偏壓對有機薄膜電晶體氣體感測器之影響以及相關機制

國立交通大學

光電工程學系 光電工程研究所

碩士論文

偏壓對有機薄膜電晶體氣體感測器之

影響以及相關機制

Influence of bias on OTFT gas sensor

and its sensing mechanism

研究生: 梁芸嘉

指導教授: 冉曉雯 博士

偏壓對有機薄膜電晶體氣體感測器之影響以及相關機制

Influence of bias on OTFT gas sensor and its sensing

mechanism

研 究 生：梁芸嘉 Student：Yun-Chia Liang

指導教授：冉曉雯 博士 Advisor：Dr. Hsiao-Wen Zan

國 立 交 通 大 學

光電工程學系 光電工程研究所

碩 士 論 文

A Thesis

Submitted to Department of Photonics & Institute of Electro-Optical Engineering College of Electrical Engineering and Computer Science

National Chiao Tung University in partial Fulfillment of the Requirements

for the Degree of Master

in

Electro-Optical Engineering June 2007

偏壓對有機薄膜電晶體氣體感測器之影響以及相關機制

研究生:梁芸嘉 指導教授:冉曉雯 博士

國立交通大學

光電工程研究所碩士班

摘要

近年來有機薄膜電晶體氣體感應器已經引起大量的研究，此成果之應用範圍 包含食物品質檢測以及醫療用自我檢測儀器。為了達成能商業量產的價值，我們 需要製造出成本低廉且高度靈敏的感測元件，為了達成這個目的，我們選用有機 電子材料來製造氣體感測器。 現有的相關研究多著重在氣體感測器電流值的變化，對於其他參數如臨界 電壓以及載子移動率卻缺乏討論。我們希望站在元件物理的基礎上更深入的探討 氣體反應的相關機制。 在氣體感應的量測中我們觀察到造成電流值變化的主因是因為臨界電壓的 漂移。隨著氨濃度的增加，臨界電壓值的移動也會更劇烈，但當氨氣移除之後臨 界電壓有恢復到初始狀態的趨勢。故我們認為氨氣與五苯環分子的作用為一種吸 附現象。在這實驗中我們發現一個很有趣的現象:當三顆特性相同的有機元件放

在同一個有氨的環境中，不論放置在這環境中多長時間所量測到的初始臨界電壓 都相近，這結果暗示我們氣體感測的行為與閘極偏壓有關。 在長時施加閘極偏壓的氣體感測中，固定施加的偏壓改變氣體濃度，當氨 氣濃度越高臨界電壓變動也增加。若固定氣體濃度施加更高偏壓也會看到相同的 情形。不論是在有無偏壓實驗條件下，氣體擴散模型可以成功解釋氨在五苯環薄 膜中的擴散行為。

Influence of bias on OTFT gas sensor and its sensing

mechanism

Student: Yun-Chia Liang Advisor: Dr. Hsiao-Wen Zan

Institute of Electro-Optical Engineering

National Chiao Tung University

Abstract

Electronic noses have attracted lots of interests in the recent years. It can be

used from food-quality monitoring to medical self-diagnosis kits. For commercial

purpose, we need to produce low-cost and high sensitivity electronic noses. To reach

these targets, we used organic material to fabricate electronic nose.

Most of the previous works focused on the electric current effect alone. In this

work we also studied the effect on threshold voltage 、 mobility, from which we

investigated the mechanism for gas sensing.

In this experiment, we observed that the electric current variation was strongly

influenced by the threshold voltage. When the ammonia concentration was increased,

the threshold voltage became larger. But we noted that the threshold voltage tend to

reaction between ammonia and pentacene is basically an absorption reaction. At the

same time, we observed an interesting phenomenon: we measured three identical

devices that were placed at the same distance from the gas source. When the

measurement was applied on the three devices one after another along with the NH3

sensing time, we observed an interesting event- these three devices had similar initial

threshold voltages no matter how long these devices had been exposed to the NH3 gas.

This result implied that the sensing behavior may have a strong correlation with the

gate bias.

In the experiment of bias stress gas sensing, we applied a fixed gate bias and

varied the ammonia concentration. The threshold voltage became larger with the

ammonia concentration was increased. We observed the same result when we applied

higher gate bias and added a fixed amount of ammonia. Our results showed that

Acknowledgment

首先誠摯的感謝指導教授冉曉雯博士，老師不論在實驗上或者生活上 都給予我很好的建議以及鼓勵，這兩年在老師身上學到很多使我個人也成 長不少，在此獻上最誠摯的謝意。 本論文的完成亦得感謝博士班學長國錫、士欽以及政偉，在實驗上給 予很多的指導以及照顧。尤其士欽學長在我實驗量測上給予相當多的幫 助，讓我得以順利完成實驗。也很高興能與文馨、而康、光明、德倫、廷 遠、皇維、睿志一起度過這難忘的兩年，感謝你們在實驗上以及各方面的 協助。 兩年的生活有苦有樂，謝謝多媒體工程所的詠恬、網路工程所的雅琳、 以及資工所的慧伶陪我一起度過。謝謝你們這一路的關懷以及陪伴，幫助 我度過許多艱辛的時刻。希望不論是出國還是留在國內念博士的你們都會 相當順利。 最後，感謝父母給我最大的空間讓我做自己想做的事。無悔的栽培讓我 順利完成碩士學位。謝謝你們這 24 年來的照顧，謝謝。

Contents

Chinese Abstract

Ⅰ

English Abstract

Ⅲ

Acknowledgment

Ⅴ

Contents

Ⅵ

Figure Captions

Ⅷ

Chapter 1. Introduction

1-1 Introduction of Organic Thin Film Transistors (OTFTs) 1 1-1.1 Organic materials 2 1-1.2 OTFT device structure 3 1-1.3 Operating mechanisms of organic thin film transistor 4 1-2 Ammonia sensor and their applications 5 1-2.1 Application areas of ammonia sensors 5 1-2.2 Different types of ammonia sensor 7 1-3 OTFT gas (ammonia) sensor 9 1-4 Motivation 1 0 1-5 Thesis Organization 1 1

Chapter 2. Experimental

2-1 The Fabrication for OTFTs 12

2-4.1. Mobility 15

2-4.2. Threshold voltage 15

2-4.3. Subthreshod swing 16

Chapter 3. Results and Discussion

3-1 The sensing phenomenon under no bias stress 17

3-1.1 The electrical properties of OTFT gas sensor 17

3-1.2 The recovery phenomenon 18

3-1.3 Gas diffusion model 18

3-1.4 Gas concentration effect 20

3-1.5 The phenomenon of different NH3 exposed time 21

3-2 The phenomenon of bias stress gas sensing 22

3-2.1 The theory of bias stress effect 22

3-2.2 The phenomenon of bias stress gas sensing 24

3-2.3 The gas sensing mechanism under bias stress 26

3-2.3.1 Charge- trapped states 26

3-2.3.2 Dipole molecules 26

Chapter 4. Conclusion

28

References

30

Figure Captions

Chapter 1

Fig. 1-1 Molecular structure of pentacene.

Fig. 1-2 (a) Top contact structure (b) Bottom contact structure.

Fig. 1-3 The work function of various metals and ionization potential of pentacene.

Chapter 2

Fig. 2-1 Conventional bottom-gate top-contact OTFTs were used in this experiment.

Chapter 3

Fig. 3-1 The electrical properties of OTFT adding 2ℓ NH3.

Fig. 3-2 The variations of subthreshold swing、threshold voltage and mobility along

with time in environment of 2ℓ NH3.

Fig. 3-3 The variations of threshold voltage along with time in environment of 2ℓ Fig. 3-4 When NH3 was removed, the threshold voltage tended to recover.

Fig. 3-5 The fitting curve of Fick‘s second law are in the condition of d=10-5 cm and

D=10-12 cm2/s、2*10-12 cm2/s、4*10-12 cm2/s、8*10-12 cm2/s

Fig. 3-6 The fitting curve of Fick‘s second law are in the condition of D=10-12 cm2/s

Fig. 3-8 The dependence of threshold voltage shift on sensing time in the

environments with different NH3 concentrations. The circles are experiment

data. The curves are the fitting result from Fick‘s second law.

Fig. 3-9 The threshold voltage variation of three identical devices measured one after

another in the environment with NH3 gas. The NH3amount injected into the

cotton mass was 2ℓ.

Fig. 3-10 The threshold voltage shift in different bias stress conditions

Fig. 3-11 The threshold voltage shift in different bias stress conditions

Fig. 3-12 The threshold voltage shift in different amount of concentrations with

Vg-Vthini= -7V bias stress.

Fig. 3-13 The threshold voltage shift in the environment of 2ℓ ammonia with different amount of bias stress.

Fig. 3-14 The threshold voltage shift in the environment of 10ℓ ammonia with different amount of bias stress.

Fig. 3-15 The dependence of threshold voltage shift on sensing time in the

environments with different gate bias strees. The circles are experiment data.

Chapter 1

Introduction

1-1. Introduction of organic thin film transistor (OTFT)

The conductance of organic molecules was discovered in 1940s [1]. It began

from a very small number of preliminary works on conjugated small molecules [2,3]

and polymers[4]. The first transistor based on an organic semiconductor was only

reported in 1986 with a device made on an electrochemically grown polythiophene

film [5]. Polythiophene belongs to the family of the conducting polymers that were

discovered in the late 1970s [6]. The inventors of polyacetylene were awarded the

Nobel Prize in chemistry in 2000. The possibility of fabricating organic thin-film

transistors (OTFTs) with small conjugated molecules was demonstrated in 1989 with

sexithiophene, an oligomer of polythiophene made of sixthiophene rings linked at

alpha position [7]. For a decade, there have been a lot of researches about

better-performing organic thin-film transistors (OTFTs). Now the performances of

OTFTs are comparable to the hydrogenated amorphous silicon TFTs (a-Si:H TFT) [8].

OTFTs have many advantages compared to other electronic devices. The low

frequency identification cards (RFIDs), sensors and electronic paper. For these

reasons, more and more industrial groups have recently initiated research programs in

the field of organic transistors. Until recently, much of the research effort has been

directed at improving the charge-carrier mobility. Several papers have reviewed this

search for better materials and device architecture [9-13].

1-1.1 Organic materials

A number of organic materials such as polythiophene, α- sexithiophene (α-6T)

have been investigated for use in field effect transistors (FETs) [14]. Polycrystalline

molecular solids such as α-sexithiophene (α-6T) or amorphous/semi-crystalline

polymers such as polythiophene or acenes such as pentacene, teracene show the

highest mobilities [15]. Dimitrakopoulos et al [16] shows the evolution of organic

materials and the improvement in their mobilities over the years. Pentacene based

FETs exhibit high mobilities and have been extensively studied. Pentacene is made up

of five benzene rings as shown in Fig. 1-1. It has a sublimation temperature of 300℃.

Well ordered pentacene films can be deposited at low temperatures, this makes it

1-1.2 OTFT device structure

In organic thin film transistor, the organic film can be deposited by evaporation,

spin-on or ink-jet printing. And the devices can be fabricated on a number of

substrateslike silicon, glass and plastic.Figure 1-2(a) and (b)illustrate a widely used

structure of OTFTs: top-contact structure (Fig.1-2(a)) and bottom-contact structure

(Fig.1-2(b)).

In top-contact structure, a heavily doped silicon substrate acts as a gate electrode

for the device. Thermally grown silicon dioxide is deposited on the substrate as gate

insulator. Next, the organic material is deposited on the insulator and the metal is

deposited on the top of the organic semiconductor as the source and the drain

electrodes. Because the organic layer is sensitive to the photolithographic patterning

and chemical etching. The source and drain electrodes are defined through the shadow

mask.

The other popular structure is the bottom-contact structure, in which the organic

layer is deposited directly on the top of the structure. The structure is commonly used

in the field of organic electronics due to the easy fabrication. The entire structure

except for the organic layer can be patterned using photolithography technology.

top-contact structure. The difficulty to obtain well-ordered organic film on metal

electrodes may be the primary reason.

1-1.3 Operating mechanisms of organic thin film transistor

The gate structure of thin-film transistors operates like a capacitor. When a

voltage is applied between source and gate, a charge is induced at the

insulator-semiconductor interface. This charge forms a conducting channel, the

conductance of which is proportional to the gate voltage. At low drain voltages, the

current increases linearly with drain voltage, following the Ohm‘s law. When the

drain voltage is compared to gate voltage, the voltage drop at drain contact falls to

zero and the conducting channel is pinched off. This corresponds to the so-called

saturation regime where the current becomes independent of the drain voltage. In the

transfer characteristic, the current is plotted as a function of the gate voltage at a

constant drain bias. Below a given threshold, the current increases exponentially. This

corresponds to the below-threshold regime. In the above threshold regime, the current

becomes proportional to the gate bias, as expected from the above description of the

operating mode of the transistor.

The Fermi level of gold and the band diagram of pentacene are shown in Fig. 1-3.

source electrode. As can be seen in Fig. 1-3, the Fermi level of gold is far away from

the LUMO level, so that electron injection is very unlikely. Accordingly, no current

passes through the pentacene layer, and the small measured current essentially comes

from leaks through the insulating layer. When the gate voltage is reversed, holes can

be injected from the source to the semiconductor, because the Fermi level of gold is

close to the HOMO level of pentacene. Accordingly, a conducting channel forms at

the insulator-semiconductor interface. The charges can be driven from the source to

the drain by applying a drain bias. For this reason, pentacene is said to be a p-type

semiconductor. However, it should be pointed out that this concept differs from that

of doping in conventional semiconductors, which can be made either n-type or p-type

by introducing a large amount of an electron donating or electron withdrawing

element. Symmetrically, an organic semiconductor will be said n-type when the

source and drain electrodes can inject electrons in its LUMO level, provided electron

transport does occur, i.e., electron mobility is not too low.

1-2. Ammonia sensors and their applications [7]

1-2.1 Application areas of ammonia sensors

Ammonia is an obvious gas in the world. Apart from its natural origin, there are

many sources of ammonia, such as the chemical industry or intensive life-stock.

Subsequently, different application areas for gaseous ammonia sensors are

investigated, such as:

(1) Automotive industry

The application for ammonia sensors in the automotive area is NOx reduction in

diesel engines. Because modern diesel engines cause large concentrations of NO or

NO2 [8.9], people use selective catalytic reduction (SCR) of NOx with NH3 to reduce

the toxic NOx concentrations. The formula is shown in equation (1-1) [10].

4NO + 4NH3 + O2 -> 4N2 + 6H2O (1-1)

But adding too much ammonia causes the other pollution, known as ammonia-slip.

Therefore, measuring the excess ammonia concentration in the exhaust system is also

important.

(2) Chemical industry

The chemical industry, fertilizer factories and refrigeration systems use almost

pure ammonia. If there is a leak in the facility, high concentrations of ammonia form a

threat to the human health. Therefore these facilities should have an alarm system

detecting and warning for dangerous ammonia concentrations.

By measuring ammonia level in exhaled air, the ammonia sensor can be a fast

diagnostic method for patients with disturbed urea balance, like kidney disorder [11]

or ulcers. These diseases are caused by H.pylori bacterial stomach infection. Now, the

common way to detect the H.pylori bacterial is an endoscopic procedure. But the

endoscopic procedure is an invasive and inconvenient test. In order to develop the

more convenient way to detect the H.pylori bacterial, the non-invasive test methods

based on measuring exhaled CO2 or NH3 level was mentioned [12,13].

1-2.2 Different types of ammonia sensor

(1) Metal- oxide gas sensors

A lot of research about metal-oxide gas sensor have been done [14], especially in

Japan [15]. These sensors are rugged and inexpensive and thus very promising for the developing gas sensors. In this kind of gas sensor, WO3 based sensing material is used

to respond to ammonia [16,17]. But it must be operated at the elevated temperature of

more than 400℃ [17].

(2) Catalytic ammonia sensors

By a change in ammonia concentration, the charge carrier concentration in the

temperature. But for this kind of gas sensor, the lower detection limit is normally in

the low-ppm range and the accuracy is limited.

(3) Optical gas analyzers

There are two main optical ammonia analyzers. One is based on a change in

color as ammonia reacts with a reagent. The other is optical absorption detection

applied as a method to sense ammonia. But the equipments for the optical gas

analyzer are large and expensive.

(4) Conducting polymer gas detectors

The polymer is supposed to be deprotonated by ammonia and the conduction of

the polymer is also changed [18].Comparing with other types of ammonia gas sensor,

there are some advantages of conducting polymer gas detectors, like lower detection

limit (0.5 ppm)、faster response time (60s-100s) and it can be operated at room

temperature. In the beginning, gas sensors are chemoresistor sensors. But using a

three- terminal TFT structure instead of two-terminal chemoresistor construction is

based on gate bias can enhance sensitivity, discrimination and repeatability [19-23].

Furthermore the transistor can offer multi-parameters for different analytes, such as

bulk conductivity、 threshold voltage、 field effect mobility and field- induced

1-3. OTFT gas (ammonia) sensor

Many experiments have demonstrated the sensing ability of OTFT by using

different organic materials and different gaseous analytes [19.24]. The reasons for

organic thin film transistor can be good gas sensors are as follow. First, OTFTs have

some advantages, such as simple process、low fabrication costs and for large area.

Second, the organic materials are made up by the carbon backbones. The carbon

backbones make the organic devices have higher chemical activity than other

inorganic devices. Moreover, by synthetic chemistry we can deposit specific active

layer to control and adjust the sensitivities and selectivity of gas sensors [25].

Due to the chemical activity of the carbon backbone, the conducting polymers

can sense a wide range of analytes, including humidity, nitrogen and other organic

compounds.

For OTFT gas sensors, the morphology of the active layer and the molecular

structure of both semiconductor and gas molecule are all important factors for the gas

response. An associated research by Torsi et al [26], it has been reported that the

result of various oligothiophene films to exposure alcohol. The response is highly

dependent on the amount of grain boundaries. But for other gas species like

analyte species. For this reason, Torsi et al. have proposed multi-parameter OTFT gas

sensors[19].OTFT exhibit many sensing index such as the on-state conductivity, the

threshold voltage, the mobility, the off-state conductivity and the subthrehsold swing,

etc. These parameters can be used as ―fingerprints‖ to response the OTFT for a given

chemical compound.

Recently, nanoscale organic thin film transistors were fabricated as gas sensors.

The sensing behavior of these small dimension devices is markedly different from that

of larger devices for the same analyte [27].

1-4. Motivation

In the previous study, OTFTs can respond with ammonia was mentioned. In

order to control the NH3-gas sensor, studying the sensing mechanism in depth is

necessary. On the other hand, some groups observed that gate bias can enhance the

sensitivity. Therefore, in our experiment, the gas sensing behaviors with and without

1-5. Thesis organization

In Chapter 1, we briefly introduce the OTFTs and the current ammonia sensors.

In Chapter 2, the device fabrication procedure, measurement environment, experiment

design and parameter extraction of OTFTs are presented. In Chapter 3, the

NH3-sensing phenomena with and without bias stress are investigated. The gas

diffusion model- Fick‘s second law is used to explain the gas sensing phenomena.

Chapter 2

Experiments

2-1. The Fabrication for OTFTs

Conventional bottom-gate top-contact OTFTs were used in this experiment.

The structure is shown in Fig.2-1.

The detail fabrication processes are as follows:

Step1. Substrate and gate electrode

Heavily-doped single crystal silicon wafer was used as substrate and gate

electrode.

Step2. Gate oxide formation

After RCA cleaning, the 1000Å thermally grown SiO2 layer was deposited in

furnace.

Step3. Pentacene film deposition through shadow mask

The pentacene material obtained from Aldrich without any purification was

directly placed in the thermal coater for the deposition. It is well known that the

deposition pressure, deposition rate, and deposition temperature are the three critical

parameters to the quality of the organic film [28]. The deposition is started at the

and better ordering of the organic molecules. The deposition temperature is also a

factor influencing the pentacene film formation. The 100-nm-thick pentacene was

deposited by thermal evaporation at a deposition rate of 0.5Ǻ/s. During the

deposition process, the substrates temperature fixed at 70C. The active region and source/deain electrodes were all defined by shadow mask.

Step4. Source/Drain deposition through the shadow mask

In order to form the Ohmic contact, pentacene and source/drain electrodes must

have similar work function [29]. The work function of gold is about 5.1eV and it can

provide a better injection. Gold was deposited 100nm on pentacene film as the

source/drain electrodes by using thermal coater. The source/drain region was

defined through shadow mask.

2-2. Measurement environment

The gas sensing characteristics were studied according to the method proposed

by A.Alec Talin et al. [30]. We used a small cotton mass to absorb different amount

of pure NH3, then placed the cotton mass one centimeter from the devices. Because

2-3. Experiment Design

The experiment was discussed in two parts: (A) The phenomenon of gas sensing.

(B) The bias stress gas sensing.

In part A, in the beginning, the sensing and recovery behavior of NH3- OTFT

sensors were investigated. Then, the threshold voltage variations with different

ammonia concentrations were observed. Diffusion model- Fick‘s second law is used

to explain the sensing mechanism. Finally, three identical devices were put in the

ammonia environment at the same time and we observed the threshold voltage

variations.

In part B, the gas sensing with gate bias was discussed. At first, we applied a

constant gate bias and observed the threshold voltage shifts with different

concentration. Next, we changed the gate bias stress and investigated the threshold

voltage variations in a fixed ammonia concentration.

2-4. Parameter Extraction

In this section, we mention about how to extract mobility (μ ), the threshold

voltage (Vth) and the subthreshod swing (s.s). These parameters can be used to

2-4.1 Mobility

Mobility (μ ) is an important parameter in most electronic devices because it is

directly related to the performance of a transistor. It is normalized for channel width,

channel length, and dielectric capacitance.

Generally, mobility can be extracted from the transconductance maximum g _m

in the linear region:

OX _D t cons V G D m V L WC V I g D             tan (2.1) ( ) m ox D g w C V L   (2.2)

2-4.2 Threshold voltage

In inorganic electric devices, threshold voltage (Vth) is defined as the point at

which the channel has been completely inverted. That means when the concentration

of inversion charge in the channel equals the equilibrium majority charge carrier

concentration in the bulk. But the most commonly used definition for threshold

voltage (Vth) of an OTFT is extracted from equation (2.3), by extrapolating the linear

ID-VG plot at ID=0 2 [( ) ] 2 ox D D G T D W C V I V V V L     (2.3)

In organic semiconductor, the Vth is determined primarily by the relative

work-function of the gate, source/drain contacts and the organic layer, as well as the

fixed charge and trap distributions in the organic layer. Because of run-to-run

variations, the parameters can be difficult to suppress. Thus, Vth control is an

important hurdle to commercialization of OTFTs.

2-4.3 Subthreshod swing

Subthreshold swing is also important characteristics for device application. It is

a measure of how rapidly the device switches from the off state to the on state in the

region of exponential current increase. Moreover, the subthreshold swing also

represents the interface quality and the defect density. Good performance TFTs

mean lower subthreshold swing of transistors.





co n stan t log   



D V D G

I

V

S _{, when V} G＜VT for p-type. (2.4)

Chapter 3

Result & Discussion

In this chapter, we investigated the gas sensing mechanism with and without

bias stress. In the following experiments, drain voltage is fixed at -6V and gate

voltages are changed from 20V to -40V.

3-1 The sensing phenomenon under no bias stress

3-1.1 The electrical properties of OTFT gas sensor

The reactions between OTFTs and the ammonia gas are illustrated in Fig.3-1,

3-2. After we added 2ℓ ammonia into a cotton, we observed the drain current and the transconductance (GM) were decreased. Then, the mobility () 、the threshold

voltage (Vth) and the subthreshold swing (S.S) were extracted. The variations of the

mobility ()、the threshold voltage (Vth) and the subthreshold swing (S.S) along

with the time in the ammonia environment were depicted in Fig. 3-2. The

field-effect mobility and the subthrehsold swing kept almost unchanged, while the

threshold voltage exhibited strong dependence on the sensing time. From the above,

we supposed the drain current decreases resulted from the variations of mobility and

In this thesis, we will discuss the gas sensing mechanism through the threshold

voltage variations in Fig.3-3.

3-1.2 The recovery phenomenon

In addition to the reactions between OTFTs and ammonia, we looked forward

to observing the phenomenon when ammonia gas was removed. We noted the

threshold voltage tended to recover after taking away the cotton with ammonia. The

phenomenon is shown in Fig. 3-4. At room temperature, the threshold voltage

recovery behavior implied that the reaction between ammonia and pentacene is

physical absorption, not chemical reactions. We consider that the variation of

threshold voltage results from the absorption of ammonia molecules absorbed onto

the pentacene surface into the pentacene film through diffusion.

3-1.3 Gas diffusion model

Oana et al. [31] used Fick‘s second law to model the one dimension diffusion of

gases into pentacene single crystals. The total quantity of gas molecules in pentacene

crystal is obtained by integrating Fick‘s second law over the length of the diffusion

pentacene. An extra term needs to be added in the solution (Nstart). This term reflects

the fact that the measurement starts at t=0 with a quantity of gas that was accumulated

in crystal. The equation of modeling the diffusion of gases in pentacene is thus:

(3-1)

where N(t) is the molar fraction of gas at time t, Nsource is the gas flow expressed as

molar fraction, d is the length of the crystal in the direction in which the one

dimension diffusion occurs, and D is the diffusion coefficient. Their result showed the

diffusion length is similar to the pentacene film thickness.

We rewrite the equation (3-1) to another form:

(3-2)

Δ N(t) represents the variation of gas in pentacene film in time = t and is expressed as molar fraction. We tried to simulate the relations between equation (3-2) and

diffusion coefficient (D)、diffusion length (d) by Matlab . In Fig.3-5, we set

diffusion length d as 1000 Ǻ and varied the value of diffusion coefficient D. As the

2 start source source N(t)-N ( ) 2 { [1 ( )] [1 exp( )]} N N 2 4 N t d d d erf Dt Dt Dt           2 start 2 N(t)=N N { [1 ( )] [1 exp( )]} 4 2 source d d d erf Dt Dt Dt        

In other words, the response became faster. In Fig.3-6, we fixed the diffusion

coefficient D as 10-12 cm2/s and changed the diffusion length d. When the diffusion

length increased, the amount of gas that diffused into the pentacene film became

larger. From these results, we found if that the value of D/d was larger, the responses

of equation (3-2) became faster.

3-1.4 Gas concentration effect

The above all discussion about the gas sensing phenomenon was with a fixed

amount of ammonia concentration. In Fig.3-7, we added different amount of

ammonia and observed the threshold voltage shifts. When the amount of ammonia

increased, the threshold voltage shift also increased. To study the gas sensing

mechanism in different ammonia concentration, we used the Fick‘s second law to fit

the experiment results in Fig.3-8. An agreement can be found when the following

relationship is proposed.

(3-3)

In this relationship, we assumed that the threshold voltage shift was due to the extra

charge that came from the gas molecules.K is the ratio of molecules which form the

effective charge. ( ) ox th source N t C V K N   

In Fig.3-8, we set diffusion length d as 10-5cm and varied the fitting

parameter D、K to obtain the simulation result. In the condition with 2ℓ ammonia, D is 1.5*10-13 cm2/s and K is 4.93*10-13. But in 10ℓ ammonia condition, the fitting parameters are D= 5*10-13 cm2/s and K=5.31*10-13. The

increase of ammonia concentration increased the diffusion coefficient ―D‖ and

enhanced the percentage of ammonia molecules which can induce the threshold

voltage shift.

3-1.5 The phenomenon of different NH

3

exposed time

In previous experiments, Fick‘s second law can successfully explain that the

ammonia enhanced the threshold voltage shift. But we observed another interesting

phenomenon in the ammonia-OTFT sensor.

We measured three identical devices that were placed at the same distance from

the cotton mass. When the measurement was applied on the three devices one after

another along with the NH3 sensing time, the threshold voltage variations of these

three devices were depicted in Fig.3-9. Interestingly, these three devices had similar

initial threshold voltages no matter how long these devices had been exposed to the

3-2 The phenomenon of bias stress gas sensing

In section 3-1.5, we observed that bias seem to play a critical role in gas

sensing. In this section we looked forward to verifying the relationship between gas

sensing and bias-stress effect

3-2.1 The theory of bias stress effect

At first, we introduce the theory of bias stress effect [32]. Bias stress effect

means applying a prolonged bias to the gate electrode and causes the threshold

voltage shifts by creating new electronic states or defects. For a-Si TFTs two major

models were developed to explain Vth shifts. [33.34] Although, the underlying

microscopic processes are different, in both models the creation of states is governed

by a dispersive process. In these models, the density of states, ∆ND, is proportional to

∆Vth, since ∆ND=Cox∆Vth. These states are the ones that must be filled before

significant conduction can occur via accumulation layer. Thus, the more defects

created during bias stress, the more holes are needed to occupy these defect states,

resulting in a threshold voltage shift. In other words, the driving force for defect

creation is charge trapping into defect-creation sites. Therefore, the rate at which

defects are created depends on the density of free holes induced by the gate bias. In

by charged defects. Based on these models it is readily shown that the rate of change

in Vth is given by differential equation [35.36]:

1 0 ( ( ) th D BT dV d N t N t dt dt t         (3-4)

where NBT (t) represents the concentration of free holes in the accumulation channel.

Solving equation (3-4) with α=1 yields the stretched exponential function as given:

( ) 1 exp th th th ini ini G th t V V V V V    _{ }    _{ }  _   _    _{ }          (3-5) where ini th

V is the initial threshold voltage.  is a dispersion parameter that can be

extracted by plotting log



ln



1VT/(VGVTini)





as a function of log(t) and 

represents the effective trapping time that can be expressed by

1 exp(E ) kT 





 (3-6)

and can be obtained by measuring the stretched exponential curves at different

temperatures. According to equation (3-6), in a plot of the form log (τ) as a function

of 1/T, the slope of this curve represents the activation energy for defect creation,

Eτ=Ea/β . In other words, the value of Eτ, is a measure of the device stability.

3-2.2

The phenomenon of bias stress gas sensing

In order to study the phenomenon of bias stress gas sensing in detail, we

designed the following experiments:

Firstly, we measured the devices with bias stress of -7V、-11V、-15V. The

results are shown in Fig.3-10. When the gate bias was increased, the threshold voltage

shift was also increased. Then, we used the log scale to illustrate the experiment result

as shown in Fig.3-11. It is found that the threshold voltage has a power law

dependence on time. In other words, the experiment result is the same with the theory

of bias stress effect.

Next we measured four identical devices with different amount of ammonia

under -7V gate bias stress. We supposed that if the gate bias stress is the most

important factor in the sensing mechanism, no matter how much ammonia we added,

the threshold voltage shifts are all the same. In Fig.3-12, although we stressed bias in

Maybe the gate bias stress and the ammonia concentration all play critical roles in the

gas sensing behavior.

Then, in Fig.3-13, 3-14 we used amounts of ammonia concentration under

different stressed biases. In Fig.3-13 and 3-14 we added 2ℓ and10ℓ ammonia,

respectively. In the same ammonia environment, the threshold voltage shift increased

when the bias stress was increased. Comparing Fig.3-13 with Fig.3-14, 10ℓ

ammonia enhanced more threshold voltage shifts than 2ℓ under the same condition

of gate bias stress.

Actually the bias stress effect also caused the threshold voltage shifts in the

above experiment. But we look forward to studying the mechanism only between gas

concentration and gate bias stress. Therefore, we subtracted the bias stress effect in

Fig.3-10 from the experiment data of Fig.3-14 and showed the result in Fig. 3-15. In

Fig. 3-15, the simulation formula comes from the equation (3-3). The fitting

parameters are shown in Fig. 3-15. When the stressing bias is increased, the diffusion

coefficient increased. The diffusion behavior became more obvious. At the same time,

3-2.3 The gas sensing mechanism under bias stress

From the above experiment, we proposed two kinds of gas sensing mechanism

under bias stress.

3-2.3.1 Charge- trapped states

We considered the NH3 molecules are absorbed onto the surface of active

material or absorbed into the film through diffusion. The absorptive behavior between

NH3 molecules and pentacene molecules could result in charge-trapped state in

pentacene film. Therefore we observed the threshold voltage shift in the environment

with ammonia gas. When we increased the concentration of ammonia, the diffusion

became more obvious. Much more ammonia molecules can react with pentacene and

enhanced the variations of threshold voltage.

3-2.3.2 Dipole molecules

In the beginning, the ammonia is absorbed onto the surface of active material.

Then the ammonia molecules diffuse into the pentacene film through the grain

boundary. Because ammonia is a dipole molecule, bias stress can enhance the NH3

molecule to absorb onto the surface of pentacene or diffused into the film. Moreover

pentacene-based OTFT is a p-type transistor. We need to bias negative voltage for

accumulation. In this condition, the ammonia molecules diffuse into the active layer

have to apply much more negative bias to overcome the ammonia-induced electronic

Chapter 4

Conclusion

In this study we showed that the NH3 molecules are absorbed onto the surface

of active material and diffuse into the film. The absorptive behavior between NH3

molecules and pentacene molecules result in the variation of threshold voltage. The

threshold voltage shift may originate from creating charge-trapped states of

pentacene and polar molecules – ammonia.

Fick‘s second law can successfully explain the variations of threshold voltage in

the present of ammonia. When the concentrations of NH3 increase, the diffusion

coefficient increase and much more ammonia molecules contribute to the threshold

voltage shift. But we also observed that three identical devices had similar initial

threshold voltages no matter how long these devices had been exposed to the NH3

gas. This result implied that the sensing behavior may have a strong correlation with

the gate bias.

In addition, we varied the value of bias stress and fixed ammonia concentration.

In order to reduce the threshold voltage shifts resulting from the bias stress effect,

we subtracted the bias stress effect from the experimental data. Then, we observed

that the bias stress increase enhanced the variation in the threshold voltage. At the

Finally, we propose that gas diffusion and gate bias both play critical roles in

Reference

[1]. M. Pope and C.E. Swenberg, Electronic Processes in Organic Crystals and polymer,Oxford University Press, New York, 1999.

[2]. D.F. Barbe and C.R. Westgate ‖Surface state parameters of metal-free phthalocyanine single crystals‖. J. Phys. Chem. Solids ,31, 2679 (1970).

[3]. M.L. Petrova and L.D. Rozenshtein ‖Field effect in the organic semiconductor‖ Chloranil Fiz. Tverd. Tela, Soviet Phys. Solid State, 12, 961 (1970).

[4]. M.L. Petrova and L.D. Rozenshtein:‖Field effect in the organic semiconductor chloranil. Fiz. Tverd. Tela ,Soviet Phys. Solid State, 12, 961 (1970).

[5]. F. Ebisawa, T. Kurokawa, and S. Nara,‖Electrical properties of polyacetylene-polysiloxane interface‖, J. Appl. Phys, 54, 3255 (1983). [6]. A. Tsumura, K. Koezuka, and T. Ando‖ Macromolecular electronic device:

Field-effect transistor with a polythiophene thin film‖, Appl. Phys. Lett.,49, 1210 (1986).

[7]. Björn Timmer, Wouter Olthuis and Albert van den Berg,‖ Ammonia sensors and their applications—a review‖, Sensors and Actuators B, 107, 666 (2005)

[8]. M. Wallin, C-J. Karlsson, M. Skoglundh,‖ Selective catalytic reduction of NOx with NH3 over zeolite H-ZSM-5: influence of transient ammonia supply‖, J.

Catal.,218, 354

[9]. X. Xuan, C. Yue, S. Li,‖ Selective catalytic reduction of NO by ammonia with fly ash catalyst‖, Fuel 82, 575 (2003).

[10].Steven G. Buckley, Christopher J. Damm, Wolfgang M. Vitovec,‖ Ammonia Detection and Monitoring with Photofragmentation Fluorescence‖, Applied Optics, 37, 8382 (1998)

[11]. L.R. Narasimhan, W. Goodman, C. Kumar,‖ Correlation of breath ammonia with blood urea nitrogen and creatine during hemodialysis‖, PNAS, 98, 4617 (2001)

[12]. D.J. Kearney, T. Hubbard, D. Putnam,‖ Breath ammonia measurement in Helicobacter pylori infection‖, Digest. Dis. Sci, 47, 2523 (2002)

[13].N.K. Jain, V. Mangal, ―Helicobactor pyroli infection in children‖, J. Nep. Med. Assoc, 38, 140 (1999)

[14]. K. Zakrzewska, ―Mixed oxides as gas sensors‖, Thin Solid Films, 391, 229 (2001)

[15].N. Yamazoe, ―Chemical Sensor Technology‖, Elsevier, Amsterdam, 1991. [16]. X. Wang, N. Miura, N. Yamazoe, ―Study of WO3-based sensing material for

NH3 and NO detection‖, Sens. Actuators B ,66, 74 (2000).

[17]. C.N. Xu, N. Miura, Y. Ishida, ―Selective detection of NH3 over NO in

combustion exhausts by using Au and MoO3 doubly promoted WO3 element‖,

Sens. Actuators B, 65163 (2000)

[18]. V.V. Chabukswar, S. Pethkar, A.A. Athawale‖, Acrylic acid doped polyaniline as an ammonia sensor‖, Sens. Actuators B , 77 657 (2001)

[19]. L. Torsi, A. Dodabalapur, L. Sabbatini, P.G. Zambonin, ―Multi-parameter gas sensors based on organic thin-film-transistors‖ Sensors and Actuators B, 67, 312, (2000).

[20]. F. Liao, C. Chen, V. Subramanian, ―Organic TFTs as gas sensors for electronic nose applications,‖ Sensors and Actuators B, 17, 849 (2005).

[21]. M.C. Tanese, D. Fine, A. Dodabalapur, L. Torsi, ―Interface and gate bias dependence responses of sensing organic thin-film transistors,‖ Biosensors and

[22]. B. Crone, A. Dodabalapur, A. Gelperin, L. Torsi, H. E. Katz, A. J. Lovinger, and Z. Bao ―Electronic sensing of vapors with organic transistors,‖ Appl. Phys. Lett.,

78, 3965, (2001).

[23]. L. Torsi, M.C. Tanese, N. Cioffia, M.C. Gallazzi,L. Sabbatini, and P.G. Zambonin, ―Alkoxysubstituted polyterthiophene thin-film-transistors as alcohol sensors,‖ Sensors and Actuators B, 98, 204,(2004)

[24]. B. Crone, A. Dodabalapur, A. Gelperin,‖ Electronic sensing of vapors with organic transistors‖ Appl Phys Lett 78, 2229, (2001)

[25] JB Chang,‖Functionalized Polythiophene Thin-film Transistors for Low-cost Gas Sensor Arrays‖,eecs.berkeley.edu, 2006

[26].Torsi L, Lovinger AJ, Crone B, ―Correlation between Oligothiophene Thin Film Transistor Mophology and Vapor Response‖, J Phys Chem B, 106, 12563 (2002)

[27]. Liang Wang, Daniel Fine, and Ananth Dodabalapur,‖ Nanoscale chemical sensor based on organic thin-film transistors‖, Appl. Phys. Lett., 85, 6386, 2004. [28]. C. D. Dimitrakopoulos, A. R. Brown, and A. Pomp, ―Molecular beam deposited

thin films of pentacene for organic field effect transistor applications‖, J. Appl. Phys.,80, 2501, (1996)

[29]. Chih-Wei Chu, Sheng-Han Li, Chieh-Wei Chen,―High-performance organic thin-film transistors with metal oxide/metal bilayer electrode‖, Appl. Phys. Lett., 87, 193508, (2005).

[30]. A.Alec Talin, Luke L. Hunter, François Léonard,‖ Large area, dense silicon nanowire array chemical sensors‖ , Appl. Phys. Lett. 89, 153102 (2006)

[31].Oana D. Jurchescu, Jacob Baas, and Thomas T. M. Palstra,‖ Electronic transport properties of pentacene single crystals upon exposure to air‖, Appl. Phys. Lett. 87, 052102 ,(2005)

[32].Henrique L. Gomes, Peter Stallinga, Franco Dinelli, ―Electrical characterization of organic based transistors: stability issues‖,Polym.Adv. Technol.16, 227,(2005).

[33]. Jackson WB, Marshall JM,‖ Role of hydrogen in the formation of metastable defects in hydrogenated amorphous silicon” Phys. Rev. B ,39,1164 (1989) [34]. Crandall RS.‖ Defect relaxation in amorphous silicon: Stretched

exponentials, the Meyer-Neldel rule, and the Staebler-Wronski effect” Phys Rev. B, 43,4057 (1991)

[35]. Wehrspohn RB, Deane SC, French ID, ―Relative importance of the Si–Si bond and Si–H bond for the stability of amorphous silicon thin film transistors.‖ J. Appl. Phys.,87, 144 (1999).

[36]. Wehrspohn RB, Deane SC, French ID, ―Effect of amorphous silicon material properties on the stability of thin film transistors: evidence for a local defect creation model ‗J. Non-Cryst. Solids, 266, 459(2000)

Fig. 1-1 Molecular structure of pentacene.

Fig. 1-2 (a) Top contact structure.

Fig. 1-3 The work function of various metals and ionization potential of pentacene.

wet oxide(100nm)

Heavily doped Si substrate

Pentacene (100nm)

Au(100nm)

0.0 1.0x10-6 2.0x10-6 3.0x10-6 4.0x10-6 5.0x10-6 6.0x10-6 10-12 10-11 10-10 10-9 10-8 10-7 10-6 10-5 10-4 -40 -30 -20 -10 0 10 0.0 5.0x10-8 1.0x10-7 1.5x10-7 2.0x10-7 2.5x10-7 I D (A ) V_G(V) I D (A ) G M pure device with 2ml NH3 60s with 2ml NH3 120s with 2ml NH 3 180s with 2ml NH 3 240s with 2ml NH3 300s -18 -17 -16 -15 -14 -13 1.00 1.25 1.50 0 50 100 150 200 250 300 0.55 0.60 0.65 Vth (V ) T i m e ( s ) S .S(V/de ca d e) m o b ilit y (cm 2 /Vs) 0.021 3.35 -18 -17 -16 -15 -14 -13 1.00 1.25 1.50 0 50 100 150 200 250 300 0.55 0.60 0.65 Vth (V ) T i m e ( s ) S .S(V/de ca d e) m o b ilit y (cm 2 /Vs) 0.021 3.35

Fig. 3-1 The electrical properties of OTFT adding 2ℓ NH3.

Fig. 3-2 The variations of subthreshold swing、threshold voltage and mobility along with time in environment of 2ℓ NH3 (after pink line).

Fig. 3-3 The variations of threshold voltage along with time in environment of 2ℓ NH3 Time (s) 0 2000 4000 6000 8000 10000 12000 14000 -V th ( V ) 16 18 20 22 24 26 28

initial

Add 20



ℓ NH

3

Remove

NH

3 Time(s) 0 50 100 150 200 250 300 350 -V th (V ) 13.5 14.0 14.5 15.0 15.5 16.0 16.5 17.0 17.5

Add 2



ℓ NH

3

Time (s) 0 500 1000 1500 2000 2500  N(t) / N so ur ce 0.0 2.0e-6 4.0e-6 6.0e-6 8.0e-6 1.0e-5 1.2e-5 D=1*10^-12 D=2*10^-12 D=4*10^-12 D=8*10^-12

Fig. 3-5 The fitting curve of Fick‘s second law are in the condition of d=10-5 cm and D=10-12 cm2/s、2*10-12 cm2/s、4*10-12 cm2/s、8*10-12 cm2/s Time (s) 0 500 1000 1500 2000 2500   t) / Nso ur ce 0 1e-5 2e-5 3e-5 4e-5 5e-5 d=1*10^-5 d=2*10^-5 d=4-10^-5 d=8*10^-5 d=9*10^-5

Fig. 3-6 The fitting curve of Fick‘s second law are in the condition of D=10-12 cm2/s

Fig. 3-7 The threshold voltage shift in different amount of NH3.

Fig.3-8

Fig. 3-8 The dependence of threshold voltage shift on sensing time in the

Time(s) 0 500 1000 1500 2000 2500 T h re sh o ld v o lta g e s h ift ( V ) -0.4 -0.2 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 no ammonia with 2L ammonia with 10L ammonia Time (s) 0 500 1000 1500 2000 2500 T h re s h ld v o lt a g e s h if t (V ) 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6

experiment result ( 2ul NH3 ) experiment result ( 10ul NH3 ) simulation result ( 2ul NH3 ) simulation result ( 10ul NH3 ) D=5*10-13 _cm2_{/s, K=5.31*10}-13

Fig. 3-9 The threshold voltage variation of three identical devices measured one after another in the environment with NH3 gas. The NH3 amount

injected into the cotton mass was 2ℓ.

Fig. 3-10 The threshold voltage shift in different bias stress conditions

Sensing time(s) 0 200 400 600 800 1000 1200 - V th (V ) 14.5 15.0 15.5 16.0 16.5 17.0 17.5 18.0 Time (s) 0 500 1000 1500 2000 2500 T h re s h o ld v o lt a g e s h if t (V ) 0.0 0.5 1.0 1.5 2.0 2.5 3.0 Vg-Vth= -7 V Vg-Vth= -11 V Vg-Vth= -15V

Fig. 3-11 The threshold voltage shift in different bias stress conditions

Fig. 3-12 The threshold voltage shift in different amount of concentrations with

Time (s) 10 100 1000 10000 T h re sh o ld v o lta g e s h ift ( V ) 0.1 1 10 Vg-Vth= -7 V Vg-Vth= -11 V Vg-Vth= -15V Time(s) 0 500 1000 1500 2000 2500 T h re sh o ld v o lta g e s h ift ( V ) 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 pure device with 2ul NH3 with 10ul NH3 with 20ul NH3

Fig. 3-13 The threshold voltage shift in the environment of 2ℓ ammonia with different amount of bias stress.

Fig. 3-14 The threshold voltage shift in the environment of 10ℓ ammonia with different amount of bias stress.

Time (s) 0 500 1000 1500 2000 2500 T h re sh o ld v o lta g e s h ift ( v) 0 1 2 3 4 5 6 Vg-Vth= -7 V Vg-Vth= -11V Vg-Vth= -15 V Time (s) 0 500 1000 1500 2000 2500 T h re sh o ld v o lta g e s h ift ( V ) 0 2 4 6 8 Vg-Vth= -7 V Vg-Vth= -11V Vg-Vth= -15 V

Fig. 3-15 The dependence of threshold voltage shift on sensing time in the environments with different gate bias strees. The circles are experiment data. The curves are the fitting result from Fick‘s second law.

Time (s) 0 500 1000 1500 2000 2500 T hre sho ld v ol tag e shi ft (V ) 0 1 2 3 4 5 Vg-Vt = -7V Vg-Vt = -11V Vg-Vt = -15V simulation result simulation result simulation result D=1.7*10-13 cm2/s, K=1.82*10-2 D=1.1*10-13 cm2/s, K=1.38*10-2 D=0.5*10-13 cm2/s, K=0.983*10-2