|Home | About | Journals | Submit | Contact Us | Français|
The separation of nanoparticles from micron size particles in high conductance buffers was achieved using an AC dielectrophoretic (DEP) microarray device with hydrogel over-coated microelectrodes. While nanoparticles could be selectively concentrated into high field regions directly over the platinum microelectrodes, micro-bubbling and electrode darkening was also observed. For similar experiments using un-coated microelectrodes, SEM analysis showed severe erosion of the platinum microelectrodes and fusion of nanoparticles due to the aggressive electrochemistry.
Dielectrophoresis (DEP) has long offered an attractive mechanism for the high-resolution separation of cells [1–2], viruses , DNA [4, 5], proteins  and non-biological nanoparticles . However, until recently DEP separations had to be carried out in low conductance (ionic strength) solutions, which greatly limited practical applications for biological research or clinical diagnostics [8–13]. Now, the DEP separation of nanoparticles and DNA has been demonstrated under high conductance conditions using hydrogel over-coated microarray devices. Unfortunately, microelectrode darkening and bubbling are also observed [14, 15]. While electrochemical effects have been non-consequential for low conductance DEP applications, their effects under high conductance conditions and at lower AC frequencies [<10 kHz] are more pronounced. Thus, in order to create new viable DEP devices for biological and clinical applications, it is critically important to better understand the complex interactions and electrochemical effects that occur at the microelectrode interfaces under high conductance conditions.
DEP experiments were carried out using both hydrogel over-coated platinum microelectrodes and un-coated platinum microelectrodes. Figure 1A shows the 100 microelectrode array device (Nanogen, San Diego, CA, USA). The circular platinum microelectrodes are 80μm in diameter and over-coated with a 10μm thick porous polyacrylamide hydrogel. Experiments involving no hydrogel layer were performed on pre-cartridge fabricated microarrays (FCOS). Only a 3×3 subset of nine microelectrodes was used for the experiments (see Figure 1B). Alternating current (AC) electric fields were applied to the nine microelectrodes in a checkerboard-addressing pattern . Figure 1C shows a composite image of the separation of 10μm microspheres from 200nm yellow-green fluorescent (505/515) nanoparticles in 0.01× TBE on a hydrogel microarray. As demonstrated previously [14, 15], the 200nm nanoparticles concentrate in the DEP field maxima (high field regions) on the center of the microelectrode, and the 10μm microspheres concentrate in the DEP field minima (low field regions) between the microelectrodes. Figure 1D shows a cross-sectional view of the microarray. To predict nanoparticle concentration on the FCOS array, COMSOL Multiphysics Modeling (COMSOL Inc., Los Angeles, CA) was used to model the electric field. Figure 1E shows the electric field at the surface, with the field intensity strongest at the edge of the microelectrodes and weakest between the microelectrodes. TBE is Tris Borate EDTA buffer pH 8.3 and PBS is phosphate buffered saline (sodium chloride) pH 7.4. For more detailed information on materials and methods see references [14, 15].
Initial experiments involved the separation of 200nm fluorescent nanoparticles from 10μm microspheres under different conductance (ionic strength) conditions on microelectrodes with hydrogels (Figure 2A–F), and without a hydrogel layers (Figure 2G–H). The DEP results for all buffers 0.01× TBE (1.81 mS/m), 1× TBE (109 mS/m), 1× PBS (1.68 S/m) show the separation of the 200nm nanoparticles into the high field regions over the microelectrodes, and the concentration of 10μm microspheres into the low field regions between the microelectrodes. The concentration of the nanoparticles is highest for 0.01× TBE, decreases as the buffer ionic strength increases (see Figure 2B, 2D, 2F, and 2H) and occurs more at the center of microelectrodes with hydrogels and at the perimeter for un-coated microelectrodes (Figure 2A, 2C, 2E, 2G). Significant microelectrode darkening occurred in 1× PBS for both the hydrogel over-coated microelectrodes (Figure 2E) and the un-coated microelectrodes (Figure 2G), and increased micro-bubbling occurred in 1× PBS for both the hydrogel over-coated and the un-coated microelectrodes after four minutes.
DEP was carried out in high conductivity 1× PBS buffer using un-coated microelectrodes with no nanoparticles present. The microarray was washed, dried and imaged by a Scanning Electron Microscope (SEM). Figure 3 shows the light microscope images of an un-activated control microelectrode (Figure 3A), and an activated microelectrode (Figure 3B) after 10 minutes of DEP at 3000Hz, 10 volts pk-pk in 1× PBS. Figure 3C and Figure 3D shows the SEM images of an un-activated microelectrode and an activated microelectrode which is partially degraded. Figures 3E and Figure 3F are higher magnification SEM images showing further degradation of the microelectrode (Figure 3F). Experiments were now carried out in high conductance 1× PBS buffer with 200nm nanoparticles present. Figure 4A shows SEM images of the un-activated control microelectrode after two minutes of DEP at 3000Hz, 10 volts pk-pk in 1× PBS. Figure 4B shows a higher magnification SEM image of the edge of a control microelectrode with some nanoparticles between the edge and the dielectric material. Figure 4C shows the SEM image of a microelectrode activated for 2 minutes, with a large number of nanoparticles concentrated at the edge. A close-up image (Figure 4D) shows clusters of nanoparticles and some degradation of the microelectrode. Figure 4E and Figure 4F show images of an activated microelectrode after 5 minutes of DEP with more nanoparticle clustering and a severely degraded microelectrode. Figure 4G is a higher magnification SEM image of the edge of the microelectrode showing clustering of the nanoparticles. Figure 4H is a higher magnification image of the degraded microelectrode showing fused nanoparticles clusters.
In earlier DEP work a significant increase in the level of micro-bubbling and darkening of the platinum microelectrodes was observed in higher conductance buffers [14, 15]. These adverse effects were suspected to be due to increased electrochemical activity. The results of the present study now clearly show the exact nature of the microelectrode/nanoparticle/electrolyte interactions under high ionic strength conditions. Initial experiments at different ionic strength conditions with a hydrogel layer (Figure 2A–F), and without a hydrogel layer (Figure 2G–H) show: (1) the concentration of 200nm nanoparticles is highest for 0.01× TBE, and decreases as the buffer ionic strength is increased; (2) the concentration of nanoparticles occurs more at the center of microelectrodes with hydrogels, and at the outside perimeter for the un-coated microelectrodes; and (3) darkening/bubbling of the microelectrodes occurs at the highest buffer conductance (1× PBS). The un-coated microelectrodes allowed use of SEM to better analyze the electrochemical effects and to verify nanoparticle concentration. In the first experiments, DEP was carried out in high conductivity 1× PBS buffer on un-coated microelectrodes with no nanoparticles present. Figure 3B, 3D and 3F show that significant damage and degradation occurs on the activated platinum microelectrode after 10 minutes. Subsequent DEP experiments carried out with the nanoparticles present show after 2 minutes that a large number of nanoparticles have concentrated and adhered to the microelectrode edge (Figure 4C). The close-up image (Figure 4D) shows the concentrated nanoparticle clusters and some degradation at the microelectrode edge. Figure 4E and 4F show, after 5 minutes of DEP, more pronounced concentration and clustering of the nanoparticles and a more severely degraded microelectrode. Such microelectrode degradation can only be produced by very aggressive electrochemical effects. Figure 4G is a higher magnification SEM image of the microelectrode edge showing clustering of nanoparticles, and Figure 4H shows nanoparticle clusters interspersed with what appears to be fused or melted nanoparticles. These fused or melted nanoparticle clusters most certainly have resulted from the aggressive electrochemical activity (heat, H+ and OH−) and the longer DEP times.
The results of this study clearly support the earlier hypothesis [14, 15] that DEP at high conductance conditions (>200 mS/m) greatly increases electrochemical activity causing micro-bubbling and microelectrode darkening. This study also shows that significant degradation of the platinum microelectrodes is occurring under these conditions and as the DEP activation time increases. In spite of DEP being an AC electrokinetic process, these results can be directly attributed to DC electrolysis reactions which would produce O2, H2, H+, OH−, heat and bubbles. The presence of high levels of sodium (Na+), potassium (K+) and chloride (Cl−) ions probably also contributes to the corrosive conditions. While the hydrogel coating ameliorates some of the adverse effects of the electrolysis products, which are produced on the surface of the platinum microelectrodes, darkening of the underlying microelectrodes and some bubbling is still observed. The hydrogel layer does reduce the direct effects of the electrolysis products on nanoparticles, as compared to the un-coated microelectrodes where the nanoparticles are literally fused into the degraded microelectrode structure (see Figure 4H). The over-coating also reduces the active bubbling, probably by allowing better gas diffusion and less bubble nucleation. These results immediately make it clear as to why classical DEP, which utilizes less robust sputtered gold electrodes, required low conductance conditions [7, 10–13]. While the hydrogel over-coated microelectrodes do allow separation of nanoparticles at high conductance conditions, they are far from optimal. Thus, the further identification and understanding of these limitations now opens the door for designing more robust DEP devices for detecting nanoparticles and disease related biomarkers directly in blood and other biological samples.
We thank Roy Lefkowitz and Jennifer Marciniak for their input. Financial support from UCSD and NIH NanoTUMOR Center (U54-CA119335) is acknowledged.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.