Interfacial Fields in Organic Field-effect Transistors and Sensors

download Interfacial Fields in Organic Field-effect Transistors and Sensors

of 158

Transcript of Interfacial Fields in Organic Field-effect Transistors and Sensors

  • 8/11/2019 Interfacial Fields in Organic Field-effect Transistors and Sensors

    1/158

    INTERFACIAL FIELDS IN ORGANIC FIELD-EFFECTTRANSISTORS AND SENSORS

    By

    Thomas J. Dawidczyk

    A dissertation submitted to Johns Hopk ins University in conformity with the

    requirements for the degree o f Doctor o f Philosophy

    Baltimore, Maryland

    M arch, 2013

    2013 Thomas J. Dawidczyk

  • 8/11/2019 Interfacial Fields in Organic Field-effect Transistors and Sensors

    2/158

    UMI Number: 3572759

    All rights reserved

    INFORMATION TO ALL USERS

    The quality of this reproduction is dependent upon the quality o f the copy submitted.

    In the unlikely event that the author did not send a complete manuscript

    and there are missing pages, these will be noted. Also, i f material had to be removed,

    a note will indicate the deletion.

    Di!ss0?t& iori Piib list Mlg

    UMI 3572759

    Published by ProQuest LLC 2013. Copyright in the Dissertation held by the Author.

    Microform Edition ProQuest LLC.

    All rights reserved. This work is protected against

    unauthorized copying under Title 17, United States Code.

    ProQuest LLC

    789 East Eisenhower Parkway

    P.O. Box 1346

    Ann Arbor, Ml 48106-1346

  • 8/11/2019 Interfacial Fields in Organic Field-effect Transistors and Sensors

    3/158

    ABSTRACT

    Organic electronics are currently being commercialized and present a v iable alternative to

    conventional electronics. These organic materials offer the ability to chemically manipulate the

    molecule, allowing for more facile mass processing techniques, which in turn reduces the cost.

    One application where organic semiconductors (OSCs) are being investigated is sensors. This

    work evaluates an assortment of n- and p-channel semiconductors as organic field-effect

    transistor (OFET) sensors. The sensor responses to dinitrotoluene (DNT) vapor and solid along

    with trinitrotoluene (TNT) solid were studied. Different semiconductor materials give different

    magnitude and direction of electrical current response upon exposure to DNT. Additional OFET

    parameters mobility and threshold voltage further refine the response to the DNT with

    each OFET sensor requiring a certain gate voltage for an optimized response to the vapor. The

    pattern o f responses has sufficient diversity to distinguish DNT from other vapors.

    To effectively use these OFET sensors in a circuit, the threshold voltage needs to be

    tuned for each transistor to increase the efficiency of the circuit and maximize the sensor

    response. The threshold voltage can be altered by embedding charges into the dielectric layer of

    the OFET. To study the quantity and energy of charges needed to alter the threshold voltage,

    charge carriers were injected into polystyrene (PS) and investigated with scanning Kelvin probe

    microscopy (SKPM) and thermally stimulated discharge current (TSDC). Lateral heterojunctions

    of pentacene/PS were scanned using SKPM, effectively observing polarization along a side view

    of a lateral nonvolatile organic field-effect transistor dielectric interface. TSDC was used to

    observe charge migration out of PS films and to estimate the trap energy level inside the PS,

    using the initial rise method.

    The process was further refined to create lateral heterojunctions that were actual working

    OFETs, consisting of a PS or poly (3-trifluoro)styrene (F-PS) gate dielectric and a pentacene

    OSC. The charge storage inside the dielectric was visualized with SKPM, correlated to a

  • 8/11/2019 Interfacial Fields in Organic Field-effect Transistors and Sensors

    4/158

  • 8/11/2019 Interfacial Fields in Organic Field-effect Transistors and Sensors

    5/158

    ACKNOW LEDGEM ENT

    I would like to thank my PhD advisor, Howard Katz, who was always been available to

    help with guidance and support; 1 am very fortunate to have had him as a mentor. He is my role

    model for a scientist, mentor and teacher.

    I am thankful of the work done by the staff members in Materials Science, specifically

    Marge Weaver, Jeanine Majewski, Dot Reagle and Mark Koontz. They are the primary resources

    for all graduate students in the department. Additionally, I want to thank the members o f my PhD

    committee: En Ma, Jonah Erlebacher, Nina Markovic, Daniel Reich and Bob Cammarata.

    The Katz research group members, past and present, have been ideal research

    companions. Their varied expertise helped further my knowledge through insightful discussions

    about research and experiments. Additionally, Recep Ozgun and Gary Johns were great

    collaborators for my thesis research. JHU Applied Physics Lab researchers Andrew Mason and

    Joan Hoffman were always available if I had questions. I am grateful to have been able to work

    with Subodh Mhaisalkar, Nripan Matthews, and Verawati Tjoa at Nanyang Technical University.

    Lastly Bill Wilson was a great resource for spectroscopy.

    I would not have been able to complete my PhD if it were not for my entire family. My

    parents have been extremely encouraging and supporting during this process and my brother ,

    Dan, has been a great sounding board. There are no words to convey the importance my wife,

    Charli, has had in my success at JHU. Her faith in me and my intellect has helped me

    tremendously during my time at JHU.

  • 8/11/2019 Interfacial Fields in Organic Field-effect Transistors and Sensors

    6/158

    CONTENTS

    ABSTRACT............................................................................................................................................ ii

    ACKNOWLEDGEMENT................................................................................................................... iv

    LIST OF TABLES.............................................................................................................................. viii

    LIST OF FIGU RES ............................................................................................................................. ix

    CHAPTER I: Introduction to Organic Semiconductors (OSCs) and Devices .......................... 1

    1. Introduction to Organic Semiconductors.................................. 1

    1.1 Organic Semiconductor Electronic Structure......................................................................... 1

    1.2 Applications of Organic Semiconductors............................................................................... 2

    1.3 Operation of Field-effect Transistors...................................................................................... 3

    1.4 Charge Density in a MIS Capacitor......................................................................................... 5

    1.5 Charge Transport in the Transistor Channel .......................................................................... 6

    1.6 Mobility in OFETs................................................................................................................... 7

    1.7 Extracting Mobility and Threshold Voltage........................................................................... 8

    1.8 Threshold Voltage in OFETs .................................................................................................. 8

    1.9 Semiconductor-Dielectric Interface and Traps...................................................................... 9

    1.10 Thin Film Deposition and Morphology.............................................................................. 10

    2. Mechanisms of Threshold Voltage Shifting............................................................................... 12

    2.1 Electrostatic Threshold Voltage Shifting............................................................................. 12

    2.2 Non Electrical Threshold Voltage M odulation .................................................................... 16

    2.3 Operational Stability............................................................................................................... 17

    CHAPTER II: Chemical Sensing with Organic Field-Effect Tran sistors ................................. 18

    1. Sensor Testing................................................................................................................................ 18

    2. Previous Work on OSC Vapor Sensing ....................................................................................... 18

    3. Enhanced Sensor Response with Receptors.............................................................................. 21

    4. Sensing Mechanism...................................................................................................................... 22

    5. DNT Vapor Sensing...................................................................................................................... 24

    5.1 Experimental Setup.............................................................................................................. 25

  • 8/11/2019 Interfacial Fields in Organic Field-effect Transistors and Sensors

    7/158

    5.2 DNT Sensor Results............................................................................................................... 25

    5.3 Proposed Sensing Mechanism s............................................................................................. 52

    6 . TNT Sensing........................................ ........................................................................................ 53

    7. Sensing Multiple Vapors.............................................................................................................. 56

    8 . Maximum Sensitivity of an OFET as a Function of Gate Voltage ........................................... 59

    9. Future Work................................................................................................................................... 59

    CHAPT ER III: Storing Charge in Dielectrics............................................................................... 61

    1. Gate Bias Stress............................................................................................................................ 61

    2. SKPM Operation........................................................................................................................... 61

    3. Previous SKPM Experiments on OFETs.................................................................................... 62

    4. Visualizing the Stored Charge ..................................................................................................... 65

    4.1 Experimental setup.................................................................................. 65

    4.2 SKPM Scans of the Interfaces............................................................................................... 67

    4.3 Analyzing the SKPM Scans.................................................................................................. 74

    5. Quantities and Energies of Injected Cha rge ............................................................................... 76

    7. Conclusions and Future Work..................................................................................................... 78

    CHAPTER IV: C orrelating the Surface Potential to T hresholdVoltage Shifts....................... 80

    1. Lateral Transistors........................................................................................................................

    80

    2. Fabricating Lateral OFETs........................................................................................................... 80

    3. Testing Lateral OFETs Containing PS ........................................................................................ 84

    4. Changing the Dielectric in Lateral OFETs................................................................................. 96

    4.1 Lateral OFETs with a F-PS Dielectric.................................................................................. 96

    4.2 Lateral OFETs with a PMMA D ielec tric........................................................................... 103

    5. Lateral OFETs with an Air Gap Dielectric ............................................................................... 106

    6 . Spectroscopy................................................................................................................................ 107

    7. Conclusions and Future W ork.................................................................................................... 109

    CHAP TER V: Refining Gate Dielectric Prop erties to W ithstan dGate Bias Stress.............. I l l

    1. Gate Bias Stress ........................................................................................................................ I l l

    vi

  • 8/11/2019 Interfacial Fields in Organic Field-effect Transistors and Sensors

    8/158

    1.1 Extrinsic Factors Contributing to Gate Bias Stress............................................................I l l

    1.2 Intrinsic Factors Contributing to Gate Bias Stress ............................................................ 112

    2. Mechanisms of Gate Bias Stress................................................................................................. 113

    3. The Influence of the Dielectric on Gate Bias Stress ................................................................. 115

    3.1 The Effect of Dielectric Addi tives...................................................................................... 119

    3.2 F-PS Dielectric Gate Bias Response................................................................................... 129

    4. Conclusions and Future W ork.................................................................................................... 130

    BIBLIOGRAPHY: ............................................................................................................................. 132

    VITA..................................................................................................................................................... 141

    vii

  • 8/11/2019 Interfacial Fields in Organic Field-effect Transistors and Sensors

    9/158

    LIST OF TABLES

    1 The response of four different electrode pairs on a PQT12 sensor to DNT and TNT transfer

    from a PDMS stamp...................................................................................................................... 56

  • 8/11/2019 Interfacial Fields in Organic Field-effect Transistors and Sensors

    10/158

    LIST OF FIGURES

    1.1 Schematic view of a bottom contact, bottom gate OFET. S refers to the source electrode while

    D refers to the drain electrode......................................................................................................... 3

    1.2 Output (left) and transfer (right) curves for an n-channel transistor. Note the y-axis in the

    right graph is for the square root of the drain current .................................................................. 4

    1.3. Energy band diagram of a p-channel MIS capacitor (a) at thermal equilibrium and (b) in

    accumulation mode with a negative gate voltage applied............................................................ 6

    1.4 A point-to-grid corona triode schematic. A corona is generated near the point and ions flow

    through the grid and impart their charge in the dielectric layer.Error! Bookmark not

    defined.4

    2.1 Schematic of an analyte delivery system for a OFET vapor sensor. The arrows symbolize the

    analyte...................................................................................... Error! Bookmark not deflned.9

    2.2 The effective voltage seen by the molecular semiconductor as the distance from the dielectric

    interface is increased. (13)............................................................................................................ 24

    2.3 Chemical structures of DNT (left) and TNT (right)...................................................................... 25

    2.4 The chemical structure of bis-CF3 NTCDI (a) and response of a 6 nm thick OSC film FET

    upon exposure to DNT vapors at various concentrations.......................................................... 27

    2.5 The chemical structure of C2PhF5 NTCDI (a) and response of a 15 run thick OSC film FET

    upon exposure to DNT vapor....................................................................................................... 30

    2.6 The chemical structure of 6 PTTP6 (top) and response of a 6 nm thick OSC film FET upon

    exposure to DNT vapor (bottom)............................................ 3Error! Bookmark not defined.

    2.7 The chemical structure of 2PTTP2 (a) and response of a 4.5 nm thick OSC film FET upon

    exposure to DNT vapor................................................................................................................. 33

    2.8 The chemical structure of CuPc (a). The mobility (b) and threshold voltage (c) response o f a

    20 nm thick OSC film FET upon exposure to 4 LPM DNT vapor........................................... 35

  • 8/11/2019 Interfacial Fields in Organic Field-effect Transistors and Sensors

    11/158

    2.9 The chemical structure of a-6T (top) with the corresponding response (bottom) upon exposure

    to 4 LPM DNT vapor.................................................................................................................... 36

    2.10 The chemical structure of PTPTP (top) with the corresponding response (bottom) upon

    exposure to 4 LPM DNT vapor.................................................................................................... 37

    2.11 The chemical structure of PQT12 (a) and response of an OFET upon exposure to DNT

    vapor. The response o f mobility (b); response o f threshold voltage (c)................................... 39

    2.12 The chemical structure of PIFTBT-10 (a) and response of an OFET upon exposure to DNT

    vapor............................................................................................................................................... 41

    2.13 The response of mobility (a) and threshold voltage (b) to DNT. The gray bars indicate air

    flowrate over DNT at 4 LPM ....................................................................................................... 43

    2.14 The response of NTCDI, pentacene and two bilayer devices to DNT. The normalized

    mobility (a) and normalized response o f the threshold voltage (b).......................................... 45

    2.15 The chemical structure of H060PT (top) and the OFET response to DNT vapor for a 3nm

    thick 6PTTP6 film with 3 nm o f H 06 0P T and 6PTTP6co-evaporated on top o f it 46

    2.16 The chemical structure of porphyrin (top) and the two different device structures that have

    porphyrin as a receptor.................................................................................................................. 48

    2.17 The normalized mobility (a) and threshold voltage (b) response to 4 LPM DNT vapor for

    different blend ratios..................................................................................................................... 49

    2.18 The normalized mobility (a) and threshold voltage (b) response to DNT o f bilayer films with

    different spin coating speeds for the PS/porphyrin layer........................................................... 50

    2.19 The mobility (a) and threshold voltage (b) response to DNT of bilayer films with different

    weight concentrations for the porphyrin layer........................................................................... 51

    2.20 Schematic of the procedure for testing TNT and DNT transfer to an OFET sensor. The

    sensor would have the PDMS stamp placed on it to get a baseline source drain cu rre nt 54

  • 8/11/2019 Interfacial Fields in Organic Field-effect Transistors and Sensors

    12/158

    2.21 The output curves before and after exposure to DNT (a) and TNT (b). The TNT consistently

    lowered the source drain current more than DNT ...................................................................... 55

    2.22 The mobility (a) and threshold voltage (b) response of a 25 nm bis CF3 NTCDI film to

    DMMP and DNT vapor................................................................................................................ 58

    2.23 The normalized current response of a 15nm C2PhF5 NTCDI film to DMMP and DNT

    vapor. Note that the delivery gas for the DMMP vapor was nitrogen while the delivery.... 59

    3.1 The SKPM scans of bottom (a & c) and top (b & d) contact pentacene FETs. The device

    geometries are shown for top and bottom contact devices ........................................................ 63

    3.2 The transfer curves at various time intervals of an OFET after a continuous gate bias of -60 V

    (a). The source drain current was -5 V during gate biasing...................................................... 64

    3.3 The procedure used to fabricate the lateral in terfaces between PS and pentacene.....................6 6

    3.4 The step-by-step procedure for measuring the surface potential of the lateral interfaces.

    During Step 1 SKPM scans are taken at each voltage step........................................................ 67

    3.5 SKPM scans of a PS-pentacene interface before (a) and after (b) +200 V charging for 10

    minutes. The interface between the PS and pentacene is where the large drop in .................68

    3.6 The SKPM line scans of a PS-pentacene interface before (a) and after (b) -200 V charging for

    10 minutes. Again, the interface between the PS and pentacene ............................................. 70

    3.7 The SKPM line scans for samples charged at +100 V (a) and -150 V (b) for 5 minutes. In

    both cases the surface potential reached the limit o f the SKPM so the flat lines .................... 71

    3.8 The SKPM line scans for gold samples before charging (a) and after being charged at +100 V

    (b) and -100 V (c) for 10 minutes................................................................................................ 73

    3.9 A schematic of the surface potential for the PS and pentacene. The PS is always on the left of

    the schematic. The surface potential of the uncharged samples with small ........................... 74

    3.10 A schematic showing the charges entering the PS during positive charging, then being

    trapped in the PS after charging................................................................................................... 75

    xi

  • 8/11/2019 Interfacial Fields in Organic Field-effect Transistors and Sensors

    13/158

    3.11 The change in surface potential difference between PS and pentacene at each incremental

    applied bias, as shown in Steps 1 and 3 from Figure 3.4........................................................... 76

    3.12 The TSDC for samples corona charged with a V^a o f 150 V. Note the different

    temperatures where the peaks occur, the glass transition temperature for PS is ~100 C. ... 77

    4.1 The output characteristics of lateral transistors shifted by application of either polarity of

    charging voltage to the gate.......................................................................................................... 81

    4.2 Schematic of the fabrication process for a lateral OFETs. First, 50 nm gold e lectrodes with a

    4-5 nm Cr adhesion layer are deposited on the Si substrate via photolithography (a) ........... 83

    4.3 Optical image of the lateral OFET showing the pentacene covering the left side of the image

    and the PS covering the right side ................................................................................................ 84

    4.4 Height scans of the two lateral OFETs in Figure 4.5. The source (S), drain (D) and gate (G)

    electrodes can be easily seen and are labeled ............................................................................. 85

    4.5 SKPM surface potential scans of lateral OFETs. Similar to the optical and height images the

    source and drain electrodes are always at the left and the gate is at the far right ..................... 86

    4.6 SKPM scans showing the surface potential o f the same lateral OFET immediately after

    transistor testing and 1 hour later................................................................................................. 88

    4.7 Transfer curves for the samples from Figure 4.4 before and after charging. The lines after

    charging to -100 V (a) and +100 V (b) are in red....................................................................... 89

    4.8 Output curves for the lateral OFETs before (a,b) and after (c,d) 100 V charging.A sample

    was tested before (a) and after (c) charging at -100 V for 10 minutes............................ 90

    4.9 SKPM surface potential scans of lateral OFETs. Similar to the previous SKPM scans the

    source and drain electrodes are always at the left and the gate is at the far right.................... 91

    4.10 Transfer curves for the samples from Figure 4.8 before and after charging. The lines after

    charging to -50 V (a) and +50 V (b) are in red ........................................................................... 92

  • 8/11/2019 Interfacial Fields in Organic Field-effect Transistors and Sensors

    14/158

  • 8/11/2019 Interfacial Fields in Organic Field-effect Transistors and Sensors

    15/158

  • 8/11/2019 Interfacial Fields in Organic Field-effect Transistors and Sensors

    16/158

  • 8/11/2019 Interfacial Fields in Organic Field-effect Transistors and Sensors

    17/158

    Chapter I: Introduction to Organic Semiconductors (OSCs) and Devices

    1. Introduction to Organic Semicond uctors

    M olecular materials for electronics applications are on the cusp of bein g com me rcially

    viable as alternatives to conventional electronics. The interest in these m olecular m aterials has

    been increasing ex po nent ially ove r the prev ious tw o decades, draw ing in terest from bo th

    academia and corporate institutions. One m ajor draw o f these organic materials is the ability to

    chemically manipulate the molecule, which will allow for more facile mass processing

    techniques. The ease of processing the molecular materials can offset the reduced performance

    compared to conventional high quality inorganic semiconductors. Som e benefits of organic

    semiconductors are low cost; low temperature and large area processing; and large scale

    manufacturing on a flexible substrate. One application where organic semicond uctors are

    currently emerging is electronic displays, while applications such as flexible solar cells, printed

    memory, and white lighting are close to comm ercialization.

    1.1 Organic Semiconductor Electronic Structure

    A conjugated bonding scheme, alternating double and single carbon-carbon bonds, has

    been shown to be a very effective de sign in molecu la r el ec tronics .( l) The ke y is de sign in g an

    aromatic system that packs face-to-face, and limits the more favorable edge-to-face interactions.

    The main method used to preferentially create face-to-face packing is to build aromatic systems

    that have a large x-surface to circumference ratio that limits perfect herringbone packing (all

    contact between molecules is edge-to-face). Generally side chains are added to increase

    solubility, but they can also force face-to-face interactions by steric hindrance that does not allow

    for efficient edge-to-face interactions. The face-to-face mole cular contact of OSCs is preferred

    because it increases the overlapping of 7i-orbitals. The charge transport in organic materials may

    have band character in high quality single crystals. In this case, the highest occupied m olecu lar

    1

  • 8/11/2019 Interfacial Fields in Organic Field-effect Transistors and Sensors

    18/158

    orbital (HOMO) can be approximated as the valence band from inorganic semiconductors, and

    the lowest unoccupied molecular orbital (LUMO) in organic semiconductors (OSCs) corresponds

    to the conduction band in inorganic semiconductors. However, thermally activated hopping is

    prop osed for p olycrystalline OSCs , which are us ed in mos t applications o f o rg an ic elec tron ics.

    The closer adjacent molecules become, the more likely there will be interactions between

    their respective molecular orbitals. This close proximity helps with charge tran sfer from

    molecule to molecule. Generally the charge mobility increases with a more o rdered an d closer

    packed organic semicon ductor system. Eve n closely pack ed organic se m ic onductor s are sti ll a

    weakly interacting system, which has been shown to display a thermally activated mobility with

    Arrhenius behavior.(2) Two models that use this Arrhenius behavior to describe charge transport

    are a multiple trapping and release model(3) and a model, which assumes hopping between

    localized states.(4)

    Just as inorganic semiconductors can transport either holes or electrons, organic

    semiconductors can be primarily p- (hole transport) or n-channel (electron transport). Their

    be havior is governed by the HOMO and LUMO overlap with the elec trodes and the

    environmental stability of the charge carriers. In general, an n-channel materia l will hav e a high

    electron affinity, while p-channel materials have lowe r ionization potentia ls.(l) Unlik e inorganic

    semiconductors which have been deliberately doped to become n-type or p-type, organic

    semiconductors have an intrinsic ability to transport holes or electrons. Ho we ver, organic

    materials that were once thought to only transport holes have exhibited n-channel behavior. (5)

    This occurred in vacuum and only when special care was taken to reduce the nu m ber o f traps at

    the dielectric/semiconductor interface.

    1.2 Applications of Organ ic Semiconductors

    Organic semiconductors are being investigated in three major fields: photovoltaics;

    emissive devices, such as light emitting diodes; and circuits. For circuits, the device architecture

    2

  • 8/11/2019 Interfacial Fields in Organic Field-effect Transistors and Sensors

    19/158

    is generally in the form of a transistor or a diode. In this work, I will be fo cusing solely on

    transistors. Organic field-effect transistors (OFETs) are the primary device s that are used in

    active circuits, performing logic and switching operations. Figure 1.1 shows a cross sectional

    view of an OFET. Most OFETs are operated in accumulation mode because a majority o f

    organic semiconductors have an inherent preference f or p- or n-channel conduction in air.

    ^Con ducting channel

    s | 1Da a Figure 1.1 Schematic view of a bottom contact, bottom gate OFET. S refers to the source electrode while D

    refers to the drain electrode.

    1.3 Operation of Field-effect Transistors

    A transistor can be considered a metal-insulator-semiconductor (MIS) combined with a

    two terminal resistor. In the resistor, the two term inals are called the sourc e and drain w ith

    the semiconductor acting as the resistive material. To modulate the resistance of the

    semiconductor layer, a voltage is applied to the gate electrode, which polarizes the dielectric and

    causes charge carriers to accumulate at the dielectric/semico nductor interface (this is noted as the

    conducting channel in Figure 1.1). As the charge carrier density increases so does the

    conductivity of the resistor. Since the resistance can be transferred from the gate terminal, the

    device was called a transfer resistor or transistor .

    The gate electrode can be patterned on an insulating substrate or the entire substrate may

    be conductive, lik e a single crystal o f highly doped sil icon . Both organic an d inorga nic in su la ting

    layers are used for the dielectric. SiC>2 and AI2O 3 are common inorganic dielectrics while

    polymeric materia ls like po lystyrene (PS) , po ly(m ethy l meth acry late) (PMMA), or Cytop (a

  • 8/11/2019 Interfacial Fields in Organic Field-effect Transistors and Sensors

    20/158

  • 8/11/2019 Interfacial Fields in Organic Field-effect Transistors and Sensors

    21/158

  • 8/11/2019 Interfacial Fields in Organic Field-effect Transistors and Sensors

    22/158

    KLUMO

    H O M O HOMO

    Metal f Semiconductor

    Insulator

    At Equilibrium (a) A tV g V g)

    occurs. This regime is the result o f a charge depletion region near the drain electrode. Id during

    the saturation regime is given in equation ( 1 .2 ):

    depicted in Figure 1.1. Charges ma y be inserted and remo ved from the cond uction channel b y the

    1,- j -nC, ( n - V r H - y ( 1 . 1 )

    . where L is the length of the channel between the sou rce and drain electrodes, W is the w idth o f

    6

  • 8/11/2019 Interfacial Fields in Organic Field-effect Transistors and Sensors

    23/158

    W / \2^ - ^ A * q v f - v r ) (1.2)

    The two terms V T and p w ill be discussed in detail later o n, but the mo bility o f a material

    can be thought of as the speed o f charges per applied field. W hile the V t can be conside red the

    gate voltage where the field-effect mobility determines the cha rge transport in the OF ET channel.

    Generally organic semiconductors have field-effect mobilities in the range of 10 ' 3 - 10 cm2/V s in

    transistors, while crystalline silicon can have a mo bility o f 1000 cm2/Vs an d am orph ous silicon

    mobility is around 1 cm 2/Vs.

    1.6 Mobility in OFETs

    The mobility of a semiconductor is the velocity of the charges per unit of electric field

    betw een the source and drain co ntacts. Th ere are nu merou s methods to ex trac t th e mob ility in

    semiconductors, such as time of flight, space charge limited current, and, as described in this

    work, field-effect transistors. How ever each test method produces a different m obility value that

    cannot be used interchangeably in the other methods. Furthermore in the case of field-effect

    transistors, the mobility value should be considered device mobility not a material mobility.

    Num erous variables in transistor prep aration can al te r the mobili ty in the OFET (e .g. the

    dielectric choice, grain boundaries in the semiconductor, traps in the semiconductor, charge

    injection into the semiconductor). Additionally, just the choice of a top contact or bottom contact

    geometry can change the device mobility. Top contact devices have the m etal for the contacts

    diffuse into the semiconductor creating a more complicated electronic interface than a simple

    metal-semiconductor junction. While the bottom contact interface may be cleaner, the packing o f

    the semiconductor on top of the source or drain electrodes m ay be poo rer than o n top o f the

    dielectric.(6 ) This results in lower device mobility for bottom con tact devices com pared to the top

    contact devices. In the work done by Kym issis et al., the higher surface energy o f the SiC>2 was

    shown to result in more intermolecular interactions compared to the lower surface energy of the

    7

  • 8/11/2019 Interfacial Fields in Organic Field-effect Transistors and Sensors

    24/158

    metal contacts. However, self-assembled mo nolayers on top of the metal contacts can improv e

    the surface morphology of the semiconductor in bottom contact geometries.(7)

    Different metal contacts can be used to alter the charge injection into the

    semiconductor.(8 ) For an n-channel semiconductor the metal work function should be close to the

    LUMO energy level in the semiconductor, while for p-channel semiconductors the metal work

    function should be closer to the HO M O energy level. W hile simply altering the metal work

    functions sounds like a simple idea, ultraviolet photoemission spectroscopy showed that there are

    chemical interactions between the metal contact and the semiconductor.(9) These chemical

    interactions have an influence on the charge injection from the me tal into the se micond uctor.

    1.7 Extracting Mobility and Threshold Voltage

    To derive the field-effect mobility and V T for a specific OFET , a transfe r curve is

    recorded. A Vd necessary to attain saturation at every V g measu red will applied to the dev ice and

    the gate voltage will be swept, recording the la. By utilizing equation (1.2) we see there are only

    two variables: p and VT. If we take the square root o f the equation and make the p lot see n in

    Figu re 1.2 the Vt can be seen as the x-intercep t and the slope is pro por tiona l to p l/2. This

    assumes that the field-effect mobility is independent of gate voltage(lO) and the contact

    resistance is n egligible.(l 1)

    1.8 Threshold V oltage in OFETs

    There is a possibility that trap states exist in the semiconductor and/or at the dielectric-

    semiconductor interface in OFETs.(12) To accumulate charges in the channel and begin to have

    charge transport between the source and drain these trap states mu st first be filled. The p resence

    of these traps will slow the transition from intrinsic to field-effect mobility. These traps states

    may have different densities and may be at various energetic levels in the semiconductor band

    gap, resulting in gate dependence on the field-effect mob ility. The traps are distributed

  • 8/11/2019 Interfacial Fields in Organic Field-effect Transistors and Sensors

    25/158

    throughout the semiconductor film, but may be more concentrated at grain boundaries or at the

    dielectric-semiconductor interface. As the gate voltage is increased the traps are f illed by the

    injected charges, this decreases the number of traps available to slow the motion of charges,

    which allows for higher field-effect mobility at higher gate voltage s.(3,4, 13, 14)

    For inorganic semiconductor-based electronics, a trap can be considered a localization of

    energy levels present in the band gap of the semiconductor (either in the bulk or at the surface),

    which belong to an impurity, a defect in the semicon ductor or both. These traps can be shallow

    or deep depending on where the energy level of the trap aligns with the semiconductor bandgap.

    Shallow traps have en ergy level differe nces with the ba nd gap ed ge of about 1 -2 tim es k BT, wh ile

    deep traps may be 3-5 times kBT. These charge traps play a major role in the transp ort o f charges

    in disordered semiconductors such as amorphou s silicon, oxide based semico nductors a nd OSCs.

    1.9 Semicond uctor-Dielectric Interface and Traps

    The interface between the semiconductor and dielectric is crucial to the performance of

    the semiconductor in OFETs because most accumulated charge is localized within the first 3-5

    monolayers of semiconductor.(13) The presence of traps, dopants, impurities and morphology

    play a major role in de term ining the mob ility an d threshold vo ltag e of the OFET. As men tion ed

    previously, ca refu l control o f the dielec tr ic-sem icon du ctor in terfac e can ev en co nt ro l th e po la rity

    of the semiconductor.(5) By utilizing a trap free, hydrophobic benz ocyclo buten e (BC B)

    dielectric, instead of the traditional Si0 2 , n-channel behavior was observed in common p-channel

    semiconductors such as regioregular poly(3-hexylthiophene) (P3HT). Silanol (Si-OH ) groups on

    the hydrophilic Si0 2 dielectric act as electron traps. Even pentacene, a very com m on p-channel

    OSC, was shown to exhibit n-channel behavior by using a poly(vinyl alcohol) (PVA) dielectric

    that created interface states which helped in electron transport in pentacene.(15)

    Both the VT and mobility in OFETs may be influenced by a variety o f factors such as

    surface dipoles, surface energy, roughness of the interface and the energetic disorder at the

    9

  • 8/11/2019 Interfacial Fields in Organic Field-effect Transistors and Sensors

    26/158

    interface.(16-18) There has been substantial work on surface dipoles and the influence o f polar

    groups at the dielectric-semiconductor interface on OFE T properties.( 19-29) T he imp act is

    mainly on threshold voltage but there may be additional traps, which influence the mobility as

    well. The surface energy can alter the orientation of the OSC molecules and correspon ding film

    morphology.(16) Lower surface energy dielectrics aid in increasing the interconnection between

    grains, yielding improved mobilities.(30-34) Charge accumulation in bottom gate transistors can

    be drastically redu ced by ro ug her dielectric surfaces. (16, 35 -39) The ro ug her surfac e leads to a

    more disordered OSC, either by a reduction in grain size, increased voids, or greater molecular

    disorder from the defects in the intermolecular packing.(38, 40, 41) The increa sed surface

    roughness o f the dielectric has been associated with in an increase in the num ber o f traps states at

    the interface.(42) Also, higher permittivity dielectric layers will alter charge transport by

    broadening the Gaussian De nsity o f States (D OS) in di sord ered OSC s.(43) H ig he r mob ility,

    lower V j and sm aller hysteresis were seen in transistors on low k dielectrics. This is attributed to

    the randomly oriented dipoles inside the dielectric.

    1.10 Thin Film Deposition and Morph ology

    With so much depending on the semiconductor-dielectric interface, the device fabrication

    plays a critical role in the eventual pe rforman ce o f the OFET. There are tw o m ain way s to

    deposit thin films onto the substrates: thermal evaporation and solution processing. In this w ork

    small molecules will generally be thermally evaporated while larger more soluble molecules and

    po lymer wi ll be solut ion processed . The main technique us ed for so lu tion pro ce ss in g is spin

    coating, but eventual large-scale production will require other fabrication techniques such as

    gravure, offset, or flexographic printing.

    Thermal evaporation is a type o f physical vap or deposition where substrates a re placed in

    a deposition chamber under a vacuum pressure on the order of 10'6 torr. The material to be

    deposited is placed in a thermally crucible. The crucible is then placed into a wire filament

    10

  • 8/11/2019 Interfacial Fields in Organic Field-effect Transistors and Sensors

    27/158

    between two electrodes. Cu rrent is pa ssed th ro ug h th e fi lamen t resulting in re sistiv e hea ting and

    the material inside the crucible is sublimed. Due to the very low pressure, m elting o f the material

    is uncomm on for symmetrical and rigid organic materials. The sublimated organ ic mo lecules

    then condense on the substrate placed above the crucible source. Most deposition rates for the

    organic and metal thermal evaporation are appro xim ately 0.5 A/s.

    The solution processing technique requires OSCs that are soluble in an appropriate

    solvent, which is not always feasible. Drop casting is where a solution o f OSC is d ropped onto a

    substrate and the solvent is allowed to evaporate away. The resulting film thickness is de pendent

    on the total amount o f solute in the solution. For spin coating a m ore viscous solution is required

    for thicker films, which is why spin coating generally requires polymers. W hile the mo rphology

    of films made via drop casting and spin coating can be refined with thermal and s olvent a nnealing

    they still never achieve the crystallinity o f thermally evapo rated films (partly due to the m olecular

    structure differences, small molecule vs. polymer).

    Small molecule OSCs form polycrystalline films with each grain being a single crystal.

    While polymers may have some crystallinity, they will never be fully crystalline and will contain

    free volume and amorphous regions. The difference in crystal structure is one reason that small

    molecules still show greater mobility values than their poly m er counterparts. The interface

    be tw een each grain is called a grain bo un da ry . The re is gr ea te r energetic disord er at th e grain

    boundaries, which lim its the mobility co mpa re d to the more ordered interio r o f the grain, so grain

    boundaries influence the OFET mob ility trem endo us ly. (44)

    For improved charge transport, highly crystalline films are preferred because they

    increase the jr-orbital overlapping of adjacent molecules. Most small molecule OSCs will orient

    themselves so that the n-orbitals overlap, which means the long axis of the m olecule will be close

    to perpendicular to the substrate. The film structure is controlled by both intermo lecular V an der

    Waals forces and interfacial energy between the substrate and OSC m olecule. Tw o approach es to

  • 8/11/2019 Interfacial Fields in Organic Field-effect Transistors and Sensors

    28/158

    improve the crystallinity of the OSC are the utilization of self-assembled mono layers (S AM s) on

    the substrate and thermal and/or solvent annealing. Alkyl silanes are very com mo n SA Ms for

    OSC films on silicon dioxide dielectric layers.(45) The silicon portion o f the SAM will

    chemically bond to the Si0 2 while the alkyl side chain will extend outw ards from the S i0 2

    interface. This SAM layer will alter the OSC surface interactions, impro ving the crystalline

    packing in the OS C film. Th ermal an ne aling o f substrates during sm al l molec ule th ermal

    evaporations can also lead to larger grain sizes, but each OSC system is different and requires an

    optimized substrate temperature.(46, 47)

    2. Mechanism s of Threshold Voltage Shifting

    For some applications, such as inverters and sensors, there is an ideal, non-zero threshold

    voltage that the OFET should have for optimum performance. The reasons for the non-zero Vt

    will be discussed later, but for these applications, the threshold voltage is optim ized e ither during

    fabrication or once the OFET has been manufactured. There are numerous mechanisms where a

    controlled shift in the threshold voltage can be achieved, but all the m echanisms can be classified

    in one o f two broader categories: electrostatic and nonelectrical threshold voltage shifting.

    2.1 Electrostatic Threshold V oltage Sh ifting

    2.1.1 S urface Dipoles a t the Dielectric Interface

    As discussed earlier, the SAM on the dielectric layer can improve the OSC film

    morphology, but a dipolar SAM can be placed at the dielectric interface with the OSC or the gate

    electrode and cause a depletion or accumulation o f charges inside the OSC. T he stored cha rges in

    the dielectric layer will cause the accumulation/depletion of charges in OSC depending on the

    po larity o f the charge. When functional en d grou ps were used on the SA M at th e dielec tr ic-O SC

    *interface, the threshold voltage of a p-channel OFE T could be shifted more n egative fo r a NH 2

    terminated SAM and more positive for a fluorinated SAM.(21) The same shifting behavior was

    12

  • 8/11/2019 Interfacial Fields in Organic Field-effect Transistors and Sensors

    29/158

    shown for an n-channel OSC, which means the p-channel device was easier to turn on with the

    fluorinated SAM (accumulation of charges in the OSC), while the n-channel device was harder to

    turn on (depletion of charges in the OSC). The amine term inated SAM had the opposite eff ect

    and made the n-channel device easier to turn on and the p-channel device hard er to turn on. Since

    the SAM layers are permanently on the devices the shift in thresho ld voltage is permanen t.

    2.1.2 Static Charge in Dielectrics

    The main method for threshold voltage shifting in this work is to store charge in the bulk

    of the dielectric. The charge may be stored in the dielectric layer for prolong p eriods o f time,

    meaning this is a fairly permanent change in the OFE T threshold voltage. One m ethod used to

    store electrostatic charge in the dielectric is corona charging.(48-51) Figure 1.4 shows a

    schematic of a corona triode charging setup. A dielectric on a conductive substrate is placed in a

    chamber. The substrate is grounded and a potential is applied to a grid above the substrate. A

    corona voltage is applied (generally around 10 kV). The blue lines represe nt the electrostatic

    field lines in air. This corona voltage generates a corona at the point due to the breakd ow n o f air.

    In the case of a negative polarity point electrode, only negative ions can reach the surface o f the

    dielectric. Once the negative ions reach the surface o f the dielectric, they donate th eir electrons

    and become neutral. This leaves a layer o f trapped electrons underneath the dielectric surface.

    (52) The grid, consisting of a metal wire mesh, is utilized to control the ch argin g o f the surfac e

    and improve uniformity of charges embedded into the dielectric. Ions will pass through the grid

    and deposit onto the surface of the dielectric until the surface potential matche s the grid potential.

    Once the surface potential of the dielectric m atches the grid po tential there is no electric field and

    the ions will just flow to the grid. In addition to the corona voltage and grid v oltage, ch arging

    temperature and time can be varied to alter the surface potential.

    13

  • 8/11/2019 Interfacial Fields in Organic Field-effect Transistors and Sensors

    30/158

    Corona i

    voltage T

    Grid

    Grid I

    voltage ^x

    Dielectric

    Conductive Substrate

    Figure 1.4 A point-to-grid corona triode schematic. A corona is generated near the point and ions flow through

    the grid and impart their charge in the dielectric layer. The blue lines show the electrostatic field lines in air. A

    metallic wire mesh grid is used for uniform charge distribution.

    2.1.3 Ferroelectric Polarization

    Polymer and ceramic ferroelectric transistors exhibit threshold voltage shifts from a

    po larization response analo gous to a ferrom ag ne tic magne tic response. (53-63)T he ferroelectrics

    show a hysteresis due to remnant polarization from an intrinsic dipole alignm ent. To achieve

    electrical polarization ferroelectrics must be insulating and cannot be metallic. The po larization

    will occur above the coercive field (Ec) and below the dielectric breakdown. The im print time,

    time under the electric field required to alter the polarization, will vary from system to system,

    with amorphous materials never actually polarizing due to the completely random dipole

    orientation. The threshold voltage shift is the remnant polarization causing eithe r depletion or

    accumulation o f charges in the OSC. Ferroelectrics are mainly used for m em ory applications,

    allowing the threshold voltage to be shifted so the OFET is either ON or OFF at a certain

    gate voltage.

    2.1.4 Charging o f Emb edded Floating Gates

    Floating gate devices are a mod ified version o f the conv entional m etal-oxide-

    semiconductor field-effect transistor (MO SFET ), where a second gate electrode is placed betw een

    14

  • 8/11/2019 Interfacial Fields in Organic Field-effect Transistors and Sensors

    31/158

  • 8/11/2019 Interfacial Fields in Organic Field-effect Transistors and Sensors

    32/158

    2.2 Non Electrical Threshold Voltage Modulation

    2.2.1 M echanical Force

    Hydrostatic pressure can reversibly shift the threshold voltage and mobility in polymer

    OSC devices.(75) The threshold voltage was seen to become more negative with increasing

    pressure, while the mobility o f the tran sistor s in crea sed four fo ld over th e same pressure rang e.

    The increase in mobility was attributed to the increase in 7t-orbital overlap, due to the shorter

    inter-a tom ic. distances while under comp ression, while the shift in thresh old v oltage was

    suggested to be from an increase in trap density.

    2.2.2 Photopolarization and Photoin duce d Charge Transfer

    Photoinduced charge transfer from a single crystal OSC to the dielectric layer has been

    shown, with the illumination time playing a majo r role in the threshold voltage shift.(79) Shorte r

    illumination times were able to alter the threshold voltage, w hile only gate bias stress wa s see n in

    the longer exposures. At the shorter time scales the rate o f thresho ld voltage shifts were correlated

    to the wavelength and intensity of the light. The idea that this is a photoconductive effec t is not

    valid, since upon illumination the current decreases instead of increasing. Podzo rov et al.

    attributed the threshold voltage shift to hot charge carriers going from the OSC into the dielectric

    layer. The hot charge carriers are generated by the absorption of photons with energies g reater

    than the band gap of the semiconductor. The emb edded charges in the dielectric layer the n shift

    the threshold voltage.

    When a SAM o f a dipolar nonlinear optical chrom ophore was used on a Si0 2 dielectric

    layer the initial un-illuminated dipole of 8 Debye could-be increased with photo-illumination to

    approximately 20 Debye in its excited, charge separated state.(80) The increase in dipole will

    accumulate or deplete the charges in the semicond uctor. How ever the directio n o f the gate sw eep

    influences the threshold voltage, indicating there may be a photoinduced charge transfer to the

    SAM.

    16

  • 8/11/2019 Interfacial Fields in Organic Field-effect Transistors and Sensors

    33/158

    2.3 Operational Stability

    For organic electronics to be a viable commercial product, the devices need to be reliable

    and stable. The intrinsic stability of the semicon ductor and the role o f the gate dielec tric are both

    critical factors in the stability of the device. Organic sem iconductors are sensitive to

    environmental factors such as oxygen, water, various polar molecules, oxidizing agents, and

    reducing agents. Additionally, some gate dielectrics may trap charge carriers over time, causing a

    shift in the threshold voltage o f the transistor; this phenomena is called the bias stress effect. The

    bias stress effect gradually will tu rn a tran sistor from ON to O FF m ak in g th is a ch al lenging

    problem for a pplication s like display dr iver s, where the tran sistor is used to op erate a pixe l.

    17

  • 8/11/2019 Interfacial Fields in Organic Field-effect Transistors and Sensors

    34/158

    Chapter II: Chemical Sensing with Organic Field-Effect Transistors

    1. Sensor Testing

    When OSCs are made into field-effect transistors they are commonly operated in

    accumulation mode, meaning induced majority charge carriers are mainly responsible for

    conduction. Like most organic materials, OSCs are insulators under ordinary ambient conditions,

    bu t have the prop erty that th eir cond uc tivity can increase marke dly on the ap plica tion o f voltage,

    electromagnetic radiation, or heat. The change in conductivity possible for an OS C is the product

    o f the charge density induced by the energetic input times the mobilities o f the charges under

    electric fields. Vapors may alter the effective voltage seen by the semiconductor or the mobility

    of the charge carrier inside the semiconductor. To create a vapor sensor from an OF ET the

    change in output signal will need to be monitored under both normal conditions and exposure to

    the vapor.

    There are two methods to measure the OFET characteristics: output or transfer curves.

    With an output curve the drain current is m easured vs. the drain voltage at stepp ed g ate voltages.

    This generates multiple curves, since there is one for each gate voltage. W hen a transfer curve is

    taken the OFET is generally operated in the saturation regime. In the saturation regim e, the drain

    current is independent o f the drain voltage, which is held fixed during the measur emen t. The gate

    voltage is swept and the drain current is measured generating only one curve compared to the

    output curve measurements that generate multiple curves. In this work, transfer curves are used

    to monitor the sensor, since the mobility and threshold voltage can be quickly m easured. This

    minimizes the gate bias stress, which will also alter the OFET output.

    2. Previous Work on OSC V apor Sensing

    Due to the reactivity of both charge carriers and domain bou ndarie s w ith extrinsic

    chemical species, OSCs can be highly responsive to environmental chemical agents (analytes).

    18

  • 8/11/2019 Interfacial Fields in Organic Field-effect Transistors and Sensors

    35/158

    When OSCs are positioned between a pair of electrodes and their native conductivities are

    modulated by chemical exposure, the device is termed a chemiresistor. W hen a third, gate

    electrode is used to preset, or scan, a range of conductivities that are further perturbed by an

    analyte, the device becomes and OFET sensor. A schematic o f an OFET sensor, integrated with a

    gaseous analyte delivery system, is shown in Figure 2.1.

    Analyte

    Flow

    meter

    Air

    OSC

    Valve

    0 $ 4

    Dielectric

    Gate

    Figure 2.1 Schematic of an analyte delivery system for a OFET vapor sensor. The arrow s sym bolize the analy te.

    OFETs that have not been sealed, or packaged, will generally have their output current

    change upon exposure to polar vapors, including relative humidity from water vapor.(57) The

    electrical contacts o f pentacene OFETs were shown to be susceptible to humidity. The shorter

    channel length OFETs resulted in higher sensitivity, while the ON/OFF ratio (conductance with a

    gate voltage versus without the gate voltage) could vary by an order of mag nitude with v arying

    humidity levels. Another pentacene-based OFET showed a current decrease o f approxim ately

    80% when the relative humidity was varied from 0% to 30%.(52) Wh en the relative humidity

    19

  • 8/11/2019 Interfacial Fields in Organic Field-effect Transistors and Sensors

    36/158

    was 75%, the OFET no longer functioned. Electron con duction materials are suscep tible to

    oxygen as well as humidity, because of the ability of both to quench, or trap, the electrons. For

    example, in perylenetetracarboxylic diimide (PT CDI), ox ygen was shown to decrease the electron

    mobility and density.(83)

    Hole transporting copper phthalocyanine (CuPc) OFETs showed a greater response for

    more polar vapors; with the response further increasing when the OSC layer thickness was

    reduced.(54) An array of CuPc and other phthalocyanines were exposed to various analytes.(55)

    By changing the copper core to cobalt, nickel, zinc, and H 2 the array was seen to have a

    sensitivity that was correlated to the Lewis base b inding enthalpy of the analyte, for the m etal

    phthalocyanines; and to the an alytes h yd ro gen bo nd ing enthalpy , for the ph th alocy an in e with no

    metal core.

    Using different OSCs and 16 analytes of various polarities, a response fingerprint was

    developed for an array of OFETs. This fingerprint gives a distinct response from the arra y for

    each of the various analytes.(86) These devices were then show n to operate in circuits and the

    sensitivity was calculated for 1 ppm .(57) Ano ther application of the fingerprin t approa ch has

    been used to de tect vo lat ile organic compo un ds for fo od quality analysis.(55) Pen tace ne, po ly (3 -

    hexylthiophene) (P3HT) and poly(3-octylthiophene) devices were exposed to acetic acid,

    octanoic acid, ethanol, propanol and other vapors giving unique responses for each vapor. The

    carbonyl group was shown to have a greater response. Wh en the side chain of the analyte w as

    lengthened, the analyes had less of an effect on the output current o f the OFET . This depe ndence

    on side chain length is due to the lower diffusivity in the OSC film. Other studies have show n that

    the sensing array method could give lower detection limits, compared to single

    me asurem ents. (59)

    Contact effects were investigated in an a,co-dihexylsexithiophene (D H a6 T) film exposed

    to butanol vapor.(90) The higher sensitivity ON state showed that the sensing enhan cem ent was

    20

  • 8/11/2019 Interfacial Fields in Organic Field-effect Transistors and Sensors

    37/158

    from the channel transport in the OFET and the contact resistance and leakage current had lesser

    roles, while the less sensitive lower gate voltages showed the contact resistance played a larger

    role. Additionally, the OSC will interact more w ith longer alkyl side chains; the lon ger side chain

    means there is higher mass absorbed onto the OSC .(97)

    A decrease in film thickness increases sensor response to the analyte, from the faster

    diffusion to the conduction channel.(92) There is a minim um thickness requ ired for a ch annel to

    form in the OFET but once that thickness is met additional OSC thickness only redu ces the sensor

    response. Further work showed that even wh en the dielectric thickness wa s reduce d to ju st the

    native oxide (~2nm S i0 2) the thin OSC layer was very sensitive.(46)

    3. Enhanced Sensor Response with Receptors

    When polar molecules are adsorbed onto the OFET they introduce traps into the OSC

    film; these traps alter the source drain current. W hile these molecules may ad here to th e grain

    boundaries or other defect sites in the film , there is no specific binding to the OSC film . I f the

    binding str ength were increased , the nu mbe r o f analytes on the fi lm would in crease , re su lt in g in a

    stronger signal. One o f the attractive features o f OSC s is the ability to covalen tly bo nd a receptor

    group to an existing semiconducting material. This allows the receptor to ge t closer to the charge

    conduction channel, which will increase the sensitivity of the OFET. By placing an ether group

    on a polythiophene side chain the response to ethanol was increased com pared to a poly thiophen e

    which just had an alkyl side chain.(97)

    To better sense dimethyl methylphosphonate (DMMP), a simulant for the nerve agent

    sarin, a hydroxylated bithiophene oligomer was used as a receptor.(93) The OS C used was 5,5-

    bis(4-he xy lpheny l)-2 ,2-bithiophene (6 PTTP6 ) in a 1:1 ratio with the receptor that seem ed to give

    a single phase film, when examined with scanning electron microscopy and x-ray diffraction

    (XRD). The current decrease occurred faster upon exposure to DM MP and w as greater in

    magnitude for the device with the incorporated receptor. The decrease in curren t can be

    21

  • 8/11/2019 Interfacial Fields in Organic Field-effect Transistors and Sensors

    38/158

    attributed to stronger bonding o f the DM MP to the receptor and a change in the local electric field

    from binding to the receptor. The sensitivity could be enhanced by incorporating a receptor with

    the 6 PTTP6 and decreasing the film thickness, reducing the detection limit from 150 ppm to 5

    ppm.(92)

    Electron transporting materials have also been used with receptors to detect DMMP. (94)

    An OSC film o f NTCD I had receptors of a hydroxphenylated NTCDI dep osited on top o f the

    film. This receptor formed islands on the OSC due to the difference in surfac e energy w ith the

    NT CD I. Eve n though the receptor was no t fu lly inco rporated in to the fi lm it st ill sh ow ed a more

    selective response to DM MP.

    To further show how receptor groups can influence the response of a sensor, two

    different side chain receptors were incorporated onto an OSC core and placed on top of an

    existing OSC film.(95) These receptors made the sensor respond differently to two enantiom ers

    of a chiral alcohol citronellol. This shows the selectivity that is achievable by ca refully selecting

    the receptor molecule. In another study, 2 mon olayers (lOnm) films of 5,50-bis-(7-do decyl-9 H-

    fluoren-2-yl)-2,20-bithiophene (DDFTTF) were exposed to various volatile organic compounds.

    (96) The receptors used were calix[8 ]arene (C[8 ]A) and c-methyl calyx[4]-resorcinarene

    (CM[4]RA), which were previously shown to work as size selective cavities in other host-guest

    sensors. C[8 ]A showed a ethyl acetate detection limit five times lower than then pure DDFTTF

    film.

    4. Sensing Mechanism

    Since the charge carriers at the boundaries between grains dominate the charge transport

    in longer channel devices, the sensing mechanism is dominated by the formation of dipole-

    induced traps at the grain boundaries when analyte is adsorbed. This analyte-in duced trap

    decreases the source-drain current for most polar analytes, because there is always a region

    around the dipole where charge carriers are at lower energy, and there is an activation barrier for

    22

  • 8/11/2019 Interfacial Fields in Organic Field-effect Transistors and Sensors

    39/158

    them to move away from the lower energy region. There are fewer grain boundarie s in short

    channel devices, so the analyte molecules that diffuse to the electrode-OSC interface will alter the

    source-drain current. The electrode-OSC interface is the region where the majority o f the

    resistance occurs for short channel devices. W hen interfacial resistances do not determ ine the

    current-voltage relationships, then dipole adsorption or diffusion into the bulk OSC results in

    current-lowering traps. Analytes with a high redox activity can act as quen chers or dop ants for

    the main semiconducting domains, lowering or raising current levels, respectively. An electron

    withdrawing analyte can act as a dopant for a p-type OSC and a trap for an n-typ e OSC. The

    doping mechanism is particularly useful for sensors because of the rareness and distinction o f a

    tum-on signaling for analytes. Two recent examples o f this are the use of n-channel OS Cs to

    obtain current increases in response to ammonia vapor.(97, 98)

    As discussed earlier, the extent to which an analyte can alter the charge transport varies

    with the analytes location in the OSC film. The charge carriers furthest awa y from the gate can

    be screened by the lower OSC layer from up to 90% o f the appl ied gate vo ltag e, which means

    that the charge carriers are operating at a voltage much lower than the applied gate voltage. (11)

    Figure 2.2 shows the drop in the effective voltage seen by the OSC as the distance from the

    dielectric interface increases. Electrostatic m odeling has suggested that a cluster o f polar analytes

    can have a potential of a few tenths of a volt, which means that carriers further from the gate can

    be trapped by analytes .(93) Analytes may be more prevalent at grain boundaries of the OS C than

    the bulk, depending on the ability of the analyte to intercalate into the bulk O SC. The m ain

    benefits o f a transistor sensor over a ch em iresis to r are the ab ility to set the in it ia l charge ca rr ie r

    density with the gate electrode, the ability for multiple parameters to be measured, instead of a

    pure current measurement as in the case o f a chem iresistor, an d the ab il ity to inco rp orate the

    transistor into cascaded logical circuits. A transistor allows for current, thresho ld voltage ( V t ) ,

    and mobility (p) to be measu red.(/7)

    23

  • 8/11/2019 Interfacial Fields in Organic Field-effect Transistors and Sensors

    40/158

    semiconductorinsulatorV

    Vo

    Vo=Vs

    Vi

    V2V = 0

    0 d n d x

    Figure 2.2 The effective voltage seen by the molecular semiconductor as the distance from the dielectric

    interface is increased. (13)

    5. DNT Vapor Sensing

    Currently there is substantial effort in sensing for compounds used in explosives.

    Dinitrotoluene (DNT), can be used as a simulant for the more hazardous trinitrotoluene (TNT), as

    shown in Figure 2.3. DNT is also a byproduct o f the manufacturing o f TNT. DN T is a yellow

    solid with very low vapor pressure at room temperature. The saturation vapor pressure of DNT at

    room tem perature is 173 parts per billion (ppb), while he vapor pressure o f TNT is orders o f

    magnitude lower. (99, 10G)

    Both n- and p-channel semiconductors are used to fabricate OFETs that are responsive to

    DNT. Different responses are obtained from devices with various semiconductor materials and

    receptor layers. The pattern of responses includes sufficient diversity to set it apart from other

    vapors that might give similar responses to som e OFE Ts in the set, but n ot all o f them. T his

    allows for the possibility o f distinguishing DNT from other vapors in a man ner that c ould be

    further developed using mechanistic principles elucidated in this work.

    24

  • 8/11/2019 Interfacial Fields in Organic Field-effect Transistors and Sensors

    41/158

    n o 2 o 2N n o 2

    n o 2 n o 2

    DNT TNT

    Figure 2 J Chemical structures of DNT (left) and TNT (right).

    5.1 Exp erimen tal Setup

    A testing setup similar to the DMMP sensor experiments previously performed in the

    group using a vapor delivery system and testing chamber was fabricated to analyze the DNT

    vapor response for individual OFET sensors. (93) Figure 2.1 show s a schem atic d iagram o f the

    test chamber and DNT vapor delivery system. The vapor delivery system can be switched

    be tween air and DNT vapor. Whe n it is sw itch ed to air-on ly ; pure , fil tered ai r at co ntro lled flow

    rate is delivered into the testing chamber. To deliver the DN T analyte to the OFE T sensor, the

    airflow is switched so the airflow is now directed above the DN T solid in a bu bbler con tainer and

    carries a certain amount o f DNT vapor into the testing chamber. The DN T v apor g enerated by

    this system should be much lower than the saturation vapor of DNT, less than or equ al to 100

    ppb.

    5.2 DNT Sensor Results

    5.2.1 n-Channel Sma ll Molecule Semiconductors

    N ,N -Bis (3 ,5 -b is (t ri fluoro m ethyl) p hen ylm eth yl)n aphth a le ne-l ,4 ,5 ,8 -te tr acarboxyli c

    diimide (Bis-CF3 NTCDI) semiconductor. Bis-CF3-NTCDI is an air stable n-channel

    semiconductor (101)with fairly high electron mobility. The increased air stability is beneficial in

    sensor applications where the OFET is not packaged and the higher mobility allows for lower

    voltage OFET operation Figure 2.4a shows the chemical structure o f the bis-CF3 NT CD I

    molecule. The electron conducting bis-CF3 NTCDI is expected to have its source-drain current

    25

  • 8/11/2019 Interfacial Fields in Organic Field-effect Transistors and Sensors

    42/158

    partially quenched whe n it is expo sed to elec tron withd rawing analytes. Add itio nal ly , the charge

    carrier mobility would be expected to decrease. Ano ther benefit for sensing applications is that

    the minimum thickness of OSC required to achieve OFET o peration with this com poun d is very

    thin, 6 nm. As previously reported for DMM P, the thickness of the OSC layer of OFETs

    effectively determines the time taken for analytes to diffuse through the semiconductor layer to

    the conduction channel, and hence affects response time o f sensors.(92) Organ ic sem iconductors

    with thinner threshold thickness are thu s'more desirable for fast sensing.

    Heavily doped silicon substrates with 300 nm o f thermally grow n S i0 2 had an

    octyltriethocylsilane (OTS) SAM on the Si0 2 surface. The bis-CF3 NT CD I was thermally

    evaporated onto the OTS surface. Top contact gold source-drain electrodes were dep osited by

    thermal evaporation. The electrodes were evaporated through a shadow mas k with a spacing of

    L= 270 pm and W = 6.5 mm. Figure 2.4b shows the norm alized source drain curr ent at a Vd and

    Vg of 100V o f an OFET with 6 nm bis-CF3 NTCDI film. Devices with 6 nm bis-CF3 NTCDI

    exhibited the best response signals upon exposure to D NT vapors, using the sam e d elivery system

    as described above, at various concentrations. The gray squares indicate w hen 4 liters per minu te

    (LPM) o f airflow was directed over the DNT . Yellow , pink, and blue denote flowrates of 2 LPM ,

    6 LPM, and 8LPM respectively.

    26

  • 8/11/2019 Interfacial Fields in Organic Field-effect Transistors and Sensors

    43/158

    1.00

    0.95

    0.90

    0.85

    0.80

    0 10 20 30 4 0 50 6 0 70 80 90 100 110 120

    Time (mins)

    0.038

    0.037

    p 0.036

    0.035

    0.034

    0.033

    10 20 30 40 50 60 70 80 90 100 110 120

    Time (mins)

    -50.5

    -51.0-

    -52.5-

    -53.010040 60

    Time (mins)

    120

    (b)

    (C)

    ( d )

    Figure 2.4 The chemical structure of bis-CF3 NTCDI (a) and response of a 6 nm thick OSC film FET upon

    exposure to DNT vapors at various concentrations. The normalized response of source-drain saturation current

    (Id at/I

  • 8/11/2019 Interfacial Fields in Organic Field-effect Transistors and Sensors

    44/158

    concentration. The source-drain current of the OFE T decreased rapidly wh en D NT vap or was

    delivered into the testing chamber, and the current leveled o ff at a min imu m valu e after

    approximately 180 seconds. At this point, the total amount o f analyte accum ulated in the

    semiconductor layer began to saturate, and the rate of analyte desorption is countered by the

    adsorption rate. Once the chamber was vented with air, the source-drain current in the device

    recovered quickly. The change o f field effect mo bility follow ed the respon se of source-drain

    current when the OFET sensor was exposed to DNT vapors and air periodically. The magnitude

    o f mobility decrease is also proportional to DNT v apor concentrations. Th e thres hold v oltage

    response, while small, for the device can also be observed as shown in Figure 2.4d. Gentle air

    flow over the device was sufficient to recover some o f the initial current. I f the reco very needed

    to be faster it could be accelerated by using a higher a ir flow rate or slightly heating the d evice.

    N ,N -B ls (p en ta flu oro phenyle th yI) naph th a lene-1 ,4 ,5 ,8 - te t ra c a rb o x y li c - d iimid e

    (C2PhF5 NTCDI) semiconductor. OFETs with this n-channel semiconductor (102) are

    relatively unstable compared to bis-CF3 NTCDI when the thickness of semiconductor layer is

    ultrathin. The instability may be due to discontinuities in the thin film, and since the cu rrent m ust

    traverse relatively small area grain boundaries with a very high resistance, the OSC will heat up

    causing possible decomposition. In an exemplary OFE T sample, 15 nm C2PhF5 N TC D I (Figure

    2.4a) was deposited on Si0 2 /Si substrates. Again, top gold electrodes were deposited through

    shadow masks with the same w:l ratio. This device exhibited stable performance in air, as

    opposed to a 10 nm thickness, which did not, indicating 15 nm is the approximate threshold

    semiconductor thickness for this material. D NT vapor response w as then assessed w ith this 15 nm

    device.

    This device displayed a large response upon exposure to DNT vapors, as shown in Figure

    2.5. The current response did show some gate voltage dependence. At lower gate voltage, the

    relative curren t Id,Sa/Id,o changed m ore, but the base line also drifts m uch mo re. The res po nse o f

    28

  • 8/11/2019 Interfacial Fields in Organic Field-effect Transistors and Sensors

    45/158

    the mobility also tracked that of the current. Upon e xposure to DNT vap or, m obility decreased

    immediately, and started to increase once the testing chamber was vented with air. Interestingly,

    the threshold voltage VT did not show a response to DNT vapor. In Figure 2.5d, the curve is a

    straight line. VT kept increasing during the entire measurem ent, no m atter whe ther DN T v apor or

    air was delivered into the testing chamber. This is notably d ifferent from wh at was observed from

    bis-CF3 NT CD I, which displayed obviou s VT respon se . The DNT seems to slo w the m otion o f

    the electrons but not trap them for periods longer than the time scale of the measurem ents. This

    is an example of OFETs desirably responding in different ways upon exposure to the sam e

    analyte vapor, to help generate a pattern o f responses associated with a particular analyte o f

    interest.

    29

  • 8/11/2019 Interfacial Fields in Organic Field-effect Transistors and Sensors

    46/158

    (S

    A/

    4 8 12 16 20 24 28 32 36 40

    Time (mins)

    0.054

    0.052

    0.050

    \0.048

    0.046Mobility

    0.044 (c)

    0 4 8 12 16 20 24 28 32 36 40

    Time (mins)

    >

    > '

    ( d )

    0 4 8 12 16 20 24 28 32 36 40

    Time (mins)

    Figure 2.5 The chemical structure of C2PhF 5 NTC DI (a) and response o f a 15 nm thic k OSC film F ET upon

    exposure to DNT vapor. The normalized response of source-dra in saturation current (b); response of m obility

    (c); response of threshold voltage (d). The gray bars a re at an air flowrate o f 4 LPM.

  • 8/11/2019 Interfacial Fields in Organic Field-effect Transistors and Sensors

    47/158

    5.2.2 p-Ch annel Small Molecule Semicondu ctors

    5,5-bis(4-hexylphenyI)-2,2-bithiophene (6 P T T P 6 ) semiconductor. 6 PTTP6 has been

    used as a p-channel semiconductor by other group m embers. {103, 104)The chem ical structure of

    6PTTP6 can be seen in Figure 2.6. Devices made from two molecular mo nolay ers (6 nm) of

    6PTTP6 were deposited by thermal evaporation, follow ed by the deposition of go ld top electrodes

    through shadow masks. Substrate temperature was nom inally kept at room tem perature. The

    botto m pa rt o f Figure 2.6 shows the respon se o f th is device up on exposure to DNT va por at a

    flowrate of 4 LPM. The response to DNT can be seen, how ever the signal to noise ratio is poorer

    than other OFET sensors. Current decreased right after the exposure to DNT vapor, and partially

    recovered after the chamber was vented with air. When the 6 PTTP6 OFET sensor is compared to

    the bis-CF3 NT CDI sensor, it is not considered as sensitive to DN T.

    1.05Vg=-4QV

    Vg=-50V

    Vg=-60VVg=-7QV

    Vg=-80V1.00

    o

  • 8/11/2019 Interfacial Fields in Organic Field-effect Transistors and Sensors

    48/158

  • 8/11/2019 Interfacial Fields in Organic Field-effect Transistors and Sensors

    49/158

    1.02-

    1.01-

    1.00-o 0.99-TJ

    0.98-

    *o~ 0.97-

    0.96

    0.95-

    0.94

    0.93-

    *82saos,

    -V =-40V - V -60V

    - Vg= -80V,

    DNT

    (b)

    CO

    I&15

    0.02230

    0.02225

    0.02220

    0.02215

    0.02210

    0.02205-

    0.02200

    0.02195

    Time (mins)

    n

    fT

    -------

    c

    r

    -----

    pcPo

    ce-

    0

    DNT

    (c)

    0 2 4 6 8 10 12 14 16

    -25.2

    -25.6-

    > ~ -26.0-

    DNT

    -26.4-

    0 2 4 6 8 10 12 14 16

    (d )

    Time (mins)

    Figure 2.7 The chemical structure of 2PTTP2 (a) and response of a 4.5 nm thick OSC film FET upon exposure

    to DNT vapor. The normalized response of source-drain saturation current (Id,sat/Id^at,o) (b); response of

    mobility (c); response of threshold voltage (d). The gray bars are at an air flowrate o f 4 LPM.

    Co pper phthalocyanine (CuPc) sem iconductor. CuPc is another p-channel small

    molecule semiconductor, but unlike the other semiconductors discussed so far it has a co ppe r core

    33

  • 8/11/2019 Interfacial Fields in Organic Field-effect Transistors and Sensors

    50/158

    inside the molecular structure as seen in Figure 2.8a. Some groups had shown that Cu Pc was

    responsive to DNT vapor but due to the better packing o f the CuPc the responses w ere muc h

    lower than other- OSCs. (105-108) 20 nm o f CuPc was deposited at an elevated substrate

    temperature of 75 C since lower temperature depositions resulted poor quality films without

    field effect behavior. The films were deposited on hexamethyldisilazane (HM DS) treated

    substrates. 50 nm gold electrodes were therm ally evaporated and defined by the same w:l ratio

    shadow mask. The mobility response and normalized current response to the DNT vapor at 4

    LPM was very small and did not recover as well compared to the bis CF3 NTCDI OFET sensors.

    The threshold voltage plot in Figure 2.8c shows gate bias stress occurring but no response to the

    DNT vapor.

    34

  • 8/11/2019 Interfacial Fields in Organic Field-effect Transistors and Sensors

    51/158

    1.32x10(/)

    4

    Eo

    CM

    -QO

    Cu

    (a)

    7 *

    (b)

    5 10 15 20Time (min)

    > -10

    _ (c)

    5 10 15 20Time (min)

    Figure 2.8 The chemical structure of CuPc (a). The mobility (b) and threshold voltage (c) respon se of a 20 nm

    thick OSC film FET upon exposure to 4 LPM DNT vapor. The gate and drain voltage are at -50V for the scans

    above.

    Alpha-sexithiophene (-6 T) and PT PT P semiconductors. Some small molecule

    organic semiconductors either show negligible or no response to DNT vapor like the previously

    discussed CuPc. Two additional examples are a-6 T and PTPTP shown in Figure 2.9 and 2.10

    35

  • 8/11/2019 Interfacial Fields in Organic Field-effect Transistors and Sensors

    52/158

    respectively. The top of the figures show the molecu lar structure of the OSC . Alp ha-

    sexithiophene is one of the earliest p-channel se micon ductors.(709) An OF ET w ith 15 nm a- 6T

    semiconductor layer only exhibited a slight decrease in source-drain current upon exposure to

    DNT vapor. PTPTP is another commonly used p-channel semiconductor. This material displayed

    no response to DNT vapor, as shown in the bottom o f Figure 2.10.

    Figure 2.9 The chemical structure of a-6T (top) with the corresponding response (bottom) upon exposure to 4

    LPM DNT vapor.

    a l p h a - s e x i t h i o p h e n e

    0 2 4 6 8 10 12T i m e ( m i n s )

    36

  • 8/11/2019 Interfacial Fields in Organic Field-effect Transistors and Sensors

    53/158

    P T P T P

    1.4

    1.2

    2 1.0

    _! 0.8^ 0.6

    0.4

    0.20.0

    0 2 4 6 8 10 12

    T i m e ( m i n s )

    Figure 2.10 The chemical structure of PTPTP ( top) with the corresponding response (bottom) upon exposure to

    4 LPM DNT vapor.

    5.2.3 p-Ch annel Polymer Semiconductors

    Poly(3,3 -didodecylquaterthiophene) (PQT12) semiconductor. OFETs based on

    the relatively stable p-channel polym er semicon ductor poly(3 ,3 -didodec ylquaterthioph ene)

    (PQT12) (110-113) were fabricated by spin coating. Fig. 2.11a shows the chemical structure of

    the PQT12 molecular repeat unit. The polymer was dissolved in ehlorobenzene at a concentration

    of 4 mg/mL. T he PQT12 solution was heated to 80 C to achieve a completely dissolved solution ,

    filtered through a 0.45 pm po ly(tetrafluoroethylene) (PTFE) filter, and sonicated w hile cooling to

    room temperature. The polym er was deposited onto the Si/S i0 2 substrates by spin coating at

    2000 RPM for one minute. Gold electrodes were then thermally evaporated using the same w:l

    ratio shadow mask as the previous OFE Ts (1 = 270 pm and w = 6.5 mm).

    | DNT Vapor

    -

    Vg=-40V- o - Vg=-60V

    Vg50V~Vg70V Vg=-80V

    37

  • 8/11/2019 Interfacial Fields in Organic Field-effect Transistors and Sensors

    54/158

  • 8/11/2019 Interfacial Fields in Organic Field-effect Transistors and Sensors

    55/158

    c 12h25

    12n 25 (a)

    CO

    0.046-

    0.045-

    0.044-Eo. 0.043-

    $0.042-

    j O

    0.041-

    0.040-

    8LPM 6LPM 4LPW 4LFM 2LPM

    20 30 40

    Time (mins)

    6LPM 6LPM 4LPM 4LPM 2LPM

    (C)

    20 30 40

    Time (mins)

    Figure 2.11 The chemical structure of PQT12 (a) and response of an OFET upon exposure to DNT vapor. The

    response of mobility (b); response of threshold vo ltage (c). The blue, pink, gray and yellow b ars indica te air

    flowrates of 8, 6, 4 , and 2 LPM respectively. The gate and drain voltages each at -100 V for the scans.

    (PIFTBT-10) semiconductor. Another p-channel polymer synthesized for solar cell

    applications was utilized as an OFET senso r.(/74 ) The chem ical structure of this polym er can be

    seen in Figure 2.12a. The PIFTBT-10 O FETs were fabricated by spin coating. The po lym er was

    39

  • 8/11/2019 Interfacial Fields in Organic Field-effect Transistors and Sensors

    56/158

    dissolved in chlorobenzene at a concentration o f 7 mg/m L an d gently heated to a chieve a

    completely dissolved solution. The solution was then filtered through a 0.45 p m PT FE filter and

    allowed to cool. The polymer was deposited onto the Si/SiC>2 substrates by spin coating at 500

    RPM for 20 seconds and 2000 RPM for 40 seconds. The device electrodes were then deposited in

    the same manner as the PQ T12.

    The normalized source-drain current at different gate voltages, field-effect mobility and

    threshold voltage response to the DNT vapor can be seen in Figure 2.12. The device showed

    reproducible response to exposure to DNT vapor, with the change in mobility slightly larger for

    PIFTB T than PQT12. How ever the PIFTBT -10 is less air stable, with grea ter bias stress

    observed. W hile the response is not as fast as the PQT 12 OFET it still is fairly quic k with a t90=

    1.5 minutes. The device was more responsive at higher gate voltages, which is somewhat more

    atypical. The response at higher gate voltages can be due the larger gate bias stress observed in

    the PIFTBT-10.

    40

  • 8/11/2019 Interfacial Fields in Organic Field-effect Transistors and Sensors

    57/158

  • 8/11/2019 Interfacial Fields in Organic Field-effect Transistors and Sensors

    58/158

    of threshold voltage (d). The gray bars are at an air flowrate o f 4 LPM. Th e gate and drain voltages are each at

    -100 V (c & d)

    5.2.4 Com binedp-Channel and n-C hann el OFETs

    PQT 12 w ith car bo n black. Pure PQT 12 devices were fabricated the way previously

    describe