ChargeBased Biosensor Using Carbon Nanotube Transistors Array Presenter

Charge-Based Biosensor Using Carbon Nanotube Transistors Array Presenter: Jui-Ping Chiang

Outline • Introduction of field effect detection of DNA hybridization • Experiments – Fabrication – Hybridization – Measurements • Results

Detection of hybridization • Hybridization: combining complementary, single-stranded nucleic acids into a single molecule • A single strand of DNA with 15 bases long has a charge of -15 q (2. 4 a. C) • Charges adsorbed on the electrolyte -insulator-silicon (EIS) structure produce a field effect to shift the flat band potential and/or change the capacitive/resistive impedance across the EIS structure 20 qe 10 qe DNA hyb. Sensor

• In a very simplified model, the threshold voltage shift is given by • Express the effective capacitance as: e. g. For 10 m. M TE buffer used during measurement, • For DNA hybridization, we get positive threshold voltage shift

Compared to fluorescent method. . . Fluorescent Field effect • • Pros: – Well-established – Very sensitive (sensitivity depends of sophistication of detection instrument) – 3 zeptomolar (10 -21 M) has been reported Cons: – Instrument-heavy: photomultiplier, electronics – Labeled Other labeled methods: Electrochemical redox… • • Pros: – Simple device structure, scalable – Probe attached to oxide: protocols are compatible – label-free – Sensitivity around n. M range Cons: – Real-monitoring not possible, unless charge-less PNA probes are used Other labeled methods: surface plasmon resonance; ellipsometry; THz-transmission analysis; capacitance measurement; SAW devices…

Device fabricaiton: carbon nanotube growth • Prepare iron catalyst: - Brief iron evaporation in UHV • CNT growth in CVD furnace 850 ºC ethylene hydrogen methane argon • Cool down • Most single walled, and contain both semiconducting and metallic type in roughly 2: 1 proportion (metallic CNTs would be burned out later) exhaust

Device fabrication: microlithography Carbon nanotube mats on silicon dioxide substrate Spin on photoresist, prebake at 90 ºC for 1 min carbon nanotubes silicon dioxide silicon UV exposure Metallization gold Photoresist development mask photoresist Final structure: Lift-off S Similar steps to mask active area during oxygen plasma etch D Al 2 O 3 Si. O 2

• The active area between source and drain electrodes where the nanotube channel sits is 10 um long and 50 um wide. D S

DNA immobilization and hybridization 1. 2. 3. 4. 5. 6. Aluminium oxide surface hydroxylation: O 2 plasma Exposed to 3 -mercapto-propyl-trimethoxy-silane (MPTMS) vapor 10 m. M of Acrydite™-modified probe oligonucleotide dissolved in TE buffer (p. H 7), was pipetted to the chip. Dry overnight Measure the transfer characteristics (IDS-VGS curves) Incubate the target DNA of on the chip up to 2 hrs in 100% humidity Wash the whole chip throughly in TE buffer

Measurement • • Applied fixed 100 m. V bias between the drain and source electrode. IDS was recorded as VGS was swept to generate transistor transfer characteristic curve.

Results 26 probe only 1 u. M hybridization dehybridization wash 24 22 IDS ( A) 20 18 16 14 fluorescent image of 1 u. M complementary target hybridization 12 10 8 -1 -0. 5 0 VGS (V) 0. 5 1 1. 5 • The shift in threshold voltage, which arises from the additional surface charges, serves as the hybridization signal fluorescent image after dehybridization wash to remove target DNA

1. 4 • Threshold voltage shift of all transistors are positive upon DNA hybridization, and the recovery upon target DNA denaturation are universal The amount of voltage shift should not depend on the initial position of the transistor’s threshold voltage 1. 3 threshold voltage (V) • 1. 2 1. 1 1 0. 9 0. 8 probe only 1 u. M hybridization dehybridization wash 0. 7 0 5 10 15 device no. 20 25 30

• For different target DNA concentrations The non-complementary signal serves as the noise floor – Sensitivity: 10~100 n. M – To be useful for diagnostic applications, the sensitivity of 1 p. M or better is needed – Compelling estimates that suggest the feasilibity of achieving this level of detection through improved probe immobilization chemistry, better hybridization protocols and improved performance by usingle CNT transistors

Thanks for your attention.