Biosensors and Bioelectronics 87 (2017) 447–452

Contents lists available at ScienceDirect

Biosensors and Bioelectronics

http://d0956-56

n CorrNew Ha

E-m

journal homepage: www.elsevier.com/locate/bios

Metal-coated microfluidic channels: An approach to eliminatestreaming potential effects in nano biosensors

Jieun Lee a, Mathias Wipf a, Luye Mu a, Chris Adams b, Jennifer Hannant b, Mark A. Reed a,c,n

a Department of Electrical Engineering, Yale University, New Haven, CT 06511, United Statesb QuantuMDx Group, Newcastle NE1 2JQ, United Kingdomc Department of Applied Physics, Yale University, New Haven, CT 06511, United States

a r t i c l e i n f o

Article history:Received 13 June 2016Received in revised form15 August 2016Accepted 19 August 2016Available online 24 August 2016

Keywords:Streaming potentialMicrofluidic channelSilicon nanowire biosensorsMetal thin filmMultiplexed electrical readout

x.doi.org/10.1016/j.bios.2016.08.06563/& 2016 Elsevier B.V. All rights reserved.

esponding author at: Department of Electricalven, CT 06511, United States.ail address: mark.reed@yale.edu (M.A. Reed).

a b s t r a c t

We report a method to suppress streaming potential using an Ag-coated microfluidic channel on a p-typesilicon nanowire (SiNW) array measured by a multiplexed electrical readout. The metal layer sets aconstant electrical potential along the microfluidic channel for a given reference electrode voltage re-gardless of the flow velocity. Without the Ag layer, the magnitude and sign of the surface potentialchange on the SiNW depends on the flow velocity, width of the microfluidic channel and the device'slocation inside the microfluidic channel with respect to the reference electrode. Noise analysis of theSiNW array with and without the Ag coating in the fluidic channel shows that noise frequency peaks,resulting from the operation of a piezoelectric micropump, are eliminated using the Ag layer with tworeference electrodes located at inlet and outlet. This strategy presents a simple platform to eliminate thestreaming potential and can become a powerful tool for nanoscale potentiometric biosensors.

& 2016 Elsevier B.V. All rights reserved.

1. Introduction

In the past decade, nano-scale field effect transistor (FET)-based devices, such as Si nanowires (SiNW), Si nanoribbons, car-bon nanotubes (CNT), graphene, and metal oxide nanowires (NW)have attracted significant attention in biological and chemicalsensing, due their ability to achieve label-free, real-time detectionat high sensitivity (Chen and Zhang, 2011; Cui et al., 2001; Guoet al., 2013; Smith et al., 2014; Stern et al., 2007a; Wipf et al., 2013;Zhang et al., 2011, 2015). In particular, silicon-based devices fab-ricated by top-down approaches have the advantage to be in-tegrated with readout circuitry on-chip with high reproducibilityfor mass production at low cost (Lee et al., 2015, 2013; Livi et al.,2013; Luo et al., 2009; Mu et al., 2014; Rigante et al., 2015; Sternet al., 2007a; Zhang et al., 2010). They have shown their potentialas promising biosensors to detect a wide range of charged biolo-gical species, such as DNAs and proteins with high sensitivity(Bunimovich et al., 2006; Duan et al., 2012). Real-time and label-free detection has been successfully demonstrated by detectingvariations in the surface potential upon binding of charged ana-lytes in the electric double layer at the sensor surface (Duan et al.,2012; Wipf et al., 2016).

Engineering, Yale University,

The common strategy for label-free biosensing is to immobilizethe receptors for the analyte on the sensor surface. For the de-tection of large molecules such as proteins or DNA, the Debyelength of the buffer solution needs to be considered to optimizethe sensing performance, as the charges from the target moleculesare screened outside of the electrical double layer (EDL) (Schas-foort et al., 1990; Stern et al., 2007b; Vacic et al., 2011). This leadsto a significant reduction in the sensor response. The Debye lengthis mainly affected by the ionic strength of the buffer solution. Toextend the sensitive spatial boundary of the sensor, the ionicstrength of the buffer solution can be reduced. This approach hasbeen used in several biosensing applications (Chen et al., 2011;Duan et al., 2012; Mu et al., 2015; Stern et al., 2010; Zhang andNing, 2012). However, reduced ionic strength can introduce ad-ditional noise as the stability of the solution's electrical potential isaffected in electrolytes with low conductivity. In particular, po-tentiometric biosensors using microfluidics and flow systems foranalyte detection are affected by streaming potential (Kim et al.,2009). In reference (Wipf et al., 2016) microfluidic channels hadlittle effect on the signal, but significantly increased the noise inprotein detection measurements in low ionic strength buffer ascompared to a larger fluidic system.

The EDL has a characteristic potential distribution given by thesurface material, as well as the ionic composition and pH of theelectrolyte. The Gouy-Chapman-Stern model is commonly used todescribe the potential and charge distribution in EDLs (vanHalet al., 1996). In short, surface charges attract counter ions from the

Fig. 1. Schematics of the electrical double layer at a SiO2 surface. At pH values above the isoelectric point of SiO2, the surface is negatively charged. Counter ions at the surfacebuild the Stern layer but cannot screen the surface charge completely, leading to an excess of counter ions (cations) in the diffuse layer. A characteristic potential drop fromthe surface to the bulk solution (Ψbulk) over the Debye length defines the surface potential (Ψs). The zeta potential (ζ) is defined at the slipping plane. Under flow (indicated byν) the ions in the diffuse layer are dragged along the flow direction, leading to an excess of cations downstream. As a result, an electric field (E) builds up along the streamcausing differences in surface potential along the flow direction. Vstr increases with distance from the reference electrode (x) and has the opposite value in the upstreamdirection.

J. Lee et al. / Biosensors and Bioelectronics 87 (2017) 447–452448

solution. The first immobile layer of counter ions, called the Sternlayer is followed by the diffuse layer as shown in Fig. 1. The Zetapotential (ζ) is defined as the potential difference between thebulk electrolyte and the slipping plane that separates mobile fluidfrom fluid that remains attached to the surface. The streamingpotential (Vstr) in a microfluidic channel is generated in the EDL ofa pressure driven flow (Kirby and Hasselbrink, 2004; Mansouriet al., 2005; Vanwagenen and Andrade 1980). The flowing solutioncarries excess counterions in the diffuse layer downstream, whichleads to a longitudinal electric field along the flow direction. Vstr isdefined as the potential difference between two points along theflow direction of the system and is usually given by their pressuredifference ΔP which is proportional to the flow rate (Q) and thefluidic channel geometry, and hence defines the local flow velocity(ν) for a laminar system.

ζση

Δ =ϵ ϵ

Δ( )

V P1str

r 0

Here, εr is the relative permittivity of the electrolyte solution, ε0is the vacuum permittivity (F m�1). ζ is the zeta potential of theEDL (V), s is the electrical conductivity of the bulk solution(S m�1) and η is the dynamic viscosity of the solution (Pa s). ForSiO2, which isoelectric point is around pH 2–3, the surface is ne-gatively charged in an electrolyte of higher pH, resulting in a dragof cations along the flow direction. As ζ and s are directly affectedby the ionic strength of the electrolyte solution, Vstr is larger at lowionic strength (Kim et al., 2009). Potentiometric sensors oftendetect analytes in low ionic strength buffers (De et al., 2013; Duanet al., 2012; Stern et al., 2007b). In such conditions fluctuations inthe flow velocity can affect the local surface potential and henceintroduce noise. Changes in conductance of a FET-based biosensorssuch as SiNWs, CNT, and InAs NWs have been reported when astreaming potential was generated by a fluidic flow in a micro-fluidic channel (Bourlon et al., 2007; Chen et al., 2013b; Kim et al.,2009; Newaz et al., 2012). This results in an undesirable change insurface potential of biosensors, interfering with the actual signalfrom the charged target molecules.

In this work we present a simple method to minimize streamingpotential and thereby reduce its effects when measuring in electro-lyte solutions with ultra-low ionic strength. Using a thin-film metalcoating in the microfluidic channel, we effectively increase its con-ductivity. This results in a significant reduction of the local Vstrthroughout the microchannel without affecting the sensor proper-ties. Moreover, we demonstrate the reduction of noise introduced byincontinuous pumping by up to four orders of magnitude.

2. Material and methods

2.1. Measurement setup

We developed a multiplexed detection setup to concurrentlymeasure up to 32 devices. Each amplification channel consists of aquad low noise operational amplifier (LT1125) configured as acurrent-to-voltage converter which feeds into a 32 channel mul-tiplexer (MUX) (ADG732BSUZ) as depicted in Fig. 2a. The resistor(R¼1 MΩ) was chosen to match typical drain currents of �100 nAsuch that the gain would be 106 (0.1 mA of input current corres-ponds to 0.1 V output voltage). The capacitor was chosen for lowpass filtering with a cutoff frequency at 1.59 Hz. In order to im-prove the portability and the noise performance of the measure-ment setup, the amplification stage was combined with the MUXin a custom designed printed circuit board (PCB) as shown inFig. 2b. The SiNW biosensor chip was wire-bonded and insertedinto a 68-pin leadless chip carrier (LCC) on the PCB. The board wasconnected to a portable USB National Instruments Data Acquisi-tion (NI USB DAQ) Card via a 68-pin VHDCI connector from NI. Acustom Labview software was written to interface and extract datafrom the 32 channel concurrently. The DAQ supplies the drainvoltage (VD, typically �0.1 V) and the gate voltage (VG, typically0 to �2 V) to the biosensor chip. The source current from eachdevice is fed into the inverting terminal of the op-amp. Mea-surement data presented in this work was taken at a rate of 4 Hz/channel for real-time measurement. Deionized (DI) water (re-sistivity 17–18Ω cm) used in this work was delivered to the chipthrough 24 gauge PTFE tubes. Solution delivery was achievedusing a syringe pump (NE-1000, New Era Pump System, Inc.) or apiezoelectric micropump (mp6, Bartels). Ag/AgCl wires were in-serted into the tubes as pseudo reference electrodes in order tocontrol the electrical potential of the electrolyte solution.

2.2. SiNW fabrication

The p-type SiNW biosensors were fabricated on 8″ silicon-on-insulator (SOI) wafers (SOITech) using a top-down method as re-ported in literature (Chua et al., 2009; Knopfmacher et al., 2010;Zhang et al., 2010). For scalability and to ensure compatibility withCMOS processing, we utilized a top-down deep UV optical litho-graphy approach combined with interferometric masking and theself-limiting oxidation of silicon to form the NW arrays. The SOIwafer has typically 150 nm buried oxide with the active silicon layerordinarily around 140 nm. We used atomic layer deposition of Al2O3

as the gate dielectric layer with typical thickness below 15 nm, and

Fig. 2. Multiplexed detection setup and electrical characteristics of SiNW biosensors. (a) Schematic diagram of the multiplexed detection setup and (b) its photo image.(c) Microscope image of a SiNW biosensor consisting of five parallel wires. An opening in the protective SiO2 layer is defined to expose the SiNWs to solution. (d) ID-VG (top)and transconductance (bottom) of 13 SiNW biosensors on the same chip.

J. Lee et al. / Biosensors and Bioelectronics 87 (2017) 447–452 449

SiO2 was used as passivation layer to protect the metal contact leads.As shown in Fig. 2c, each SiNW device has 5 parallel wires for betterdevice uniformity, stability and less performance variation in com-parison with single NWs (Tian et al., 2011). The total length of thesensing area of a SiNW is 12 mm. Fig. 2d shows the transfer char-acteristics of the SiNW biosensors on the same chip measured by themultiplexed detection setup. Real-time measurement in this work

Fig. 3. SiNW surface potential changes due to changes in flow velocity in a microfluidconfiguration strongly influence the effect of the streaming potential. (a) Fabrication p1000 mm wide and 10 mm long channels). (b) Illustration of the three different measurelarge channel 1 mm). The total length of the fluidic channel is 10 mm. The SiNW channelreference electrodes at inlet and outlet. Case C: Ag-coated PDMS with potential appliedsurface potential at a given flow rate with respect to the base flow rate of 1 ml/min (ΔΨcalculated from the mean values at a given flow rate. The error bars represent the stan

was conducted in the linear region of the SiNW biosensors where thesignal to noise ratio is maximized (Rajan et al., 2011).

2.3. Noise analysis

Fluctuations in the drain current were measured at 512 Hz/channel for 3000 s using the multiplexed measurement setup. The

ic channel with two different widths. The SiNW position and reference electroderocess and the photo image for the Ag-coated PDMS microfluidic channel (100–ment configurations used. Solution flows from left to right (small channel 100 mm,No. increases from inlet to outlet. (c)–(e) Case A: reference electrode at inlet. Case B:to reference electrodes at inlet and outlet. Each inset shows the relative change inS¼ΨS�ΨS|Q¼1 ml/min) vs. the SiNW channel No. for different flow velocities. ΔΨS isdard deviation.

J. Lee et al. / Biosensors and Bioelectronics 87 (2017) 447–452450

data was translated into a current noise power spectral density(SI(f)) via fast Fourier transform, using the Matlab analysis algo-rithm pwelch which implies smaller variance in the final powerspectrum (Welch, 1967). The power spectral density of the gatereferred noise (SV(f)) was calculated by the formal relation : SV(f)¼SI(f)/gm2, where gm is the transconductance (Ghibaudo, 1989).

Fig. 4. Power spectral density of the gate referred noise vs. frequency of the SiNWsensors while DI was delivered by a piezoelectric micropump with 1 Hz operationfrequency in (a) PDMS channel with inlet reference electrode and (b) Ag-coatedPDMS channel with inlet and outlet reference electrodes. Each inset shows thesurface potential change of the SiNW sensor.

3. Results and discussion

3.1. Streaming potential measurements

Fig. 3a shows the fabrication process flow and the photo imagefor the Ag-coated PDMS microfluidic channel (see Supplementarymaterial). Fig. 3b illustrates 3 different measurement configura-tions regarding the position of the reference electrodes and thepresence of the Ag layer in the PDMS fluidic channel. The SiNWdevices are located in the center of the fluidic channel, and eachmeasured device has 600 mm spacing. We conducted real-timecurrent measurements of the SiNW biosensors as a function offlow velocity and extracted the surface potential (ΨS) of the sensorsurface by dividing the drain current by gm for each device.

Fig. 3c shows the effect of changing streaming potential (ΔVstr)on Ψs of the SiNW channels with changing flow rate when thereference electrode is placed at the inlet (Case A). Since the ma-jority of the sensor chip is covered with a SiO2 protection layer,which has an isoelectric point of �2–3, the surface is negativelycharged in DI water (pH �5–6). The pressure-driven flow dragsthe excess counterions in the EDL downstream, building a positiveVstr. The magnitude of the surface potential for each device in-creases as the flow rate gets higher since the streaming potential isproportional to the flow velocity. In the inset in Fig. 4b, the surfacepotential of the SiNWs in the small channel (W¼100 mm) in-creases linearly along the fluidic channel at fixed flow rate. Theslope of ΔΨs of the SiNWs in the wide channel (W¼1000 mm) issmaller than that of the small channel because the velocity be-comes smaller due to 10 times wider width. Consequently, theobserved surface potential change strongly suggests that theSiNWs are responding to the streaming potential change in respectto the flow rate, device location, and fluidic channel size. Theability to detect the local streaming potential enables the SiNWs toact as nanoscale flow sensors (Kim et al., 2009). In addition, theoverall behavior of the streaming potential is observed when thereference electrode is located in the outlet as shown in Fig. S1 inthe supporting information. In this case, the SiNWs are locatedupstream to the reference electrode. Hence, the surface potential isshifted negatively.

When two reference electrodes are applied at both inlet andoutlet (Case B), the streaming potential can be reduced, as re-ported by Chen et al. (2013a). However, the SiNWs are affected byeach electrode on either side, which means the electrostatic po-tential in the fluidic channel can adopt both upstream anddownstream characteristics. As a result, the downstream potentialchange can be compensated by the upstream potential with re-spect to the outlet electrode. Fig. 3d shows that the surface po-tential of the SiNWs shows smaller fluctuations with changingflow rate while both electrodes are applied. The absolute value ofthe maximumΔΨs at a flow rate of 11 ml/min is almost 4 mV in themiddle of the fluidic channel, approximately 40 times less thanthat of Case A. However, the effect of the streaming potential isstill significant. A superposition of upstream and downstreampotential could be expected along the microfluidic channel. Theasymmetric channel structure and a small offset in electro-chemical potential of the reference electrodes can explain the localdifferences in the streaming potential.

Fig. 3e shows the surface potential change of the SiNW in the

Ag-coated microfluidic channel. The results clearly show that ΔΨs

is further suppressed as compared to Case B. The Ag layer betweenthe reference electrodes increases the conductance and equalizedthe potential along the microfluidic channel. We emphasize thatthe Ag layer is not connected to the reference electrodes and itspotential is solely defined by the solution in the microfluidicchannel. The maximum fluctuation in surface potential is ap-proximately 25% of Case B and almost 100 times smaller than CaseA. The results imply that the Ag-coated microfluidic channel caneliminate the streaming potential regardless of the geometricstructure of the microfluidics and small offsets in the electro-chemical potential of the reference electrode.

3.2. Reduction of pump-induced noise

The use of micropumps is a promising approach for the reali-zation of a portable biosensor. Piezoelectric micropumps enablefluid delivery at a wide range of flow rates, by applying an ACvoltage to drive the expansion and compression strokes as thesignal changes in polarity (Iverson and Garimella, 2008). However,even though piezoelectric micropumps have the advantage ofminiaturization, low power consumption, and low cost, their op-eration principle generates large fluctuations in flow velocity. This

J. Lee et al. / Biosensors and Bioelectronics 87 (2017) 447–452 451

leads to additional noise associated with streaming potentialfluctuations when measuring at low ionic strengths. To demon-strate the advantages of the Ag-coated microfluidic channel inalleviating these effects, we use a piezoelectric micropump todeliver DI water to the SiNWs and compare the performance of thenoise reduction.

Fig. 4 shows the power spectral density of the gate referrednoise SV(f) of the different SiNW channels while DI water waspumped through the microfluidic channel by applying a 1 Hztrapezoid signal to the piezoelectric micropump controller (mp6-OEM, Bartels). The inset in Fig. 4a shows clear oscillations of thesurface potential at 1 Hz induced by the pump. Accordingly, thepower spectral density of the gate referred noise has its highestpeak at the fundamental frequency (1 Hz) and contains both evenand odd harmonics (Fig. 4a). The peak heights increase as thechanges in streaming potential become more pronounced withincreasing distance between the SiNW channel and the referenceelectrode. The maximum peak of SV(f) is 1.45�10�2 (V2/Hz) in theSiNW channel No.15 which is located furthest from the referenceelectrode as shown in Fig. 3b.

Fig. 4b shows that the use of an Ag-coated microfluidic channelsignificantly reduces the pump-induced noise. The inset indicatesthe surface potential fluctuation of the Ag-coated microfluidicchannel to be approximately 102 times smaller than that of themicrofluidic channel without the Ag layer, as the effect of thestreaming potential on the surface potential of the individualSiNW channels is eliminated. The averaged maximum peak of SV(f)at f¼1 Hz is 1.35�10�6 (V2/Hz) which is 104 times smaller ascompared to Fig. 4a. It is important to note that the demonstratedthin metal layer in the microfluidic channel can simply achieve adramatic reduction of the gate noise induced by fluctuations instreaming potential, caused by the piezoelectric micropump,without additional signal processing and/or off-chip circuitry.

4. Conclusions

In conclusion, we have demonstrated a simple yet effectivemethod to minimize streaming potential in microfluidic channelsand thereby reduce noise and superimposed signals of SiNWsensors when measuring in electrolyte solutions with ultra-lowionic strength. By depositing a thin Ag film on the microchannelsurface, the fluidic channel's conductivity is increased and therebythe electrical potential along the fluidic channel is equalized, re-sulting in an effective reduction of the local streaming potentialthroughout the microfluidic channel without affecting the sensorproperties. No surface treatment of the metal layer, such aschlorination, is needed as its potential is defined by the Ag/AgClreference electrode via the solution. Furthermore, we demon-strated the reduction of the noise introduced by a piezoelectricmicropump by up to four orders of magnitude. It should be notedthat the reduction of the streaming potential is very important forpotentiometric biosensors, as the streaming potential can easilyexceed signals from a specific analyte-receptor binding event. Thesimple approach presented in this work is an effective method foreliminating the streaming potential and may find uses as a pow-erful tool in nanoscale potentiometric biosensor applications.

Acknowledgments

This work was supported by the Basic Science Research Pro-gram through the National Research Foundation of Korea (NRF)funded by the Ministry of Education (2015R1A6A3A03020298),South Korea. L.M. acknowledges support from the Natural Sciences

and Engineering Research Council of Canada (NSERC) graduatefellowship, Canada. M.W. acknowledges support from the SwissNational Science Foundation (P2BSP2_158772), Switzerland. Par-tial support was provided by the Army Research Office underMURI (W911NF-11-1-0024), United States.

Appendix A. Supporting information

Supplementary data associated with this article can be found inthe online version at http://dx.doi.org/10.1016/j.bios.2016.08.065.

References

Bourlon, B., Wong, J., Miko, C., Forro, L., Bockrath, M., 2007. A nanoscale probe forfluidic and ionic transport. Nat. Nano 2 (2), 104–107.

Bunimovich, Y.L., Shin, Y.S., Yeo, W.S., Amori, M., Kwong, G., Heath, J.R., 2006.Quantitative real-time measurements of DNA hybridization with alkylatednonoxidized silicon nanowires in electrolyte solution. J. Am. Chem. Soc. 128(50), 16323–16331.

Chen, K.-I., Li, B.-R., Chen, Y.-T., 2011. Silicon nanowire field-effect transistor-basedbiosensors for biomedical diagnosis and cellular recording investigation. NanoToday 6 (2), 131–154.

Chen, S., Zhang, S.L., 2011. Contacting versus insulated gate electrode for Si na-noribbon field-effect sensors operating in electrolyte. Anal. Chem. 83 (24),9546–9551.

Chen, S., Xie, Y., De, A., van den Berg, A., Carlen, E.T., 2013a. Electrical shielding forsilicon nanowire biosensor in microchannels. Appl. Phys. Lett. 103 (17), 173702.

Chen, Y., Liang, D., Gao, X.P.A., Alexander, J.I.D., 2013b. Sensing and energy har-vesting of fluidic flow by InAs nanowires. Nano Lett. 13 (8), 3953–3957.

Chua, J.H., Chee, R.-E., Agarwal, A., Wong, S.M., Zhang, G.-J., 2009. Label-free elec-trical detection of cardiac biomarker with complementary metal-oxide semi-conductor-compatible silicon nanowire sensor arrays. Anal. Chem. 81 (15),6266–6271.

Cui, Y., Wei, Q., Park, H., Lieber, C.M., 2001. Nanowire nanosensors for highly sen-sitive and selective detection of biological and chemical species. Science 293(5533), 1289–1292.

De, A., van Nieuwkasteele, J., Carlen, E.T., van den Berg, A., 2013. Integrated label-free silicon nanowire sensor arrays for (bio)chemical analysis. Analyst 138 (11),3221–3229.

Duan, X., Li, Y., Rajan, N.K., Routenberg, D.A., Modis, Y., Reed, M.A., 2012. Quantifi-cation of the affinities and kinetics of protein interactions using silicon nano-wire biosensors. Nat. Nanotechnol. 7 (6), 401–407.

Ghibaudo, G., 1989. On the theory of carrier number fluctuations in MOS devices.Solid-State Electron. 32 (7), 563–565.

Guo, D., Zhuo, M., Zhang, X., Xu, C., Jiang, J., Gao, F., Wan, Q., Li, Q., Wang, T., 2013.Indium-tin-oxide thin film transistor biosensors for label-free detection ofavian influenza virus H5N1. Anal. Chim. Acta 773, 83–88.

Iverson, B.D., Garimella, S.V., 2008. Recent advances in microscale pumping tech-nologies: a review and evaluation. Microfluid. Nanofluid. 5 (2), 145–174.

Kim, D.R., Lee, C.H., Zheng, X., 2009. Probing flow velocity with silicon nanowiresensors. Nano Lett. 9 (5), 1984–1988.

Kirby, B.J., Hasselbrink, E.F., 2004. Zeta potential of microfluidic substrates: 1.Theory, experimental techniques, and effects on separations. Electrophoresis25, 187–202.

Knopfmacher, O., Tarasov, A., Fu, W., Wipf, M., Niesen, B., Calame, M., Scho-nenberger, C., 2010. Nernst limit in dual-gated Si-nanowire FET sensors. NanoLett. 10 (6), 2268–2274.

Lee, J., Jang, J., Choi, B., Yoon, J., Kim, J.-Y., Choi, Y.-K., Myong Kim, D., Hwan Kim, D.,Choi, S.-J., 2015. A highly responsive silicon nanowire/amplifier MOSFET hybridbiosensor. Sci. Rep. 5, 12286.

Lee, J., Lee, J.M., Lee, J.H., Uhm, M., Lee, W.H., Hwang, S., Chung, I.Y., Park, B.G., Kim,D.M., Jeong, Y.J., Kim, D.H., 2013. SiNW-CMOS hybrid common-source amplifieras a voltage-readout hydrogen ion sensor. IEEE Electron Device Lett. 34 (1),135–137.

Livi, P., Wipf, M., Tarasov, A., Stoop, R., Bedner, K., Rothe, J., Chen, Y., Stettler, A.,Schonenberger, C., Hierlemann, A., 2013. Silicon nanowire ion-sensitive field-effect transistor array integrated with a CMOS-based readout chip, in: Pro-ceedings of the 17th International Conference on Solid-State Sensors, ActuatorsAnd Microsystems (Transducers & Eurosensors XXVII), 2013, pp. 1751–1754.

Luo, L., Jie, J., Zhang, W., He, Z., Wang, J., Yuan, G., Zhang, W., Wu, L.C.M., Lee, S.-T.,2009. Silicon nanowire sensors for Hg[sup 2þ] and Cd[sup 2þ] ions. Appl.Phys. Lett. 94 (19), 193101.

Mansouri, A., Scheuerman, C., Bhattacharjee, S., Kwok, D.Y., Kostiuk, L.W., 2005.Transient streaming potential in a finite length microchannel. J. Colloid Inter-face Sci. 292 (2), 567–580.

Mu, L., Chang, Y., Sawtelle, S.D., Wipf, M., Duan, X., Reed, M.A., 2015. Silicon na-nowire field-effect transistors – a versatile class of potentiometric nanobio-sensors. IEEE Access 3, 287–302.

Mu, L.Y., Droujinine, I.A., Rajan, N.K., Sawtelle, S.D., Reed, M.A., 2014. Direct, rapid,

J. Lee et al. / Biosensors and Bioelectronics 87 (2017) 447–452452

and label-free detection of enzyme substrate interactions in physiologicalbuffers using CMOS-compatible nanoribbon sensors. Nano Lett. 14 (9),5315–5322.

Newaz, A.K.M., Markov, D.A., Prasai, D., Bolotin, K.I., 2012. Graphene transistor as aprobe for streaming potential. Nano Lett. 12 (6), 2931–2935.

Rajan, N.K., Routenberg, D.A., Reed, M.A., 2011. Optimal signal-to-noise ratio forsilicon nanowire biochemical sensors. Appl. Phys. Lett. 98 (26),264107–2641073.

Rigante, S., Scarbolo, P., Wipf, M., Stoop, R.L., Bedner, K., Buitrago, E., Bazigos, A.,Bouvet, D., Calame, M., Schonenberger, C., Ionescu, A.M., 2015. Sensing withadvanced computing technology: Fin field effect transistors with high-K gatestack on bulk silicon. ACS Nano.

Schasfoort, R.B.M., Bergveld, P., Kooyman, R.P.H., Greve, J., 1990. Possibilities andlimitations of direct detection of protein charges by means of an immunologicalfield-effect transistor. Anal. Chim. Acta 238 (2), 323–329.

Smith, J.T., Shah, S.S., Goryll, M., Stowell, J.R., Allee, D.R., 2014. Flexible ISFET bio-sensor using IGZO metal oxide TFTs and an ITO sensing layer. IEEE Sens. J. 14(4), 937–938.

Stern, E., Wagner, R., Sigworth, F.J., Breaker, R., Fahmy, T.M., Reed, M.A., 2007b.Importance of the Debye screening length on nanowire field effect transistorsensors. Nano Lett. 7 (11), 3405–3409.

Stern, E., Klemic, J.F., Routenberg, D.A., Wyrembak, P.N., Turner-Evans, D.B., Ha-milton, A.D., LaVan, D.A., Fahmy, T.M., Reed, M.A., 2007a. Nature 445 (7127),519.

Stern, E., Vacic, A., Rajan, N.K., Criscione, J.M., Park, J., Ilic, B.R., Mooney, D.J., Reed, M.A., Fahmy, T.M., 2010. Label-free biomarker detection from whole blood. Nat.Nanotechnol. 5 (2), 138–142.

Tian, R., Regonda, S., Gao, J., Liu, Y., Hu, W., 2011. Ultrasensitive protein detectionusing lithographically defined Si multi-nanowire field effect transistors.Lab Chip 11 (11), 1952–1961.

Vacic, A., Criscione, J.M., Rajan, N.K., Stern, E., Fahmy, T.M., Reed, M.A., 2011.

Determination of molecular configuration by debye length modulation. J. Am.Chem. Soc. 133 (35), 13886–13889.

vanHal, R.E.G., Eijkel, J.C.T., Bergveld, P., 1996. A general model to describe theelectrostatic potential at electrolyte oxide interfaces. Adv. Colloid Interface Sci.69, 31–62.

Vanwagenen, R.A., Andrade, J.D., 1980. Flat plate streaming potential investigations:hydrodynamics and electrokinetic equivalency. J. Colloid Interface Sci. 76 (2),305.

Welch, P.D., 1967. The use of fast Fourier transform for the estimation of powerspectra: a method based on time averaging over short, modified periodograms.IEEE Trans. Audio Electroacoust. 15 (2), 70–73.

Wipf, M., Stoop, R.L., Navarra, G., Rabbani, S., Ernst, B., Bedner, K., Schönenberger, C.,Calame, M., 2016. Label-free FimH protein interaction analysis using siliconnanoribbon BioFETs. ACS Sens. 1 (6), 781–788.

Wipf, M., Stoop, R.L., Tarasov, A., Bedner, K., Fu, W., Wright, I.A., Martin, C.J., Con-stable, E.C., Calame, M., Schönenberger, C., 2013. Selective sodium sensing withgold-coated silicon nanowire field-effect transistors in a differential setup. ACSNano 7 (7), 5978–5983.

Zhang, G.-J., Ning, Y., 2012. Silicon nanowire biosensor and its applications in dis-ease diagnostics: a review. Anal. Chim. Acta 749, 1–15.

Zhang, G.-J., Zhang, L., Huang, M.J., Luo, Z.H.H., Tay, G.K.I., Lim, E.-J.A., Kang, T.G.,Chen, Y., 2010. Silicon nanowire biosensor for highly sensitive and rapid de-tection of Dengue virus. Sens. Actuators B: Chem. 146 (1), 138–144.

Zhang, R., Curreli, M., Thompson, M.E., 2011. Selective, electrochemically activatedbiofunctionalization of In2O3 nanowires using an air-stable surface modifier.ACS Appl. Mater. Interfaces 3 (12), 4765–4769.

Zhang, Y., Arugula, M.A., Wales, M., Wild, J., Simonian, A.L., 2015. A novel layer-by-layer assembled multi-enzyme/CNT biosensor for discriminative detection be-tween organophosphorus and non-organophosphorus pesticides. Biosens.Bioelectron. 67, 287–295.