Subscriber access provided by The University of Texas at El Paso (UTEP)
Article
Resonance Frequency Modulation for Rapid Point-of-care Ebola Glycoprotein Diagnosis with a Graphene-based Field-effect Bio-transistor Arnab Maity, Xiaoyu Sui, Bing Jin, Haihui Pu, Kai Bottum, Xingkang Huang, Jingbo Chang, Guihua Zhou, Ganhua Lu, and Junhong Chen Anal. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.analchem.8b03226 • Publication Date (Web): 06 Nov 2018 Downloaded from http://pubs.acs.org on November 7, 2018
Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.
Page 1 of 10 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Analytical Chemistry
Resonance Frequency Modulation for Rapid Point-of-care Ebola Glycoprotein Diagnosis with a Graphene-based Field-effect Biotransistor Arnab Maity, ‡ Xiaoyu Sui, ‡ Bing Jin, ‡ Haihui Pu, Kai J. Bottum, Xingkang Huang, Jingbo Chang, Guihua Zhou, Ganhua Lu, Junhong Chen* Department of Mechanical Engineering, University of Wisconsin-Milwaukee, 3200 North Cramer Street, Milwaukee, WI 53211, USA ABSTRACT: Recent outbreaks of the Ebola virus infection in several countries demand a rapid point-of-care (POC) detection strategy. This paper reports on an innovative pathway founded on electronic resonance frequency modulation to detect Ebola glycoprotein (GP), based on carrier injection-trapping-release-transfer mechanism and standard antibody-antigen interaction principle within a dielectric-gated reduced graphene oxide (rGO) field-effect transistor (GFET). The sensitivity of the current Ebola detection can be significantly enhanced by monitoring the device’s electronic resonance frequency, such as inflection frequency (fi), where phase angle reaches maximum (θmax). In addition to excellent selectivity, a sensitivity of ~36-160 % and ~17-40 % for 0.001-3.401 mg/L Ebola GP can be achieved at high and low inflection resonance frequencies, respectively, which are several orders of magnitude higher than the sensitivity from other electronic parameters (e.g., resistance-based sensitivity). Using equivalent circuit modelling on contributions of channel and contact, analytical equations for resonance shift have been generalized. When matching with the incoming ac measurement signal, electronic resonance from the phase angle spectrum is evolved from various relaxation processes (e.g., trap and release of injected charge at surface trap sites of channel-gate oxide and channel-source/drain interface) that are associated with a characteristic emission frequency. Using charge relaxation dynamics, a high-performance bio-FET sensing platform for healthcare and bio-electronic applications is realized through resonance shifting.
1. Introduction The Ebola epidemic commenced in west Africa (Guinea, Liberia, Nigeria, Senegal, and Sierra Leone) during December 2013, and the World Health Organization (WHO) declared it to be a “public health emergency of international concern”. 1 The number of confirmed and possible cases of Ebola Virus Disease is at 28,603, with 11,301 resulting in death. As per recent outbreak in Congo, WHO reported 39 suspected, probable or confirmed cases of Ebola between April 4 -May 13, 2018, including 19 deaths. The Ebola virus is commonly spread through contact of bodily fluids. The outbreak can only be stopped by a combination of early diagnosis, patient segregation with effective care, and safe burial. 2 Ebola virus infection and confirmation are generally made using the following investigations: antibody-capture, enzyme-linked immunosorbent assay (ELISA), antigen-capture detection tests, serum neutralization tests, reverse transcriptase polymerase chain reaction (RT-PCR) assay, electron microscopy, and virus isolation by cell culture.1 Although widely used for clinical diagnosis, these methods do not allow for quick, easy, or effective onsite point-of-care (POC) detection; e.g., ELISA is limited by its low sensitivity, whereas RT-PCR results can vary with external environment, temperature, and sample preparation. 3-5 Several bio-assays regarding Ebola detection can be found in the current literature, and most of the reported works are based on the antigen−antibody reaction through photo/electro-luminescence, nanoparticle plasmonics -based technology, interferometry imaging, and fluorescence
resonance energy transfer. 6-9 However, most of these methods are inadequate for low cost and miniaturized POC detection because they require expensive equipment, considerable amount of time, and trained professionals. To terminate the spreading of the deadly Ebola virus before an epidemic occurs, an effective, rapid, reliable, and user-friendly sensing platform for onsite detection of the Ebola virus is greatly needed. In recent years, the use of label-free biosensors has emerged as a technique for detecting various infectious diseases and pathogens. 10-19, 48-50 Such biosensors require suitable biomarkers or probes to attract targeted species, as well as a variety of resonance-based transduction mechanisms such as surface plasmon resonance (SPR), 15 mechanical resonance, 16 optical resonance, 17 and acoustic resonance. 18,19 There are several advantages and disadvantages to these approaches. For example, optical sensors can be operated without any physical interaction between detected species with the light source and the detection sensor. In addition, they are independent of the solution’s pH, ionic strength and other properties. 20 In the presence of test species, the optical signal can be transformed into the electrical analogue, from which the sensitivity can be correlated with changes in resonance frequency shifting or intensity quenching. Yanik et al. demonstrated a successful pseudo-typed Ebola detection using the optical resonance method. 17 Optical sensors can also be unfavourable, as they require high-precision light alignment to capture the successful pairing of probes and target species, expensive hardware for
ACS Paragon Plus Environment
Analytical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
measurements and calculations, and have other drawbacks (such as uncertainty in measurements, conformational changes and scattering) that could make them unsuitable for point-ofcare diagnosis. 21,45 Mechanical and acoustic sensors are largely based on quartz crystal microbalance (QCM), cantilever, and surface acoustic wave (SAW) sensors. 18,19 The fundamental mechanism for these types of sensors is based upon monitoring the mechanical resonant frequency shift as the probes-modified cantilevers or quartz crystals interact with biological species. This interaction occurs through a specific binding event by bending the cantilever or a change in the mass of the system; however, mass resolution of this type of sensors is insufficient when operated under solution. 22 These sensors are also sensitive to external noise and thermal drift, which is another significant problem for practical implementation in POC diagnosis. Field-effect transistors (FET) are an excellent choice, as they can be operated by simple resistance measurements. Many interesting reports can be found for graphene-based FET sensors (GFET) for biomolecule and heavy metal detection due to the monoatomic 2D layer structure, high specific surface area, and low electronic noise 23-28 The sensing mechanism is based on the change in the resistance or current of the probemodified transistor channel. This is accomplished by inducing short-range electrostatic fields when interaction with specifically targeted species occurs through a known biorecognition procedure. When the transistor channel is passivated with a dielectric oxide top gate, possible influences from medium and other conducting species can be segregated from the measurement electrodes. 46 This important feature gives insulated, gated FET-based sensors a significant advantage over the non-insulated FET structure; however, there are several technical challenges (e.g., lower sensitivity and moderate selectivity due to very weak, short-range electrostatic force) that still need to be resolved for this type of sensing platform to be used reliably as a POC diagnostic tool. 29-31 For these reasons, dielectric-gated FET structures need to be investigated with proper understanding of the device, its circuit components that are directly measured, and its impact on the Ebola sensing. In this report, we have discovered that the device’s electronic resonance frequency can be exploited as a promising transducing mechanism for enhancing the sensitivity and selectivity of Ebola Glycoprotein (GP). Unlike optical spectralresonance-based techniques, it does not need highly sophisticated instruments to capture the successful pairing of antigen-antibody that causes a resonance shift. 17, 45 This is a potential problem for optical resonance to obtain a reasonable resonance wavelength shift from a very low Ebola GP concentration. 17 Our electronic resonance-based system can realize a faster (~1-2 min) 43 and stable detection avoiding external noise or disturbance, which is a significant challenge for other resonance-based techniques, as previously discussed. We conducted the wide frequency range impedance measurements to evaluate the various circuit components such as contact and channel impedance in terms of the pair of RC components and electronic resonance frequencies to investigate their influences on the sensing performance. For an RC circuit, resonance mainly comes from the characteristic frequency (fc, where the imaginary part of impedance is maximum and related to the RC time constant) and inflection frequency (fi, where the phase angle reaches maximum (θmax), i.e., dθ/dω= 0, where θ and ω are phase angle and angular frequency, respectively). For
Page 2 of 10
insulated gated FET structure, the frequency-dependent channel impedance is governed by charge carrier trapping-release at interface states between the channel - gate oxide and channelsource/drain contacts. 36, 51 Each of these influences are designated by a certain relaxation process with specific time constant and can be represented by RC pair as an electronic equivalent. Through a variable frequency input (impedance spectroscopy), these relaxation processes can be identified by monitoring the local maxima (resonance peak) of the respective spectrum. The effect of trapped charge carriers on MOSFET electronics is important as they control the transport properties of the device by creation of an internal field. 52 The trap charge internal field further controls the charge injection properties from source-drain and can show hysteresis due to the faster rate of bias voltage sweeping. 53-55 The sweep rate of bias voltage is important as it controls the charging/discharging time scale. For a conventional dc measurement (ω~0), this trapping-release time is faster (τtrap ~1 μs to τdetrap ~250 μs) 55; however, it is slower than the high frequency impedance measurement. Therefore, it is expected that channel could show a frequency dependent behaviour as the trapped charges may not release during this high frequency ac measurement. In other words, the relaxation of trapped charge should show some frequency dependent behavior that directly affects the device conductance which can be identified by the measured phase shift. Practically, phase shift 0˚ signifies an ideal resistive system; whereas, it transits to capacitive behavior when phase angle starts to increase ( 90˚ for perfect capacitive/insulating system). When charges are trapped with a significant delay in emission as compared with the measurement frequency, the channel shows an insulating (capacitive) effect by increasing the phase shift. Therefore, it is interesting to explore how this trapped charge relaxation dynamics, identified through resonance phase shift, is affected by top gate induced field through Ebola antigenantibody interaction and how it impacts on various electronic parameters that are used for measuring the sensitivity of the GFET device. 2. Results and Discussions 2.1 Fabrication of GFETs, Immobilization of Antibodies, and Experimental Setup A self-assembly electrostatic method was used to deposit single-layer commercial Graphene oxide (GO) sheets on the Cysteamine (AET) functionalized gold electroded silicon wafer.46-47 A rapid annealing for 10 min at 400 °C with flowing Ar gas was used to reduce the GO to reduced Graphene oxide (rGO) to lower the contact resistance between the rGO sheet and gold electrodes. Next, the fabricated device was transferred to an atomic layer deposition (ALD) chamber for 3 nm thick Al2O3 deposition followed by deposition of gold nanoparticles (NPs) in a sputtering chamber. After that AET solution was dropped on the sensor surface to functionalize gold NPs with thiol groups through (-SH) -Au NP interactions for 30 min. After cleaning with pure water, glutaraldehyde solution (5%, diluted with ultra-pure water) was drop-casted on AET functionalized sensor surface for 30 min. Again, after washing with pure water, Ebola antibodies (KZ 52, IBT Bio services, USA) diluted with 0.01 M PBS buffer (1 M, pH~7.4, Sigma Aldrich) were immobilized on the sensor surface through the functionalized linker molecules glutaraldehyde for 30 min. Finally, the antibody functionalized device was cleaned with ultra-pure water and Tween 20 (0.1% Super Block Tween 20 in PBS) solution was drop-casted on the sensor to avoid the non-
ACS Paragon Plus Environment
Page 3 of 10
specific binding of analytes. After 30 min, the device was cleaned with ultrapure water, dried with N2 gas and ready for
Figure 1. Schematic diagram of (a) insulator gated FET based Ebola sensor device, (b) charge injection-trapping-release-transfer mechanism at channel-oxide and channel-electrode interfaces. (c) Trapping and release time constants (or relaxation frequency) correspond to different activation energy which can vary due to the varying position of the charge trap inside the oxide from the channel. In AC measurement, phase angle vs. frequency spectrum is a superposition of possible relaxation processes with specific resonance centre (frequency where phase angle becomes maximum). Resonance frequency shift occurs when top gate field is generated by antigen-antibody interactions.
high magnifications. It is clearly seen that single-layer rGO channels are formed between gold electrodes without further flocculation. This suggests a high quality of the device, which resembles to perfect 2D nano-materials based measurement platform. Figure 2 c shows the atomic force microscopy (AFM) (a)
(b)
10 µm
2 µm (d)
3 2 1 0 -1 0
Ids (A)
0.325
500
0.320
30 20 10 0 -10 -20 -30
Ids (A)
(c)
1.1 nm
use. Here, actual Ebola virus could not be used for sensing due to bio-safety issues, the glycoprotein was used because it has the equivalent chemical and biological signature of the surface of Ebola’s viral membrane. Further details about experiment and fabrication can be found in Supporting Information 1. 2.2 Device configuration and sensing mechanism Figure 1a shows the device configuration and circuit diagram of the FET sensor. The Ebola glycoprotein detection was accomplished using standard antigen-antibody conjugation principles.41 Further details can be found in Supporting Information 1. Figure 1b shows schematic configuration of the GFET device to realize the charge transport using ac signal inside the rGO channel that involves injection, trapping-release at channel-oxide and channel-electrode interface. Due to the variation of the position of the charge trap inside the oxide from the channel, trapping-release time constants can vary. Due to this reason, each trap corresponds to different relaxation frequency in the presence of a wide frequency measurement range of ac signal. Practically, the phase angle is near zero at low frequency; however, systematically increases when the measurement frequency is swept from low to high until the resonance occurs. The phase shift increases as the trappingrelease time of charge carrier seems slower in comparison with high frequency AC measurements through an interface capacitance effect. Due to the distribution of relaxation time, the measured phase angle vs. frequency spectrum, which is a superposition of these possible relaxation processes, becomes broad with a specific local maxima that represents the resonance center (Figure 1c). Due to generation of top gate electric field by specific interaction from Ebola antigen-antibody, the charge carrier concentration inside the channel is modulated and the associated relaxation processes influence the electronic parameters (channel and contact impedance) by shifting the resonance frequency position. Using equivalent circuit modelling the resonance frequency shift has been correlated with channel-contact impedance change to generalize analytical equations and is discussed in details later. It is to be noted that interface trap can also form at the bottom gate dielectric (SiO2); however, sensing event mainly occurs on the top gate oxide (Al2O3). Therefore, influence from the back-gate is not considered and assumed as a background. 2.3 Microstructure and electrical characterization of the sensor device Figure 2 shows a field-emission scanning electron microscopy (FESEM) image of the single-layer rGO with (a) low and (b)
Topography (nm)
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Analytical Chemistry
-0.8 -0.4
0.0 0.4 Vds(volt)
0.8
Distance (nm) 1000
0.315
-20
0 Vgs(volt)
20
Figure 2. FESEM image of rGO nanosheets distributed on the electrode in (a) low and (b) high magnifications. (c) AFM image of a single layer rGO nanosheet with the thickness profile shown in the image (distance scale (nm) is same for both image). (d) Ids-Vgs curve for Vgs = -40 to 40 V with scanning step =0.2 V for the rGO based sensor. Ids-Vds characteristics for Vds= -1 to 1 V is shown in the inset.
image of a single layer rGO nanosheet bridging two electrodes. The thickness profile is shown in the inset of Figure 2c, which suggests a single layer of rGO nanosheet (~1 nm). To study the FET property, the drain current (Ids) was measured as a function of varying gate voltage (Vgs) on the backside of the silicon wafer (Figure 2d). The measurement was made from -40 V to 40 V (scanning step = 0.2 V), which resulted in a p-type smooth FET curve with an on-off current ratio ~1.29. A linear Ids-Vds measurement curve for drain-source voltage (Vds) ranging from -1 to 1 V suggests ohmic contact between the rGO nanosheet and gold electrodes (shown in the inset of Figure 2d). 2. 4 Ebola glycoprotein sensing measurement for electronic parameter evaluation After the sensor device is fabricated, a sensor chip cell is connected to the impedance measurement system with its drainsource interface. Although live Ebola virus could not be used
ACS Paragon Plus Environment
Analytical Chemistry due to biosafety issues, the glycoprotein, which has the equivalent chemical and biological signature of the external surface of Ebola viral membrane, was procured (see Figure 1a). A small amplitude sinusoidal voltage (0.01 V) was applied at the drain-source interface and the impedance spectra were recorded in air.
(a)
100
Z" (kohm)
80 60 40
Air Buffer 0.001 mg/L 0.007 mg/L 0.051 mg/L 0.410 mg/L 3.401 mg/L
(b)
4
Z" (kohm)
120
3 2 1 0 0.0
0.5 1.0 Z' (kohm)
20
(c)
114 112 110 108 0.8 0.6 0.4 0.2 1E-3
20 40 60 80 100 120 Z' (kohm)
R2 R1
0.01 0.1 1 Concentration (mg/L)
(d) Capacitance (nF)
00
Resistance (k)
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
8 7
C2
4.0
C1
3.6 3.2
1E-3 0.01 0.1 1 Concentration (mg/L)
Figure 3. (a) Nyquist plot for the sensors measured in air, buffer, and various Ebola protein concentrations; (b) The equivalent circuit for the evaluation of circuit parameters; Plots of calculated (c) resistances and (d) capacitances from RC circuit fitting for buffer and various Ebola glycoprotein concentrations.
Sequentially, the buffer (volume ~ 2.5 µl) was injected inside the cell for producing a baseline, and the impedance spectra were recorded again. Different concentrations of Ebola glycoprotein solutions (0.001- 3.401 mg/L, volume ~2.5 µl) were then injected, and similar impedance measurements were consecutively performed for each concentration. Figure 3a shows the Nyquist plot for the sensor in air, buffer, and Ebola glycoprotein test solutions with various concentrations. From the plot, it is apparent that only one semi-circle is found in air; however, in the presence of buffer and test solution, a new semicircle at a high frequency arises (shown in the inset). Figure 3b shows the equivalent circuit, which is a series combination of two RC parallel networks (R1C1 and R2C2), with a series combination of external resistance (R3, from external measurement wires and cables). All the data were fitted to calculate the resistance and capacitance change in buffer and various Ebola glycoprotein solutions using the LevenburgMarquardt algorithm. Here R1C1 and R2C2 represent contact and channel resistance- capacitance, respectively. From the Nyquist plot, it appears that low-frequency semi-circles in buffer and protein solutions are suppressed. The suppressed semi-circle signifies the presence of an imperfect capacitive element in the system and is generally replaced by the constant phase element (CPE), which arises due to the presence of the double-layer capacitor. 34,35 Therefore, the capacitance (C2) is replaced with CPE during the fitting. The details of the double layer capacitance formation and its influence on the sensing mechanism are discussed later. The experimental (dotted) and fitted (line) curve in buffer, 0.001 mg/L and 3.401 mg/L protein solutions, are shown in Supporting Information 1 Figure 1 a c, respectively. The extracted parameters for all cases are shown in Table S1 (Supporting Information 2). It was found that measured channel resistances (R2) are higher than contact
Page 4 of 10
resistances (R1), and change linearly with increasing protein concentrations (Figure 3c). The value of external resistance R3 (comes from external wires or cables) remains constant and is not affected by the sensing performance. The calculated values of C1 and C2 in buffer and various protein concentrations are also shown in Figure 3d. C2 was derived from CPE parameters, e.g., the pre-exponential factor (Q) and exponent (n) (see Table S1 in Supporting Information 2). The unit of Q is nano-siemens (designated as nS). A systematic linear increase in C2 was found for different protein concentrations, whereas change in C1 value was not so significant. The value of n, as obtained from fitting for all cases, is close to 1, and thus CPE behaves as a capacitor. Here, the value of C2 is calculated from two different models (see Supporting Information 2). The calculated capacitance (C2) from both models are shown in Figure S2, which shows that both calculated values exhibit almost equal orders of magnitude, suggesting the validity of the calculation. It should be noted that, due to the one semi-circle in air, it is impossible to separate the relative contribution from the channel and the contact resistance in air. When the sensor is exposed to an aqueous environment (e.g., buffer), the sensing environment is switched from a lower dielectric constant medium (~1) to a higher dielectric medium (~80). This influences the second capacitance (C2), and eventually the channel resistance (R2) is separately coupled with C2. This produces a time constant (τ2=R2C2) different from the contact resistance-capacitance time constant (τ1=R1C1). Consequently, two semicircles are found in the aqueous environment. Figure 4a shows the phase angle change versus the frequency plot for all cases discussed above. The maximum peak of phase angle (θmax) in the presence of the aqueous environment and protein solutions shows two distinct peaks at the low frequency (3.2-5.5 kHz) and the high frequency (>100 kHz) range. This signifies the presence of two different time constants, as previously discussed, due to the channel and the contact RC pair, respectively. At a lower frequency range, the shifting θmax values can also be related to different concentrations. It should also be noted that the resonance frequency for θmax (inflection frequency (fil)) is also systematically shifted to a higher frequency when the concentration of the test Ebola glycoprotein solution increases sequentially (indicated by slanted upward solid arrow). Therefore, the resonance frequency shift (Δfil) at θmax could be used for calibrating the different concentrations. It is also noted that a satellite resonance peak (fih) at a high frequency due to contact influence can also be found (indicated by slanted downward dotted arrow in Figure 4a). It resembles that it is also shifted to a higher frequency when Glycoprotein is added and seems much more sensitive than the low frequency one. But this peak position is not strong enough due to large influence from low frequency resonance tale-edge and is not considered for calculating sensitivity. Details of this relation of individual resonance peak position from various circuit components and simulations are discussed later. To calculate the characteristic frequency (fc), the imaginary part of the impedance (Zʹʹ) is plotted against frequency shown in Figure 4b. A decrease in fc is observed upon injection of the Ebola glycoprotein solution and is also used for relating to the change in concentrations. Additionally, the channel resistance change (ΔR) and capacitance change (ΔC) are considered for correlating different concentrations from Figure 4a and Figure 4b, which is generally followed for various FET biosensors.23-28 Therefore, the sensitivity values based on these five different parameters (ΔR, ΔC, Δθ, Δfil, and Δfc) are compared with various Ebola
ACS Paragon Plus Environment
Page 5 of 10
glycoprotein concentrations. Figure 4c - d shows the θmax, fil, and fc versus the concentration plot. The sensitivity for each case is evaluated by the following relations. Here, R0, C0, θ0,
2
3
4
0 0 10
5
(f)
30 15 0
-3
10
-2
-1
60
Frequency (kHz)
74 1
2
3
10 10 10 Frequency (Hz)
4
10
10
-3
-2
-1
30
60
Buffer 0.1 mg/L 0.7 mg/L 5.1 mg/L
30
0.20
Characteristic frequency fc
4.6 0.18
4.4 4.2
0.16
4.0 3.8
0
10 10 10 Concentration (mg/L)
10-3
10-2
10-1
0.14
100
Concentration (mg/L)
(h)
(g)
Buffer 0.1 mg/L 0.7 mg/L 5.1 mg/L
0 0 1 2 3 4 5 10 10 10 10 10 10 Frequency (Hz)
0
10 10 10 Concentration (mg/L)
75
i_l
1
10 10 10 10 10 Frequency (Hz) (e) 60 fc f R/R0 0 C/C0 f l f0 45 i /
76
Inflection frequency f i l
4.8
40
l
0
77
Sensitivity (f i f ) %
30
20
78 -Phase (deg)
-Z'' (kohm)
40
(d)
(c)
Buffer 0.001 mg/L 0.007 mg/L 0.051 mgL 0.410 mg/L 3.401 mg/L
-Phase (Deg)
Buffer 0.001 mg/L 0.007 mg/L 0.051 mgL 0.410 mg/L 3.401 mg/L
-Phase (Deg)
-Phase (deg)
60
f0i_l, and f0c are the initial values of channel resistance capacitance, maximum phase angle (θmax), inflection resonance
(b)
(a)
Sensitivity (%)
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Analytical Chemistry
0 0 1 2 3 4 5 6 10 10 10 10 10 10 10 Frequency (Hz)
Ebola Glycoprotein Human IgG protein Ferritin
20
0
-3
10
10
-2
10
-1
10
0
Concentration (mg/L)
Figure 4. (a) Phase angle and (b) Zʹʹ vs. frequency plot for buffer and various Ebola glycoprotein concentrations for evaluating inflection (fil) and characteristic (fc) frequency, respectively. Plots of change in (c) maximum phase angle (θmax), (d) inflection frequency (fil) and characteristic frequency (fc) vs. various Ebola GP concentrations. (e) Comparison of sensitivity for all parameters. Phase angle vs. frequency plot for (f) human IgG and (g) ferritin. (h) The sensitivity vs. concentration plot for selectivity study for Ebola GP sensor.
frequency at low frequency region, and characteristic frequency at buffer solution, respectively. Rp, Cp, θp, fpi_l, and fpc are the corresponding values in the presence of the injected Ebola glycoprotein solution. Calculated sensitivity values from the changes of all these parameters are compared in Figure 4e. As found from the plot, the sensitivity related to resonance shifting is higher than the sensitivity based on other calculated parameters. Moreover, the change in fil for various Ebola glycoprotein concentrations provides the highest sensitivity among all. Therefore, change in fil is used as an optimized parameter for comparing the selectivity of the sensors. Figure 4f and Figure 4g shows the phase angle versus the frequency plot of human IgG protein and ferritin protein, respectively. It was found that there is no significant shift of the inflection resonant frequency (indicated by vertical arrow). Figure 4h shows a comparison of the sensitivity shift of fil for Ebola glycoprotein, human IgG and ferritin. This suggests our sensor has also excellent selectivity due to the very specific interaction between Ebola glycoprotein and antibody probe. Overall, this could be an interesting finding for rGO-based, impedmetric dielectric- gated FET biosensors, as they are generally limited by lower sensitivity. 29-33, 44 3. 4 Sensing Mechanism and Modelling of The ResonanceBased Sensing Performance It is known that for dielectric-oxide-gated FET-based sensors, the capacitance of the oxide layer is influenced by the external gate voltage and modifies the channel conductivity by creating an opposite charge in the channel through relation (1). 36 W
V2ds
Ids = μCox L [(Vgs - Vth)Vds - 2 ] (1) Where, Ids is the drain-source current, while Vgs, Vds, and Vth are the gate voltage, drain-source voltage, and threshold voltage, respectively. μ, Cox, W and L are the linear mobility, gate oxide capacitance, and the width and the length of the transistor
channel, respectively. Thus, the dielectric constant of the oxide layer and its thickness control the effective capacitance of the passivation layer and influence the current in the channel. It was reported in literature that electric double-layer capacitance (EDL) can be formed in deposited aluminium oxide and in the graphene channel. 37,38 The charge transfer across the rGO/Al2O3 interface creates the interface layer (CS) which is associated with trap and diffused positive charges from the channel are further diffused through the gate oxide layer form diffusion capacitance (CD). Here, CS and CD capacitors are connected in series and the gate capacitance (COX ~ C2) can be expressed as, C2 = CS.CD/(CS+CD). (2) Figure 5a shows the schematic of the equivalent circuit of the proposed model. Since, the signal is applied at coplanar configuration at source-drain, the capacitive effect mainly comes from charge trapping phenomena at rGO-Al2O3 interface as discussed before. When the sensor is exposed to the buffer, the sensing environment is switched to an aqueous environment with a dielectric constant (~80) much larger than that of air (~1). This influences the gate capacitance of the sensor under testing and might be the reason for the increased capacitance.39 This modified capacitance (C2) is then coupled with gate voltage induced modified channel trans-resistance (R2) by forming a different time constant (τ2). The contact resistance- capacitance forms a separate time constant (τ1). Due to the presence of double time constants, two separate semicircles in the Nyquist plot and two separate peaks in the phase angle versus frequency plots are observed for the buffer solutions. When various Ebola glycoprotein solutions (diluted in buffer) of different concentrations are injected systematically, the specific binding of the immobilized antibody probe and the Ebola antigen creates a localized electrostatic field from the top gate oxide layer. As found from the literature, KZ52 human antibody can interact with Ebola GP through its two subunits (GP1 and GP2).
ACS Paragon Plus Environment
Analytical Chemistry Using the protein sequence, the calculated isoelectric points of these subunits from the literature were found as 6.26 and 5.25, respectively. 42 However, during the GP-antibody interaction, it is reported that most of the GP surface, buried by 41
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 6 of 10
KZ52, belongs to GP2; 41 therefore, it is presumed that the charge state of other subunits can be an important factor for creating the field. The estimated isoelectric points of other
Figure 5. (a) The equivalent circuit of sensing platform, (b) Schematic model for phase shift relation with charge capture-release at dielectricchannel interfaces for low, high and very high frequency, (c) The electronic circuit model of resonant frequency based Ebola GP detection.
subunits of Ebola GP can vary from 5 to 12. When compared with the pH of the buffer (~7) it is hard to estimate the overall charge distribution of the protein; however, due to the presence of many amino acid groups in the protein structure, a net positive electric field can appear across the sensor surface as a gating effect and increases the channel resistance (R2) by inducing the electronic charge in the p-type rGO channel. Thus, the increased top gate potential due to this specific antigenantibody binding modifies the time constant (τ2 = R2 C2) which is faced by the incoming periodic signal when measured for a higher Ebola GP concentration, resulting in a relative change the characteristic frequency (fc =1/2πτ2). The systematic decrease in contact resistance at metal-rGO is related with the gated Schottky behaviour that is caused by transition from thermionic to field emission (tunnelling) due to increase of top gate field. 36 As the antibody in the probe is not specific to mismatching proteins (e.g., IgG and ferritin), no specific binding phenomena occurred. Due to absence of field, the negligible change for θmax and fil confirms that the sensors do not experience interference. In the phase spectrum, the observed resonance peaks at low and high frequency region signify characteristic emission frequency of the surface trap level from channel-oxide and channel to source-drain access region. As mentioned before, when the carriers are injected through source-drain, the carriers are trapped across the multiple defect/trap sites at the channeloxide interface and again emitted back to the channel. This trapping and release has some distribution of time constants that depends on the distance from the channel. Due to ac voltage perturbation the mentioned trapping-release time of the carrier is slower as compared with the high frequency impedance measurement, resulting in a capacitive effect. This may eventually create the observed phase lag between voltage and current through the interface capacitive influence. Figure 5b demonstrates typical cases for channel-oxide trapping mechanism at different ac frequencies and their effect on modulating the phase shift. Thus, measured phase angle is close to zero at a low frequency and systematically increases when frequency increases until the resonance occurs. At higher frequency (beyond resonance), it is probably so fast that the device does not feel any capacitance by minimizing the trapping and phase angle starts to reduce again. For the same reason the second resonance peak appears due to different relaxation processes or time constants with the source-drain contact influence that arises from trapped charge at channel to sourcedrain access region. 36 This relationship between resonance
frequencies and resistance parameters have been investigated to make a generalized equation for a resonance shift mechanism. Figure S6 shows the schematic of the equivalent circuit model used for the calculation of inflection resonance frequency position for (i) higher and (ii) lower frequencies, respectively. For the first case, the total impedance and phase angle can be estimated as, 𝑍=
𝑅3(1 + 𝜔2𝐶21𝑅21) + 𝑅1 ― 𝑗𝜔𝐶1𝑅21 1 + 𝜔2𝐶21𝑅21
(3) 𝜔𝐶1𝑅21
𝜃 = 𝑡𝑎𝑛 ―1[ ( ] 𝑅3 1 + 𝜔2𝐶21𝑅21) + 𝑅1
(4) 𝑑θ
Now maximum phase angle (or resonance) occurs when 𝑑𝜔 = 0 . Then the inflection resonance frequency (𝜔ℎi = 2𝜋𝑓ℎi ) at 𝜃𝑚𝑎𝑥 for the high frequency region can be estimated as, 𝜔ℎi =
1 R1C1
R1
1 + R3
(5)
Similarly, adopting the second circuit schematic of Figure S6 (ii), the resonance frequency (𝜔𝑙i = 2𝜋𝑓𝑙i) at 𝜃𝑚𝑎𝑥 for the low frequency region can be estimated as, 1
R2
ωli = R2C2 1 + R1 + R3
(6)
Here, R1, R2, and R3 are the contact, channel, and external resistance (that comes from external measurement wires and cables), respectively. To verify the relevance of the low and high frequency resonance positions with measured circuit parameters, phase angle vs. frequency for inflection frequency resonance positions was simulated using the estimated circuit parameters and equivalent circuit model (Figure S6 (i) and (ii)). Typical simulated profiles for inflection resonance frequencies at low and high frequencies for buffer (0 mg/L) and 0.001 mg/L Ebola GP are shown in Figure S7. The superposition of these two resonance contributions with the respective peak position at low and high frequencies follow the same trend as the experimental resonance peak position. From the fitting, it is evident that high frequency resonance position is more sensitive (~ 36-160 %) than low frequency resonance (~17-40 %) (See Figure S3). When compared with resistance, both resonance mechanisms showed a much higher sensitivity value than a conventional resistance-based measurement. In Figure S4, a comparison is shown for a typical 0.001 mg/L Ebola GP concentration. It is evident that, the response direction for R1 and R2 is opposite to each other. Hence the total resistance change is affected by compensative effect. Whereas, for
ACS Paragon Plus Environment
6
Page 7 of 10 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Analytical Chemistry
resonance, the contributing circuit elements in both high frequency and low frequency are not related with a simple addition, rather a much complex way. In Supporting Information 5, both of these equations are simplified. Separate contributions of each term are calculated and plotted in Figure S5 a-d for high and low frequencies. It is likely that high frequency position is more sensitive than the low frequency one due to absence of the competitive circuit term. However, the low frequency peak resonance position is so prominent that the high frequency peak is suppressed in the superposed spectrum. Therefore, low frequency resonance peak is easier to estimate the resonance shift for a practical use. In Supporting Information 7 (Table S2), comparison of Ebola sensing performance has been made between the current resonance technique and various literature reports/commercial products. Considering these results, a micro AC measurement system could be developed for point-of-care detection using advanced microcontrollers and high-speed ADCs (Figure 5c). Using a closely spaced frequency interval, these mechanisms calculate the change in resonance frequency in a confined frequency range of interest, one that excludes the complex fitting of the entire frequency measurement spectra as required for evaluating contact or channel impedance. Also, by tuning the resistance parameters through variation of fabrication process, it might be possible to separate the prominence of high frequency resonance peak position which seems more sensitive. Therefore, this could be an interesting direction for future work for FET bio-sensor research. 4. Conclusion Through a specific antigen-antibody bio-recognition process, a high-performance, rapid, electronic resonance-frequencymodulation-based, dielectric-gated rGO bio-electronic FET platform was demonstrated for Ebola detection. With good selectivity, the calculated sensitivity based on inflection resonance frequency from the phase spectrum was found to be the maximum as compared with other electronic parameters. A carrier charge injection-trapping-release relaxation process at channel-oxide and channel-electrode interfaces have been utilized to explain the phase shift. The measured impedance spectra were modelled through a pair of RC circuits, adopting the contributions from channel and contact impedance to understand the sensing mechanism for correlating resonance shift equations. In this regard, the use of advanced microcontrollers to measure these values within a specific narrow frequency range in real-time computation was proposed. It is believed that this resonance-frequency-based strategy with dielectric gated GFET system not only can be used for POC diagnosis of Ebola glycoprotein, but also can be a better alternative measurement platform for FET biosensor/bio-electronics research for military surveillance and biosafety issues in healthcare.
ASSOCIATED CONTENT Supporting Information The Supporting Information is available free of charge on the ACS Publications website. The impedance fitting data and Capacitance parameters from two different models are shown in Fig. S1 and S2. Table S1 is provided for RC circuit fitting parameters from impedance data. Fig.S3 provides the value of fih from calculated circuit elements. Fig.S4 shows comparison of calculated sensitivity from R1, R2, Rtot with ωil and ωih shifting. SI 5 shows resonant shift equations and relative contributions from various circuit components. Fig. S5 shows calculated terms from resonance frequency equations. Table S2 shows the performance comparison
for present Ebola sensor with literature reports. Figure S6 and S7 shows individual circuit term and simulated profile for low and high frequency.
AUTHOR INFORMATION Corresponding Author * E-mail:
[email protected] Author Contributions ‡ A.M., X.S and B.J contributed equally
Notes The authors declare no competing financial interest
ACKNOWLEDGMENT This work was supported by the U.S. National Science Foundation under grant number IIP-1516207. The authors also acknowledge the technical support and instrumentation facilities at the Advanced Analysis Facility of University of Wisconsin - Milwaukee.
REFERENCES (1) www.who.int/mediacentre/factsheets/fs103/en/, Ebola virus disease, accessed on November 2017. (2) L.V. Stamm, Ebola Virus Disease: Rapid Diagnosis and Timely Case Reporting are Critical to the Early Response for Outbreak Control, Am. J. Trop. Med. Hyg. 2015, 93(3), 438440. (3) N. J. Sullivan, T. W. Geisbert, J. B. Geisbert, L. Xu, Z. Yang, M. Roederer, R. A. Koup, P. B. Jahrling, G. J. Nabel, Accelerated vaccination for Ebola virus haemorrhagic fever in non-human primates, Nature 2003, 424, 681-684. (4) D. Tseng, O. Mudanyali, C. Oztoprak, S. O. Isikman, I. Sencan, O. Yaglidere, A. Ozcan, Lens free microscopy on a cellphone, Lab Chip 2010, 10, 1787-1792. (5) J. H. Henkel, S. W. Aberle, M. Kundi, T. Popow-Kraupp, Improved detection of respiratory syncytial virus in nasal aspirates by seminested RT‐PCR, J. Med.Virol. 1997, 53, 366-371. (6) Y. Li, Y. T. H. Cu, D. Luo, Multiplexed detection of pathogen DNA with DNA-based fluorescence nanobarcodes, Nat. Biotechnol. 2005, 23, 885-889. (7) A. A. Yanik, M. Huang, O. Kamohara, A. Artar, T. W. Geisbert, J. H. Connor, H. Altug, An optofluidic nanoplasmonic biosensor for direct detection of live viruses from biological media, Nano Lett. 2010, 10, 4962-4969. (8) G. G. Daaboul, C. A. Lopez, J. Chinnala, B. B. Goldberg, J. H. Connor, M. S. Unlü, Digital sensing and sizing of vesicular stomatitis virus pseudotypes in complex media: a model for Ebola and Marburg detection, ACS Nano 2014, 8, 6047-6055. (9) E. A. Jares-Erijman, T. M. Jovin, FRET imaging, Nat. Biotechnol. 2003, 21, 1387-1395. (10) L. Surbhi, L. Stephan, J. H. Naomi, Nano-optics from sensing to waveguiding, Nat. Photonics 2007, 1, 641-648. (11) B. Sandra, S. L. Carly, G. Muhammed, J. Bruce, C. Rozell, D. H. Johnson, J. H. Naomi, All-Optical Nanoscale pH Meter, Nano Lett. 2006, 6, 1687-1692. (12) J. N. Anker, W. P. Hall, O. Lyandres, N. C. Shah, J. Zhao, R. P. Van Duyne, Biosensing with plasmonic nanosensors, Nat. Mater. 2008, 7, 442-453. (13) J. Homola, Present and future of surface plasmon resonance biosensors, Anal. Bioanal. Chem. 2003, 377, 528-539. (14) S. Arnold, M. Khoshsima, I. Teraoka, S. Holler, F. Vollmer, Shift of whispering-gallery modes in microspheres by protein adsorption, Opt. Lett. 2003, 28, 272–274. (15) X. Fan, I. M. White, S. I. Shopova, H. Zhu, J. D. Suter, Y. Sun, Sensitive optical biosensors for unlabeled targets: A review, Anal. Chim. Acta. 2008, 620, 8–26.
ACS Paragon Plus Environment
7
Analytical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
(16) T. P. Burg, M. Godin, S.M. Knudsen, W. Shen, G. Carlson, J. S. Foster, K. Babcock, S. R. Manalis, Weighing of biomolecules, single cells and single nanoparticles in fluid, Nature 2007, 446 (7139), 1066–1069. (17) A. A. Yanik, M. Huang, O. Kamohara, A. Artar, T. W. Geisbert, J. H. Connor, H. Altug, An Optofluidic Nanoplasmonic Biosensor for Direct Detection of Live Viruses from Biological Media Nano Lett., 2010, 10 (12), 4962-4969. (18) K. A. Marx, Quartz Crystal Microbalance: A Useful Tool for Studying Thin Polymer Films and Complex Biomolecular Systems at the Solution−Surface Interface, Biomacromolecules, 2003, 4 (5), 1099-1120. (19) K. Länge, B. E. Rapp, Rapp, M., Surface acoustic wave biosensors: a review, Anal. Bioanal. Chem. 2008, 391, 1509– 1519. (20) M. F. Pineda, L. L.-Y. Chan, T. Kuhlenschmidt, C. J. Choi, M. Kuhlenschmidt, B. T. Cunningham, Rapid specific and label-free detection of porcine rotavirus using photonic crystal biosensors, IEEE Sens. J. 2009, 9, 470–477. (21) M. Terrel, M. J. F. Digonnet, S. Fan, Ring-coupled Mach– Zehnder interferometer optimized for sensing, Appl. Opt. 2009, 48, 4874–4879. (22) R. Datar, S. Kim, S. Jeon, P. Hesketh, S. Manalis, A. Boisen, T. Thund, Cantilever sensors: nanomechanical tools for diagnostic,s MRS Bulletin. 2009, 34, 449-454. (23) J. H. Ahn, S. J. Choi, J. W. Han, T. J. Park, S. E. Lee, Y. K. Choi, Double-Gate Nanowire Field Effect Transistor for a Biosensor, Nano Lett. 2010, 10, 2934-2938. (24) Y. Ohno, K. Maehashi, Y. Yamashiro, K. Matsumoto, Electrolyte-gated graphene field-effect transistors for detecting pH and protein adsorption, Nano Lett. 2009, 9, 3318-3322. (25) K. Chen, G. Lu, J. Chang, S. Mao, K. Yu, S. Cui, J. Chen, Hg (II) ion detection using thermally reduced graphene oxide decorated with functionalized gold nanoparticles, Anal. Chem. 2012, 84(9), 4057-4062. (26) B. Xu, M. Zhu, W. Zhang, X. Zhen, Z. Pei, Q. Xue, C. Zhi, P. Shi, Ultrathin MXene‐Micropattern‐Based Field‐Effect Transistor for Probing Neural Activity, Adv. Mater. 2016, 28(17), 3333-3339. (27) M. Hasegawa, Y. Hirayama, Y. Ohno, K. Maehashi, K. Matsumoto, Characterization of reduced graphene oxide fieldeffect transistor and its application to biosensor, Jpn. J. Appl. Phys. 2014, 53(5S1), 05FD05. (28) Y.H. Kwak, D.S. Choi, Y. N. Kim, H. Kim, D. H. Yoon, S. Ahn, W. S. Yang, S. Seo, Flexible glucose sensor using CVDgrown graphene-based field effect transistor, Biosens. Bioelectron. 2012, 37(1), 82-87. (29) Q. Yin, X. He, Huang, Zhang, S. Wu, P. Chen, G. Lu, P. Chen, Q. Zhang, Q. Yan H. Zhang, Real-time DNA detection using Pt nanoparticle-decorated reduced graphene oxide field-effect transistors, Nanoscale. 2012, 4(1), 293-297. (30) O. Knopfmacher, M. L. Hammock, A. L. Appleton, G. Schwartz, J. Mei, T. Lei, J. Pei, Z. Bao, Highly stable organic polymer field-effect transistor sensor for selective detection in the marine environment, Nat. Commun. 2014, 2954, 1-9. (31) J. W. Park, S. J. Park, O. S. Kwon, C. Lee, J. Jang, Polypyrrole Nanotube Embedded Reduced Graphene Oxide Transducer for Field-Effect Transistor-Type H2O2 Biosensor, Anal. Chem. 2014, 86(3), 1822-1828. (32) A. Bonanni, M. Pumera, Graphene platform for hairpin-DNAbased impedimetric genosensing, ACS Nano, 2011, 5, 23562361. (33) M. Pumera, A. Ambrosi, A. Bonanni, E. Lay, K. Chng. H. L. Poh, Graphene for electrochemical sensing and biosensing. Trends Anal Chem, 2010, 29, 954-965.
Page 8 of 10
(34) https://electrochemistryresources.com/the-constant-phaseelement-cpe/, Electrochemistry Resources, accessed on November 2017. (35) P. Córdoba-Torres, T. J. Mesquita, R. P. Nogueira, J. Phys. Chem. C, Relationship between the Origin of Constant-Phase Element Behavior in Electrochemical Impedance Spectroscopy and Electrode Surface Structure, 2015, 119 (8), 4136-4147. (36) E. G. Bittle, J. I. Basham, T. N. Jackson, O. D. Jurchescu,; D. J. Gundlach, Mobility overestimation due to gated contacts in organic field-effect transistors, Nature Communications, 2016, 7:10908, 1-5. (37) B. K. Min, S. K. Kim, S. J. Kim, S. H. Kim, M. A. Kang, C. Y. Park, W. Song, S. Myung, J. Lim, K. A. Seok, Electrical Double Layer Capacitance in a Graphene-embedded Al2O3 Gate Dielectric, Scientific Reports, 2015, 5:16001, 1-5. (38) B. E. Conway, Electrochemical Supercapacitors - Scientific Fundamentals, Springer Science Business Media, NY, 1999. (39) J. L. Xia, F. Chen, P. Wiktor, D. K. Ferry, N. J. Tao, Effect of top dielectric medium on gate capacitance of graphene field effect transistors: Implications in mobility measurements and sensor applications, Nano Lett. 2010, 10, 5060-5064. (40) C. Staii, A. T. Johnson, R. Shao, D. A. Bonnell, High frequency scanning gate microscopy and local memory effect of carbon nanotube transistors, Nano Lett. 2005 5, 893-896. (41) J. E. Lee, M. L. Fusco, A. J. Hessell, W. B. Oswald, D. R. Burton, E. O. Saphire, Structure of the Ebola virus glycoprotein bound to an antibody from a human survivor, Nature. 2008, 454, 177-182. (42) http://www.uniprot.org/ accessed on October 2017. (43) C. D. Ahrberg, A. Manz, P. Neužil, Palm-Sized Device for Point-of-Care Ebola Detection, Anal. Chem. 2016, 88, 48034807. (44) Y. Chen, R. Ren, H. Pu, X. Guo, J. Chang, G. Zhou, S. Mao, M. Kron, Field-effect transistor biosensor for rapid detection of Ebola antigen, J. Chen, Sci. Rep. 2017, 7, 10974,1-7. (45) C. Rosman, J. Prasad, A. Neiser, A. Henkel, J. Edgar, C. Sönnichsen, Multiplexed plasmon sensor for rapid label-free analyte detection, Nano Lett. 2013, 13, 3243-3247. (46) J. Chang, G. Zhou, X. Gao, S. Mao, S. Cui, L. E. Ocola; C. Yuan, J. Chen, Real-time detection of mercury ions in water using a reduced graphene oxide/DNA field-effect transistor with assistance of a passivation layer, Sens. Biosens. Res. 2015, 5, 97-104. (47) J. Chang, S. Mao, Y. Zhang, S. Cui, G. Zhou, X. Wu, C. H. Yang, J. Chen, Ultrasonic-assisted self-assembly of monolayer graphene oxide for rapid detection of Escherichia coli bacteria, Nanoscale, 2013, 5, 3620-3626. (48) B. Zhan, C. Li, J. Yang, G. Jenkins, W. Huang, X. Dong, Graphene Field‐Effect Transistor and Its Application for Electronic Sensing, Small, 2014, 10, 4042-4065. (49) S. Mao, J. Chang, G. Zhou, J. Chen, Nanomaterial‐enabled Rapid Detection of Water Contaminants, Small, 11, 2015, 5336-5359. (50) M. S. Makowski, Albena Ivanisevic, Molecular Analysis of Blood with Micro‐/Nanoscale Field‐Effect‐Transistor Biosensors, Small ,7, 2011, 1863-1875. (51) C. Goldmann, C. Krellner, K. P. Pernstich, S. Haas, D. J. Gundlach, B. Batlogg, Determination of the interface trap density of rubrene single-crystal field-effect transistors and comparison to the bulk trap density, J. Appl. Phys., 99, 2006, 034507. (52) L. Chua, J. Zaumseil, J. Chang, E. C. Ou, P. K. Ho, H. Sirringhaus, R. H. Friend, General observation of n-type fieldeffect behaviour in organic semiconductors, Nature 434, 2005, 194-199. (53) P. Joshi ,H. E. Romero, A. T. Neal,V. K.Toutam and S. A .Tadigadapa, ntrinsic doping and gate hysteresis in graphene
ACS Paragon Plus Environment
8
Page 9 of 10 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Analytical Chemistry
field effect devices fabricated on SiO2 substrates, J. Phys. Condens. Matter , 22, 2010 , 334214. (54) H. Wang, Y.Wu, C. Cong, J. Shang, and T.Yu, Hysteresis of Electronic Transport in Graphene Transistors, ACS Nano, 2010, 4 (12), 7221-7228.
(55) Y. Lee, C. G. Kang, U. J. Jung, J. J. Kim, H. J.Hwang, H. J. Chung, S. Seo, R. Choi, and B. H.Lee, Fast transient charging at the graphene/ interface causing hysteretic device characteristics, Appl. Phys. Lett., 2011, 98 , 183508.
ACS Paragon Plus Environment
9
Analytical Chemistry
Table of Contents Sample drop
Ebola GP Al2O3
40
-Phase angle(deg)
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 10 of 10
30 20
rGO
10 Ebola GP
AC Source
Antibody@Au
0 mg/L 0.001 mg/L 0.007 mg/L 0.051 mg/L 0.410 mg/L 3.401 mg/L
Resonance shift
0 3
10
4
10
10
5
Frequency (Hz)
10
6
rGO channel Al2O3
Source trapping
Trap site
Drain De-trapping
ACS Paragon Plus Environment
10