1. Introduction
SARS-CoV-2 has caused the COVID-19 pandemic with a significant global impact. Coronaviruses cause mild to moderate upper respiratory tract illnesses in both humans and animals [
1]. To contain the outbreak and manage the pandemic, early diagnosis and screening of COVID-19 patients are necessary since rapid isolation of COVID-19 patients prevents spreading events and facilitates the timely treatment at the early state of the illness. Real-time reverse transcription-polymerase chain reaction (RT-PCR) has been approved by the US Food and Drug Administration (FDA) to detect SARS-CoV-2. However, molecular diagnosis based on RT-PCR usually requires expensive reagents, skilled staff, and specific equipment. Besides, the preparation step is time-consuming, complicated, and easily influences diagnostic accuracy [
2]. Thus, molecular diagnostic tests are not suited for point of care testing for COVID-19 or other diseases.
By developing specific immunoassays to detect antigen proteins, the SARS-CoV-2 antigen can be directly detected based on immunological diagnostic tests, which are reliable, cheap, timely, and widely used for the diagnosis of acute infection. Immunoassays, depending on specific antigen–antibody binding reactions, have attracted enormous attention in various fields, such as medical diagnostics and environmental and food monitoring [
3]. Conventional immunoassays involve complex detection protocols and require qualified professionals, limiting the wide range of applications due to inherent diffusion-limited reaction kinetics, and an incubation step of hours or even longer for achieving a detectable level [
4]. With no signal amplification technique for proteins and peptides, limit of detection (LOD) can merely be decreased by sacrificing detection sensitivity. Recently, new formats for miniaturized immunoassays have gained extensive attention. Microfluidic technology [
5], a nascent technology, seeks to alleviate the above dilemmas, and the performance of analysis can be upgraded by requiring small amounts of reagents, reducing the analysis time, improving the sensitivity and reliability through automation, and integrating multiple processes in a simple chip. Immunoassays are ported onto microfluidic formats increasingly due to enormous challenges and unmet medical needs such as early diagnosis of disease by detecting ultra-low concentration analytes, clinical trials of new drugs, and the rapid detection of disease in resource poor areas [
6]. The combination of microarray technologies and microfluidic chips provides the capability to simultaneously detect many different samples in a small area, which is quite useful in proteomics (protein microarrays) and genomics (DNA microarrays). However, they are still limited, arising from diffusion limitations of the analyte in the laminar flow regime, where large analyte depletion layers act as a resistance to analyte detection, and a long time is required for delivering an analyte to the biosensor at lower concentrations [
7]. To address this problem and further reduce LOD, analytes can be transported to the sensor by active delivery with an external force, such as electric [
4], magnetic, or acoustic force. Some scholars have utilized pumping techniques to improve the binding performance. A passive mixing structure is placed over the capture spot by Lynn et al. [
7], to improve the binding rate of analyte to the sensing surface.
Many passive mixers have been used to improve the binding performance and enhance the signal generation, but more samples are required, and the geometry is much more complex [
8,
9]. Although the method of flow confinement [
10] can confine the sample into a thin layer above the sensor surface, the depletion layer is not eliminated completely. AC electrokinetics, including dielectrophoresis (DEP), AC electrothermal (ACET) and AC electroosmosis (ACEO), have been widely used to enhance the performance of heterogeneous immunoassays thanks to the advantages of requiring low AC power and alleviating electrolysis reactions [
3,
11]. DEP [
12,
13] is often used to manipulate particles based on the difference in polarizability between a particle and the medium. Cheng et al. use DEP to trap the concentrated molecules at the surface of an antibody-immobilized electrode. Li et al. [
14,
15] presents a sensitive affinity sensor by integrating DEP into label free capacitive measurements. Nevertheless, DEP is size-dependent and short-ranged, limiting the detection and manipulation of small biomolecules [
16]. ACET is caused by the movement of induced charges due to the joule heating in the fluid under the influence of electric fields [
17]. Many scholars [
3,
18,
19] have demonstrated the capability for ACET effect [
17] to improve the detection of diluted and small molecule targets. Sigurdson et al. [
3,
19] demonstrate the ability of ACET flow to enhance the performance of heterogeneous assays. Selmi et al. [
20] numerically study the effect of the electrothermal flow on the binding reaction of C-reactive protein (CRP). Though ACET flow can transport small molecules to the sensor surface and enhance binding assays, it usually causes large temperature rise. ACEO flow, arising from the interaction of the electric field with the charge in the diffuse double layer [
21], can easily generate large velocities using low voltages and has also been investigated to enhance the binding performance [
22]. Wu et al. integrate ACEO flow and label-free electrochemical impedance spectroscopy in DNA sensing chips [
23]. The reaction surface is usually located at the driving electrode surface due the circulating ACEO flow near the electrode, which is not desirable for analyzing immunoassays in multiple or different reaction regions. Recently, based on the optically induced ACEO flow, Han et al. [
24,
25] reported a novel optoelectrofluidic immunoreaction system for enhancing antibody analyte binding performance, permitting the detection of target analyte in low sample volumes. The microarray-integrated optoelectrofluidic immunoassay system can not only result in the consumption of small amounts of both precious samples and expensive antibodies but also enable the assays to be run in parallel efficiently [
26]. However, the optoelectrofluidic system requires complex fabrication processes (the preparation of the photoconductive substrate) and illumination components, limiting its practical application in diverse fields [
27]. Merkoci et al. [
28] explores micromotors-assisted microarray technology for immunoassays; however, the use of hydrogen peroxide might degrade the involved proteins, and the accumulation of bubbles occurs in the chip.
We herein take advantage of asymmetric induced charged EO (ICEO) flow at a sinusoidal bipolar electrode (BPE) to improve the rate of antibody binding to the reaction surface based on finite element modeling. Ren et al. [
29] exploited ICEO flow in the rotating field for enhancing immunoassays, but the use of a rotating electric field limits its wide application, and it is difficult to enhance the binding reactions in a specific region. In contrast, the introduction of fixed-potential ICEO flow at a sinusoidal BPE in this work can permit the enhancement of assays in a selected reaction region and enable efficient immunoassays in an array wirelessly. This work aims to understand how the positions of the reaction region affect assay improvement and which optimum frequency and voltage to use, and to investigate the influence of the gate voltage of the sinusoidal electrode on antibody binding performance. Asymmetric induced charge electroosmosis (ICEO) flow is utilized to stir the flow field above the BPE electrodes, accelerate the transport of analyte to the functionalized surface, and simultaneously minimize the localized target depletion. An optimized design of the proposed microfluidic chip is proposed based on the immunoassay enhancement. As a result, the sensor target interaction can be improved, and it owns the capability of the efficient enhancement of assays in a selected reaction region within the microchip. The current asymmetric-ICEO-flow-assisted microarray technology can also be extended to other proteins, DNA, and cell analyses.
3. Results and Discussion
To evaluate the effect of some critical parameters on the kinetics of the microfluidic immunoassays, we conduct some simulations using the proposed designs. The parameters for SARS-CoV-2 antigen and antibody used in this work are shown as follows: an immobilized antibody concentration
B = 3.3 × 10
−8 mol/m
2, the association rate constant
kads = 197 m
3/(s·mol), the dissociation rate constant
kdes = 2.58 × 10
−4 (1/s), the concentration of the SARS-CoV-2 antigen in the bulk
C0 = 1 × 10
−7 mol/m
3, and the diffusion coefficient of the antigen
D = 2 × 10
−12 m
2 s
−1, which are comparable to the values reported in other studies [
38]. We analyze the grid independence in
Supporting Information. As shown in
Figure S2, surface slip velocity versus different grid elements is investigated. When element numbers are 35,121 and 112,223, the surface slip velocities are 1.5103 mm/s and 1.5302 mm/s, respectively, and the standard deviation is 1.4074 × 10
−5. Above all, we can consider that the grid division method is independent after element number exceed 35,121. In this work, we divide 35,121 meshes for the following numerical simulations.
3.1. Binding Enhancement by a Sinusoidal Bipolar Electrode
Swirling ICEO flow pattern can be used to circulate suspended antigen past the binding surface, offering more binding opportunities for the suspended molecules. Flow velocity and electric field distribution at different cross-sections predicted by the numerical model are shown in
Figure 2, when the top and bottom electrodes are energized with AC signals of
V1 = 19 V,
V2 = 0 V, and
F = 80 Hz and the sinusoidal bipolar electrode is floating.
Figure 2a,c indicate that a higher AC field strength and stronger asymmetric ICEO flow vortexes occur near the region where the driving electrode is located nearest to the sinusoidal electrode. Symmetric ICEO flow vortexes also appear near the middle region of the microcavity in
Figure 2b. Because the high flow velocity can be generated near the region where the driving electrode is close to the sinusoidal electrode, we at first expect the opposite surface of the region to be an optimum sensor location. The sinusoidal electrode offers asymmetric vortexes for enhancing antigen–antibody binding by micro-stirring.
We investigate the effect of this flow pattern on the binding response of a heterogeneous immunoassay in which antibody has been immobilized at the surface reaction region. Simulation results in
Figure 3 show the surface concentration field of bound antigen–antibody complexes and suspended antigen concentration field in the micro-cavity over time. In a standard and passive condition where the chip is not energized with an AC signal (
Figure 3a,c), diffusion is the only transport mechanism and binding process depletes the suspended analyte concentration, ensuring that the depleted region near the reaction surface grows with time and decreases the binding rate of the immunoassay.
Figure 3a,c show the surface concentration distribution of bound antigen–antibody complexes and suspended antigen concentration distribution before applying an AC signal, indicating that the immunoassay is implemented in the mass-transport-limited regime and the majority of the surface binding of antigen occurs near the reaction region. As the antigen binds to the reaction surface, a “depletion zone” quickly appears above the binding surface. Thus, a decreasing in the assay time requires the overcoming of the sharp reduction in the antigen binding near the reaction surface.
Figure 3b,d display the surface concentration field of bound antigen–antibody complexes and suspended antigen concentration field over time after applying an AC signal of
V1 = 19 V,
V2 = 0 V and
f = 80 Hz. The reason for the apparent change in bound antigen–antibody complexes is mainly due to the accelerated transmission of antigen to the surface reaction areas after applying an electric field. With the application of an electric field, ICEO vortexes are generated at the bipolar electrode surface and the micro-stirring over the BPE surface accelerates the fluid flow over the reaction surface, redistributing the depleted antigen concentration (
Figure 3d). The near-wall deleted concentration is replenished with fresh antigen, thus promoting a uniform and rapid distribution of antigen concentration throughout the chamber.
Figure 3b shows a greater amount of antigen–antibody complexes bound at the reaction surface, which can be attributed to efficient transport of suspended antigens to the reaction surface.
3.2. Effect of the Position of the Reaction Surface, Damkohler Number, Applied Voltage, and Frequency
As the ICEO flow driven circulation redistributes the depleted concentration throughput the micro-cavity, the position of the functionalized reaction surface needs to be investigated for achieving a higher binding rate. As illustrated in
Figure 4c, we designed three microfluidic chips with different locations of reaction surface regions. The coordinates of the central position of circular reaction surfaces are given in
Table 2. Simulations have been carried out for these three structures under different conditions.
To demonstrate the effectiveness of ICEO flow stirring above sinusoidal BPE surface in a variety of applications, we first study the influence of the Damkohler number (Da) upon binding rate. As for large Da, slow diffusion limits the binding rate, and any improvement in transport of analyte to the reaction surface or through ICEO-flow driven micro-stirring will enhance the binding performance. A series of numerical simulations are carried out for different Da. The binding enhancement factor Be= Cs1/Cs0 is defined as the ratio of bound antigen–antibody complexes from ICEO micro-stirring to the bound complexes without micro-stirring.
For each parameter set, the binding enhancement factor
Be is calculated at
t = 120 s (
Figure 4d,f). According to the parameters of SARS-CoV-2 antigen and antibody, here we choose
D = 3 × 10
−12 m
2 s
−1 as the diffusion coefficient for the following simulation analysis (
Da = 260.04). We also further studied the binding performance when
Da is larger than 500 by changing the diffusion coefficients as shown in
Figure 4d. It is noted that the effect of ICEO flow for enhancing immunoassays using these three structures is becoming more significant. As shown in
Figure 4d, the antigen–antibody binding efficiency increases with increasing
Da number, especially for Structure B. According to this plot, we can predict how much binding enhancement can be generated by ICEO flow vortexes at a sinusoidal electrode. The ICEO flow micro-stirring in Structure B yields an enhancement factor of 10 higher binding for
Da = 500. The results show that for all values of
Da, a reasonably significant enhancement factor can be obtained in Structure B compared to other structures including Structure A and C.
The applied voltage and frequency will both influence the fluid velocity and binding performance. To find the appropriate voltage, simulations to determine the suitable voltage are conducted by varying voltage ranging from 2.5 V to 25 V based on the Structure B. The binding enhancement factor
Be, as a function of the applied voltage in the three structures when
Da is 260.04 at 120 s, is shown in
Figure 4e, illustrating
Be generally increases with increased applied voltage, and an extreme value of
Be (about 7.5) occurs when the applied voltage is 20 V in Structure B. When the applied voltage is small, the transportation of antigen to the reaction surface is accelerated as the voltage magnitude increases, thus enhancing the binding performance. However, when the voltage amplitude is too high, the value of
Be increases slowly or even decreases because the reaction rate and the ICEO flow induced convection rate cannot be matched. The antigen–antibody binding reaction regime changes from being transport-limited to being reaction-limited due to the high ICEO flow velocity at high voltages. To obtain the ideal operating conditions for the immunoassays, an optimum frequency is investigated by varying frequency ranging from 1 Hz to 9 kHz based on these three structures. The dependence of the
Be on the applied frequency is shown in
Figure 4f for these three microfluidic chips, and
Be at
t = 120 s decreases with increasing frequency because the applied potential drops completely across the electrolyte and the induced surface charges in double layer tend to zero at higher frequencies, leading to weaker ICEO vortexes. The lowest frequency used is 80 Hz, to avoid hydrolysis and electrode damage. From the comparison of
Be for the three structures in
Figure 4d,g, we choose structure B for the subsequent simulation analysis. The predicted transient process of the antigen–antibody binding at the reaction region using three different designs is depicted in
Figure 4g, indicating the average improvement in antigen–antibody binding during different assay durations. The improvement factors are 7.2, 4.5, and 4 for one minute of ICEO flow stirring in Structure B, Structure C, and Structure A, respectively, and increase steadily for longer assay times. After about 80 s, the increases in binding efficiency approach a maximum value and remain fairly constant for these three structures, demonstrating the effectiveness of ICEO flow stirring above a sinusoidal bipolar electrode to enhance binding rates.
Figure 4h,i show the effect of applied AC signal frequency and amplitude on surface-averaged slip velocity at
Da = 260.04 for the Structure B. The binding efficiency versus the inlet flow rate when
Da = 260.04 and
V1 = 19 V,
f = 80 Hz at
t = 120 s can also be found in
Figure S3. When the flow velocity is under 50μm/s, the binding efficiency improves significantly because the interaction of the ICEO flow and the inlet pressure flow facilitates the transport of antigen to the surface reaction areas. Note that when a higher flow rate of the inlet pressure flow (larger than 100 μm/s) is introduced, the pressure flow dominates over the ICEO flow, which decreases the binding enhancement factor of ICEO flow.
3.3. Effect of Gate Voltage at the Sinusoidal Bipolar Electrode
We previously have demonstrated the influence of ICEO flow above the sinusoidal floating electrode on the binding enhancement. Here, we also investigate how the AC signal exerted on the sinusoidal bipolar electrode will influence the binding performance when the driving electrodes are energized with AC potentials of
V1 = 19 V at
f = 80 Hz.
Figure 5 shows the asymmetrical distribution of vortex flow and electric field within the microchannel at different cross-sections when the driving electrodes and sinusoidal bipolar electrode are driven with AC signals of
V1 = 19 V,
V2 = 0 V,
Vg = 19 V, and
f = 80 Hz.
Figure 5a illustrates that the cross-sections of the flow field and electric field distributions at x = 250 μm, displaying that a high electric field and strong asymmetric ICEO flow vortexes form near the region where the sinusoidal electrode located nearest the ground electrode at
V1 = 19 V,
V2 = 0 V,
Vg = 19 V, and
f = 80 Hz. In
Figure 5b, a large distance between the sinusoidal electrode and ground electrode exists and results in a lower electric field and weaker ICEO flow velocity field. In contrast, a much higher ICEO flow field and electric field can be generated near the region where the ground electrode is close to the sinusoidal electrode in
Figure 5c, contributing to the enhancement of antigen–antibody binding by micro-stirring.
With the ability to actively change the potential of the gate sinusoidal electrode, the asymmetric ICEO vortex can be used to flexibly generate effective mixing, further improving the binding rate of assays in a selected reaction region. The influence of the gate voltage of the sinusoidal bipolar electrode on the enhancement performance of the heterogeneous immunoassay is investigated in detail, as shown in
Figure 6.
Simulation results in
Figure 6a,c show suspended antigen concentration field and the surface concentration field of bound antigen–antibody complexes in the micro-cavity over time after applying an AC signal of
Vg = 19 V (
Vg >
V1/2) to the sinusoidal electrode at
V1 = 19 V,
V2 = 0 V and
f = 80 Hz. A large electric field forms and produces a strong asymmetric ICEO flow (
Figure 5a) near the region where the sinusoidal electrode is located closest to the ground electrode at
V1 = 19 V,
V2 = 0 V,
Vg = 19 V, and
f = 80 Hz, facilitating a strong, continuous, and non-invasive mixing of the target molecule with the immobilized antibody in the reaction area 2. In this case, compared to the reaction area 1, strong ICEO vortexes generated at the bipolar electrode surface near the reaction area 2 redistribute the depleted antigen concentration, and the majority of the surface binding of antigen occurs near the reaction region 2. On the other hand, when the sinusoidal bipolar electrode is driven with an AC signal of
Vg = 19 V (
Vg <
V1/2) at
V1 = 19 V,
V2 = 0 V, and
f = 80 Hz, a strong asymmetric ICEO flow will appear near the region where the sinusoidal electrode is located furthest the ground electrode. As a result, antigen concentration distribution and the surface concentration distribution of bound antigen–antibody complexes within the cavity are shown in
Figure 5b,d. The near-wall deleted concentration is replenished with fresh antigen in surface reaction area 1 due to strong ICEO flow micro-stirring, thus generating a uniform and rapid distribution of antigen concentration near surface reaction area 1. We finally explore the binding enhancement factor Be versus times at each surface reaction area by flexibly changing the voltages applied to the sinusoidal electrode, as depicted in
Figure 6e,f. When
Vg <
V1/2, the binding enhancement can be improved in surface reaction area 1 with the decreasing voltage applied to the sinusoidal electrode. When the sinusoidal bipolar electrode is driven with an AC signal of
Vg = 0 V (
Vg <
V1/2) at
V1 = 19 V,
V2 = 0 V, and
f = 80 Hz, the binding enhancement factor Be in surface reaction area 1 can reach about 7.5 at 120 s. In contrast to reaction area 1, the depletion effect near surface reaction region 2 can be alleviated due to the strong ICEO micro-stirring as the sinusoidal electrode is driven by a gate voltage
Vg (
Vg >
V1/2) and the binding performance is enhanced greatly with the increasing voltage exerted upon the sinusoidal electrode.
Above all, the electric field distribution can be adjusted by energizing the sinusoidal bipolar electrode with different voltages, further enabling efficient and flexible enhancement of heterogeneous immunoassays in a specific reaction area and contributing to the reasonable selection of the location of binding surface. This proposed device using coplanar the sinusoidal BPE electrodes has several advantages, such as low cost, being easy to fabricate, better controllability, and high efficiency, since other methods usually require laser [
39], micromotor [
28], or complicated 3D structure [
7,
40].