Visual and Quantitative Analysis of the Trapping Volume in Dielectrophoresis of Nanoparticles

Abstract

Nanoparticle manipulation requires careful analysis of the forces at play. Unfortunately, traditional force measurement techniques based on the particle velocity do not provide sufficient resolution, while balancing approaches involving counteracting forces are often cumbersome. Here, we demonstrate that a nanoparticle dielectrophoretic response can be quantitatively studied by a straightforward visual delineation of the dielectrophoretic trapping volume. We reveal this volume by detecting the width of the region depleted of gold nanoparticles by the dielectrophoretic force. Comparison of the measured widths for various nanoparticle sizes with numerical simulations obtained by solving the particle-conservation equation shows excellent agreement, thus providing access to the particle physical properties, such as polarizability and size. These findings can be further extended to investigate various types of nano-objects, including bio- and molecular aggregates, and offer a robust characterization tool that can enhance the control of matter at the nanoscale.

Keywords: dielectrophoresis; electrokinetic effects; force; nanoparticles; polarizability; trapping volume.

Figure 1

(a) DEP device design showing the unit cell for the sawtooth metal electrode array and (b) schematic representation of the microfluidic chamber utilized for the DEP experiments. (c) Optical microscope and (d) SEM images of a sawtooth metal electrode array (top view). (e) Schematic representation of the DEP device preparation (cross-sectional view) and working principle utilized to visualize the trapping region: (i) DEP device surface cleaning and hydro–oxidation by oxygen plasma treatment; (ii) gas phase (3-aminopropyl)triethoxysilane (APTES) deposition on top of the OH–rich DEP device surface; (iii) the experimental system before AC voltage application, after addition of Au nanoparticles and microfluidic chamber assembly; (iv.a) top and (iv.b) cross-sectional views of the experimental system during the DEP experiment. Au nanoparticles outside the trapping region indicated by the red circle attach to the primary amine (NH₂–) of APTES molecules through a diffusion–limited process. Au nanoparticles inside the red circle region move toward and accumulate near the sawtooth electrode apexes. This produces two distinct areas on the surface with high and low concentrations, which may be observed by dark-field microscopy.

Figure 2

(a–c) Dark–field images acquired for Au nanoparticles with the radius of (a, d) 25 nm, (b, e) 50 nm, and (c, f) 75 nm after DEP at 15 V_p–p and 3 MHz. (g–i) Dark–field scattering intensity profiles obtained by integrating within a rectangle shown in cyan in panel (d) (see text for details). An exponential fit, shown in purple, is obtained for the intensity profiles in panels (g–i) after subtracting the data near the electrode gap. All scale bars are 50 μm.

Figure 3

(a) 3D geometry of the DEP device used to simulate (b) the electric field strength distribution near sawtooth metal electrodes. (c–h) 2D simulation results of the concentration distributions for (c, f) 25 nm, (d, g) 50 nm, and (e, h) 75 nm radius Au nanoparticles after applying a sinusoidal electric signal with 15 V_p-p peak–to–peak voltage and 3 MHz frequency. The concentration distribution profiles in (f) and (h) were calculated along the yellow line crossing the middle of the gap between adjacent electrode pairs. The red contours in (c–e) and bands in (f–h) depict the area where the DEP potential energy is larger than the thermal diffusion energy, .

Figure 4

Quantitative analysis of the dark–field scattering intensity profiles acquired from approximately 50 different electrode pairs for (a) 25 nm, (b) 50 nm, and (c) 75 nm Au nanoparticles at 15 V_p–p and 3 MHz. The gray lines represent the experimental exponential fits obtained with the procedure outlined in the main text and are similar to those depicted in Figure 2g–i. The purple lines correspond to the simulated concentration profiles shown in Figure 3f–h. Each experimental profile has been normalized between its minimum and its maximum.