Field Effect Transistors using Carbon Nanotubes Biosensor David

Field Effect Transistors using Carbon Nanotubes Biosensor David Hecht

Why Study NTFET’s? Size!!! • Diameter ≈ 1 nm, length ≈ 1 micron • Quasi 1 -D object Bacteria 1 m 100 nm Virus Proteins 10 nm 1 nm DNA 0. 1 nm Biosensor High Mobility Semiconductors • Moore’s Law • Can be p-type or n-type • Flexible Transistors/Electronics? Sensor • Biosensor (walls can be functionalized) • Chemical Sensor • high sensitivity/large surface area

How a NTFET Works p-type in air Source Drain Insulating Layer (Si. O 2) Conducting “Gate” (p type Si) VRprot Vgate Vsd Vg = 0 Isd

How a NTFET Works p-type in air S D Insulating Layer (Si. O 2) Conducting “Gate” (p type Si) VRprot Vsd Rprot Vg Isd Vg = positive, Holes are Depleted

How a NTFET Works p-type in air S D Insulating Layer (Si. O 2) Conducting “Gate” (p type Si) VRprot Vsd Rprot Vg Isd Vg = negative, Holes are enhanced

Nanotube FET transistor S D Si. O 2 AFM image Si back gate Vsd Vg Isd Ideal NTFET Device Conductance ( S) A p-type A. Max Conductance B. Modulation – Signal to Noise B C C. Transconductance (slope at zero gate) Mobility of Carriers (electrons or holes) D Gate Voltage (V) D. Threshold Shift – Changes in Doping

Effect of charge transfer on the device electronics Detection of gases 1 NH 3 el. donor NO 2 el. acceptor 1 Bradley, K. ; Gabriel, J. -C. P. ; Briman, M. ; Star, A. ; Grüner, “Charge Transfer from Ammonia Physisorbed on Nanotubes Phys. Rev. Lett. 2003, 91, 218301.

Novel active electronic devices Single Nanotube vs. Network Single tube/fiber channel • greater sensitivity • individual device fabrication (basic research) Network channel • easier/more consistent device fabrication • Applications

Metallic vs. Semiconducting Determines geometry and diameter Armchair: (n, n) Zig-Zag: (n, 0) If n – m is a multiple of 3, the nanotube is metallic. 1/3 of NT’s are metallic, 2/3 are semi-conducting!!!!

Metallic Tubes are the Enemy I I SD SD On/Off Ratio Low On/Off Ratio High VG VG Vg Few Metallic Tubes D Lots of Metallic Tubes NT Film S x Si gate Si. O 2 V Metallic Tubes Act To Screen Potential from outermost tubes of Film!!!

Deposition Techniques Want: Uniform Film of individually separated NT’s n Direct Deposition Drop Casting -- Flocculation due to Van der Waals between tubes limits uniformity. n Spin Coating – work in progress n Langmuir-Blodgett/Quasi-Langmuir-Blodgett n n Separate Tubes using Solubilization Agents n n Starch/Enzymes Pm. PV

Quasi-Langmuir-Blodgett Film DEPOSITION METHOD 1. Dissolve NT’s in Solvent by Sonication. Single tube dissolution is ideal. For a solvent we used a 10: 1 mixture of ortho-xylene and 1, 2 -dichlorobenzene.

Quasi-Langmuir-Blodgett Film DEPOSITION METHOD 2. Quickly suck fluid through a porous alumina filter (pore size = 20 nm) Alumina Filter (pore size = 20 nm)* * From Whatman website

Quasi-Langmuir-Blodgett Film DEPOSITION METHOD 3. While film is still slightly damp with solvent, wash water over filter. Film will break off as a “raft” and float to top.

Quasi-Langmuir-Blodgett Film DEPOSITION METHOD 4. Put in substrate and suck out water to redeposit the film. Substrate

Why 10: 1 solvent mixture? n Ortho-xylene: Dichlorobenzene solvent mixture used for 3 reasons 1) High nanotube solubility ≈ 15 mg/L n 2) Specific Gravity < 1 (so rafts can float) n 3) Immiscibile in water n

Quasi-Langmuir-Blodgett Film on Glass Slide ADVANTAGES • Film fairly uniform over large area • Thickness of Film controllable. (20 nm 1 um) • Room Temperature Technique vs. CVD at DISADVANTAGE • Film’s too thick. Can’t get monolayer.

Does water Immersion affect Films? 200 nm SEM of Nanotube Film 1 • MECHANICAL • • Film consists of well separated ropes of ≈10 nm before and after immersion ELECTRICAL • 1 N. Temperature Dependence of Resistivity 1 Immersion in water does not affect DC resistivity. Peter Armitage

Quasi-LB Film: Are They Uniform? 30 nm thick film 80 nm thick film 50 microns Average “roughness” = 36 nm Average “roughness” = 14 nm “Roughness” = Σ|(xi - xave)|/n *Data taken in UCLA Nanolab using Nano-Or 3 -D Scope 2000 for 2 D profiles, and Dektak 8 profiler to measure film thickness

Device Fabrication Si. O 2 Steps to making Device 1) 5000 Ao Si. O 2 on doped Silicon • commercially bought, HF remove Si. O 2 from one side 2) Deposit NT Film 3) Evaporate Gold source and drain through shadow masking 4) Use Silver epoxy to attach wires 5) Clamp onto metal chuck to apply Vg • Dielectric Breakdown Electric Field in Si. O 2 = 1 x 107 V/cm

Measurement Setup p-type in air Output (Vg ) ± 100 V quasi-AC Sawtooth Waveform. n 200 V S D Si. O 2 p type Si back gate VRprot Output bias voltage (100 m. V) across SD, and measure Isd n n Vsd Measure Voltage across Rprot to get Ileakage Rprot Vg Isd

Transistor Characteristics Film = 4000 Angstroms thick (≈400 NT layers) => 2% Modulation!!

Transistor Characteristics Vg = 0 Vg = 50 Vg = 100 Blew on Sample Notice the slow drift…Need to stabilize temperature for future measurements

Transistor Characteristics Film = 800 Angstroms thick (≈80 NT layers) => 20% Modulation!!

Vsd vs Isd

Transistor Characteristics Film = 300 Angstroms thick (≈30 NT layers) => 40% Modulation!!

Exponentially Better with Film thinness

Calculation of Mobility n Quadratic Model of MOSFET: n ISD = (μCox. W)[(VGS – VT)VDS – VDS 2/2] for VDS << VGS – VT L n Slope of IVg curve = μCox. WVDS L Plugging in the numbers yields mobility of 0. 9 cm 2/V*s Mobility's: Single Carbon NT = 105 Silicon = 102 -103 NT Network = 101 Organic Semiconductor = 10 -4 - 10 -1 High Mobility means device can operate at Higher frequency!!!

Liquid Gating 50 % Modulation!! Alumina Filter on glass Liquid Gating Setup Data for 1000 Angstrom thick sample 1 • Larger Modulation than Bottom Gating => liquid penetrates porous film • Liquid Gating useful for protein/biomolecule detection. 1 Data taken with aid of Mikhail Briman

Nanotube Reflux in Nitric Acid O H Original Nanotube After 20 Hour Reflux in HNO 3 O OH O Carboxylic Acid Group After filtering and rinsing in water O- At PH 7 Becomes Polar

Did Refluxing Work? 1 Minute after Sonication 10 Minute after Sonication

Did the Reflux Work? 24 Hours after Sonication

Future Work n Improve Device Characteristics n n n Improve Probe Station n n True AC Setup Gold Pogo Pin probe/Micrometer positioner Temperature/Humidity Control Chamber Characteristics vs. Network Density n n Thinner Films More Dispersed Tubes Separate Semiconducting and Metallic Tubes Mobility vs. applied pressure Photo-lithographic Mask for micro scale geometries Study 2 -D Percolation problem of random array of semiconducting rods E-Beam Lithography for nano scale geometries Protein detection

Conclusion n Created NT network transistor using room temperature fabrication process n Film too thick to get good characteristics.

Thanks The Gruner Group: Peter Armitage, M. Briman, Erika Artukovic, Liangbing Hu, George Gruner n The Chemists: Erik Richman, Will Molenkamp (Tolbert); Matt Spotnitz (Kaner). n Steve Franz (Nanolab). n MCTP n