Wet Organic Field Effect Transistor as DNA sensor

(1)

Department of Physics, Chemistry and Biology

Wet Organic Field Effect Transistor as DNA sensor

Yu-Jui Chiu

2008-04-23

Supervisor

Mattias Andersson

Examiner

Olle Inganäs

(2)

(3)

Datum

Date 2008-04-23

Avdelning, institution

Division, Department

Biomolecular and Organic Electronics

Department of Physics, Chemistry and Biology Linköping University

URL för elektronisk version

Rapporttyp Report category Licentiatavhandling Examensarbete C-uppsats D-uppsats Övrig rapport _____________ Språk Language Svenska/Swedish Engelska/English ________________ ISBN ISRN: LITH-IFM-EX--08/1932--SE _________________________________________________________________

Serietitel och serienummer ISSN

Title of series, numbering ______________________________

Titel

Title

Wet Organic Field Effect Transistor as DNA Sensor

Författare

Author

Yu-Jui Chiu

Sammanfattning

Abstract

Label-free detection of DNA has been successfully demonstrated on field effect transistor (FET) based devices. Since conducting organic materials was discovered and have attracted more and more research efforts by their profound advantages, this work will focus on utilizing an organic field effect transistor (OFET) as DNA sensor.

An OFET constructed with a transporting fluidic channel, WetOFET, forms a fluid-polymer (active layer) interface where the probe DNA can be introduced. DNA hybridization and non-hybridization after injecting target DNA and non-target DNA were monitored by transistor characteristics. The Hysteresis area of transfer curve increased after DNA hybridization which may be caused by the increasing electrostatic screening induced by the increasing negative charge from target DNA. The different morphology of coating surface could also influence the OFET response.

Nyckelord

Keyword

Organic field effect transistor (OFET), Wet OFET, Poly(3-hexylthiophene) (P3HT), DNA sensor, Conjugated polymer, Micro fluidic channel.

http://urn.kb.se/resolve?urn= urn:nbn:se:liu:diva-11761

(4)

(5)

Upphovsrätt

Detta dokument hålls tillgängligt på Internet – eller dess framtida ersättare – under 25 år från publiceringsdatum under förutsättning att inga extraordinära omständigheter uppstår.

Tillgång till dokumentet innebär tillstånd för var och en att läsa, ladda ner, skriva ut enstaka kopior för enskilt bruk och att använda det oförändrat för ickekommersiell forskning och för undervisning. Överföring av upphovsrätten vid en senare tidpunkt kan inte upphäva detta tillstånd. All annan användning av dokumentet kräver upphovsmannens medgivande. För att garantera äktheten, säkerheten och tillgängligheten finns lösningar av teknisk och administrativ art.

Upphovsmannens ideella rätt innefattar rätt att bli nämnd som upphovsman i den omfattning som god sed kräver vid användning av dokumentet på ovan beskrivna sätt samt skydd mot att dokumentet ändras eller presenteras i sådan form eller i sådant sammanhang som är kränkande för upphovsmannens litterära eller konstnärliga anseende eller egenart.

För ytterligare information om Linköping University Electronic Press se förlagets hemsida

http://www.ep.liu.se/.

Copyright

The publishers will keep this document online on the Internet – or its possible replacement – for a period of 25 years starting from the date of publication barring exceptional circumstances.

The online availability of the document implies permanent permission for anyone to read, to download, or to print out single copies for his/her own use and to use it unchanged for non-commercial research and educational purposes. Subsequent transfers of copyright cannot revoke this permission. All other uses of the document are conditional upon the consent of the copyright owner. The publisher has taken technical and administrative measures to assure authenticity, security and accessibility.

According to intellectual property law the author has the right to be mentioned when his/her work is accessed as described above and to be protected against infringement.

For additional information about Linköping University Electronic Press and its procedures for publication and for assurance of document integrity, please refer to its www home page:

http://www.ep.liu.se/.

(6)

Abstract

Label-free detection of DNA has been successfully demonstrated on field effect transistor (FET) based devices. Since conducting organic materials was discovered and have attracted more and more research efforts by their profound advantages, this work will focus on utilizing an organic field effect transistor (OFET) as DNA sensor.

An OFET constructed with a transporting fluidic channel, WetOFET, forms a fluid-polymer (active layer) interface where the probe DNA can be introduced. DNA hybridization and non-hybridization after injecting target DNA and non-target DNA were monitored by transistor characteristics. The Hysteresis area of transfer curve increased after DNA hybridization which may be caused by the increasing electrostatic screening induced by the increasing negative charge from target DNA. The different morphology of coating surface could also influence the OFET response.

(7)

Acknowledgement

Over the years, I would like to express my sincere gratitude to the following people.

First of all, I want to thank Olle Inganäs. Thank you for providing me an opportunity to do this thesis work in your group. Your humor and your considerable inspiration impress and enlighten me a lot. You are such a real scientist in my mind, and it is my honor to work in your group. I would like to thank my supervisor Mattias Andersson, who is the most important person during these months. Thank you for teaching me a lot from the basic theories to the attitudes a scientist should have. Thanks for your patience and accurate guidance. I really learn a lot from you. Thanks to all the members in Biomolecular and Organic Electronics. Karin Magnusson and Jens Wigenius gave me very useful supports of DNA samples, Mahiar Hamedi always carries out inspired discussion, Yinhua Zhou and Fengling Zhang taught me many about polymer properties, Bo Thuner gave me engineering advisement, and, of course, all of you who gave me any supports and response during the group meeting and in the laboratory. I feel enjoyable and comfortable to work in this group. And I will always remember the exciting-brainstorming time. Thanks to my family, my mother, father, and sister. You always accept me and support me to chase my dreams. Because of you, I can face the unknown future with brave and optimistic attitude. Although Sweden is far from Taiwan, I can still your encouragement.

(8)

I would like to thank Yun-Hsuan Chen, who is thoughtful and considerate, and always gives me many mental and physical supports. Your happiness can always sweep away the whole day tiredness. Thank to Rosa Huang, who supplied professional and talented skills of illustration. Also, I would like to thank all the friends from my country, Taiwan, and who I met in Sweden. All of you enrich my life and leave sweet memories of Sweden.

Last but not least, I would like to thank Leif Johansson, who is always pleased to help me to deal with troublesome problems of everyday life. You are always shown up as a hero who takes care of us when needed.

Again, thank you all!

Linköping, April 2008 Yu-Jui Chiu

(9)

Chapter 1 Introduction

1.1 Background and Goal

1.1.1 Background

DNA or sequences on it is the smallest unit that contains hereditary information. Researchers have done huge amounts of works on DNA, expecting to solve the mystery of life since DNA is not only the blueprint but also like a historical record of life. Several applications of DNA highlight its important role: Genetic engineering open the doors of improving medical researches. Detection of DNA gives forensic scientists a trustable way to get clues, and also give historians and anthropologists an accessible clue to discover the relationship between different lives or people. DNA has been applied in computer science and nanotechnology. No matter what the application is, the most important step is to detect or sense the DNA sequence. After conducting polymers were discovered in 1977, they were used frequently in research and its application. Flexible, lightweight, bio-friendly, low cost, and so on are some wide known advantages of polymers. Compared to inorganic materials, polymers are less understood and theorized, however, their potential explains why people has done much research on it so far, and will also continue to do so in the future.

Application of polymers or polymer based devices into biotechnology, such as for DNA sensor is undoubtedly one of the most interesting topics today and in the future. Field effect transistor (FET) has been proven as an appropriate device to detect label-free DNA in the recent two decades [1]-[6]. The DNA immobilized on the FET structure influences the surface potential by

(12)

2

its intrinsic molecular charge [7]. The theoretical description of the observed effects, however, is still under discussion [8] [9]. Most DNA sensors based on FET are of extended-gate structure, and require complex manufacturing and evaluation. Organic material based DNA sensors are less reported.

1.1.2 Aim of This Work

This work is aimed to evaluate the possibility of using a wet organic field effect transistor (WetOFET) to detect the hybridization of DNA chains. Differing from the extended-gate FET, a more convenient design, with a transporting fluidic channel, is used in this work. A micro fluidic interface will be built on an OFET allowing WetOFET measurements in the presence of fluid. Single strand DNA with a cholesterol tag will be introduced on the fluid-polymer (active layer) interface. The characteristics changes after injecting non-complementary DNA and complementary DNA will be compared and discussed with respect to possible DNA sensing application.

1.1.3 Outline

The basic concept of organic field effect transistors and DNA will be given in Chapter1. The main goal of this work and the structure of WetOFET will be presented in Chapter2. All the experimental details are in Chapter3, and the results and discussions are in Chapter4. Chapter5 is the summary of this work.

1.2 Theoretical Background

1.2.1 Conjugated Polymers

In 1977, Heeger, McDarmid, and Shirakawa demonstrated that a certain polymer, polyacetylene, can be converted into a metallic-like conducting material if exposed to chemical dopants. This discovery was awarded the Nobel Prize in Chemistry (2000), and opened a fantastic new era of research on polymers. The basic knowledge of conducting polymers will be described from the density of states via carbon-carbon bonds, the most seen bond in polymer, to the polymeric semiconductor and its charge carriers in the following.

1.2.1.1 Density of States

The number of allowed energy levels for an electron as a function of energy is called the density of states. The simplest case is an isolated atom where the electrons have discrete energy levels.

(13)

3

For example, the energy levels of an isolated hydrogen atom are given by the Bohr model [10]. For two identical atoms, when they are far apart, the allowed energy levels for a given principal quantum number consist of one doubly degenerate level resulting in both atoms having the same energy. When the two isolated atoms are brought closer, the doubly degenerate energy levels will split into two levels by interaction between the atoms. When a number, N, of isolated atoms are brought together and bound to each other, the interactions include the forces of attraction and repulsion, causing a shift in the energy levels. When N is large, the result is an essentially continuous energy band. The electrons always obey the Pauli exclusion principle, that two electrons can occupy one energy level. If all atoms contribute one electron, at a temperature of absolute zero (0K), the electrons will occupy all states in the valance band and the conduction band will be empty. The bottom of the conduction band is called Ec, and the top of the valence band is called Ev. The difference between Ec and Ev is the band gap energy Eg. The size of Eg will decide whether the material is an insulator or semiconductor. Conductors have no Eg.

1.2.1.2 Hybridization and Conjugation

The electron configuration can be described as a wave function, and the electrons of an atom are spatially distributed over a set of orbitals defined by probability. Regarding the single electron approximation, carbon has the ground state electronic configuration 1s2_2s2_2p2_{, denoting that}

there are two electrons in each 1s-orbital, 2s-orbital, and 2p-orbital respectively. 2s22p2 represents especially the four valence electrons in carbon. Furthermore, the wave function of the electrons can be described by hybridized bonding orbital, which represents the total charge distribution in a given bonding configuration depending on different interactions with surrounding atoms. The three hybridized orbitals of carbon are shown in Figure1.1.

(14)

4

z y x sp：Linear sp2_：_Triangular sp3_：_Tetrahedral y z x 120° z y x 109.5° z y x z y x y x sp：Linear sp2_：_Triangular sp3_：_Tetrahedral y z x 120° _y z x y z x 120° z y x 109.5° z y x z y x 109.5°

Figure 1.1. Hybridization forms of carbon.

For example, in Figure 1.2, the bonding between carbons in polyacetylene is formed by sp2-hybridized orbitals. Three electrons in the 2p- and 2s-orbital are hybridized into a sp2-orbital, which form σ-bonds, while the last 2p-electron forms the π-bond. The extra bond between two carbon atoms appears as a double bond (C=C) in chemical notation.

Sp2 Energy 2p 2s 1s 2p 1s Hybridize Sp2 Energy 2p 2s 1s 2p 1s Hybridize Energy 2p 2s 1s 2p 2s 1s 2p 1s Hybridize (a)

(15)

5

π-bonding C H C pσ-bonding C H C sp2_p z π-bonding C H C pσ-bonding C H C sp2_p z (b)

Figure 1.2. sp2 hybridization is shown by (a) energy levels and (b) 3-dimensional model of carbon-carbon bond in polyacetylene. The π-bonding forms the extra bond.

The π-orbital will be focused on now, since it influences the electrical properties. Again, when two 2p-orbitals are brought close, they are described by new wave function. Orbitals having the same energy level will split into two different orbitals with different energy levels. One is bonding and marked as π, the other is anti-bonding and marked as π*. When there is a huge amount of 2p-orbitals such as in a polyacetylene polymer chain, the energy levels will form two continuous energy bands, which are shown in Figure 1.3. The orbital that is occupied by the highest energy electrons is called the Highest Occupied Molecular Orbital (HOMO) and the lowest unoccupied orbital is called Lowest Unoccupied Molecular Orbital (LUMO).

E

1 2 4 8 ∞

2p

π

π*

Number of carbon atom in molecule

E

1 2 4 8 ∞

2p

π

π*

Number of carbon atom in molecule

Figure 1.3. The overlap of π-orbitals generate split energy states. The number of states are dependent on the number of overlapping π-orbitals. If the number is large, the energy states will form the energy bands.

(16)

6

1.2.1.3 Charge Carriers

Polymers can be divided into two categories depending on the available energy states of the system. One is called degenerate ground state systems, and the other is called non-degenerate ground state systems. If single and double bonds can be interchanged without changing the ground state energy, the polymer system is denoted as a degenerate ground state system. If the interchange of single and double bonds is associated with two states of energy levels, the system is said to have a non-degenerate ground state. Polyacetylene is found with two different forms, trans-polyacetylene or cis-polyacetylene, as shown in Figure 1.4. Trans-polyacetylene has degenerate ground state while cis-polyacetylene has non-degenerate ground state. Cis-polyacetylene has two forms with two different energies, aromatic and quinoid; the quinoid form is a higher energy state.

trans-polyacetylene cis-polyacetylene aromatic quinoid ＝ trans-polyacetylene cis-polyacetylene aromatic quinoid trans-polyacetylene cis-polyacetylene aromatic quinoid ＝

Figure 1.4. Trans-polyacetylene and cis-polyacetylene which represent degenerate- and non-degenerate ground state systems respectively.

In order to have a conducting polymer, there must be charge carriers in the polymer chain. Charge carriers can be created through oxidative or reductive doping processes. The oxidative doping, p-doping, is the result of withdrawing an electron from the π-bond system, which produces a positive charge unit on the conjugated polymer:

Equation 1

P-doping: P + yA-→ Py+A-y + ye-

Where P denotes the polymer chain, A denotes the charge-compensating counter ion, e-is the electron and y is the number of counter ions. Instead of oxidation, reductive doping, n-doping, is caused by introducing electrons into the π-bond system which produces a negative charge unit in the conjugated polymer:

(17)

7

Equation 2

N-doping: P + ye- + yA+ → Py-A+y

Charge carriers are transported along the π-bonds of the polymer chain. The charge carrier can either be soliton, polaron, or bipolaron. For example, combining two segments of trans-polyacetylene within different bond order, a defect in the form of an unpaired electron is created. This defect is called a neutral soliton. Solitons can either be neutral, positive, or negative depending on the number of carried electrons, zero, one, or two. Different solitons are shown in Figure 1.5. Neutral soliton Positive soliton Negative soliton Neutral soliton Positive soliton Negative soliton (a) (b)

Figure 1.5. Three classes of soliton in polyacetylene. (a) Number of electron, one, zero, or two, define the neutral, positive, or negative soliton. (b) Neutral, positive, and negative soliton shown by band structures.

(18)

8

For non-degenerate polymers, the charge carriers are called polarons or bipolarons [11]. By oxidizing the polymer, an electron is removed and a cation-radical pair is created. In between the cation and radical, a transformation from aromatic form to quinoid form occurs. This confined change and associated charge is called a polaron. The polaron occupies an energy level which is represented by poly(3-hexylthiophene) P3HT in Figure 1.6. The quinoid structure has a higher energy compared to the aromatic structure. So in contrast to the soliton, the polaron must overcome an energy activation barrier associated with quinoid and aromatic structure differences while moving. If two electrons are withdrawn from the polymer, it forms a bipolaron. And if the polymer has been oxidized even further, bipolaron energy bands will be generated.

(19)

9

Conduction Band

Valence Band

Neutral Polymer Polaron Bipolaron Bipolaron Band Conduction

Band

Valence Band

Neutral Polymer Polaron Bipolaron Bipolaron Band

(b)

Figure 1.6. (a) Generation of positive polaron and bipolaron in P3HT. (b) Energy level of the neutral polymer, a polaron, a bipolaron, and a polymer with bipolaron energy band are shown.

1.2.2 Organic Field-Effect Transistor

Field-effect transistor (FET), which is nowadays the most common transistor type, has been presented for lots of application such as logical circuits and gas sensors. FETs using conducting polymers are normally called organic field-effect transistor (OFET). It has started attracting considerable attention in this decade due to potential application of inexpensive and flexible electronic devices [12]. In normal situation, there are many similarities between FET and OFET, which imply that people can operate both of them in the same way. However, the physical explanations may be different, most of which are not understood well enough. The term OFET or any analogue term, is commonly used for devices where the active part of the transistor is an organic material. The typical structure and operation of OFET will be discussed in the following.

1.2.2.1 Basic Structure of OFET

Most OFET [13] contains four essential parts. Firstly, the gate is used to generate the modulating field. Secondly, two electrodes, source and drain, are used to define the carrier channel. The third is the dielectric layer. Last but not least is the active layer, which can be

(20)

10

chosen from an appropriate organic material. The characteristics will change due to different modulation of gate, source, and drain.

When applying a gate voltage, charges will start to accumulate at the interface of the dielectric layer and the active layer. When a certain number of charge carriers are created and the potential generated by the source and drain is enough to overcome the activation barrier, the transistor has been turned on.

The potential difference between source and drain will result in a gradient in the carrier concentration distribution in the channel. If the potential between source and drain is smaller than gate bias after the channel has been created, the current through the channel can be controlled and is proportional to the potential of source and drain. This is called linear mode or ohmic mode. While the potential between drain and source is close to the gate voltage (gate potential minus threshold voltage), the asymmetrical distribution of carriers will create a “pinch off” near the drain end of the channel which means there are no carriers at or near the drain end. During pinch off, the current through the channel is no longer dependent to the potential difference of drain and source, this is called saturation mode.

Gate Source Drain Active layer Insulator (dielectric layer) Gate Source Drain Active layer Insulator (dielectric layer) (a) Gate Insulator

(dielectric layer) Source and Drain _{Active layer}

Gate Insulator

(dielectric layer) Source and Drain _{Active layer}

(b)

(21)

11

1.2.2.2 Transfer, Output, and Time Measurements

There are many methods to extract information from the transistor. Transfer characteristics is one of the most common methods. In a transfer characteristics measurement, the voltage bias of source and drain are kept at a certain value, and the gate voltage swept through a range of suitable values. Ideally, the saturation current that flows from source to drain can be described by Equation 3 [14]: Equation 3

(

2 T G 0 n sat V -V 2L C W I ⎟ ⎠ ⎞ ⎜ ⎝ ⎛ ≅ μ

)

Output characteristics are obtained by stepping the gate voltage and sweeping the drain or source voltage between certain values for each gate step. Time measurements are done by adding time steps to measurement and monitoring the source-drain current. In this work, the source and drain are usually kept at constant value and the current dependence on time and with changing gate bias is investigated.

1.2.3 Fluorescence

Fluorescence is a luminescence phenomenon observed while suitable excitation and emission processes of a specific material occur under exposure of certain energy photons. In scientific applications, the specific material is usually called a fluorophore. Fluorophores can be attached to non-fluorescent molecules to make them fluorescent. The fluorescence can then tell whether the attached molecule is definitely present. The fluorescent dye used in this thesis work is Alexa350, which has an excitation peak at 350nm and an emission peak at 441nm, see Figure 1.7 [15].

(22)

12

1.2.4 DNA and Hybridization

Chemically, DNA or Deoxyribonucleic acid is a long polymer chain composed of a sequence of simple units called nucleotides. The Backbone of DNA is made of sugar and phosphate groups, see Figure 1.8(a). Attached to each sugar is one of four types of nitrogenous bases; adenine, thymine, cytosine, or guanine abbreviated as A, T, C, and G respectively [16][17]. The phosphate groups on the backbone contain negative charges, which are assumed to be detectable by the OFET in this thesis work. The DNA double helix is formed by hydrogen bonds between certain base pairs, which are A-T and C-G, foremost because the specific chemical structure of these base pairs are advantageous for forming hydrogen bonds with each other, the hydrogen bonding of A-T and C-G are shown in Figure 1.7(b). Commonly, the DNA will be shown by letter combination instead of a drawn structure. For example, 5’-CCC TCC GTC GTG CCT or 3’-CCC AGG CAG CAC GGA is shown by a certain number with each abbreviation of bases. The 5’(five prime) and 3’(three prime), which means a certain carbon on the sugar, are referring to the sequence of the DNA chain. Usually, for convenience, 5’ is called the head side and 3’ is called the tail side.

(23)

13

N N N N N H₂ N N N N H O N H₂ N N N H₂ O N _N O O H O O P O O O O O O P O O O O O P O O O O O P O O O N N O NH₂ N N N NH O NH₂ N N N N NH₂ N N O O H O O P O O O O O O P O O O O O P O O O O O P O O O Adenine Thymine Guanine Cytosine 5’end 3’end 3’end 5’end N N N N N H₂ N N N N H O N H₂ N N N H₂ O N _N O O H O O P O O O O O O P O O O O O P O O O O O P O O O N N O NH₂ N N N NH O NH₂ N N N N NH₂ N N O O H O O P O O O O O O P O O O O O P O O O O O P O O O N N N N N H₂ N N N N H O N H₂ N N N H₂ O N _N O O H O O P O O O O O O P O O O O O P O O O O O P O O O N N O NH₂ N N N NH O NH₂ N N N N NH₂ N N O O H O O P O O O O O O P O O O O O P O O O O O P O O O Adenine Thymine Guanine Cytosine 5’end 3’end 3’end 5’end (a) N N N N O H N H H N N N H O H Guanine Cytosine N N O O H N N N N N H H Adenine Thymine N N N N O H N H H N N N H O H Guanine Cytosine N N N N O H N H H N N N H O H N N N N O H N H H N N N H O H Guanine Cytosine N N O O H N N N N N H H Adenine Thymine N N O O H N N N N N H H N N O O H N N N N N H H Adenine Thymine (b)

Figure 1.8. DNA structure. (a) Double strand DNA. The oxygen on the phosphate contains one negative charge. (b) Hydrogen bond between the nitrogenous bases, adenine, thymine, cytosine, and guanine.

(24)

(25)

Chapter 2 Architecture of Wet Organic Field-Effect Transistor

Wet organic field effect transistor (WetOFET) contains three parts: main body (OFET), injector holder, and injector which are shown in Figure 2.1 (c), (b), and (a) respectively. While injecting fluid by a syringe, the fluid will go form the injector via the holder and through the fluid channel of the main body. Filter papers will collect the waste fluid at the other side.

(a) (b) (c) (a) (b) (c)

Figure 2.1. Separated parts of WetOFET (a) injector (b) injector holder (c) transistor The main body of the WetOFET is constructed as following: An ITO coated glass is appropriately etched and used as the base. The length and width of carrier channel were about 200μm and 1cm respectively. The reason for etching the ITO at the sides is that sometimes the polymer coating is not good at the edge of the substrate, and it is also convenient for processing

(26)

16

the device without damaging the polymer in the channel. After spin coating the polymer, a spacer, usually double sided tape, is used to define a fluid channel, ideally as wide as the current channel, which is shown in Figure 2.2. The gate electrode finally caps the top, creating a fluid channel with the same width as the length of the transistor channel and same length as that of the glass base . Be careful, the spacer should be well attached to both polymer surface and gate in order to avoid leakage current from either the gate or ITO. It is important to avoid shear forces when attaching the gate; otherwise the spacer may move and destroy the polymer film or channel alignment. A side view of a WeTOFET is shown in Figure 2.3.

ITO Spacer Substrate ITO Spacer Substrate

Figure 2.2. The double side tape define the fluid channel.

Spacer Gate Substrate Spacer Gate Substrate

(27)

17

The injector holder is made of a thin PDMS film that is 0.5 to 1 mm and can be easily produced by putting a PDMS droplet on a smooth plastic surface and covering it with a piece of glass before heating. An injector of appropriate size, wider than the thickness of OFET and longer than the side of the gate electrode can be cut from the thin film. A hole which is few times larger than the fluid channel is made at the center of thin film. This holder can be attached on the side of the OFET, and can support the OFET when standing on a heater of 75 degree Celsius, with the holder side downward as is shown in Figure 2.4. The holder and OFET can be sealed by PDMS on the heater. The holder is not only for attaching the injector but can also prevent the PDMS from going from the edge into the fluid channel and sealing the entrance.

PDMS

Heater PDMS

Heater

(28)

18

The injector is made by PDMS with an IC-socket pattern, see Figure 2.5. The diameter of the socket is slightly smaller than the syringe head, so the syringe head can be well covered by the PDMS without fluid leakage while injecting. The smooth side of the injector can be attached to the holder, and can be sealed by PDMS on the heater, too. Finally, all the contacting edge of PDMS and glass need to be sealed by instant adhesive to make sure the fluid will not leak out due to the syringe pressure.

(29)

Chapter 3 Experimental Details

3.1 Material

3.1.1 Poly(3-hexylthiophene)

P3HT, shown in Figure 3.1, has been demonstrated as semiconducting layer of FET since 1988 [18] The charge carriers are holes which could be polaron or bipolarons. P3HT has excellent electronic and mechanical properties, and a high environmental stability [12]. A quite small threshold voltage was reported in P3HT based FET having a hygroscopic polymer as gate dielectric [19]. P3HT is thus undoubtedly an appropriate choice as the semiconducting layer (active layer) for this application. P3HT is dissolved in chloroform with a 5mg/ml concentration. The solution was heated to 70 degree Celsius and stirred at 2500rpm by a vortex for few seconds before spin coating.

S

C

₆

H

₁₃

C

₆

H

₁₃

C

₆

H

₁₃

...

n

Figure 3.1. Structure of poly(3-hexylthiophene) (P3HT)

(30)

20

3.1.2 Cholesterol Tag

Cholesterol, shown in Figure 3.2, is a lipid that can be found in cell membranes of all animal tissue. In this work, its hydrophobic character make it attach to the P3HT which is also hydrophobic, and introduces the DNA strand close to the P3HT surface if the DNA has been synthesized with a cholesterol tag.

CH₃ O H CH₃ C H₃ CH₃ C H₃

Figure 3.2. The chemical structure of cholesterol which is used as tag on DNA and which can be attached to hydrophobic surface, for example P3HT.

3.1.3 DNA samples

There are five kinds of DNA used in this work. DNAa-Alexa350 with cholesterol tag, which is used to check whether the DNA with a cholesterol tag can attach to the P3HT surface because of its hydrophobic property. DNAa with cholesterol tag which is introduced into the fluid channel as the first single strand and which can be hybridized with its complementary DNA strand. DNAx with cholesterol tag can hybridize with the tail side of DNAa and form DNAxa to increase the chance of DNA to attach to the P3HT surface since there is one more cholesterol tag. DNAa’ that is used to hybridize with DNAa, and it is also called complementary part of DNAa. DNAb’ is used as the reference for DNAa’ because it will not hybridized with DNAa. The DNA configuration is shown below:

DNAa-Alexa350 3’CCC AGG CAG CAC GGA ACC TGT AGT CTT TAT

[5’]Alexa350[3’]Cholesterol-Tag DNAx 5’-CCC TCC GTC GTG CCT

[5’]Cholesterol-Tag DNAa 3’-CCC AGG CAG CAC GGA ACC TGT AGT CTT TAT

(31)

21

DNAa’ 5’-TGG ACA TCA GAA ATA CCC \

DNAb’ 5’-AAC CTA GTG AGA AAT CCC

3.1.4 PDMS

Polydimethylsiloxane (PDMS) is a widely used silicon-based polymer which is shown in Figure 3.3. It is well known for its unusual rheological properties. In this thesis work, PDMS was supplied as a viscous liquid, and solidified by heating at 75 degrees Celsius.

Si

O

Si

O

Si

n

Figure 3.3. Chemical structure of Polydimethylsiloxane (PDMS) 3.1.5 Phosphate Buffered Saline (PBS)

Buffer solutions are solutions that resist change of pH upon addition of small amounts of acids or bases, or upon dilution. PBS is a general buffer used for dilutions and washing. All the kinds of different DNA in this thesis work are dissolved in PBS, and PBS is also used to rinse the fluid channel in order to wash away superfluous DNA. PBS contains sodium chloride (NaCl), dibasic sodium phosphate (Na2HPO4), potassium chloride (KCl), and monobasic potassium

(32)

22

3.1.6 Experimental set-up

Syringes and sample holder are shown in Figure 3.4(a) and (b) respectively. The box is used to connect the WetOFET to a Keithley 4200 parameter analyzer. At the side of box, there is a hole for the syringe. A transparent holder inside the box is used to fix the WetOFET, and is easy to set up.

(a)

(b)

Figure3.4. Experimntal setup. (a) Syringes. (b) Analyzing box which is designed for coaxial connections.

(33)

23

Double-sided tape from 3M-Scotch is used as spacer, and Loctite-401 instant adhesive is used as sealant. It is recommended to leave the instant adhesive in contact with air for thirteen minutes to one hour before use. Concerning instant adhesive needs one or two hour to dry up in order to seal the edge of the PDMS and glass.

3.2 Spin Coating

Spin coating is a common process in the semiconductor industry to spread photoresist on wafers. Since polymer materials can be dissolved in certain solvents such as water and chloroform. Spin coating offers a convenient and acceptable way to produce thin uniform films near room temperature without a vacuum surrounding. Spin coating process follows a few steps: (1) Dispense suitable amount of solution on the substrate. If needed, cover the area that is supposed to be coated. (2) Switch on the rotation and ramp-up the speed in order to spread out the solution to cover the whole substrate. (3) Keep the rotation speed constant while removing the solvent until dry. See Figure 3.5. Typically, the speed of spin coating is between 1000rpm (revolutions per minute) and 3000rpm. The thickness depends on the rotation speed, the concentration of the solution, and the wetting compatibility of the substrate.

DIspense

Ramp-up,

Spreading

Constant

ω

,

Drying

DIspense

Ramp-up,

Spreading

Constant

ω

,

Drying

(34)

24 3.3 Measurements

Fluorescence measurements of single stranded DNA with the Alexa350 dye was made on a Zeiss Axiovert inverted microscope A200Mot with a Hg lamp (HBO 100) as light source. Photos were taken before and after injection of the DNA strand with Alexa350. The double-sided tape was substituted by PDMS film in order to avoid strong background luminescence from the tape.

(35)

25

The electrical measurements, transfer characteristics, output characteristics, and time measurement, were done with a Keithley 4200 parameter analyzer. Different measuring parameters have been adjusted during experiments in this work; Hold time, measurement corresponds to the time that electrode bias is held before sweeping. The time that the analyzer waits before getting a data point from a certain bias is called delay time. For example, if hold time and delay time are 15seonds and 2 seconds respectively in a transfer characteristic measurement, the drain and source will be biased 15seconds before sweeping the gate voltage. During the sweeping, the instrument will wait for 2 seconds at each gate voltage, and then collect the data.

(36)

(37)

Chapter 4 Results and Discussion

4.1 General results

4.1.1 Transfer characteristics

A normal transfer characteristic curve of a WetOFET is shown in Figure 4.1. Hysteresis phenomena, when the current of on-sweeping is smaller than current of off-sweeping, almost always appears in WetOFETs. Because the spacer do not completely cover the source and drain electrodes, there is always a leakage current, depends on how well the spacer is constructed. It goes from the electrodes through the polymer thin film via the dielectric liquid to the gate or vice versa. The leakage can be seen more clearly in the diagram of gate current versus gate voltage which is shown in Figure 4.2. When the source was set as 0V (VS = 0V) and drain was

set as -0.5V (VD = -0.5V). At the beginning, most of the leakage current is between drain and

gate electrodes. Decreasing the gate voltage (from 0V to -0.5V) increases the current leakage between source and gate and at the time decrease that between drain and gate. In the same way, the leakage current present a more important role at Vg = 0 in drain current and Vg = -0.5 in source current respectively. This is why the I-V curve of source versus gate usually goes from zero, and the I-V curve of drain versus gate will decrease slightly at the beginning. This will also make the hysteresis of the source versus gate curve look smaller than the drain versus gate curve. The data is more reliable near VG = 0V region and near VG = -0.5V region in diagrams of

source current versus gate voltage and diagrams of drain current versus gate voltage respectively. On the other hand, the gate current should always be taken care of in order to

(38)

28

double-clarify if is a well fabricated fluid channel.

-0.5 -0.4 -0.3 -0.2 -0.1 0.0 0.0 2.0x10-8 4.0x10-8 6.0x10-8 8.0x10-8

Source current (A)

Gate Voltage (V)

(a) -0.5 -0.4 -0.3 -0.2 -0.1 0.0 0.0 1.0x10-8 2.0x10-8 3.0x10-8 4.0x10-8 5.0x10-8 6.0x10-8 7.0x10-8 8.0x10-8

Drain Current

(

A

)

Gate Voltage (V)

(b)

Figure4.1. Typical transfer characteristics with dual sweep. VS = 0V and VG = -0.5V. (a) Source current versus gate voltage. (b) Drain current versus gate voltage.

(39)

29

-0.5 -0.4 -0.3 -0.2 -0.1 0.0 -2.0x10-9 -1.0x10-9 0.0 1.0x10-9 2.0x10-9 3.0x10-9 4.0x10-9 5.0x10-9 6.0x10-9

Gate Curr

ent

(

A

)

Gate Voltage (V)

Figure 4.2. Sum of leakage current through the gate electrode. VS = 0V and VG = -0.5V.

The square root of source and drain current can be calculated and plotted with gate voltage, and mobility and threshold voltage can be extracted. But not only hysteresis plays a trick during mobility and threshold voltage evaluation, the number of times of measurement and delay times also play troublesome roles. Figure 4.3 is a good example for the measurement of number effect. A device has been analyzed sixteen times, including first data and fifteen extra data swept with the same 0.1 second delay time. Usually, current will increase with the measurement number as shown by fifth, tenth, and sixteenth data unless the current goes to some saturation value. But in some cases, the current will go down for some unknown reason such as second data. The word “usually” means the phenomena does not always appear. Threshold voltage may also change with measurement number, for example, threshold voltage was smaller in sixteenth cycle than in second cycle. In addition, if mobility was roughly evaluated by drawing a line through the top of curve and center of hysteresis area, it was larger in sixteenth cycle than in second cycle. The most important thing is that this effect may lead to the wrong conclusion.

(40)

30

-0.5 -0.4 -0.3 -0.2 -0.1 0.0 2.0x10-5 4.0x10-5 6.0x10-5 8.0x10-5 1.0x10-4 1.2x10-4 1.4x10-4 1.6x10-4 1.8x10-4 2.0x10-4 2.2x10-4 2.4x10-4

(D

rain current

)1/2

(A)

1/2

Gate voltage (V)

first cycle

second cycle

fifth cycle

tenth cycle

sixteenth cycle

Figure4.3. Influences of measurement number on current, mobility, and threshold voltage. Figure 4.4 shows the influence of delay time. The device was measured twice for each delay time, and the second data is shown in this diagram. First of all, hysteresis was decreasing while increasing delay time. For instance, mobility and current increased with increasing delay time if the rough evaluation mentioned above was used. However, the sequence of measuring is still concerned here, so the increase of mobility and current may be caused by either effect. Again, the most important thing is that the evaluated results can be misinterpreted which will be taken into account in the following DNA analysis.

(41)

31

-0.5 -0.4 -0.3 -0.2 -0.1 0.0 0.0 5.0x10-5 1.0x10-4 1.5x10-4 2.0x10-4 2.5x10-4 3.0x10-4 3.5x10-4

(Dr

ai

n

cur

rent)

1/2

(A)

1/2

Gate voltage (V)

0 sec 0.5 sec 1 sec 1.5 sec 2 sec 2.5 sec (a) -0.5 -0.4 -0.3 -0.2 -0.1 0.0 0.0 1.0x10-8 2.0x10-8 3.0x10-8 4.0x10-8 5.0x10-8 6.0x10-8 7.0x10-8 8.0x10-8 9.0x10-8 1.0x10-7 1.1x10-7

|d

rain

cu

rren

t| (

A

)

Gate voltage (V)

0 sec 0.5 sec 1 sec 1.5 sec 2 sec 2.5 sec (b)

Figure 4.4. Influences by delay time. (a) Mobility and threshold changes. (b) Hysteresis and current changes.

(42)

32

4.1.2 Output characteristics

From the discussion above, the current may change depending on number of times of measurement and delay time, On/Off ratio obtained from output characteristics may not be a suitable way to evaluate data. Figure 4.5 is an example, two output measurements were done one after the other, but the drain current was not stable. However output characteristic is a simple method to verify whether a transistor is working or not. In addition, in Figure 4.6, large drain current at zero gate voltage shows that there was a huge leakage between gate and drain, and the characteristic was much worse compared to Figure 4.5.

-0.5 -0.4 -0.3 -0.2 -0.1 0.0 -5.0x10-8 -4.0x10-8 -3.0x10-8 -2.0x10-8 -1.0x10-8 0.0

Drain current (

A

)

Drain voltage (V)

first cycle second cycle

Figure 4.5. Typical output characteristics of WetOFET. Instability between measurements suggests using On/Off ratio is an inappropriate way of evaluation.

(43)

33

-0.5 -0.4 -0.3 -0.2 -0.1 0.0 -5.0x10-8 -4.5x10-8 -4.0x10-8 -3.5x10-8 -3.0x10-8 -2.5x10-8 -2.0x10-8 -1.5x10-8 -1.0x10-8 -5.0x10-9 0.0 5.0x10-9 1.0x10-8

Drain current

(

A

)

Drain voltage (V)

V_G=0 V_G=0.1 V_G=0.2 V_G=0.3 V_G=0.4 V_G=0.5

Figure 4.6. Output characteristics from a WetOFET with a huge leakage current between gate and drain electrodes.

4.1.3 Water and PBS

The characteristics of WetOFET can be changed by using different fluids as dielectric layer. For example, in the current versus time measurement shown in Figure 4.7, pure water and PBS were injected into the fluid channel one by one and repeatedly. Both the current from the water and the PBS follows a trend. An unknown noise appeared while injecting water, but was not shown while injecting PBS. On the other hand, a higher leakage current is detected from the gate electrode when the PBS is introduced, which may be caused by the abundance of ions in the PBS compared to the water.

(44)

34

0 500 1000 1500 2000 2500 1.0x10-8 2.0x10-8 3.0x10-8 4.0x10-8 5.0x10-8 6.0x10-8

|D

rain

curren

t| (A)

Time (sec)

PBS Water Water PBS Water

Figure 4.7. Current versus time measurement. Current changes with different dielectric fluids.

4.2 Fluorescence Measurement

A fluorescence image taken from a device that has double-sided tape as spacer is shown in Figure 4.8. Obviously, the background fluorescence from the double-sided tape is too strong to see the fluorescence emitted from Alexa350, which is only coming from the fluid channel. Compared with the above, a device having a PDMS spacer can successfully reduce the background. Before the DNA is injected, the fluid channel reflected only a color of light purple which is not the fluorescent color of Alexa350. The blue light coming from the Alexa350 can be seen after injecting DNA and rinsing with PBS immediately. Since ITO coated glass has hydrophilic surface, the blue light should definitely come from the DNA that is attached to the P3HT surface and on the PDMS edges. Furthermore, the color in the P3HT region looks quite homogeneous, so the light should come from P3HT region itself and not from the PDMS edges.

(45)

35

Double-sided tape Fluid Channel Double-sided tape Fluid Channel

Figure4.8. Fluorescence picture from a device with double-sided tape as spacer. The background form the double-sided tape, blue area, is too strong.

PDMS film Fluid Channel PDMS film Fluid Channel (a) PDMS film Fluid Channel PDMS film Fluid Channel (b)

Figure 4.9. Fluorescence photos comparison between (a) before injecting DNA and (b) after injecting DNA with the Alexa350 dye.

(46)

36 4.3 DNA Detection

4.3.1 Overview

As described in chapter3, the DNA which is to be detected is the complementary DNA strand. Complementary DNA will be focused on and discussed below, although DNA with a cholesterol tag will also be mentioned. There are four parameters from transfer characteristics that will be used to evaluate data in this chapter. The maximum current (IMAX), the relative

mobility (μrel), the threshold voltage (VTh), and the hysteresis area (HA), will be introduced in

individual sections. The purpose is to discover a parameter which will not change after injecting non-complementary DNA but that will change after injecting complementary DNA. As a compromise between the uncertainties associated with the measurement number and the time for a device to “switch on”, all the plotted data are collected from the second scan. The impact of delay time will also be included.

4.3.1.1 Maximum current (IMAX)

In a dual sweep of gate voltage, there will be two current data points at Vg = -0.5V. The first point will be used here.

4.3.1.2 Relative mobility (μrel)

Because of the difficulty in measuring fluid capacitance, and since what is interesting in a sensor is relative change, relative mobility has been used for evaluation instead of real mobility. Relative mobility can be accessed by comparing slopes of the square root of the transfer characteristics. Since the mobility will change during sweeps of the gate voltage, the slope from the average value of on-sweep and off-sweep (between VG = -0.4V and VG = -0.5V) will be

used, and the difference between average value and on-sweeping value (between VG = -0.4V

(47)

37

-0.5 -0.4 -0.3 -0.2 -0.1 0.0 0.0 5.0x10-5 1.0x10-4 1.5x10-4 2.0x10-4 2.5x10-4 3.0x10-4

(Drain

cu

rren

t)

1/2

(

A)

1/2

Gate Voltage

(1) (2)

Figure 4.10. Mobility evaluation. (1) Line plotted from the average of on-sweep and off-sweep data between VG = -0.4V and VG = -0.5V (2) Line plotted from the on-sweeping data between VG

= -0.4V and VG = -0.5V. The slope of (1) is used to evaluate mobility. The slope difference

between (1) and (2) is used as error.

4.3.1.3 Threshold voltage (VTh)

Similar to 4.3.1.2, threshold voltage will change during sweeping. Threshold voltage can be extracted from the intersection of a gate voltage axis through zero current and the straight line of the square root of current. The line plotted from the average value of on-sweep and off-sweep (between VG = -0.4V and VG = -0.5V) will be used to extract threshold voltage, and the

difference between average value and on-sweep value (between VG = -0.4V and VG = -0.5V)

(48)

38

-0.5 -0.4 -0.3 -0.2 -0.1 0.0 0.0 5.0x10-5 1.0x10-4 1.5x10-4 2.0x10-4 2.5x10-4 3.0x10-4

(D

rain current

)1/2

(A)

1/2

Gate voltage (V)

(1) (2)

Figure 4.11. Threshold voltage evaluation. (1) Line plotted from the average of on-sweeping and off-sweep data between VG = -0.4V and VG = -0.5V (2) Line plotted from the on-sweeping data

between VG = -0.4V and VG = -0.5V. Intersect of (1) and gate voltage axis is used to evaluate

threshold voltage. The difference between (1) and (2) is used as error.

4.3.1.4 Hysteresis Area (HA)

The off-sweep current is higher than on-sweep current, as described in 4.1.1, and is described as hysteresis. While the ions close to the P3HT surface are screened by negative charges on the DNA during relaxation, the hysteresis may increase, see Figure 4.12. Similar result has been shown theoretically in literature, to analyze the electrostatic screening caused by the electrolyte, and can increase the average incubation times in order to limit sensor response [20]. The characteristic time constant of transfer function of FET device increased after DNA hybridization as presented in literature [6]. It is suggested that after the hybridization of DNA, the increasing negative charge from the complementary DNA strand may increase the screening phenomenon and the hysteresis area.

(49)

39 Gate

P3HT

DNA with Cholesterol Tag

Ions in PBS

Gate

P3HT

DNA with Cholesterol Tag

Ions in PBS

Figure 4.12. Screening by the negative charges on DNA. When the gate voltage decreases during off-sweep, the negative ions in solution will diffuse away from the bottom, P3HT surface, but negative charges on the DNA may impede the ion diffusion.

In Figure 4.13 (a), hysteresis is obtained from the difference between VG = -0.4 and the

off-sweep (VG from -0.5V to 0V) gate voltage which induces the same current as induced by Vg

= -0.4V of on-sweep (VG from 0V to -0.5V.) The reason for using -0.4V is that the WetOFET is

sufficiently turned on when VG = -0.4V, and at the same time, the leakage current between drain

and gate is small enough (as discussed in 4.1.1) to use the drain current hysteresis data. The hysteresis measure obtained in this way will be called V-0.4 in the following.

The other method to compare hysteresis area is to normalize the data and integrate, as shown in Figure 4.13 (b). This area will be called HAnorm in the following.

(50)

40 -0.5

-0.4

-0.3

-0.2

-0.1

0.0

0.0 1.0x10

-8

2.0x10

-8

3.0x10

-8

4.0x10

-8

5.0x10

-8

6.0x10

-8

7.0x10

-8

Drain current

(A)

Gate voltage (V)

(a) -0.5 -0.4 -0.3 -0.2 -0.1 0.0 0.0 0.2 0.4 0.6 0.8 1.0

N

o

rm

aliz

ed drain

cu

rrent

Gate voltage (V)

(b)

Figure 4.13. Hysteresis area evaluation. (a) hysteresis is evaluated by the on-to-off gate voltage which produces the same current as when the off-to-on gate voltage equals -0.4V (b) hysteresis is evaluated by the normalized covered area.

(51)

41

4.3.2 Non-complementary DNA versus Complementary DNA

4.3.2.1 Maximum current (IMAX)

Usually, after injecting both DNAb’ and DNAa’, IMAX will decrease, but it seems like the

fraction it decrease is similar (Figure 4.14.) The maximum current does therefore not discriminate between complementary or non-complementary DNA. IMAX sometimes recovers to

the original value after a few times measurements, indicating the instability of IMAX.

Xa b' a' 5.5x10-8 6.0x10-8 6.5x10-8 7.0x10-8 7.5x10-8 8.0x10-8 8.5x10-8

|Drain

cu

rrent

| (

A)

DNA

(52)

42

4.3.2.2 Comparison of Relative mobility (μrel)

One WetOFET was analyzed during the fluid sample injection sequence – DNAXa, DNAb’, DNAa’. It means that the device was injected with DNAXa solution for a certain time, rinsed with PBS buffer, and then analyzed with transfer characteristics. After analyzing DNAXa, the same procedure was repeated with DNAb’ and DNAa’ one by one. The data was evaluated by the method mentioned in 4.3.1 and is shown in Figure 4.15. The relative slopes in Figure4.15 has been normalized for clarity.

From Figure 4.15 it can be seen that mobility decreases after injecting non-complementary DNA (DNAb’) or complementary DNA (DNAa’) for all different delay times, except DNAa’ made mobility increase slightly when delay the time was 1.5 second. On the other hand, the error from one DNA data covers part of the same region as the others, so this can be caused by a misjudged mobility. -0.5 -0.4 -0.3 -0.2 -0.1 0.0 0.0 5.0x10-5 1.0x10-4 1.5x10-4 2.0x10-4 2.5x10-4 (D ra in c u rr en t) 1/2 (A) 1/2 Gate voltage (V) DNA Xa DNA b' DNA a' Xa b' a' 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 1.2 Re lativ e slop e DNA (a) -0.5 -0.4 -0.3 -0.2 -0.1 0.0 0.0 5.0x10-5 1.0x10-4 1.5x10-4 2.0x10-4 2.5x10-4 3.0x10-4 (Drai n cu rrent )1/2 (A) 1/2 Gate voltage (V) DNA Xa DNA b' DNA a' Xa b' a' 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 1.2 R elat ive sl ope DNA (b)

(53)

43

-0.5 -0.4 -0.3 -0.2 -0.1 0.0 0.0 5.0x10-5 1.0x10-4 1.5x10-4 2.0x10-4 2.5x10-4 3.0x10-4 3.5x10-4 (D rain cu rrent )1/2 ( A )1/2 Gate voltage (V) DNA Xa DNA b' DNA a' Xa b' a' 0.5 0.6 0.7 0.8 0.9 1.0 1.1 1.2 Re lative sl ope DNA (c) -0.5 -0.4 -0.3 -0.2 -0.1 0.0 5.0x10-5 1.0x10-4 1.5x10-4 2.0x10-4 2.5x10-4 3.0x10-4 3.5x10-4 (Drain curren t) 1/ 2 (A) 1/ 2 Gate voltage (V) DNA Xa DNA b' DNA a' Xa b' a' 0.55 0.60 0.65 0.70 0.75 0.80 0.85 0.90 0.95 1.00 1.05 1.10 R el ati ve sl op e DNA (d) -0.5 -0.4 -0.3 -0.2 -0.1 0.0 5.0x10-5 1.0x10-4 1.5x10-4 2.0x10-4 2.5x10-4 3.0x10-4 3.5x10-4 (D ra in cu rr ent )1/2 ( A )1/2 Gate voltage (V) DNA Xa DNA b' DNA a' Xa b' a' 0.65 0.70 0.75 0.80 0.85 0.90 0.95 1.00 1.05 1.10 Rel ative s lope DNA (e)

(54)

44

different delay time (a) 0 second (b) 0.5 seconds (c) 1 second (d) 1.5seconds (e) 2 seconds. Mobility may not decrease even after injecting DNAb’ or DNAa’, Figure 4.16 is an example. Figure 4.16(a) is plotted for another sample, and the solid line in Figure 4.16(b) shows the non-decreasing mobility after injecting DNAb’ and DNAa’. The dotted line in Figure 4.16(b) is from Figure 4.15(a) as a reference. As a conclusion, these tells us that mobility cannot be used to evaluate whether the DNAa’ has been injected.

-0.5 -0.4 -0.3 -0.2 -0.1 0.0 0.0 5.0x10-5 1.0x10-4 1.5x10-4 2.0x10-4 2.5x10-4 3.0x10-4 3.5x10-4 (Dr ain curr ent )1/2 (A) 1/ 2 Gate voltage (V) DNA Xa DNA b' DNA a' Xa b' a' 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 1.2 Re lati ve s lope Gate voltage (V) No slope change Reference (a) (b)

Figure 4.16. Mobility comparison between non-complementary and complementary DNA. (a) Square root of drain current versus gate voltage. (b) Solid line is the same as (a), and the dotted line is from Figure 4.15. This shows that the mobility may not change after injecting non-complementary and complementary DNA.

4.3.2.3 Comparison of Threshold Voltage (VTh)

Threshold voltage changes after injecting DNA with different delay time is shown in Figure 4.17. Again, results from both non-complementary (DNAb’) and complementary DNA (DNAa’) strand went toward the same direction with increasing threshold voltage (toward positive direction). The step it increases is not enough to distinguish DNAb’ and DNAa’, even though, when delay time is 2 seconds, the threshold voltage decreased slightly after injecting DNAa’, the uncertainty still covers the threshold voltage region of DNAb’. Threshold voltage is therefore not an appropriate way to distinguish non-complementary DNA and complementary DNA. Concluding the discussions above, the negative charge on DNA, in this work, may be not sufficient to cause a change of the surface potential.

(55)

45

-0.5 -0.4 -0.3 -0.2 -0.1 0.0 0.0 5.0x10-5 1.0x10-4 1.5x10-4 2.0x10-4 2.5x10-4 (D ra in c u rr en t) 1/2 (A) 1/2 Gate voltage (V) DNA Xa DNA b' DNA a' Xa b' a' -0.20 -0.16 -0.12 -0.08 -0.04 0.00 0.04 0.08 Th resh o ld v o lt ag e (V) DNA (a) -0.5 -0.4 -0.3 -0.2 -0.1 0.0 0.0 5.0x10-5 1.0x10-4 1.5x10-4 2.0x10-4 2.5x10-4 3.0x10-4 (Dr a in curr ent) 1/ 2 (A) 1/2 Gate voltage (V) DNA Xa DNA b' DNA a' Xa b' a' -0.16 -0.12 -0.08 -0.04 0.00 0.04 0.08 T hr e shold v o lta ge (V) DNA (b) -0.5 -0.4 -0.3 -0.2 -0.1 0.0 0.0 5.0x10-5 1.0x10-4 1.5x10-4 2.0x10-4 2.5x10-4 3.0x10-4 3.5x10-4 (D rain cu rrent )1/2 ( A )1/2 Gate voltage (V) DNA Xa DNA b' DNA a' Xa b' a' -0.12 -0.08 -0.04 0.00 0.04 0.08 Thr es hol d vo ltag e (V ) DNA (c)

(56)

46

-0.5 -0.4 -0.3 -0.2 -0.1 0.0 5.0x10-5 1.0x10-4 1.5x10-4 2.0x10-4 2.5x10-4 3.0x10-4 3.5x10-4 (D rain c u rr en t) 1/2 (A) 1/2 Gate voltage (V) DNA Xa DNA b' DNA a' Xa b' a' -0.12 -0.08 -0.04 0.00 0.04 Th res h old volt ag e (V) DNA (d) -0.5 -0.4 -0.3 -0.2 -0.1 0.0 5.0x10-5 1.0x10-4 1.5x10-4 2.0x10-4 2.5x10-4 3.0x10-4 3.5x10-4 (D ra in cu rr ent )1/2 ( A )1/2 Gate voltage (V) DNA Xa DNA b' DNA a' Xa b' a' -0.12 -0.08 -0.04 0.00 Thre shol d v o lta g e (V ) DNA (e)

Figure 4.17. Threshold voltage comparisons of non-complementary and complementary DNA with different delay time (a) 0 second (b) 0.5 seconds (c) 1 second (d) 1.5seconds (e) 2 seconds.

4.3.2.4 Comparison of hysteresis area (HA)

The influence of delay time has been mentioned in the above discussions. Shorter delay time will cause larger hysteresis, which implies a long response time of WetOFET. After applying gate voltage, ions in the fluid may need some time to reach a new stable equilibrium position. This phenomenon can be seen more clearly in time measurements that have a certain value of gate, drain, and source bias and are measured with respect to time. Figure 4.18 is an example, when drain and source are biased with -0.5V and 0V respectively, if a gate voltage is suddenly applied or removed, there will appear a large current in just 0.1 second, and it may be a

(57)

47

capacitive current caused by a sudden redistribution of ions. This figure also implies that the transistor may need some time to reach saturation and may relax to the original state in just a few seconds. Figure 4.18 is under a huge bias-changing situation which may over amplify the results. Shorter delay time may make the WetOFET unsaturated or introduce some noise.

0

50

100

150

200

250

300

0.0 2.0x10

-8

4.0x10

-8

6.0x10

-8

8.0x10

-8

|Dr

ain curr

ent|

(A)

Time (sec)

Gate=-0.5V Gate=0V

Figure 4.18. Current versus time measurement with different gate bias. Huge leakage current appear with sudden changes of gate voltage.

On the other hand, the smaller the hysteresis is, the less the leakage influence on the drain current is when the gate voltage is near -0.4V. Combining the above reasons, longer delay times, 1.5 seconds to 2.5 seconds, have been used in this section in order to detect the slight change of negative charges on the DNA.

A device was analyzed with 2.0 second delay time, and the diagram of relative drain current versus gate voltage is shown in Figure 4.19. V-0.4 did not change after introducing DNAb’, but

increases (toward positive) after introducing DNAa’. HAnorm decreased after introducing DNAb’,

and increased after introducing DNAa’. This result supports the suggestion in 4.3.1.4, the screening from the negative charge on the DNA may increase the hysteresis, and the hysteresis may be an appropriate way to detect the complementary DNA strand.

(58)

48 -0.5

-0.4

-0.3

-0.2

-0.1

0.0

0.2

0.4

0.6

0.8

1.0 Relative drain cur

rent

Gate voltage (V)

DNA Xa

DNA b'

DNA a'

(a) Xa b' a' -0.386 -0.384 -0.382 -0.380 -0.378 -0.376 -0.374 -0.372 -0.370 -0.368 G ate vo lt age (V) DNA Xa b' a' 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 R e

lative normalized hysteresis

ar

ea

DNA (b) (c)

Figure 4.19 Hysteresis comparison between non-complementary (DNAb’) and complementary (DNAa’) DNA. (a) Transfer characteristics comparison. (b) V-0.4 change corresponding to type

of DNA. V-0.4 increased after injecting complementary DNA. (c) HAnorm change corresponding

Wet Organic Field Effect Transistor as DNA sensor

Department of Physics, Chemistry and Biology