Interferometric Electrohydrodynamic Tweezers (IET)

• IET are an integrated optofluidic platform that traps, images, and spectroscopically characterizes nanoscale objects in solution.
• The platform employs AC electro-osmotic flows to concentrate nanoparticles in under three seconds, enabling precise size and shape determination through interferometric imaging.
• Integrated Raman spectroscopy provides chemical fingerprinting while advanced surface passivation ensures reversible trapping and minimal fouling.

Interferometric Electrohydrodynamic Tweezers (IET) are an integrated optofluidic platform for trapping and characterizing single nanoscale objects in solution by combining AC electro-osmotic confinement with interferometric imaging and, in the reported implementation, Raman spectroscopy. The platform was introduced as a method that rapidly traps single nanoparticles in parallel within three seconds, enables label-free characterization of particle size and shape via interferometric imaging, and identifies molecular composition through Raman spectroscopy, all without fluorescent labeling (Hong et al., 2024). Subsequent work on the gold-based IET device focused on surface passivation chemistries required for reversible trapping and release, especially for negatively charged polystyrene nanoparticles and extracellular vesicles, and established antifouling design rules for reusable nanotweezer interfaces (Ugwu et al., 18 Jun 2026).

1. Definition and reported scope

IET denotes a trapping architecture in which an alternating electric field is applied across a patterned gold electrode and an indium–tin oxide counter-electrode to generate AC electro-osmotic flows that concentrate nanoscale objects into discrete trapping sites. In the reported studies, the platform was used for colloidal polymer beads, nanoscale extracellular vesicles (EVs), and extracellular nanoparticles known as supermeres, with label-free optical readout and optional Raman-based molecular fingerprinting at single-particle sites (Hong et al., 2024).

The reported scope of IET is multiparameter characterization rather than trapping alone. The 2024 study describes a single integrated workflow in which nanoparticles are trapped, imaged interferometrically to extract size and shape information, and then interrogated spectroscopically for chemical composition (Hong et al., 2024). The 2026 passivation study extends the platform at the interface level by examining how surface chemistry controls reversible operation, quantitative reliability, and reusability on a previously reported gold-based IET device (Ugwu et al., 18 Jun 2026).

A common misunderstanding is to treat IET as a purely optical tweezer. In the reported operating principle, confinement is attributed to electrohydrodynamic transport and surface interaction forces, while interferometric scattering supplies the principal label-free imaging modality; optical fields serve readout and spectroscopic functions rather than being described as the trapping mechanism itself (Hong et al., 2024).

2. Electrohydrodynamic confinement mechanism

At the core of IET is AC electro-osmosis (ACEO) generated by a sinusoidal voltage applied between a gold electrode and an ITO counter-electrode. When a voltage of the form $V_0 \sin(\omega t)$ is applied, normal and tangential electric-field components develop at the metal–fluid interface. The tangential field acts on induced double-layer charge and drives surface slip with Helmholtz–Smoluchowski velocity

$u_s = -(\epsilon_w \zeta / \eta) E_t ,$

where $\epsilon_w$ and $\eta$ are the permittivity and viscosity of the fluid and $\zeta$ is the zeta potential (Hong et al., 2024).

Finite-element simulations in the 2024 report show that these ACEO flows radiate inward toward the unpatterned centers of adjacent holes and converge to form in-plane stagnation zones at the center of each square of four holes. A trapped nanoparticle experiences two dominant forces: hydrodynamic drag from ACEO, reported as $F_{\mathrm{ACEO}} \approx 6\pi \eta a u_s$ for a particle of radius $a$, and an out-of-plane “surface interaction” force $F_s$ arising from image-charge attraction and double-layer repulsion that keeps the particle approximately 100 nm above the gold film. Together, these forces confine nanoparticles in three dimensions without any physical contact or chemical binding (Hong et al., 2024).

The passivation study presents the same mechanism in interface-centric terms. There, the tangential field $E_t$ acts on the diffuse part of the electric double layer, driving ACEO surface slip with

$u_s = (\epsilon_w \zeta / \eta)\cdot E_t ,$

and the resulting flows converge at the interstitial sites between adjacent holes, creating recirculating vortices that reversibly trap sub-100 nm objects at those sites (Ugwu et al., 18 Jun 2026).

These descriptions jointly indicate that IET depends on a controlled coupling among electrode geometry, double-layer electrokinetics, and near-surface hydrodynamics. A plausible implication is that any modification of zeta potential, conductivity, or interfacial chemistry can affect not only fouling but also the stability of the electrohydrodynamic trap itself.

3. Interferometric and spectroscopic readout

Once trapped, particles are characterized optically by interferometric scattering. In the 2024 device description, a continuous-wave 520 nm laser illuminates the array from above; the transmitted field $u_s = -(\epsilon_w \zeta / \eta) E_t ,$0, having passed through the 15 nm gold film, interferes at the camera with the forward-scattered field $u_s = -(\epsilon_w \zeta / \eta) E_t ,$1 from each particle, producing

$u_s = -(\epsilon_w \zeta / \eta) E_t ,$2

After background normalization, the normalized interferometric contrast is reported as

$u_s = -(\epsilon_w \zeta / \eta) E_t ,$3

Because forward Mie scattering dominates over backscattering for particles whose diameter approaches half the illumination wavelength, the magnitude of $u_s = -(\epsilon_w \zeta / \eta) E_t ,$4 scales nearly linearly with particle size $u_s = -(\epsilon_w \zeta / \eta) E_t ,$5 when plotted as $u_s = -(\epsilon_w \zeta / \eta) E_t ,$6 (Hong et al., 2024).

Finite-difference time-domain simulations reported in the same work show that spherical and ellipsoidal particles of identical volume produce distinct interferometric signatures, enabling shape identification through analysis of the two-dimensional intensity pattern around each trap. For 100 nm, 200 nm, and 300 nm polystyrene beads, the effective contrast $u_s = -(\epsilon_w \zeta / \eta) E_t ,$7 plotted versus known diameter yielded the calibration line

$u_s = -(\epsilon_w \zeta / \eta) E_t ,$8

and Brownian-diffusion validation gave diameters 96.6, 207.2, and 302.2 nm (Hong et al., 2024).

The 2026 passivation study describes interferometric scattering in a related but differently formulated geometry: a coherent laser beam illuminates the array from below, a portion reflects off the gold to form a reference field $u_s = -(\epsilon_w \zeta / \eta) E_t ,$9, and a portion scatters from a trapped nanoparticle as $\epsilon_w$0. The detected intensity is

$\epsilon_w$1

with the cross-term scaling linearly with particle volume for sufficiently small particles and producing high-contrast, label-free images of single 100 nm vesicles or beads (Ugwu et al., 18 Jun 2026).

Raman spectroscopy is integrated into the 2024 platform through the same high-NA objective used for imaging. A 785 nm Raman excitation beam is focused onto a selected trap, and inelastically scattered light from the single nanoparticle is routed through a dichroic mirror and notch or long-pass filter into a fiber-coupled spectrometer. Typical acquisition times of 1–10 s yield characteristic bands for polystyrene beads and EVs, enabling simultaneous size and chemistry readout at single-particle sites (Hong et al., 2024).

4. Reported device implementations and operating regimes

The reported IET architecture consists of a patterned gold electrode, a semitransparent ITO counter-electrode, and a thin spacer that defines a closed microfluidic channel. In the 2024 report, the microfluidic chip is fabricated by photolithographically patterning an 8 mm $\epsilon_w$2 4 mm area of 15 nm Au on $\epsilon_w$3 with 15 µm-diameter circular holes, spaced on an 18 µm square lattice to yield $\epsilon_w$4 stagnation-zone traps. A dielectric spacer defines the channel height at approximately 20 µm between the gold film and an ITO-coated coverslip (Hong et al., 2024).

The 2026 passivation study describes the gold-based IET device as a 100 nm-thick gold film patterned lithographically on glass into a 15 µm-diameter hole array with 2 µm edge-to-edge spacing. In that configuration, a semitransparent ITO counter-electrode is separated from the gold by a thin dielectric spacer, creating a closed microfluidic channel, and an AC voltage, typically 10 Vpp at 3 kHz, is applied between gold and ITO (Ugwu et al., 18 Jun 2026).

Under a 10 Vpp AC bias at 3 kHz, the 2024 study reports ACEO slip velocities of approximately 5–10 µm/s, sufficient to sweep nanoparticles from millimeter-scale distances into each trap in under 3 s. The same report states a parallel throughput of up to 60,000 traps on chip and approximately 25 traps in a typical 80 $\epsilon_w$5 80 µm field of view with a 60$\epsilon_w$6 objective. The reported detectable diameter range is 50 nm to 400 nm, with size accuracy of $\epsilon_w$7 when calibrated via Brownian-motion analysis, and a sensitivity limit at detectable diameter $\epsilon_w$8 nm, limited by camera noise and background suppression (Hong et al., 2024).

These implementation details are significant because they place IET between single-particle analytical microscopy and massively parallel microfluidics. The reported throughput and trapping time imply that the platform is intended for many-site acquisition rather than sequential manipulation of isolated targets.

5. Surface passivation and antifouling interfaces

Surface passivation in the gold-based IET device is introduced to suppress irreversible adhesion of negatively charged particles by creating a steric and/or electrostatic barrier on the gold surface. The 2026 study compares three passivation states: 11-mercaptoundecanoic acid (MUA), poly(sodium styrene sulfonate) (PSS) grown by atom transfer radical polymerization (ATRP), and poly(methacryloyloxyethyl phosphorylcholine) (PMPC) grown by ATRP (Ugwu et al., 18 Jun 2026).

MUA is formed by cleaning chips in ethanol/DI water, UV/Ozone for 30 min, then soaking in 10 mM MUA in ethanol/H$\epsilon_w$9O (1:1) for 24 h, followed by rinsing in ethanol and DI water and drying under N$\eta$0. Its thiol headgroup chemisorbs to Au to form a dense self-assembled monolayer, and the terminal COOH groups deprotonate at pH $\eta$1, imparting a negative surface charge. Characterization in the study shows that the contact angle decreases from approximately 69$\eta$2 for bare Au to approximately 26$\eta$3 for MUA, and ATR-FTIR bands at 1740 cm$\eta$4 and 2850 cm$\eta$5 verify MUA on Au (Ugwu et al., 18 Jun 2026).

PSS is prepared by first assembling 11-mercaptoundecyl 2-bromo-2-methylpropanoate initiator at 10 mM on Au for 24 h, then carrying out “grafting from” ATRP under N$\eta$6 with Cu(I)Br, 2,2′-bipyridyl ligand, and a N$\eta$7-sparged 1:1 methanol/H$\eta$8O solution of sodium 4-vinylbenzenesulfonate at 0.1 M. After 24 h at room temperature, the chip is rinsed sequentially in ethanol, PBS, and DI water. The resulting film is a multilayer brush of polystyrene sulfonate chains bearing negatively charged SO$\eta$9 groups, and ATR-FTIR shows features at approximately 1,200–1,300 cm$\zeta$0 and approximately 2,700–3,200 cm$\zeta$1 (Ugwu et al., 18 Jun 2026).

PMPC follows the same initiator grafting and ATRP procedure as PSS, substituting the monomer with 2-methacryloyloxyethyl phosphorylcholine at 0.1 M. The product is a zwitterionic polymer brush with tightly bound hydration layers but net-neutral charge, and ATR-FTIR confirms P–O and N–C vibrations characteristic of phosphorylcholine (Ugwu et al., 18 Jun 2026).

Surface state	Reported interfacial feature	Reported fouling outcome
Bare Au	Unpassivated gold surface	$\zeta$2 beads adhered within 27 min; $\zeta$3 EVs adhered in 40 min
MUA	Dense SAM with terminal COOH groups	$\zeta$4 beads adhered in 40 min
PSS	Multilayer brush with negatively charged SO$\zeta$5 groups	Only one bead in 40 min; for EVs, $\zeta$6 adhered in 40 min
PMPC	Zwitterionic polymer brush with tightly bound hydration layers	For EVs, $\zeta$7 adhered in 40 min

The comparative result reported for 100 nm polystyrene beads is PSS $\zeta$8 MUA $\zeta$9 bare Au in antifouling performance, whereas for extracellular vesicles the reported relation is PSS $F_{\mathrm{ACEO}} \approx 6\pi \eta a u_s$0 PMPC $F_{\mathrm{ACEO}} \approx 6\pi \eta a u_s$1 bare Au (Ugwu et al., 18 Jun 2026). The study attributes MUA’s weaker performance to its single self-assembled monolayer and attributes the stronger behavior of PSS to a thick, multi-charged brush delivering stronger long-range repulsion. For EVs, both negatively charged PSS and highly hydrated zwitterionic PMPC resist fouling equally well (Ugwu et al., 18 Jun 2026).

6. Measurements, interpretation, and design implications

The reported experimental workflow integrates trapping with release-based assessment of fouling. In the passivation study, fluorescence images of 100 nm fluorescent polystyrene beads and interferometric scattering images of label-free EVs are recorded at 3 kHz AC trapping, and the voltage is then turned off to observe particle release. Release efficiency is interpreted operationally: after switching AC off, only non-adhered particles diffuse away, whereas adhered particles remain and are counted as fouling (Ugwu et al., 18 Jun 2026).

The particle populations used in those measurements are specified. The bead concentration is $F_{\mathrm{ACEO}} \approx 6\pi \eta a u_s$2 particles/mL for 100 nm beads with zeta potential approximately $F_{\mathrm{ACEO}} \approx 6\pi \eta a u_s$3 mV, whereas EVs have a broader size distribution of approximately 50–200 nm and zeta potential approximately $F_{\mathrm{ACEO}} \approx 6\pi \eta a u_s$4 to $F_{\mathrm{ACEO}} \approx 6\pi \eta a u_s$5 mV (Ugwu et al., 18 Jun 2026). In the 2024 study, three EVs extracted from human plasma with diffusion-derived radii of 91 nm, 63.5 nm, and 38.4 nm exhibited cube-root contrasts of approximately 0.33, 0.25, and 0.18, while a trapped supermere of estimated diameter 58 nm yielded $F_{\mathrm{ACEO}} \approx 6\pi \eta a u_s$6; Raman spectra from single EVs showed different peak groups, consistent with heterogeneous cargo (Hong et al., 2024).

Several design recommendations are stated explicitly in the passivation work. Antifouling chemistry should be selected based on particle zeta potential and operating pH. ATRP “grafting from” should be employed for high grafting densities and tunable film thickness in the tens of nm. Solution conductivity should be monitored because high ionic strength compresses double layers and can weaken electrostatic or hydration repulsion (Ugwu et al., 18 Jun 2026).

The broader implication stated in that work is that ATRP enables independent choice of headgroups for substrate adhesion and terminal groups for target-specific antifouling, and can be extended to silicon or ITO by changing the initiator chemistry. The same source further states that the demonstrated ATRP-based strategy is generalizable to any gold-based nano- or microfluidic trapping substrate and, by swapping in silane or phosphate ATRP initiators, to silicon, $F_{\mathrm{ACEO}} \approx 6\pi \eta a u_s$7, or ITO platforms (Ugwu et al., 18 Jun 2026). This suggests that, within the reported framework, IET should be understood not only as a trapping-and-imaging platform but also as an interface-engineering problem in which electrokinetics, optical readout, and surface chemistry are jointly determinative.