**5. ACEO sensor integrated with microcantilever**

We have investigated another microstructural design for ACEO devices, which is similar to a pair of parallel plates (Fig. 15). In this configuration, electrodes are facing each other, similar to a pair of parallel plates. The two face-to-face electrodes are asymmetric in design, so they produce non-uniform electric field. For mechanism identical to that of planar electrodes, surface EO flows are generated from the electrode edges inwards, and slow down to stagnation at the center, where particles are expected to trap.

MEMS Microfluidics for Lab-on-a-Chip Applications 57

E

**++++ + +**

(b)

**+++++++++++**

Fig. 15. Concept of Parallel plate particle trapping for an AC cycle; (a) during the half cycle when the bottom electrode has positive polarity and ITO coated top electrode is negative; (b) next half cycle with opposite polarity. Flow motion and induced charges is also shown

For the next half cycle as in Figure 15b, applied potentials switches polarity and the bottom electrode is at negative potential. Here the induced positive ions will interact with the electric field and again produces two counter-rotating vortices from electrode edges inward, and fluid motions are sustained thus the particles are trapped at the centre of the electrode. As the bottom plate is smaller than the top plate, the electric field is almost always normal to

The tangential electric field for the asymmetric electrode pattern induces electro-osmotic fluid motion in the bottom plate. It is the microfluidic flow that conveys particles from the bulk of the fluid onto the fluid surface. The stagnation point created at the centre of the

This section explains the novel particle trapping method using microcantilever. Here we have presented the first integration of the microcantilever with the ACEO particle trapping mechanism. Recently microcantilever sensor technology has boomed and become a promising sensor technology. Microcantilever sensors have several advantages over many other sensor technologies, including faster response time, lower cost of fabrication, the ability to explore microenvironments, and improved portability. Cantilever resonance responses, such as frequency, deflection, Q-factor, and amplitude, undergo changes due to adsorption or changes in environment. Resonance frequency of a microcantilever can be used to detect particles. When the target is loaded on the microcantilever, the resonance frequency of microcantilever is going to change. That means for the mass loading on the

The parallel plate design has been used to attract particles onto cantilevers for high sensitivity detection [14]. Because particle trapping/concentrating effect is more obvious with the smaller electrode, in our design the metal-coated cantilevers are facing one large electrode (covering a whole fluid chamber), so that particles will aggregate on the cantilevers. The tangential electric field of parallel plate configuration is generated for the asymmetric electrode pattern, which induces electro-osmosis fluid motion. In our design of microcantilever trap, the metal-coated cantilevers substitute the patterned bottom electrode, so that particles will aggregate on the cantilevers [15]. Figure 16 shows the experimental

the top electrode, hence tangential force can be neglected.

En Et E

**++++**

(a)

**+++++++++++**

**5.3 Microcantilever particle trapping using ACEO** 

cantilever the resonance frequency is supposed to go down.

setup of cantilever particle trap.

bottom plate. The particles are trapped at the stagnation point of the fluid.

### **5.1 ACEO particle trap**

ACEO can transport the particles from a large region in the bulk fluid to the electrode surface. The flow velocity is important for optimizing the micropump and particle transportation. In contrast to electrophoretic and dielectrophoretic (DEP) velocity, which are typically limited to less than 20 microns per second [9], the ACEO velocity exceed 100 micron/sec. Figure 14(a) shows the initial distribution of particles when no signals are applied over the electrodes. Figure 14(b) shows that particles accumulated from both sides into lines at approximately 1 <sup>2</sup> of electrode width [12]. This corroborates the theoretic prediction, since fluid velocity reduces at the null points of electric fields, and particles

become trapped to the electrodes due to surface forces of between particles and electrodes. For the biased ACEO experiments, applied voltage exceeds the threshold for reactions at V0=1.5V (i.e. high level & low level of biased voltage is 3V & 0V respectively). At the same voltage, the maximum flow velocity shifts to higher frequency compared with symmetric AC signals. This is because Faradaic polarization becomes suppressed at high frequency. Beyond 500Hz, microflows from capacitive charging are much stronger than those from Faradaic charging, so that the stagnation point on the left electrode disappears. At 100Hz, streamlines from capacitive charging and Faradaic charging become connected, forming a large vortex over the electrode pair and the particles aligned on the right electrode. Figure 14b also validates the null point formation by Comsol Multiphysics simulation in Figure5.

(a) 1/√2 of electrode width (null point) (b)

Fig. 14. (a) Particle without the supply voltage (b) Experimental picture of the particles accumulating at the 1/√2 electrode width;

#### **5.2 Electric field analysis of parallel plate particle trap**

Most ACEO devices reported so far adopt a side-by-side (interdigitated) configuration. To integrate such design on microcantilevers would call for sophisticated microfabrication. We use a face-to-face configuration, very much like a pair of parallel plates, with one plate having smaller electrode area than the other, as shown in Figure 15. As the top and bottom electrodes are asymmetric, the tangential electric fields are generated which induces electroosmotic fluid motion. For the first half cycle as in Figure 15a, the bottom electrode is at positive potential, and negative ions are induced at the interface of the electrode and fluid. These negative ions interact with the electric field and produce two counter-rotating vortices from electrode edges inward on the electrode surface, creating null point at the center of the electrode.

ACEO can transport the particles from a large region in the bulk fluid to the electrode surface. The flow velocity is important for optimizing the micropump and particle transportation. In contrast to electrophoretic and dielectrophoretic (DEP) velocity, which are typically limited to less than 20 microns per second [9], the ACEO velocity exceed 100 micron/sec. Figure 14(a) shows the initial distribution of particles when no signals are applied over the electrodes. Figure 14(b) shows that particles accumulated from both sides

prediction, since fluid velocity reduces at the null points of electric fields, and particles become trapped to the electrodes due to surface forces of between particles and electrodes. For the biased ACEO experiments, applied voltage exceeds the threshold for reactions at V0=1.5V (i.e. high level & low level of biased voltage is 3V & 0V respectively). At the same voltage, the maximum flow velocity shifts to higher frequency compared with symmetric AC signals. This is because Faradaic polarization becomes suppressed at high frequency. Beyond 500Hz, microflows from capacitive charging are much stronger than those from Faradaic charging, so that the stagnation point on the left electrode disappears. At 100Hz, streamlines from capacitive charging and Faradaic charging become connected, forming a large vortex over the electrode pair and the particles aligned on the right electrode. Figure 14b also validates the null point formation by Comsol Multiphysics simulation in Figure5.

Fig. 14. (a) Particle without the supply voltage (b) Experimental picture of the particles

(a) 1/√2 of electrode width (null point) (b)

Most ACEO devices reported so far adopt a side-by-side (interdigitated) configuration. To integrate such design on microcantilevers would call for sophisticated microfabrication. We use a face-to-face configuration, very much like a pair of parallel plates, with one plate having smaller electrode area than the other, as shown in Figure 15. As the top and bottom electrodes are asymmetric, the tangential electric fields are generated which induces electroosmotic fluid motion. For the first half cycle as in Figure 15a, the bottom electrode is at positive potential, and negative ions are induced at the interface of the electrode and fluid. These negative ions interact with the electric field and produce two counter-rotating vortices from electrode edges inward on the electrode surface, creating null point at the center of the

<sup>2</sup> of electrode width [12]. This corroborates the theoretic

**5.1 ACEO particle trap** 

into lines at approximately 1

accumulating at the 1/√2 electrode width;

electrode.

**5.2 Electric field analysis of parallel plate particle trap** 

Fig. 15. Concept of Parallel plate particle trapping for an AC cycle; (a) during the half cycle when the bottom electrode has positive polarity and ITO coated top electrode is negative; (b) next half cycle with opposite polarity. Flow motion and induced charges is also shown

For the next half cycle as in Figure 15b, applied potentials switches polarity and the bottom electrode is at negative potential. Here the induced positive ions will interact with the electric field and again produces two counter-rotating vortices from electrode edges inward, and fluid motions are sustained thus the particles are trapped at the centre of the electrode. As the bottom plate is smaller than the top plate, the electric field is almost always normal to the top electrode, hence tangential force can be neglected.

The tangential electric field for the asymmetric electrode pattern induces electro-osmotic fluid motion in the bottom plate. It is the microfluidic flow that conveys particles from the bulk of the fluid onto the fluid surface. The stagnation point created at the centre of the bottom plate. The particles are trapped at the stagnation point of the fluid.

#### **5.3 Microcantilever particle trapping using ACEO**

This section explains the novel particle trapping method using microcantilever. Here we have presented the first integration of the microcantilever with the ACEO particle trapping mechanism. Recently microcantilever sensor technology has boomed and become a promising sensor technology. Microcantilever sensors have several advantages over many other sensor technologies, including faster response time, lower cost of fabrication, the ability to explore microenvironments, and improved portability. Cantilever resonance responses, such as frequency, deflection, Q-factor, and amplitude, undergo changes due to adsorption or changes in environment. Resonance frequency of a microcantilever can be used to detect particles. When the target is loaded on the microcantilever, the resonance frequency of microcantilever is going to change. That means for the mass loading on the cantilever the resonance frequency is supposed to go down.

The parallel plate design has been used to attract particles onto cantilevers for high sensitivity detection [14]. Because particle trapping/concentrating effect is more obvious with the smaller electrode, in our design the metal-coated cantilevers are facing one large electrode (covering a whole fluid chamber), so that particles will aggregate on the cantilevers. The tangential electric field of parallel plate configuration is generated for the asymmetric electrode pattern, which induces electro-osmosis fluid motion. In our design of microcantilever trap, the metal-coated cantilevers substitute the patterned bottom electrode, so that particles will aggregate on the cantilevers [15]. Figure 16 shows the experimental setup of cantilever particle trap.

MEMS Microfluidics for Lab-on-a-Chip Applications 59

4 *K*

*m*

 f = resonance frequency before mass adsorption; And f' is the resonance frequency during mass adsorption.

> 1 2

(a) (b)

signal applied, 279.52 kHz.

Fig. 18. Resonance frequencies of the MCs (measured with multimode AFM.)

on the cantilever beam surface forms the basis of the particle detection scheme.

(a) After the particle trapping by ACEO, 276.07 kHz, (b) Control experiment with no electric

Microcantilever dimension is 125m x 30m x 4m. The volume of the MC is 1.5e-14 m3. Microcantilever is Si based, and the density of Si is 2330 kg/m3. So the mass of microcantilever is 3.495e-8 gm. By putting the frequency and mass values in equation (4.2), we have got an increase of 2.52% increase of mass for the frequency change from 279.52 KHz to 276.07 KHz. The change in resonant frequency as a function of the particle mass binding

Where, Δm is the mass change; K=sensor spring constant;

we get,

22 2 1 1

Changes in the mass and surface properties of the microcantilever through binding or hybridization of analytes to receptor molecules will directly influence its surface stress. This causes the microcantilever to deflect and the deflection is proportional to the analyte concentration and inversely proportional to mass loading. The more the particle concentration on ACEO-cantilever, the more is the bending. So the more mass on the cantilever means the lower resonance frequency. From our experimental result (Fig. 18) we have got the resonance frequency of the MC goes down to 276.07 KHz after particle trapped on the MC for ACEO, which translates to a change of mass. For a change in mass

*eff*

*f m m f m*

*eff*

 

*f f*

Fig. 16. Experimental Setup of Cantilever particle trap

As shown in Figure 16, photoresists (dielectrics) are coated on the conductive areas other than the cantilevers to suppress unwanted local EO flows. The ITO glass slide works as the top electrode, which is covering the whole fluid chamber. We have used Au-coated AFM probes as the MC, which has the dimension of 125µm X 30µm X 4µm. Tipped MC side was not used to avoid sharp electric fields.

Figure 17 shows the experimental results of trapping 200nm fluorescent particles on MC. After applying the AC signal (100Hz, 400mVp-p), suspended particles accumulate at the center of the cantilever from all directions. As time passes, more fluorescent particles from the surrounding area accumulated and formed bright object pattern. After the particle trapping on the surface, the MC was dried with AC signals applied, so that particles will not get dispersed by diffusion, surface tension, etc. Then the particle trapping effect was verified with AFM resonance measurement.

Fig. 17. Image sequence of 200 nm fluorescent particles trapped on the micro-cantilever;

#### **5.3.1 Microcantilever particle trapping validation**

To verify the concentrated particle trapping on MC, we also measure the resonance frequency of MC before and after trapping experiments. MC resonance frequency is inversely proportional to the differential mass of cantilever [21]. The sensitivity of a cantilever to mass loading is mainly determined by the excited cantilever resonance frequency.

$$
\Delta m = \frac{K}{4\pi^2} \left( \frac{1}{f'^2} - \frac{1}{f^2} \right)
$$

As shown in Figure 16, photoresists (dielectrics) are coated on the conductive areas other than the cantilevers to suppress unwanted local EO flows. The ITO glass slide works as the top electrode, which is covering the whole fluid chamber. We have used Au-coated AFM probes as the MC, which has the dimension of 125µm X 30µm X 4µm. Tipped MC side was

Figure 17 shows the experimental results of trapping 200nm fluorescent particles on MC. After applying the AC signal (100Hz, 400mVp-p), suspended particles accumulate at the center of the cantilever from all directions. As time passes, more fluorescent particles from the surrounding area accumulated and formed bright object pattern. After the particle trapping on the surface, the MC was dried with AC signals applied, so that particles will not get dispersed by diffusion, surface tension, etc. Then the particle trapping effect was verified

(a) (b) (c) Fig. 17. Image sequence of 200 nm fluorescent particles trapped on the micro-cantilever;

To verify the concentrated particle trapping on MC, we also measure the resonance frequency of MC before and after trapping experiments. MC resonance frequency is inversely proportional to the differential mass of cantilever [21]. The sensitivity of a cantilever to mass loading is mainly determined by the excited cantilever resonance

Fig. 16. Experimental Setup of Cantilever particle trap

not used to avoid sharp electric fields.

with AFM resonance measurement.

frequency.

**5.3.1 Microcantilever particle trapping validation** 

Where, Δm is the mass change;

K=sensor spring constant;

 f = resonance frequency before mass adsorption; And f' is the resonance frequency during mass adsorption.

Changes in the mass and surface properties of the microcantilever through binding or hybridization of analytes to receptor molecules will directly influence its surface stress. This causes the microcantilever to deflect and the deflection is proportional to the analyte concentration and inversely proportional to mass loading. The more the particle concentration on ACEO-cantilever, the more is the bending. So the more mass on the cantilever means the lower resonance frequency. From our experimental result (Fig. 18) we have got the resonance frequency of the MC goes down to 276.07 KHz after particle trapped on the MC for ACEO, which translates to a change of mass. For a change in mass we get,

$$\frac{f\_1}{f\_2} = \frac{\sqrt{m\_{eff} + \Delta m}}{\sqrt{m\_{eff}}}$$

Fig. 18. Resonance frequencies of the MCs (measured with multimode AFM.) (a) After the particle trapping by ACEO, 276.07 kHz, (b) Control experiment with no electric signal applied, 279.52 kHz.

Microcantilever dimension is 125m x 30m x 4m. The volume of the MC is 1.5e-14 m3. Microcantilever is Si based, and the density of Si is 2330 kg/m3. So the mass of microcantilever is 3.495e-8 gm. By putting the frequency and mass values in equation (4.2), we have got an increase of 2.52% increase of mass for the frequency change from 279.52 KHz to 276.07 KHz. The change in resonant frequency as a function of the particle mass binding on the cantilever beam surface forms the basis of the particle detection scheme.

MEMS Microfluidics for Lab-on-a-Chip Applications 61

trap the particles and pump the fluid. Interfacing the micro-cantilever with ACEO mechanism has expanded its capability for biological, physical and chemical detection and makes the whole system ultra sensitive. The research work substantially enriches the portfolio of transducers, lab-on-a-chip (LOC) and MEMS that can be used in high

This work is supported in part by the Nanotechnology for Undergraduate Education Program of the National Science Foundation (NSF-NUE) under Award Number EEC-1138205. The work is also supported from the internal fellowship grant from the University of Texas at Brownsville. We also want to thank Arizona State University (ASU) Nanofab

[1] Helene Andersson, Albert ven den Berg, " Microfluidic devices for cellomics : A review",

[2] Vincent Studer, Anne Pépin, Yong Chen, and Armand Ajdari, "An integrated ac

[3] C.-H. Chen and J.G. Santiago, "A planar electroosmotic micropump," J.

[4] Lastochkin, D., Zhou, R., Wang, P., Ben, Y. and Chang, H.-C., "Electrokinetic Micropump

[5] K. H. Bhatt, S. Grego and O. D. Velev, "An AC Electrokinetic Technique for Collection

[6] M. R. Brown, and C. D. Meinhart, "AC electroosmotic flow in a DNA concentrator," Microfluidics & Nanofluidics, ISSN:1613-4982 (Print) 1613-4990 (Online), 2006. [7] Ronald F. Probstein, Physicochemical Hydrodynamics : An Introduction, 2nd Editon,

[8] Stone, H.A. and S. Kim, "Microfluidics: Basic issues, applications, and challenges,"

[9] A. Ramos, H. Morgan, N. G. Green and A. Castellanos, "AC electrokinetics: a

[10] A. Ramos, H. Morgan, N. G. Green and A. Castellanos, "AC electric-field-induced fluid

[11] J. Wu, "Electrokinetic Microfluidics for On-Chip Bioparticle Processing," IEEE Trans.

[12] J. Wu, Y. Ben and H.-C. Chang, "Particle Detection by Micro- Electrical Impedance

flow in microelectrodes", J. Colloid Interface Sci. 217(1999) 420–2

Microfluidics & Nanofluidics, 1(2), pp. 161-167, 2005

review of forces in microelectrode structures", J. Phys. D: Appl. Phys. 31(1998)

Spectroscopy with Asymmetric-Polarization AC Electroosmotic Trapping,"

electrokinetic pump in a microfluidic loop for fast and tunable flow control",

and Micromixer Design Based on AC Faradaic Polarization", J. of Applied Physics ,

and Concentration of Particles and Cells on Patterned Electrodes," *Langmuir* 21, pp.

performance miniaturized analytical systems.

Laboratory for assistance with microfabrication.

Sensors and Actuators, B 92, (2003), 315-325.

Microelectromech. Syst., vol. 11, pp. 672-683, 2002.

Analyst, vol. 129, pp. 944-949, 2004.

New York : John Wiley and Sons, c 1994

AIChE J., 47(6), 1250-1253 (2001).

96 , 1730 (2004).

6603-6612, 2005.

2338–53

Nanotech., Mar. 2006.

**7. Acknowledgment** 

**8. References** 
