4-1 Characterization of surface molecular on gold electrode
4-1-1 Electrochemical detection for monolayer coverage on gold electrode
The CV method is capable of analyzing the SAM formation on gold electrode via investigation of moving ability in electrons transfer. Before processing the modification, CV recorded the ions transfer current between the reduction and oxidization at the cleaning gold electrode. It shows the redox species Fe2+/3+ converts easily in black line of Figure 4-1. However, the CV curve varies as the electrode is SAM modified. A 2 mM MUA with the spacer 8mM MH monolayer in absolute ethanol were absorbed for 12 hour on the pattern gold electrode with size of inner diameter in 500 um. CV curve for MUA/MH SAM on gold are given in red line of Figure 4-1. During the CV scan, the MUA/MH monolayer insulate the gold surface against the electrons transfer with Fe2+/3+ molecule in solution. The anodic and cathodic peak currnet was deceased compared to bare gold electrode. The results effectively identify the MUA/MH monolayer formation on biosensing area.
50
Figure 4-1 Comparison of CV scan (a) before modification with (b) after modification.
51
4-1-2 Binding of fluorescent dye to Protein on MUA/MH-covered gold electrode
The CV measurement revealed the 11-MUA/3-MPOH thiol layer modified on gold electrode. The COOH-terminal thiols can covalently bind antibody after activating it to NHS-group. To present the antibody of Vibrio Parahaemolyticus binding to NHS group on patterned gold electrode, a fluorescent dye performed a visualized study of antibody binding. Rhodamine interaction with and without modified gold electrode were characterized by fluorescence microscopy. Unmodified chip act as a control data.
Figure 4-2 and Figure 4-3 show the patterned electrode fixed rhodamine with modified and unmodified antibody. Though the rhodamine on the antibody-modified electrode compared with the control can clearly be observed, the SiO2 substrate still have unstable absorption of Rhodamine.
Figure 4-2 Rhodamine fixed on electrode with antibody modification.
(a) (b)
52
Figure 4-3 Rhodamine fixed on electrode without antibody modification.
4-1-3 HRP development using second antibody
Figure 4-4 depicts the TMB chromogenic resulting of thiol-modified electrode immobilized HRP-conjugated secondary antibody. Under light exposure, the changed color could be visualized. The absorbance at 450 nm was further measured to compare the modification
Figure 4-4 (a) TMB chromogenic result on gold electrode without and with antibody modification (b) absorbance at 450nm.
(a) (b)
53
4-2 AC electrokinetics and microfluidic system in bacteria trapping
4-2-1 The behavior of AC electrokinetics in microfluidic channel
Locating bacteria on the inner surface of electrode is the ideal concentration.
Appearance of AC electrokinetics results from combination of DEP with ACEO.
Figure 4-5(a) depicts the profile of simulated ACEO velocity in the 0.01xPB solution.
It may produce ACEO on the surface of electrode(y=0) even if the velocity on gap inside inner interdigitated electrode is zero. Figure 4-5(b) shows profile of . E is electric field and is proportional to DEP force due to the same parameters of bacteria. We still can get the notice that the DEP is available under this electrode type.
In fact, the bacteria can be manipulated on the surface of inner electrode efficiently with 2 Vp-p and 200 Hz. When applied voltage with 2 Vp-p and 30 kHz, the DEP mainly trap the bacteria at electrode edge. The outcome of bacteria manipulation could be set as a function of frequency, in Figure 4-6 and Figure 4-7. At low frequency, the bacteria could be trapped on the inner electrode because of the combination of DEP and ACEO. At high frequency, DEP force dominated in the microfludic flow. Bacteria were dilution in 0.01xPB and non-uniform electric field was applied by 2 V from peak to peak. The sample flow rate was at 5 uL/min.
54
Figure 4-5 Profile of (a) ACEO (b) gradient of quadratic electric field.
Figure 4-6 Bacteria trapping with 2Vp-p and 200Hz (a)at 0min(b) after 10min.
Figure 4-7 Bacteria trapping with 2Vp-p and 30kHz (a)at 0min(b) after 10min.
55
4-2-2 The influence of flow rate on bacteria trapping
Before bacteria detection, the bacterial cells delivered by tube to microfluidic channel with a rate of flow. The flow rate can affect on the efficiency of bacteria trapping. We use pump to deliver the bacteria in 0.01xPB to microfluidic chamber with different rate and observe the effect on the bacteria concentration on the inner circular electrode. Flow rate at 5, 50, and 100 uL per minute was chosen for monitoring. The results in Figure 4-8 show that the best efficiency of bacteria concentration occurred at flow rate of 5 uL/min. The time for bacteria trapping is 10min by applied voltage 2 Vp-p and frequency 200 Hz.
Figure 4-8 Flow rate at 100uL/min (a) before trapping (b)after trapping , flow rate at 50uL/min (c) before trapping (d) after trapping, flow rate at 5uL/min (e) before trapping (f) after trapping.
56
4-3 Bacteria detection
The AC electrokinetics for Vibrio Parahaemolyticus concentration on surface of inner electrode was manipulated successfully by the data above. The impedance change was measured for immobilization of bacteria and accounted for the physical model based on electrical circuit.
4-3-1 Impedance analysis for three types of electrodes
The sizes of electrodes were fabricated for efficient trapping and detection of bacteria. Three sizes of electrodes with different diameter of inner circular electrodes(100 um, 250 um, and 500 um) was investigated for the impedance change.
We present electrodes with inner diameter 500 um as electrode A, inner diameter 250um as electrode B, and inner diameter 100 um as electrode C(Figure 4-9).
Detection of Vibrio Parahaemolyticus at concentration of 108cfu/mL was performed for comparison. Applying 2 Vp-p with 200 Hz can concentrate bacteria efficiently for electrode A, 2 Vp-p with 400 Hz for electrode B, and 2 Vp-p with 750 Hz for electrode C. One of the electrodes has largest variation of Normalized impedance change (NIC). The electrode with greatest change allowed it to use for following detection. Figure 4-10, Figure 4-12, and Figure 4-14 show impedance spectra as a function of frequency for individual electrode. Figure 4-11, Figure 4-13, and Figure 4-15 results of Normalized impedance change for each electrode. Experiments of bacteria binding with AC electrokinetics are set against without AC electrokinetics for NIC analysis.
Figure 4-16(a) and (c) which are the SEM images tell the difference between bacteria immobilization bound to functionalized electrode with and without AC electrokinetics.
Figure 4-16(b) shows that few bacteria attach to the SiO2 substrate resulting from antibodies not being immobilized on the oxide surface.
57
Figure 4-9 Different sizes of electrode. (a) inner diameter: 500 um (b) inner diameter: 250 um (c) inner diameter: 100 um.
100 1k 10k 100k
Bacteria binding without AC electrokinetics enhancement
at 108 cfu/mL concentration
100 1k 10k 100k
Bacteria binding with AC electrokinetics enhancement
electrode diameter: 500um
at 108 cfu/mL concentration
Figure 4-10 Impedance spectra using electrode A.
(left: without AC electrokinetics /right: with AC electrokinetics)
100 1k 10k 100k
at 108 cfu/mL concentration
Figure 4-11 Normalized impedance change(NIC) for electrode A . (blank line: without AC electrokinetics;
red line: with AC electrokinetics)
58
Bacteria binding without AC electrokinetics enhancement
electrode diameter: 250um
at 108 cfu/mL concentration
100 1k 10k 100k
Bacteria binding with AC electrokinetics enhancement
electrode diameter: 250um
at 108 cfu/mL concentration
Figure 4-12 Impedance spectra using electrode B.
(left: without AC electrokinetics /right: with AC electrokinetics)
at 108 cfu/mL concentration
Figure 4-13 Normalized impedance change (NIC) for electrode B.
(blank line: without AC electrokinetics;
red line: with AC electrokinetics)
Bacteria binding without AC electrokinetics enhancement
at 108 cfu/mL concentration
100 1k 10k 100k
Bacteria binding with AC electrokinetics enhancement
electrode diameter with 100um
at 108 cfu/mL concentration
Figure 4-14 Impedance spectra using electrode C.
(left: without AC electrokinetics /right: with AC electrokinetics)
59
at 108 cfu/mL concentration
Figure 4-15 Normalized impedance change(NIC) for electrode C.
(blank line: without AC electrokinetics;
red line: with AC electrokinetics)
Figure 4-16 SEM images of Vibrio Parahaemolyticus bound to antibody-modified electrode surface (a) trapping with AC electrokinetics enhancement (b) edge of electrode (c) trapping without AC electrokinetics enhancement.
60
All of the impedance spectra from three different sizes of electrodes have the same trend of decreasing value in low frequency after binding with the bacteria. It can be attributed to the change of dielectric properties on electrode surface. The surface of bacteria has many proteins with high conductivity.[40] These charges can gain the conductance around the electrode surface causing total impedance reducing.
Compared with value of detection(red line) from Figure 4-10, Figure 4-12, and Figure 4-14, the electrode A gives a greater change in bacteria detection. The AC electrokinetics trapping can actually detect the bacteria in limited time. Figure 4-17 illustrates NIC in three sizes of electrodes. The NIC can obviously tell a major difference on electrode with inner diameter 500 um (electrode A).
Figure 4-17 NIC of bacteria detection with AC electrokinetics in three sizes of electrodes.
We fitted the impedance data to the electrical circuit model. The value of the element in the circuit with and without bacteria binding can explain the behaviors inside the chamber. The circuit model can fit the measured impedance under 5% error.
Figure 4-18(a) Figure 4-19(a), and Figure 4-20(a) are the fitting of the equivalent circuit of 500 um, 250 um, and 100 um of inner diameter electrode. In one hand, thick
61
gray (antibody-modified) and blue (bacteria binding) line perform the measured data.
On the other hand, Thin cyan (antibody-modified) and red (bacteria binding) line are fitting spectra. The impedance spectra of antibody-modified electrodes are fitted by impedance of distinctive elements in Figure 4-18(b), Figure 4-19(b), and Figure 4-20(b). We figure out that the CPEdl at low frequency dominating the impedance value. Frequency ranges from 50 to 1 kHz can play a role for electrode A, 50 to 550Hz for electrode B, and 50Hz to 300Hz for electrode C, collected in Table 4-1.
The electrical element from the model in different electrodes is followed by Figure 4-21, Figure 4-22, and Figure 4-23. CPEdl represents capacitance of the electrode surface and its change in three elements is largest due to the attachment of bacteria onto the electrode. The value of each element for different electrode and with/without bacteria binding is showed in Table 4-2. The difference of CPEdl-T before and after bacteria binding for electrode A is evaluated as 2.4 times. Others could be 1.07 times for electrode B and 0.9 times for electrode C.
62
after bacteria binding via AC electrokineticsfitting electrode inner diameter:500um
at 108 cfu/mL concentration of Vibrio parahaemolyticus
at 108 cfu/mL concentration of Vibrio parahaemolyticus
Figure 4-18 Impedance spectroscopy of inner diameter: 500 um (a) Fitting spectra of the measured data (b) Magnitude of three elements in impedance analysis.
(a)
(b)
63
after bacteria binding via AC electrokineticsfitting electrode inner diameter: 250um
Concentration of 108 cfu/mL
100 1k 10k 100k measured data (b) Magnitude of three elements in impedance analysis.
(a)
(b)
64
after bacteria binding via AC electrokineticsfitting electrode inner diameter: 100um
Concentration of 108 cfu/mL
100 1k 10k 100k measured data (b)Magnitude of three elements in impedance analysis.
(a)
(b)
65
Table 4-1 Influence of frequency for each element.
Inner diameter:500 um Inner diameter:250 um Inner diameter:100 um
CPEdl 50 ~ 1k Hz 50 ~ 550 Hz
50 ~ 300 Hz
Rsol 1k~ 200k Hz 550 ~ 60k Hz
300 ~ 10k Hz
CPEsol 200k ~ 500k Hz 60k ~ 500k Hz
30k ~ 500k Hz
Table 4-2 Value of each element for the electrical circuit fitting.
CPEdl-T CPEdl-p Rsol CPEsol-T CPEsol-p
Inner electrode diameter : 500 um before
Inner electrode diameter : 250 um before
Inner electrode diameter : 100 um before
66
Figure 4-21 Impedance spectroscopy of electrode A (a) impedance of CPE on electrode surface (b) resistance of solution(c) impedance of CPE in the solution.
(a) (b)
(c)
67
Figure 4-22 Impedance spectroscopy of electrode B (a) impedance of CPE on electrode surface (b) impedance of solution(c) impedance of CPE in the solution.
(a) (b)
(c)
68
Figure 4-23 Impedance spectroscopy of electrode C (a)impedance of CPE on electrode surface (b) impedance of solution(c) impedance of CPE in the solution.
4-3-2 Impedance analysis on antibody-unmodified chip
The efficiency of bacteria binding was characterized by detection on the antibody-unmodified chip. The detection was gotten by trapping Vibrio
Parahaemolyticus at concentration of 10
8 cfu/mL with AC electrokinetics compared with trapping without it on electrode A. Figure 4-24 show the results of measurement for antibody-unmodified process. Figure 4-24(a) is the impedance spectra for bacteria binding with excitation. Figure 4-24(b) is control without excitation. The pictures in Figure 4-25 reveal few bacteria under microscopy are still binding after AC electrokintices enhancement and wash. The remaining attachment of bacteria is caused by the polarized effect from the DEP force. The impedance normalized(a) (b)
(c)
69
change(NIC) between the bacteria binding in Figure 4-26 states similar variation, so the gold electrode with antibody modification has beneficial to bind bacteria. We also take the NIC data from the detection of Vibrio Parahaemolyticus at concentration of 108 cfu/mL on antibody-modified chip for comparison at 50 Hz, shown in Figure
Bacteria detection with AC electrokinetics
Impedance ()
Frequency (Hz)
electrode diameter: 500um
at 108 cfu/mL concentration
100 1k 10k 100k
Bacteria detection without AC electrokinetics
Frequency (Hz)
electrode diameter: 500um
at 108 cfu/mL concentration
Figure 4-24 Impedance spectra (a)with AC electrokinetics (b) without AC
electrokinetics on antibody-unmodified electrode.
Figure 4-25 Optical images of Vibrio Parahaemolyticus attaching on electrode after binding (a) without AC electrokinetics (b) with AC electrokinetics and
wash.
70
at 108 cfu/mL concentration
The electrode was no modified by antibodies
Figure 4-26 NIC spectra for Vibrio Parahaemolyticus binding with and without AC electrokinetics enhancement.
Figure 4-27 NIC between bacteria binding with and without modification of antibodies.
Appling the impedance spectra to the equivalent circuit model, the data fitting of three elements is collected by Table 4-3. The electrode with antibodies modification has biggest change on surface capacitance (CPEdl-T) after bacteria binding, 2.4 times of value from surface capacitance before binding, however, the chip without antibodies modification gives a change of 0.9 times in CPEdl-T. The impedance
71
spectra of each element for antibodies unmodification are shown in Figure 4-28. The magnitude of capacitance of electrode surface, resistance of solution, or capacitance of solution demonstrates a negligible change before and after bacteria binding.
Table 4-3 Data fitting for bacteria detection with and without antibodies modification.
CPEdl-T CPEdl-p Rsol CPEsol-T CPEsol-p
With antibodies
before
detection 8.7701×10
-9 0.72793 68620 1.6052×10-10 0.91754
After
detection 2.0848×10
-8 0.6613 44204 2.3348×10-10 0.8948 Without antibodies
before
detection 1.2653×10
-8 0.76104 94521 9.4923×10
-11 0.99394
After
detection 1.1471×10
-8 0.76894 07305 9.5325×10-11 0.99378
72
Figure 4-28 Impedance spectra of each element (a)impedance of CPE on electrode surface (b) impedance of solution(c) impedance of CPE in the solution.
(a) (b)
(c)
73
4-3-3 Impedance analysis of various concentrations of Vibrio Parahaemolyticus
The study also utilized the functionalized chip, the electrode A, detect the Vibrio
Parahaemolyticus in concentration ranging from 10
5 to 107cfu/mL, binding on inner electrode. The analysis was achieved in a comparison based on bacteria manipulation with AC electrokinetics without AC electrokinetics(control).Figure 4-29(a), Figure 4-30(a), and Figure 4-31(a) stated the detection of Vibrio
Parahaemolyticus at a concentration ranging from 10
5cfu/mL to 107cfu/mL. Figure 4-29(b)(c), Figure 4-30(b)(c), and Figure 4-31(b)(c) shows different concentration of the bacterial cells before and after binding with AC electrokinetics enhancement. Less concentration of bacteria means less bacteria binding on the electrode surface. In the same way, the immobilized bacteria have effect of the impedance change on the electrode surface at low frequency. Its changes indicate a negative value of Normalized impedance change (NIC) due to the decrease of the magnitude of impedance as well. Furthermore, the variation of the Normalized impedance change (NIC) is proportional to number of the bacterial reduction. Figure 4-32 thus depicts the impedance change for different number of Vibrio Parahaemolyticus at a fixed frequency of 50 Hz. The limitation of this system for low concentration of bacteria is greater than 105 cfu/mL. However, the infection dose of Vibrio Parahaemolyticus giving rise to symptoms of illness is at 106 cfu/mL.74
at 107 cfu/mL concentration
(a)
Figure 4-29 Optical images of Vibrio Parahaemolyticus at a concentration of
107cfu/mL (a) NIC spectra for binding with and without AC electrokinetics enhancement (b) optical microscopy before binding of bacteria (c) after binding with AC electrokinetics enhancement.
75
at 106 cfu/mL concentration
Figure 4-30 Optical images of Vibrio Parahaemolyticus at a concentration of 106cfu/mL (a) NIC spectra for binding with and without AC
electrokinetics enhancement (b) optical microscopy before binding of bacteria (c) after binding with AC electrokinetics enhancement.
76
at 105 cfu/mL concentration
Figure 4-31 Optical images of Vibrio Parahaemolyticus at a concentration of 105cfu/mL (a) NIC spectra for binding with and without AC
Impedance change at 50Hz for various concentrations
Normalized Impedance Change
106 cfu/mL
Figure 4-32 Impedance change of various concentrations of Vibrio
Parahaemolyticus at 50 Hz
77
4-3-4 Selective detection between Vibrio Parahaemolyticus and XL1 blue E. coli
Bacteria detection has to be distinguishable in clinical diagnosis whereas the pathogenic bacteria always go along with many other non-pathogenic bacterial cells.
In this experiments, ELISA method was utilized to perform the Vibrio
Parahaemolyticus recognition by its specific antibodies. The assay also introduced
XL1 blue E coli strain as nonpathogenic bacteria in the system. The indirect type ELISA was thus carried out. Two sorts of bacteria were incubated in plate separately.Figure 4-33 showed the TMB chromogenic result by using different dilution of concentration of antibodies immobilized in the plate. The darker color changes obviously in the columns (rightside) with Vibrio Parahaemolyticus differ from the columns with E coli (Figure 4-33(a)). The recognition of the antibodies is clearly specific for Vibrio Parahaemolyticus, enable the distinguish detection of bacteria selectively. The absorbency for each plate well measured at 450nm decrease progressively depends on concentration of antibodies, shown in Figure 4-33(b).
Figure 4-34 is including the negative and positive control.
Figure 4-33 ELISA chromogenic result for XL blue E coli and Vibrio
Parahaemolyticus detection with Vibrio Parahaemolyticus-specific
antibodies (a) color change from TMB development (b) absorbency78
Figure 4-34 Negative/positive Control for ELISA.
The selective detection was further demonstrated by two assay systems. First, the mixed solution with Vibrio Parahaemolyticus and XL blue E coli at the same concentration of 5107 cfu/mL was measured after bacteria binding with AC electrokinetics enhancement. The operational parameters for AC electrokinetics rely on Vibrio Parahaemolyticus trapping. Second, observed the performance of the antibodies’ recognition, only XL blue E coli at concentration of 108 cfu/mL was trapped and measured.
The impedance spectra of mixed solution were shown in Figure 4-35. Bacteria binding can be also observed by optical microscopy in Figure 4-36. The shape of XL1 blue E coli and Vibrio Parahaemolyticus cannot tell difference apparently, so the AC electrokinetics for trapping may force on both bacteria. The measurement after bacteria binding with AC electrokinetics(Figure 4-35 (a)) has different scale compared to the binding without AC elctrokinetics(Figure 4-35(b)). The NIC for mixed solution is shown in Figure 4-37. However, the detections of E coli individually with and with AC electrokinetics depicted in Figure 4-38 vary no difference after bacteria binding and wash. The NIC of E coli measurement shown in Figure 4-40 also report the same drift. In Figure 4-39, few E coli attach to the electrode firmly due to DEP force even washed by buffer. The polarized force induced by DEP bring bacteria to touch the
79
edge of electrodes(Figure 4-39(b)). All the selective detection of bacteria including
Vibrio Parahaemolyticus, E. coli, and Vibrio Parahaemolyticus+ E. coli compared
together with impedance change at 50 Hz in Figure 4-41. The selectivity was particularly for Vibrio Parahaemolyticus, which show a slightly larger measurement relative to XL1 Blue E coli100 1k 10k 100k
of 5x107 cfu/mL nonpathogeinc E. coli and 5x107 cfu/mLVibrio parahaemolyticus
Impedance ()
Antibody-modified electrode
Selective detection with AC electrokinetics enhancement
electrode diameter: 500um
Selective detection without AC electrokinetics enhancement
electrode diameter: 500um
at concentration
of 5x107 cfu/mL nonpathogeinc E. coli and 5x107 cfu/mLVibrio parahaemolyticus
Figure 4-35 Impedance spectra (a)with AC electrokinetics (b) without AC electrokinetics.
Figure 4-36 Optical images of Vibrio Parahaemolyticus (5*107cfu/mL) with XL1 blue E. coli (5*107cfu/mL) binding on electrode with AC electrokinetics enhancement. (a)before binding of bacteria (b) after binding.
80
5x107cfu/mL nonpathogenic E. coli and 5x107cfu/mLVibrio parahaemolyticus
Figure 4-37 NIC spectra of Vibrio Parahaemolyticus (7*105cfu/mL) with XL1 blue
E. coli (7*10
5cfu/mL) binding.100 1k 10k 100k
at 108 cfu/mL concentration of nonpathogeinc E. coli
at 5x107 cfu/mL concentration of nonpathogeinc E. coli
Figure 4-38 Impedance spectra (a)with AC electrokinetics (b) without AC electrokinetics.
Figure 4-39 Optical images of XL1 blue E. coli(108cfu/mL) binding on electrode with AC electrokinetics enhancement. (a)before binding of bacteria (b) after binding.
81
at 108 cfu/mL concentration of nonpathogenic E. coli
Figure 4-40 NIC spectra of XL1 blue E. coli (108cfu/mL) binding.
Figure 4-41 NIC related to two bacterial species at 50 Hz.
Appling the impedance spectra to the equivalent circuit model, the fitting data of three elements is collected by Table 4-4. Measurement of the Vibrio
Parahaemolyticus binding in the presence of Ecoli change surface capacitance
(CPEdl-T) as usual after bacteria binding both, 2.2 times of value from surface capacitance before binding, however, the chip for another detection of E coli gives a change of 1.1 times in CPEdl-T. The impedance spectra of each element for selective detection are shown in Figure 4-42 and Figure 4-43. For detection the magnitude of82
electrode surface capacitance demonstrates a change, relative to detection of E coli, before and after bacteria binding at low frequency.
Table 4-4 Data fitting for selective detection.
CPEdl-T CPEdl-p Rsol CPEsol-T CPEsol-p Vibrio parahaemolyticus 10
8
cfu/mL before detection 8.7701×10
-9 0.72793 68620 1.6052×10-10 0.91754
After detection 2.0848×10
-8 0.6613 44204 2.3348×10
-10 0.8948
Vibrio parahaemolyticus 5×10
7
cfu/mL+ nonpathogenic E. coli 5×10
7
-9 0.78692 86021 1.0268×10-10 0.98189
83
Figure 4-42 Impedance of each element (a)impedance of CPE on electrode surface (b) impedance of solution(c) impedance of CPE in the solution for
V.P.+E coli detection.
84
Figure 4-43 Impedance of each element (a)impedance of CPE on electrode surface (b) impedance of solution(c) impedance of CPE in the solution for E