., “capture”) zone formation. There is an opportunity to extend this approach to the determination of these forces to regions of the sawtooth design where full capture events do not occur. This is accomplished here by quantifying the deviations of particle trajectories caused by gradient-related forces in order to deduce the total magnitude of those forces. These measurements are especially important for particles smaller than one micron in diameter as there is emerging experimental evidence supported by evolving theory indicating the presence of much larger gradient-related forces than the classical Clausius-Mossotti factor (K_CM) predicts. On theoretical grounds, the total effective polarizability K of a particle-solvent system is generally expected to be larger than K_CM, the latter of which does not account for permanent dipoles nor the particle-solvent interfacial polarizability. Starting with a general expression for all gradient-related forces in our system as F=1/2 Kv∇〖|E|〗^2, where v and E are the volume and electric field, respectively, we extract empirical bounds on K from experimental data by leveraging numerical simulations of the detailed electrostatics, microfluidics, and particle dynamics in the sawtooth channel geometry. There are several ways to extract K values utilizing using our combined experimental-computational approach; one strategy, which we take here, is to first compute the linear forces induced by the electric field and then quantify the deviation of particle trajectories induced by the non-linear forces. Particle tracking is used to extract trajectories from experimental data, while the value of K is varied in simulations until the similarity between simulated and real trajectories crosses a quantifiable threshold. Using a test data set, we establish the viability of this approach for obtaining estimates for K values of arbitrary target analytes. This investigation can help more understand the motion of bioparticles such as proteins, exosomes, and virions subjected to linear and non-linear forces inside a sawtooth designed microfluidic device.

86

Please keep updated at

www.dep2020.org