Cd

YVMa

b

c

d

e

Pf

a

ARRAA

KCDCBH

1

bpHthw[afib

v

h0

Sensors and Actuators B 249 (2017) 691–699

Contents lists available at ScienceDirect

Sensors and Actuators B: Chemical

jo ur nal home page: www.elsev ier .com/ locate /snb

hemiresistive biosensors based on carbon nanotubes for label-freeetection of DNA sequences derived from avian influenza virus H5N1

angxi Fua,∗, Victor Romaya, Ye Liua, Bergoi Ibarluceaa,b, Larysa Barabana,b,yacheslav Khavrusa,c,d, Steffen Oswaldc, Alicja Bachmatiukc,e,f, Imad Ibrahimc,ark Rümmeli c,e,f, Thomas Gemmingc, Viktor Bezuglya,b,c,∗, Gianaurelio Cuniberti a,b

Institute for Materials Science and Max Bergmann Center for Biomaterials, Technische Universität Dresden, 01062 Dresden, GermanyCenter for Advancing Electronics Dresden (CfAED), TU Dresden, 01062 Dresden, GermanyLeibniz Institute for Solid State and Materials Research (IFW), Helmholtzstraße 20, 01069 Dresden, GermanyProNT GmbH, Gitterseetraße 2, 01187 Dresden, GermanyCollege of Physics, Optoelectronics and Energy & Collaborative Innovation Center of Suzhou Nano Science and Technology, Soochow University, Suzhou,R ChinaCentre of Polymer and Carbon Materials, Polish Academy of Sciences, M.Curie-Sklodowskiej 34, Zabrze 41-819, Poland

r t i c l e i n f o

rticle history:eceived 14 October 2016eceived in revised form 11 April 2017ccepted 12 April 2017vailable online 18 April 2017

eywords:

a b s t r a c t

We developed chemiresistor-type biosensors based on carbon nanotubes for highly efficient and fastdetection of avian influenza virus (AIV) subtype H5N1 DNA sequences. Semiconducting single-walledcarbon nanotubes (sc-SWCNTs) or nitrogen-doped multi-walled carbon nanotubes (N-MWCNTs) wereused as two alternative active sensing elements, and their sensitivity to different concentrations of DNAtarget were compared. In these sensors long nanotubes (>5 �m) were placed between interdigitatedmetal electrodes so that individual nanotubes connect the electrodes. The nanotubes were functional-

arbon nanotubeNAhemiresistoriosensorybridization

ized with DNA probe sequences non-covalently attached to the sidewalls. Such functionalized-nanotubesensors could reliably detect complementary DNA target sequences of the AIV H5N1 with concentrationranging from 2 pM to 2 nM in 15 min and at room temperature. Our nanotube-based biosensors are small,flexible, disposable and easy-to-fabricate that makes them promising for point-of-care applications andclinical diagnostics.

© 2017 Elsevier B.V. All rights reserved.

. Introduction

Avian influenza virus (AIV), especially subtype H5N1 hasecome nowadays a problem not only for poultry because of itsathogenicity and relatively high lethality rate among human.ighly pathogenic AIV H5N1 can infect people and can be transmit-

ed from human to human with disastrous implications for publicealth [1,2]. High-sensitive and rapid detection of H5N1 infectionould allow early antiviral therapy and control of the outbreaks

3–6]. However, the most commonly used detection techniquesre laborious, time-consuming, or require specialized laboratory

acilities and well-trained technical personnel, which greatly lim-ts their application in clinical tests [3–5]. Direct DNA hybridizationiosensing has become recently a very promising approach, not

∗ Corresponding authors.E-mail addresses: yfu@nano.tu-dresden.de (Y. Fu),

bezugly@nano.tu-dresden.de (V. Bezugly).

ttp://dx.doi.org/10.1016/j.snb.2017.04.080925-4005/© 2017 Elsevier B.V. All rights reserved.

only because of its high sensitivity, but also of label-free detec-tion and easy integration with other, in particular portable devices,which makes it a promising candidate for the development of anintegrated, high-throughput, low-cost and tiny AIV biosensor for“point-of-care” diagnostics [7–10].

Nowadays, DNA hybridization chemiresistive nanosensors havebeen developed, which are very attractive because of their reducedsize, high sensitivity and a simple detection principle based onthe change of resistance in response to the binding of DNA target(DNA T) to active nanomaterials [10–13]. Recently, there has beena rise in the use of one-dimensional nanostructures such as siliconnanowires (Si-NW) [8,14], conducting polymer nanowires [11,15]and carbon nanotubes (CNTs) [16–19] as transducing elements inchemiresistive sensors. Among these materials, CNTs, especiallysingle-walled CNTs (SWCNTs), have increasingly garnered consid-

erable interest from the DNA sensor research due to their highaspect ratio, good environmental stability, excellent mechanicaland electronic properties as well as the convenience for label-freesensing, which hold a great potential for integrating them into com-

6 ctuators B 249 (2017) 691–699

pHltabs

tntfCiciSsawwHotnfbtqs

2

2

foG(iswNsp5THupt(An

2

cplwn1

Table 1N doping content in N-MWCNTs produced from varying N-containing feedstocks.

Nr. Acetonitrile percentage in ethanol (vol%) N doping content (at%)

1 50 9.34

92 Y. Fu et al. / Sensors and A

act, low-power and portable miniaturized sensors [13,17–20].owever, relatively strong van der Waals interaction between CNTs

eads to agglomeration of CNTs in both powder and dispersion formhank to their small size and high aspect ratio. This makes individu-lization, processing and integration of CNTs quite challenging thatecomes finally a bottleneck in a scalable fabrication of CNT-basedensing devices [21,22].

Herein we report on the development of stable chemiresistor-ype sensors based on semiconducting SWCNTs (sc-SWCNTs) anditrogen-doped multi-walled CNTs (N-MWCNTs) for highly sensi-ive and rapid label-free detection of AIV H5N1 virus. Our sensorabrication technology overcomes the difficulty of integration ofNTs into devices; relatively long nanotubes with lengths exceed-

ng 5 �m directly connect metallic electrodes. Since DNA areharged molecules, field effect is expected to play the major rolen influencing electronic structure of nanotubes. Therefore, sc-WCNTs were applied in this work as they showed a higherensitivity than pristine SWCNTs when gas and bio-molecules aredsorbed to their surface [23–25]. On the other hand, N-MWCNTsere employed as they have amine groups on their sidewalls whichould facilitate functionalization of nanotubes with DNA probe [8].owever, doping with nitrogen modifies electronic performancef nanotubes compared to undoped ones [26–28] which may leado a decrease in sensitivity. In case of N-MWCNT, well-alignedanotube arrays placed between interdigitated electrodes were

abricated on both rigid and flexible substrates that was achievedy contact-printing of vertically-grown N-MWCNTs onto a cleanarget substrate. The sensors with sc-SWCNTs were fabricated onuartz substrates only as these substrates served as a support forc-SWCNT growth.

. Experimental

.1. Materials

All the reagents were of analytical grade and were used withouturther purification. Three series of the samples were fabricated:n glass and quartz substrates of about 1 mm thick (QSIL GmbH,ermany) as well as on 100 �m thick polyimide foils Kapton

®

Dupont HN100). The latter were applied for the fabrication of flex-ble device. Acetonitrile and ethanol (Sigma–Aldrich) were used forynthesis of N-MWCNTs. ProNTTM sc-SWCNTs on quartz substratesith a length from 5 to 20 �m were produced by ProNT GmbH.o surfactants or catalyst particles were present on ProNTTM

c-SWCNTs. Single-stranded DNA of AIV H5N1 sequences wereurchased from Eurofins Genomics, with the base pair sequences:′-CAA ATC TGC ATT GGT TAT CA-3′ for DNA probe, and 5′-TGAAA CCA ATG CAG ATT TG-3′ for DNA T. For negative control AIV1N1 DNA T sequence 5′-GTA GGT TGA CAG AGT GTG-3′ wassed which was non-complementary to the H5N1 DNA probe. Therobe and DNA T were diluted with phosphate buffer (PB) solu-ion to a final concentration of 20 �M for further use. TritonTM

4-(1,1,3,3-Tetramethylbutyl)phenyl-polyethylene glycol, Sigma-ldrich) 0.01% in buffer was used to block the uncovered sites atanotube sidewalls.

.2. N-MWCNT growth, characterization and contact printing

Vertically aligned N-MWCNTs were synthesized by means ofhemical vapor deposition (CVD) on a Si/SiO2 wafer surface withre-deposited Fe/Al2O3 metallic-layer serving as the source of cata-

yst. 1 nm thick Fe layer was deposited after 10 nm thick Al2O3 layeras built on the 400 nm thick SiO2 surface of the wafer by mag-etron sputter deposition in high vacuum chamber (base pressure:0−7 mbar; Ar sputter pressure: 10−3 mbar). Acetonitrile/ethanol

2 30 8.203 20 6.14

mixtures were employed as C/N source, the doping content of N-MWCNT structure can be controlled through varying the proportionof acetonitrile to ethanol [28–30]. The CVD was performed between760 and 910 ◦C (synthesis temperature was closely related to ace-tonitrile/ethanol ratio in a precursor solution and was optimizedfor each acetonitrile/ethanol ratio) for 15 min with a pressure of100 mbar. After the growth, the morphology and density of verticalnanotube “forest” was analyzed with scanning electron microscope(SEM), see Fig. 1(a). X-ray photoelectron spectroscopy (XPS) mea-surements were performed to determine the chemical constitutionof N-MWCNT samples (a PHI 5600 CI system with a hemisphericalenergy analyzer was used). According to XPS investigation, the Ndoping concentration was varied from 6.14% to 9.34% when thefraction of acetonitrile in ethanol in the precursor was varied from20% to 50% (see Table 1). Samples of vertically grown N-MWCNTswith different content of nitrogen have been synthesized.

In order to form horizontally aligned N-MWCNT arrays, directcontact printing was applied for transferring as-grown verticallyaligned nanotubes onto various clean target substrates similar tocontact-printing of Si-NW reported elsewhere [8]. The “ink” forthe direct contact printing is a dense “forest” of vertically alignedN-MWCNT arrays with the height of 5–10 �m. Target substrateapplied here were glass, quartz and Kapton

®polyimide foil, which

were first cut into pieces with a size of 20 × 20 mm and thencleaned with ultrasonic washer in an acetone, ethanol and deion-ized water bath successively for 10 min. Octadecyltrichlorosilane(C18H37SiCl3) (ODTS) was applied to chemically modify the sur-face of rigid target substrates and to increase the adhesion ofN-MWCNTs. The main steps of direct contact printing are schemat-ically illustrated in Fig. 1(b). During the contact printing process,an intimate contact at the interface between the growth and tar-get substrates causes a relatively strong van der Waals interactionbetween nanotubes and target substrate that leads to the detach-ment of nanotubes from the growth substrate and binding them tothe surface of the target substrate. Besides adhesion strength, thesurface density of horizontal N-MWCNT arrays was controlled byadjusting CVD process parameters to synthesize vertically alignedN-MWCNTs with optimal length and surface density as well asexternal pressure applied to the substrate with vertically alignedN-MWCNTs during contact printing onto the target substrate. Theresults of direct contact printing for external pressure ranging from0.5 to 3.0 kPa can be seen in Figure S1 of Supplementary data. Forthe fabrication of our N-MWCNT-based sensors external pressureof 1 kPa was utilized.

In case of sc-SWCNT, no horizontal alignment was achieved. Inthese samples sc-SWCNTs of 5 to 20 �m in length were obtainedby CVD-based epitaxial elongation of sc-SWCNT short fragments(seeds) chaotically distributed on a quartz substrate and weredirectly used for device fabrication. It is to emphasize, that nosurfactants were present on sidewalls of nanotubes in all sam-ples (M-NWCNTs and sc-SWCNTs) before functionalization step(described below).

2.3. Sensor fabrication and device layout

The chemiresistor-type sensors were fabricated using a stan-dard microfabrication procedure. After forming CNTs arrays onthe target substrates, Cr/Au (3 nm Cr and 50 nm Au) interdigitated

Y. Fu et al. / Sensors and Actuators B 249 (2017) 691–699 693

Fig. 1. (a) Scanning electron micrographs of close-packed vertical arrays of N-MWCNTs grown on a Si/SiO2 substrate with CVD. (b) Sketch of the contact-printing process forvertically aligned CNTs. (c) SEM image of N-MWCNT based chemiresistor sensor device on Kapton® substrate; the higher magnified image in the inset shows a regular arrayo terspC pticalf mide fi

ebmsutfiwt(

nnFsaS

f horizontally aligned N-MWCNTs bridging Cr/Au interdigitated electrodes. (The inr/Au interdigitated electrodes on quartz substrate bridged with sc-SWCNTs. (e) Oabricated on various substrates (from left to right: quartz, glass and Kapton® polyi

lectrode fingers with an interspacing of 3 �m were depositedy means of thermal evaporation onto the patterned photoresistold, and a lift-off technique was used to remove the photore-

ist, as described elsewhere [8,31–33]. Finally the electrodes wereltrasonically rinsed in acetone, isopropanol and deionized waterhoroughly, and then dried in nitrogen flow. Before use, all these as-abricated devices were annealed at 200 ◦C for 15 min in vacuum formproving their electrical performance. The functionality of devices

as checked through electrical characterization in a probe sta-ion and corroborated with SEM characterization of their structuresPhillips XL30 ESEM-FEG).

Cr/Au interdigitated electrodes were placed in the directionormal to the alignment direction of N-MWCNT arrays so thatanotubes bridged the electrodes and acted as conductor channels.

ig. 1(c) and (d) show the channel region of the N-MWCNT andc-SWCNT based chemiresistor devices on Kapton

®polyimide film

nd quartz, respectively. In Fig. 1(d), randomly distributed long sc-WCNTs are visible in the areas between the electrodes as shadows

acing distance of the electrodes is 3 �m.) (d) Low-voltage SEM image of patterned images of CNT chemiresistor biosensor chips each containing 400 sensing deviceslm).

on a quartz substrate. Here, low-voltage SEM measurements (accel-erating voltage 1 kV) were employed to visualize SWCNTs betweeninterdigitated electrodes. It is apparent that SWCNTs may be par-tially agglomerated in some areas, however the major part of theelectrode comb structure has individualized sc-SWCNTs bridgingmetallic electrodes. Fig. 1(c) shows the whole comb structure ofelectrodes each having a dimension of 4.5 �m width and 75.3 �mlength and an interspacing distance of 3 �m. As shown in the inset,high-dense and uniform parallel array of individual N-MWCNTswith a well-defined direction was covered by Cr/Au interdigitatedelectrodes. Each nanotube is not in a contact with others and is longenough to bridge two electrodes as a conducting channel. Manyefforts in recent have focused on tackling the challenge of a control-lable integration of CNTs into microelectronic circuits. Significant

progress has been made, but the majority of presented approaches,like spray coating [34,35], soft lithography [36], and inkjet printing[37–39], still lack the desired scalable positioning of uniform andwell-aligned individual CNTs. Direct contact printing offers a great

6 ctuato

pfdopbnprtpmCs

2

s1eaieoftuscwfloiwoep0r2arwc0ednssTc

2

scimsa

94 Y. Fu et al. / Sensors and A

otential for a scalable transfer of parallel arrays of CNTs onto dif-erent target substrates with a large and uniform coverage. Fig. 1(e)emonstrates that CNT-based nanosensor arrays can be fabricatedn both rigid and flexible substrates, like quartz, glass and Kapton

®

olyimide foils. Electronic devices on flexible substrates recentlyecame extremely attractive owing to a high mechanical robust-ess and to a large demand in wearable, bio-compatible, hand-held,ortable consumer electronics as well as the compatibility witholl-to-roll fabrication [40–42]. Ishikawa et al. [40] reported thathe use of CNTs bridging electrodes directly would enable theroduction of flexible devices which would exhibit higher perfor-ance in contrast to devices based on random networks of short

NT, including higher mobility, lower power consumption, higherwitching speed and better mechanical flexibility.

.4. Functionalization and DNA sensing

As the first step, a droplet of PB 200 mM was placed onto theensor chip to cover the entire electrodes area and incubated for5 min at room temperature. This was followed by rinsing thentire device with deionized water and drying in nitrogen flownd then in vacuum at room temperature for 15 min. Such prelim-nary incubation is crucial to assess the effect of the buffer on thelectrical characteristics of the sensing devices. Then, DNA probeligonucleotides were immobilized on CNT sidewalls that was per-ormed with drop-casting of 100 �L of DNA probe (20 �M) ontohe surface of the CNT sensor chip and kept in an oven at 60 ◦Cntil complete drying. This amount of DNA probe solution wasufficient to functionalize dozens of sensing devices in the sensorhip. Afterwards, the chips were immersed in PB buffer 200 mM,ashed thoroughly with deionized water and dried in nitrogenow and then in vacuum for 15 minutes. This way DNA probeligonucleotides were non-covalently attached to CNTs via �-�nteraction between a honeycomb-like structure of nanotube side-

alls and the heterocyclic rings of the nucleotides [43]. The excessf DNA probe molecules was washed away. To block the uncov-red sites on the CNT sidewalls that DNA probe did not occupy, therobe-functionalized CNT sensors were incubated in a solution of.01% TritonTM X-100 dissolved in PB buffer 200 mM for 15 min atoom temperature. Afterwards the chips were rinsed with PB buffer00 mM, followed by deionized water and dried in nitrogen flownd then in vacuum. For DNA T sensing experiments, hybridizationeaction was carried out by incubating the functionalized sensorsith 90 �L of H5N1 DNA T in PB buffer with successively increasing

oncentrations for 15 minutes each at room temperature, which are pM, 2 pM, 20 pM, 200 pM, 2 nM, 20 nM, 200 nM and 2 �M. Afterach concentration step the chip was washed with PB buffer andeionized water to remove the un-hybridized DNA and dried initrogen flow and then in vacuum for 15 minutes. In order to proveelectivity of the functionalized CNT-based genosenor the sameensing experiments were performed with solutions of H1N1 DNA

which was non-complementary to H5N1 DNA probe (negativeontrol).

.5. Electrical characterization

After functionalization and each hybridization step, CNT-basedensing devices were characterized with a probe station under dryonditions by recording the direct current–voltage (I–V) character-

stics from −2.0 to +2.0 V. The probe station consisted of an optical

icroscope (Olympus Europe), two micro-positioners with tung-ten needles, an electrically contactable sample holder (chuck), and

sourcemeter Keithley 2604B from Tektronix.

rs B 249 (2017) 691–699

3. Results and discussion

Device preparation and working principle of our chemiresistor-type DNA sensor are schematically presented in Fig. 2(a). The firststep depicts device fabrication when CNTs are placed betweenmetallic electrodes. In the second step (functionalization), DNAprobe molecules were non-covalently attached to the CNTs side-walls. It was demonstrated both experimentally and theoreticallythat DNA nucleobases have a strong van der Waals attractionto hexagonal carbon structures (characteristic to graphene andnanotube walls) in a �–� stacking configuration [44]. Shortsingle-stranded DNA can easily adopt a flat conformation whichmaximizes van der Waals interaction and leads to DNA probeadsorption on CNT sidewalls. The attachment of DNA probe to CNTsled to a dramatic increase of the device resistance, which was mea-sured with the probe station (Fig. 2(b)) after rinsing and dryingthe device. In the third step, TritonTM X-100 was applied to blockthe uncovered sites on the CNT sidewalls that DNA probe did notoccupy. This was done to avoid non-specific binding of DNA T toCNTs in a similar way like DNA probe. A minor change of deviceresistance was associated with covering of CNTs with Triton, mea-sured in a dry condition. When a DNA T solution was drop-castedfor the DNA sensing experiment, step four (DNA T), the DNA Tmolecules were hybridized with their complementary DNA probemolecules removing them from the surface of CNTs since double-stranded DNA have a helical structure which is not optimal for thenoncovalent binding with CNT sidewalls. A similar principle wasused for producing optical DNA sensors with graphene [45]. In oursensors, the removal of DNA probe after hybridization with DNAT was manifested by a decrease of device resistance (measuredagain after rinsing and drying) toward initial resistance value. Thisdecrease of resistance depends on the concentration of DNA T insolution, namely, the higher concentration is the more DNA probemolecules are detached from the CNT sidewalls during 15 min ofincubation. By repeating step 4 several times with increasing DNAT concentrations it was possible to determine the range of con-centrations for which the sensing device can be used. It is worthnoticing that for concentrations higher than the one at which com-plete detachment of DNA probe is achieved, no reliable sensing ispossible. This is a specific feature of the sensing principle realizedin presented sensing devices.

The sensing mechanism of chemiresistor DNA sensors attributesthe electrical conductance change either to the electron doping byDNA attached to the CNT sidewall suggested previously by Staret al. [17] or to the change in Schottky barrier at the CNT/metallic-lead interface, suggested by Tang et al. [18] The latter should bevalid for semiconducting nanotubes only whereas the former effectshould influence both metallic and semiconducting CNTs. In ourexperiments, some devices had N-MWCNTs, which showed metal-lic behavior, and some were based on sc-SWCNT where Schottkybarrier was formed at the CNT/metallic-lead interface [46,47]. Bycomparing sensing response of different devices we were able todistinguish the two effects suggested by the two groups. A typi-cal electronic response of our CNT based chemisistor DNA sensorsto the CNT functionalization steps is shown in Fig. 2(c) and (d)where device resistance measured at voltage 2 V is presented. Thedevice resistance at “Initial” step corresponds to newly fabricateddevice before any functionalization (step 1 in Fig. 2(a)). The initialresistance of N-MWCNT-based sensor was of 40.1 M�, while for sc-SWCNT-based sensor it was of 22.8 M�. The initial values of deviceresistance varied from device to devicedepending on the numberof nanotubes bridging the electrodes in each device as well as on

the electronic properties of active materials.

In order to estimate the influence of pure buffer on CNTs thedevices were immersed into the buffer for 15 min, rinsed and dried.The measured resistance is marked “Buffer”. It is apparent that

Y. Fu et al. / Sensors and Actuators B 249 (2017) 691–699 695

Fig. 2. (a) Schematic illustration of the CNT-based chemiresistor biosensor functionalization and sensing steps. (b) Image of the probe station during electric characterizationof a sensing device. (c) Resistance change of the sensors based on N-MWCNTs (N/C ratio of 9.34%) and (d) on sc-SWCNTs after the CNT functionalization experiments withrespect to resistance of as-fabricated device (Initial) measured at voltage V = 2.0 V for every step. Sensing experiments: resistance change of (e) N-MWCNT and (f) sc-SWCNT-b centraT esentm

bStC(ftrief(ipn

ased sensors as the response to processing of the same sensor with increasing con). Lines connecting points in plots (c)-(f) are used to guide an eye; the points reprean value for three independent measurements at 2.0 V.

uffer had no effect on N-MWCNTs, however the device with sc-WCNTs showed an increase of resistance which can be attributedo the modulation of Schottky barrier by ions attached close toNT/lead interface. When DNA probe was used for fuctionalizationstep 2 in Fig. 2(a)), a significant increase of the relative resistanceor both N-MWCNT and sc-SWCNT sensors was observed. The resis-ance of N-MWCNT sensors increased by roughly 160% and theesistance of sc-SWCNT-based device by 430%, exhibiting a strongernfluence of absorbed DNA probe molecules on carbon nanotubelectronic structure. After that, the blocking of non-functionalizedree areas of CNT sidewalls with TritonTM X-100 was performedstep 3 in Fig. 2(a)), which did not lead to significant changes

n device resistances since 4-(1,1,3,3-Tetramethylbutyl)phenyl-olyethylene glycol are neutral organic molecules and they couldot noticeably affect conductance of nanotubes.

tions of DNA T (at V = 2.0 V) count with respect to resistance value at 0 pM (no DNA the mean values and error bars are calculated as the standard deviation from the

After treatment with Triton, the sensors were prepared forDNA T sensing experiments. The sensing experiments (step 4 inFig. 2(a)) for all devices were performed repeatedly with suc-cessively increasing concentrations ranging from 0 to 2000 pM.Fig. 2(e) shows the sensing response of N-MWCNT-based deviceto different concentrations as the ratio R/R0 of the resistance afterincubation with DNA T solution of given concentration to thatwith 0 pM concentration (in order to compare sensing responseof different devices with different initial resistance); the pointsrepresent the mean values at 2.0 V for three independent mea-surements performed with a time interval of several minutes, andthe full IV-curves are shown in Figure S2a in Supplementary data.

The error bars are calculated as the standard deviation from themean value and show a good reproducibility of sensing measure-ments. It is apparent that no noticeable change in resistance for2 pM DNA T have been detected. However, starting with concen-

6 ctuators B 249 (2017) 691–699

toewcctsrwoNfsftc

iiatutAccuotohtfsad

faaisoIltpamsttTChrofau

sfni

Fig. 3. Comparison of sensing responses of sensors based on N-doped MWCNTswith N content 9.34 (black), 8.20 (green) and 6.14 at% (blue) and on semiconductingSWCNTs (purple). All sensors show similar downwards trend of resistance changeR/R0 for concentrations up to 200 pM. Increase of R/R0 at 2 nM is attributed to the

96 Y. Fu et al. / Sensors and A

ration of 20 pM, a significant decrease of device resistance wasbserved. As explained above, such resistance reduction can bexplained by the hybridization of complementary DNA T moleculesith the DNA probe molecules and detachment of the latter from

arbon nanotube sidewalls. The resistance change maintained aontinuous downward trend with increasing of DNA T concentra-ion up to 2 nM. Further increase of concentration led to unreliableensing response as two competing processes influenced the deviceesistance, namely detachment of DNA probe after hybridizationith DNA T and attachment of access DNA T to the free places

f CNT sidewalls left after removing DNA probe. Interestingly,-MWCNTs had metallic behavior, and no Schottky barrier was

ormed at CNT/lead interface. Therefore, the sensing response pre-ented in Fig. 2(e) as well as the increase of the resistance afterunctionalization with DNA probe in Fig. 2(c) was solely based onhe modulation of electronic states in N-MWCNTs as conductinghannels by attachment or detachment of DNA probe molecules.

A qualitatively similar response to DNA T could also be observedn sc-SWCNT-based sensors, Fig. 2(f) (full IV-curves can be foundn Supplementary data, Figure S2b). However, sc-SWCNTs showed

strong response already for 2 pM of DNA T. The device resis-ance kept a systematic drop with increasing DNA T concentrationp to 200 pM. Sensing at higher concentrations was unreliable forhe same reasons described above for N-MWCNT-based sensors.n overall stronger sensing response of sc-SWCNT-based devicesompared to N-MWCNT ones (especially at concentration of 2 pM)ould be attributed to an additional effect of Schottky barrier mod-lation at the sc-SWCNT/metallic lead interface. The attachmentf charged DNA probe molecules on nanotube sidewalls close tohe metallic lead reshaped Schottky barrier and led to the increasef resistance. Consequently, when DNA probe were detached afterybridization with DNA T, Schottky barrier was returned closero its original form, and the device resistance decreased. There-ore, from the comparison of sensing response of N-MWCNT andc-SWCNT based devices we concluded that both electron dopingnd modulation of Schottky barrier effects were causing change ofevice resistance after functionalization and sensing steps.

A number of sensing devices based on N-MWCNTs with dif-erent amount of nitrogen as dopant (varied from 6.14% to 9.34%ccording to XPS data) and on different substrates (quartz, glassnd Kapton) were fabricated and tested. No noticeable differencen performance of these devices with respect to dopant content orubstrate type was found: different sensors show a similar decreasef relative resistance after exposure to DNA T as can be seen in Fig. 3.n case of sc-SWCNT based sensors, the downwards trend starts at aower concentration and is more pronounced. However, at concen-ration of 2 nM the effect of the attachment of DNA T to the emptylaces left on CNT sidewalls after previous sensing experiment with

lower concentration start to compete with effects from detach-ent of the rest of DNA probe that leads to a backwards trend (also

een for N-MWCNTs with 6.14 at% of N). Still, at lower concentra-ions of DNA T where the competing effect is absent or negligible,he decrease of relative resistance caused by hybridization of DNA

and DNA probe and the consequent detachment of the latter fromNT sidewalls is discernible for all sensors. It should be emphasizedowever, that our experiments were performed to determine theage of concentrations at which DNA T can be reliably detected withur sensors (from 2 to 200 pM for sc-SWCNT based sensors androm 20 pM to 2 nM for N-MWCNT based ones). Calibration curvesnd determination of the limits of detection will be the topic of ourpcoming publication.

Finally, the selectivity of sensors to AIV H5N1 subtypes DNA

equences was assessed by testing sensing response of devicesunctionalized with H5N1 DNA probe to H1N1 DNA T which is aon-complementary sequence in the same way as sensing exper-

ments for H5N1 DNA T was performed. No noticeable sensing

competing effect described in the text. Lines connecting points are used to guide aneye.

response to H1N1 DNA T was detected for concentrations rangingfrom 20 pM to 2 �M that showed a perfect selectivity of our sen-sors to a specified DNA T sequence based on the hybridization withcomplementary DNA probe (see corresponding section and FigureS3 in Supplementary data).

AIV H5N1 virus can be found in the blood as well as in noseor throat swabs of infected individuals. The viral load in bloodmay reach 105 copies per ml and in throat/nose swabs 108 copiesper ml [48]. For the detection of AIV H5N1, RNA must be firstextracted from the virus and a reverse transcription to DNA shouldbe performed. Concentration of 108 virus copies per ml corre-sponds to 0.2 pM, being recalculated as the concentration of DNAsingle-stranded sequences. However, the final concentration of theanalyte can be still increased by reducing the volume since only alittle droplet of the DNA solution (less than 0.1 �l), which can coverthe 0.2 × 0.1 mm2 area of interdigitated microelectrodes, is neededfor sensing with our CNT-based sensors (2 to 200 pM). Therefore,the amount of H5H1 DNA in a real sample is sufficient for the detec-tion using proposed sensors without the need of amplification ofDNA copies. Further, it should be taken into account that real sam-ples contain proteins typical to serum like immunoglobulin, serumalbumin, transferrin, fibrinogen, lysozyme and other. However, aswas shown for human �-thrombin sensor based on hybridizationof complementary nucleobase sequences, the presence of proteinsmay partially reduce the sensitivity with respect to the pure buffersolution but does not show any interference with the analyte assay[49,50].

Sensors for the detection of AIV single-stranded DNA basedon different principles were demonstrated in recent years. Thedetection ranges are summarized in Table 2 and compared to theresults of this work. As can be seen from this comparison, CNT-based chemiresistor-type sensors proposed here have a sensitivitycomparable to the majority of listed ones, however a poorer sensi-tivity compared to the sensors reported in [7,53,]. ChemiresistiveDNA sensors based on carbon nanotubes were also demonstrated

recently. For example, Singh et al. [55] presented SWCNT-basedsensor for the detection of human rheumatic heart disease basedon hybridization of the target S. pyogenes single stranded genomic

Y. Fu et al. / Sensors and Actuators B 249 (2017) 691–699 697

Table 2Comparison of detection range of different sensors for AIV viruses.

Detection method DNA Sensing materials Range Ref.

Voltammetric detection AIV H5N1 HS-ssDNA Probe modified gold electrodes 10–100 pM [51]Electrochemistry based on an ion barrier switch-off AIV H5N1 Redox-active monolayers 1–10 pM [52]Voltammetric detection AIV H5N1 Amino-ssDNA probe modified gold electrodes 10–100 fM [53]Electrochemistry AIV H5N1 Probe modified single gold electrode 20–100 nM [54]Electrochemical impedance spectroscopy AIV H1N1 Tubular nanomembrane 20 aM–20 pM [7]

i nano-MWC-SWC

DpwSwdpmdmDtnw

4

doeonwfidwa2tawrCsais

A

lAfiTt(uRtA

[

[

[

[

[

FET AIV H1N1 SChemiresistor AIV H5N1 NChemiresistor AIV H5N1 sc

NA (ssG-DNA) to its complementary 24-mer single-stranded DNArobe, and the estimated lowest detection limit was 22 pM. In theirork, pristine SWCNTs (a mixture of metallic and semiconducting

WCNTs) were placed between gold electrodes and functionalizedith covalently attached DNA probe sequences. The detection wasemonstrated for DNA T sequences, which were attached to DNArobe and stayed, and the measurements were performed in liquidedium directly. In our case, all measurements were done under

ry condition, and the detection principle was based on the detach-ent of non-covalently bound DNA probe after hybridization withNA T sequences. Such an approach had an advantage with respect

o the detection in liquids in a simplified device fabrication sinceo microfluidic channels as well as no isolation of metallic leadsere needed.

. Conclusions

In this study CNT based chemiresistor-type sensors for theetection of the AIV H5N1 subtype DNA sequences were devel-ped. Two different types of CNT were used as active sensinglements, namely N-MWCNTs and sc-SWCNTs. The sensors basedn N-MWCNTs were fabricated with direct contact printing ofanotubes where both rigid and flexible substrates were employedhereas sc-SWCNTs were produced by epitaxial elongation of short

ragments of selected sc-SWCNTs on quartz substrates. The work-ng principle is based on the change of devices resistance afteretachment of DNA probe non-covalently bound to CNT sidewallshen they hybridize with complementary DNA T. The lowest reli-

bly detected concentration of DNA T was 2 pM for sc-SWCNT and0 pM for N-MWCNT sensor after 15 min of incubation. This meanshe sc-SWCNT-based sensor exhibited higher sensitivity that wasttributed to a modulation of Schottky barrier as an additional effecthich was not present in N-MWCNT-based devices. No sensing

esponse to non-complementary H1N1 type DNA T was observed.NT based DNA sensors are small, flexible, easy-to-use and highlyensitive that makes them promising in clinical diagnostics as wells for portable applications. Taking into account relative simplic-ty and low cost of fabrication, chips with integrated CNT-basedensors can be used as one-way sensing devices.

cknowledgements

Authors thank Dr. Denys Makarov for help in preparing catalystayers used for CVD synthesis of N-MWCNTs. Y.F. and V.R. thank Luisntonio Panes-Ruiz for an assistance. V. K. gratefully acknowledgesnancial support from the European Social Fund (Nr. 100234682).his work is partly supported by the European Union (ERDF) viahe FP7 projects “CARbon nanoTube photonic devices on silicon”CARTOON) and “Nano-carbons for versatile power supply mod-

les” (NanoCaTe). We also acknowledge the support by the Germanesearch Foundation (DFG) within the Cluster of Excellence “Cen-er for Advancing Electronics Dresden” EXC 1056, by the Germancademic Exchange Service (DAAD), the Mexican National Council

[

wire 40–102 pM [8]NT 20–2000 pM This workNT 2–200 pM This work

for Science and Technology (CONACYT) and the China ScholarshipCouncil (CSC).

Appendix A. Supplementary data

Supplementary data associated with this article can be found, inthe online version, at http://dx.doi.org/10.1016/j.snb.2017.04.080.

References

[1] WHO, Avian and other zoonotic influenza: fact sheet, 2016,http://www.who.int/mediacentre/factsheets/avian influenza/en/.

[2] J.H. Beigel, J. Farrar, A.M. Han, G. Hayden, R. Frederick, M.D. Hyer, S. de Jong,N.T.K. Lochindarat, N.T. Tien, T.T. Hien, A. Hien, K. Nicoll, Yuen, Avian InfluenzaA (H5N1) Infection in Humans, N. Engl. J. Med. 353 (2005) 1374–1385.

[3] K. Malecka, A. Stachyra, A. Góra-Sochacka, A. Sirko, W. Zagórski-Ostoja, H.Radecka, J. Radecki, Electrochemical genosensor based on disc and screenprinted gold electrodes for detection of specific DNA and RNA sequencesderived from Avian Influenza Virus H5N1, Sens. Actuators B Chem. 224 (2016)290–297, http://dx.doi.org/10.1016/j.snb.2015.10.044.

[4] L. Xu, R. Wang, L.C. Kelso, Y. Ying, Y. Li, A target-responsive andsize-dependent hydrogel aptasensor embedded with QD fluorescentreporters for rapid detection of avian influenza virus H5N1, Sens. Actuators B:Chem. 234 (2016) 98–108, http://dx.doi.org/10.1016/j.snb.2016.04.156.

[5] S.P. Layne, T.J. Beugelsdijk, C. Kumar, N. Patel, J.K. Taubenberger, N.J. Cox, I.D.Gust, A.J. Hay, M. Tashiro, D. Lavanchy, A Global lab against influenza, Science293 (2001) 1729, http://dx.doi.org/10.1126/science.293.5536.1729.

[6] Y. Fu, Z. Callaway, J. Lum, R. Wang, J. Lin, Y. Li, Exploiting enzyme catalysis inultra-low ion strength media for impedance biosensing of avian influenzavirus using a bare interdigitated electrode, Anal. Chem. 86 (2014) 1965–1971,http://dx.doi.org/10.1021/ac402550f.

[7] M. Medina-Sánchez, B. Ibarlucea, N. Pérez, D.D. Karnaushenko, S.M. Weiz, L.Baraban, G. Cuniberti, O.G. Schmidt, High-performance three-dimensionaltubular nanomembrane sensor for DNA detection, Nano Lett. 16 (2016)4288–4296, http://dx.doi.org/10.1021/acs.nanolett.6b01337.

[8] D. Karnaushenko, B. Ibarlucea, S. Lee, G. Lin, L. Baraban, S. Pregl, M. Melzer, D.Makarov, W.M. Weber, T. Mikolajick, O.G. Schmidt, G. Cuniberti, Light weightand flexible high-performance diagnostic platform, Adv. Healthc. Mater. 4(2015) 1517–1525, http://dx.doi.org/10.1002/adhm.201500128.

[9] W.-A. Lai, C.-H. Lin, Y.-S. Yang, M.S.-C. Lu, Ultrasensitive and label-freedetection of pathogenic avian influenza DNA by using CMOS impedimetricsensors, Biosens. Bioelectron. 35 (2012) 456–460, http://dx.doi.org/10.1016/j.bios.2012.02.049.

10] J. Thavanathan, N.M. Huang, K.L. Thong, Colorimetric detection of DNAhybridization based on a dual platform of gold nanoparticles and grapheneoxide, Biosens. Bioelectron. 55C (2013) 91–98, http://dx.doi.org/10.1016/j.bios.2013.11.072.

11] C. García-Aljaro, L.N. Cella, D.J. Shirale, M. Park, F.J. Munoz, M.V. Yates, A.Mulchandani, Carbon nanotubes-based chemiresistive biosensors fordetection of microorganisms, Biosens. Bioelectron. 26 (2010) 1437–1441,http://dx.doi.org/10.1016/j.bios.2010.07.077.

12] C. García-Aljaro, M.A. Bangar, E. Baldrich, F.J. Munoz, A. Mulchandani,Conducting polymer nanowire-based chemiresistive biosensor for thedetection of bacterial spores, Biosens. Bioelectron. 25 (2010) 2309–2312,http://dx.doi.org/10.1016/j.bios.2010.03.021.

13] L. Wei, D. Lu, J. Wang, H. Wei, J. Zhao, H. Geng, Y. Zhang, Highly sensitivedetection of trinitrotoluene in water by chemiresistive sensor based onnoncovalently amino functionalized single-walled carbon nanotube, Sens.Actuators B: Chem. 190 (2014) 529–534, http://dx.doi.org/10.1016/j.snb.2013.09.017.

14] C.H. Lin, C.H. Hung, C.Y. Hsiao, H.C. Lin, F.H. Ko, Y.S. Yang, Poly-siliconnanowire field-effect transistor for ultrasensitive and label-free detection ofpathogenic avian influenza DNA, Biosens. Bioelectron. 24 (2009) 3019–3024,

http://dx.doi.org/10.1016/j.bios.2009.03.014.

15] M.A. Bangar, D.J. Shirale, W. Chen, N.V. Myung, A. Mulchandani, Singleconducting polymer nanowire chemiresistive label-free immunosensor forcancer biomarker, Anal. Chem. 81 (2009) 2168–2175, http://dx.doi.org/10.1021/ac802319f.

6 ctuato

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

98 Y. Fu et al. / Sensors and A

16] K. Maehashi, K. Matsumoto, K. Kerman, Y. Takamura, E. Tamiya, Ultrasensitivedetection of DNA hybridization using carbon nanotube field-effect transistors,Japanese J. Appl. Phys. Part 2-Letters Express Lett. 43 (2004) L1558–L1560,http://dx.doi.org/10.1143/JJAP.43.L1558.

17] A. Star, E. Tu, J. Niemann, J.-C.P. Gabriel, C.S. Joiner, C. Valcke, Label-freedetection of DNA hybridization using carbon nanotube network field-effecttransistors, Proc. Natl. Acad. Sci. 103 (2006) 921–926, http://dx.doi.org/10.1073/pnas.0504146103.

18] X.W. Tang, S. Bansaruntip, N. Nakayama, E. Yenilmez, Y.L. Chang, Q. Wang,Carbon nanotube DNA sensor and sensing mechanism, Nano Lett. 6 (2006)1632–1636, http://dx.doi.org/10.1021/nl060613v.

19] B.K. Rajesh, S. Das, A. Srinives, Mulchandani, ZnS nanocrystals decoratedsingle-walled carbon nanotube based chemiresistive label-free DNA sensor,Appl. Phys. Lett. 98 (2011) 30–33, http://dx.doi.org/10.1063/1.3529951.

20] T.T. Tran, A. Mulchandani, Carbon nanotubes and graphene nano field-effecttransistor-based biosensors, TrAC - Trends Anal. Chem. 79 (2016) 222–232,http://dx.doi.org/10.1016/j.trac.2015.12.002.

21] L. Hu, D.S. Hecht, G. Grüner, Carbon nanotube thin films: fabrication,properties, and applications, Chem. Rev. 110 (2010) 5790–5844.

22] A. Béduer, F. Seichepine, E. Flahaut, C. Vieu, A simple and versatile microcontact printing method for generating carbon nanotubes patterns on varioussubstrates, Microelectron. Eng. 97 (2012) 301–305, http://dx.doi.org/10.1016/j.mee.2012.04.011.

23] Y. Battie, O. Ducloux, P. Thobois, N. Dorval, J.S. Lauret, B. Attal-Trétout, A.Loiseau, Gas sensors based on thick films of semi-conducting single walledcarbon nanotubes, Carbon 49 (2011) 3544–3552, http://dx.doi.org/10.1016/j.carbon.2011.04.054.

24] R.J. Chen, S. Bangsaruntip, K.A. Drouvalakis, N.W.S. Kam, M. Shim, Y. Li, W.Kim, P.J. Utz, H. Dai, Noncovalent functionalization of carbon nanotubes forhighly specific electronic biosensors, Proc. Natl. Acad. Sci. U. S. A. 100 (2003)4984–4989, http://dx.doi.org/10.1073/pnas.0837064100.

25] P.W. Barone, S. Baik, D.A. Heller, M.S. Strano, Near-infrared optical sensorsbased on single-walled carbon nanotubes, Nat. Mater. 4 (2005) 86–92, http://dx.doi.org/10.1038/nmat1276.

26] A. Javey, R. Tu, D.B. Farmer, J. Guo, R.G. Gordon, H. Dai, Transistors withchemically doped contacts high performance n-type carbon nanotubefield-effect transistors with chemically doped contacts, Nano Lett. 5 (2005)345–348, http://dx.doi.org/10.1021/nl047931j.

27] L.S. Panchakarla, A. Govindaraj, C.N.R. Rao, Boron- and nitrogen-doped carbonnanotubes and graphene, Inorganica Chim. Acta 363 (2010) 4163–4174,http://dx.doi.org/10.1016/j.ica.2010.07.057.

28] P. Ayala, R. Arenal, M. Rümmeli, A. Rubio, T. Pichler, The doping of carbonnanotubes with nitrogen and their potential applications, Carbon 48 (2010)575–586, http://dx.doi.org/10.1016/j.carbon.2009.10.009.

29] P. Ayala, A. Grüneis, C. Kramberger, M.H. Rümmeli, I.G. Solórzano, F.L. Freire,T. Pichler, Effects of the reaction atmosphere composition on the synthesis ofsingle and multiwalled nitrogen-doped nanotubes, J. Chem. Phys. 127 (2007),http://dx.doi.org/10.1063/1.2781509.

30] E.M.M. Ibrahim, V.O. Khavrus, A. Leonhardt, S. Hampel, S. Oswald, M.H.Rümmeli, B. Büchner, Synthesis, characterization, and electrical properties ofnitrogen-doped single-walled carbon nanotubes with different nitrogencontent, Diam. Relat. Mater. 19 (2010) 1199–1206, http://dx.doi.org/10.1016/j.diamond.2010.05.008.

31] S. Pregl, W.M. Weber, D. Nozaki, J. Kunstmann, L. Baraban, J. Opitz, T.Mikolajick, G. Cuniberti, Parallel arrays of Schottky barrier nanowire fieldeffect transistors: Nanoscopic effects for macroscopic current output, NanoRes. 6 (2013) 381–388, http://dx.doi.org/10.1007/s12274-013-0315-9.

32] F.M. Zörgiebel, S. Pregl, L. Römhildt, J. Opitz, W. Weber, T. Mikolajick, L.Baraban, G. Cuniberti, Schottky barrier-based silicon nanowire pH sensorwith live sensitivity control, Nano Res. 7 (2014) 263–271, http://dx.doi.org/10.1007/s12274-013-0393-8.

33] J. Schütt, B. Ibarlucea, R. Illing, F. Zörgiebel, S. Pregl, D. Nozaki, W.M. Weber, T.Mikolajick, L. Baraban, G. Cuniberti, Compact nanowire sensors probemicrodroplets, Nano Lett. 16 (2016) 4991–5000, http://dx.doi.org/10.1021/acs.nanolett.6b01707.

34] C. Park, S.W. Kim, Y.-S. Lee, S.H. Lee, K.H. Song, L.S. Park, Spray coating ofcarbon nanotube on polyethylene terephthalate film for touch panelapplication, J. Nanosci. Nanotechnol. 12 (2012) 5351–5355, http://dx.doi.org/10.1166/jnn.2012.6343.

35] H. Zhang, Y. Liu, M. Kuwata, E. Bilotti, T. Peijs, Improved fracture toughnessand integrated damage sensing capability by spray coated CNTs on carbonfibre prepreg, Compos. Part A Appl. Sci. Manuf. 70 (2015) 102–110, http://dx.doi.org/10.1016/j.compositesa.2014.11.029.

36] J.A. Rogers, R.G. Nuzzo, Recent progress in soft lithography, Mater. Today 8(2005) 50–56, http://dx.doi.org/10.1016/S1369-7021(05)00702-9.

37] H. Ago, K. Murata, M. Yumura, J. Yotani, S. Uemura, Ink-jet printing ofnanoparticle catalyst for site-selective carbon nanotube growth, Appl. Phys.Lett. 82 (2003) 811–813, http://dx.doi.org/10.1063/1.1540726.

38] A. Denneulin, J. Bras, F. Carcone, C. Neuman, A. Blayo, Impact of inkformulation on carbon nanotube network organization within inkjet printedconductive films, Carbon 49 (2011) 2603–2614, http://dx.doi.org/10.1016/j.

carbon.2011.02.012.

39] A. Shimoni, S. Azoubel, S. Magdassi, A. Kumar, C. Zhou, M. Layani, A.Kamyshny, S. Magdassi, T. Tokuno, M. Nogi, J. Jiu, T. Sugahara, K. Suganuma,M. Layani, S. Magdassi, H. Shimoda, S. Oh, H. Geng, R. Walker, X. Zhang, L.McNeil, O. Zhou, C. Liu, X. Yu, L. Hu, H. Kim, J. Lee, P. Peumans, Y. Cui, G. Eda, G.

[

rs B 249 (2017) 691–699

Fanchini, M. Chhowalla, S. De, T. Higgins, P. Lyons, E. Doherty, P. Nirmalraj, W.Blau, J. Boland, J. Coleman, H. Becerril, J. Mao, Z. Liu, R. Stoltenberg, Z. Bao, Y.Chen, D.S. Leem, A. Edwards, M. Faist, J. Nelson, D.D.C. Bradley, J.C. de Mello, L.Hu, D. Hecht, G. Gruner, D. Hecht, L. Hu, G. Irvin, F. Mirri, A. Ma, T. Hsu, N.Behabtu, S. Eichmann, C. Young, D. Tsentalovich, M. Pasquali, C. Park, S.W.Kim, Y.-S. Lee, S.H. Lee, K.H. Song, L.S. Park, J. Jo, J. Jung, J. Lee, W. Jo, S. De, P.Lyons, S. Sorel, E. Doherty, P. King, W. Blau, P. Nirmalraj, J. Boland, V. Scardaci,J. Joimel, J. Coleman, M. Castro, A. Lasagni, H. Schmidt, F. Mucklich, Y. Zhou, L.Hu, G. Gruner, Z. Wu, Z. Chen, X. Du, J. Logan, J. Sippel, M. Nikolou, K. Kamaras,J. Reynolds, D. Tanner, A. Hebard, A. Rinzler, M. Lima, M. de Andrade, C.Bergmann, S. Roth, T. Mustonen, K. Kordas, S. Saukko, G. Toth, J. Penttilla, P.Helisto, H. Seppa, H. Jantunen, Y. Lee, S. Kim, K. Lee, N. Myung, Y. Choa, R.P.Tortorich, J.-W. Choi, S. Azoubel, S. Shemesh, S. Magdassi, R. Deegan, O.Bakajin, T. Dupont, G. Huber, S. Nagel, T. Witten, R.D. Deegan, L. Cui, J. Zhang,X. Zhang, L. Huang, Z. Wang, Y. Li, H. Gao, S. Zhu, T. Wang, B. Yang, J. Perelaer,P. Smith, C. Hendriks, A. van den Berg, U. Schubert, M. Layani, M. Gruchko, O.Milo, I. Balberg, D. Azulay, S. Magdassi, M. Majumder, C. Rendall, M. Li, N.Behabtu, J.A. Eukel, R.H. Hauge, H.K. Schmidt, M. Pasquali, A. Denneulin, J.Bras, F. Carcone, C. Neuman, A. Blayo, H. Geng, K. Kim, K. So, Y. Lee, Y. Chang,Y. Lee, D. Shin, J. Lee, Y. Kim, S. Yu, S. Park, J. Yoo, B. Liu, C. Hsu, W. Wang, Z. Li,H. Kandel, E. Dervishi, V. Saini, Y. Xu, A. Biris, D. Lupu, G. Salamo, A. Biris,Inkjet printing of flexible high-performance carbon nanotube transparentconductive films by “coffee ring effect”, Nanoscale 6 (2014) 11084–11089,http://dx.doi.org/10.1039/C4NR02133A.

40] F.N. Ishikawa, H. Chang, K. Ryu, P. Chen, A. Badmaev, L. Gomez, D. Arco, G.Shen, C. Zhou, Transparent Electronics Based on, ACS Nano. 3 (2009) 73–79,http://dx.doi.org/10.1021/nn800434d.

41] D. Shi, L. Wei, J. Wang, J. Zhao, C. Chen, D. Xu, H. Geng, Y. Zhang, Solid organicacid tetrafluorohydroquinone functionalized single-walled carbon nanotubechemiresistive sensors for highly sensitive and selective formaldehydedetection, Sens. Actuators B Chem. 177 (2013) 370–375, http://dx.doi.org/10.1016/j.snb.2012.11.022.

42] S.R. Shin, R. Farzad, A. Tamayol, V. Manoharan, P. Mostafalu, Y.S. Zhang, M.Akbari, S.M. Jung, D. Kim, M. Comotto, N. Annabi, F.E. Al-Hazmi, M.R. Dokmeci,A. Khademhosseini, A Bioactive carbon nanotube-based ink for printing 2Dand 3D flexible electronics, Adv. Mater. 28 (2016) 3280–3289, http://dx.doi.org/10.1002/adma.201506420.

43] M. Zheng, A. Jagota, E.D. Semke, B.A. Diner, R.S. McLean, S.R. Lustig, R.E.Richardson, N.G. Tassi, DNA-assisted dispersion and separation of carbonnanotubes, Nat. Mater. 2 (2003) 338–342, http://dx.doi.org/10.1038/nmat877 M4 - Citavi.

44] N. Varghese, U. Mogera, A. Govindaraj, A. Das, P.K. Maiti, A.K. Sood, C.N.R. Rao,Binding of DNA nucleobases and nucleosides with graphene, ChemPhysChem10 (2009) 206–210, http://dx.doi.org/10.1002/cphc.200800459.

45] C.H. Lu, H.H. Yang, C.L. Zhu, X. Chen, G.N. Chen, A graphene platform forsensing biomolecules, Angew. Chemie - Int. Ed. 48 (2009) 4785–4787, http://dx.doi.org/10.1002/anie.200901479.

46] A. Fediai, D.A. Ryndyk, G. Seifert, S. Mothes, M. Claus, M. Schröter, G.Cuniberti, Towards an optimal contact metal for CNTFETs, Nanoscale 8 (2016)10240–10251, http://dx.doi.org/10.1039/C6NR01012A.

47] A. Fediai, D.A. Ryndyk, G. Seifert, S. Mothes, M. Schroter, M. Claus, G.Cuniberti, Impact of incomplete metal coverage on the electrical properties ofmetal-CNT contacts: A large-scale ab initio study, Appl. Phys. Lett. 109 (2016)103101, http://dx.doi.org/10.1063/1.4962439.

48] M.D. de Jong, C.P. Simmons, T.T. Thanh, V.M. Hien, G.J.D. Smith, T.N.B. Chau,D.M. Hoang, N. Van Vinh Chau, T.H. Khanh, V.C. Dong, P.T. Qui, B. Van Cam,D.Q. Ha, Y. Guan, J.S.M. Peiris, N.T. Chinh, T.T. Hien, J. Farrar, Fatal outcome ofhuman influenza A (H5N1) is associated with high viral load andhypercytokinemia, Nat. Med. 12 (2006) 1203–1207, http://dx.doi.org/10.1038/nm1477.

49] Q. Zhao, X. Lu, C.-G. Yuan, X.-F. Li, X.C. Le, Aptamer-linked assay for thrombinusing gold nanoparticle amplification and inductively coupled plasma−massspectrometry detection, Anal. Chem. 81 (2009) 7484–7489, http://dx.doi.org/10.1021/ac900961y.

50] L. Römhildt, C. Pahlke, F. Zörgiebel, H.G. Braun, J. Opitz, L. Baraban, G.Cuniberti, Patterned biochemical functionalization improves aptamer-baseddetection of unlabeled thrombin in a sandwich assay, ACS Appl. Mater.Interfaces 5 (2013) 12029–12035, http://dx.doi.org/10.1021/am4038245.

51] K. Malecka, I. Grabowska, J. Radecki, A. Stachyra, A. Góra-Sochacka, A. Sirko, H.Radecka, Voltammetric detection of a specific DNA sequence of avianinfluenza virus H5N1 using HS-ssDNA probe deposited onto gold electrode,Electroanalysis 24 (2012) 439–446, http://dx.doi.org/10.1002/elan.201100566.

52] K. Kurzatkowska, A. Sirko, W. Zagórski-Ostoja, W. Dehaen, H. Radecka, J.Radecki, Electrochemical label-free and reagentless genosensor based on anion barrier switch-off system for DNA sequence-specific detection of theavian influenza virus, Anal. Chem. 87 (2015) 9702–9709, http://dx.doi.org/10.1021/acs.analchem.5b01988.

53] K. Malecka, I. Grabowska, J. Radecki, A. Stachyra, A. Góra-Sochacka, A. Sirko, H.Radecka, Electrochemical detection of avian influenza virus genotype usingamino-ssDNA probe modified gold electrodes, Electroanalysis 25 (2013)

1871–1878, http://dx.doi.org/10.1002/elan.201300113.

54] I. Grabowska, K. Malecka, A. Stachyra, A. Góra-Sochacka, A. Sirko, W.Zagórski-Ostoja, H. Radecka, J. Radecki, Single electrode genosensor forsimultaneous determination of sequences encoding hemagglutinin and

ctuato

[

B

YniHn

Vnt

YiPrn

BNa

LSd

Nanotechnology at TU Dresden and the Max Bergmann Center of Biomaterials Dres-den, Germany. He is a member of the TU Dresden School of Engineering Science(Materials Science) and of the School of Mathematics and Natural Sciences (Physics).His research work addresses molecular and organic electronics, bionanotechnology,

Y. Fu et al. / Sensors and A

neuraminidase of avian influenza virus type H5N1, Anal. Chem. 85 (2013)10167–10173, http://dx.doi.org/10.1021/ac401547h.

55] S. Singh, A. Kumar, S. Khare, A. Mulchandani, Rajesh, Single-walled carbonnanotubes based chemiresistive genosensor for label-free detection of humanrheumatic heart disease, Appl. Phys. Lett. 105 (2014), http://dx.doi.org/10.1063/1.4902447.

iographies

angxi Fu received her M. Eng. degree in the School of Materials Science and Engi-eering from North Western Polytechnical University, China, in 2013. Currently she

s a Ph.D. student at the chair Materials Science and Nanotechnology, TU Dresden.er research interests include synthesis, characterization and processing of carbonanotubes as well as fabrication of CNT based gas- and bio-sensors.

ictor Romay is a master student at the chair of Materials Science and Nanotech-ology at TU Dresden. His research interest is the application of nanotechnology inhe healthcare diagnostics and therapeutics.

e Liu received his M. Eng. degree in the School of Materials Science and Engineer-ng from North Western Polytechnical University, China, in 2014. Currently he is ah.D. student at the chair Materials Science and Nanotechnology, TU Dresden. Hisesearch interests include functionalization and characterization of carbon basedano-structures for energy and sensing application.

ergoi Ibarlucea is a postdoctoral researcher at the chair Materials Science andanotechnology, TU Dresden. With background in biology, his research interests

re biosensors, lab on a chip and surface chemistry.

arysa Baraban currently leads BioNanoSensoricsgroup at the chair Materialscience and Nanotechnology, TU Dresden. Her major research interests are theevelopment of high throughput biosensors and lab on chip systems.

rs B 249 (2017) 691–699 699

Vyacheslav Khavrus is a scientific researcher at the Leibniz Institute for SolidState and Materials Research Dresden (IFW-Dresden). He has rich experience in thedevelopment of methods for synthesis and functionalization of tailor-made carbonnanostructures for their application in nanomedicine or as components of sensorsand thermoelectric devices.

Prof. Alicja Bachmatiuk currently heads the electron microscopy laboratories atthe Centre of Polymer and Carbon Materials, Polish Academy of Sciences. She isalso a visiting professor at Soochow University, China. She has a strong interestin developing electron microscopy of 2D materials as well as their synthesis andfunctionalization.

Prof. Mark Rummeli is currently a professor at the College of Physics, Opto-electronics and Energy, Soochow University, Suzhou, China. His research focuseson understanding structure property relationships in nanomaterials, nanoma-terials synthesis and nanomaterials applications in ion insertion batteries andbio-applictions. He also heads laboratories at the Polish Academy of Sciences andthe Leibniz Institute for Solid State and Materials Research Dresden in Germany.

Viktor Bezugly is a group leader at the chair Materials Science and Nanotechnology,TU Dresden and a researcher at the Leibniz Institute for Solid State and MaterialsResearch Dresden. His research interests are synthesis, modification and applicationof nanostructured materials.

Prof. Gianaurelio Cuniberti holds since 2007 the chair of Materials Science and

nanostructures, methods development and technology transfer.