Citation
Behavior of Light-Emitting Polymers: Film Formation, Optical Properties, and Degradation

Material Information

Title:
Behavior of Light-Emitting Polymers: Film Formation, Optical Properties, and Degradation
Copyright Date:
2008

Subjects

Subjects / Keywords:
Annealing ( jstor )
Bubbles ( jstor )
Cathodes ( jstor )
Electrical bias ( jstor )
Film water ( jstor )
Lead ( jstor )
Polymers ( jstor )
Solvents ( jstor )
Thin films ( jstor )
Viscosity ( jstor )

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Embargo Date:
7/30/2007

Downloads

This item has the following downloads:


Full Text












BEHAVIOR OF LIGHT-EMITTING POLYMERS: FILM FORMATION, OPTICAL
PROPERTIES, AND DEGRADATION















By

SHYH-CHYANG LUO


A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA


2005

































Copyright 2005

by

Shyh-Chyang Luo


































This dissertation is dedicated to my parents, my brothers, and Inga for their love, support
and encouragement.















ACKNOWLEDGMENTS

First of all, I would like to thank my advisor and my committee chair, Dr. Elliot

Douglas, for lending an unwavering supportive presence to my studies in the United

States throughout these years. Without a doubt, I feel very fortunate to have received the

benefits of his guidance. His welcomed openness and confidence in my ideas have made

working in his group a pleasant and enriching experience. I would like to express sincere

gratitude to my committee members, Dr. Paul Holloway, Dr. Ronald Baney, Dr. David

Norton, and Dr. Kirk Schanze, for their invaluable help and support of my research.

I would like to thank Dr. Valentin Craciun, Dr. Eric Lambers, Andrew Gerger, and

Juhyun Woo in the Major Analytical and Instrumentation Center for their assistance in

the measurements of X-ray reflectivity, Auger electron spectroscopy, and atomic force

microscopy. I also wish to thank Dr. Mark Davidson in Microfabritech for helping with

the photoluminescence measurement. Without their aid and input, this work could not

have been done.

I would like to extend a special "thank you" to all of the members in Dr. Douglas'

and Dr. Holloway's group and my classmates, especially Dr. Wenling Jia, Lewei Bu,

Brendan Collins, Hsiu-Hsin Chung, Phil Chung, and Jia- Hau Yen, for their help and

support. I am so lucky to have worked with all of them. Many thanks go to Jennifer

Wrighton for her attention to all the details that made my working in this group all the

more enlightening.









I especially thank Ying-Chih Wang. Her support and encouragement enabled me

to overcome the challenges of this project. Finally, I would like to thank my parents and

my younger brothers. They are always there for me. Without them, I could not have

gotten this far.
















TABLE OF CONTENTS

Page

A C K N O W L E D G M E N T S .................................................................... ......... .............. iv

TA BLE OF CON TEN TS................................................................ .......................... vi

L IST O F T A B L E S ........ ................................................................. ...... ix

LIST OF FIGURES ............................... ... ...... ... ................. .x

ABSTRACT .............. .......................................... xvi

CHAPTER

1 INTRODUCTION ............... ................. ........... ................. ... .... 1

2 REVIEW OF LITERATURE ......................................................... .............. 4

2.1 Fundamentals of t-Conjugated Polymers....................... ...... ..............4
2.1.1 A Brief History of 7t-Conjugated Polymers...............................................4
2 .1.2 P h otop hy sics..................................... ............... ................. 5
2.2 Polym er Light-Em hitting D iodes (PLED s) ........................................ ...................6
2 .2 .1 S o lv en t E ffects.................................................................... .......... .. .. .8
2.2.2 A nnealing Effects ......................................................... .............. 11
2.2.3 Degradation of PLED Devices ............ ............................................13
2.3 Defects of Spin-Coated Polymer Thin-Film ................. .................................... 17
2.3.1 Benard-M arangoni Convection ...................................... ............... 18
2.3.2 D ew getting Phenom enon......................................... .......... ............... 20

3 MARANGONI EFFECTS ON POLYMER THIN FILM STRUCTURE .................23

3 .1 In tro d u ctio n ..................................................................................................... 2 3
3.2 Experiment.......................................... 24
3.2.1 Sam ple Preparation .......................................................... ............... 24
3.2.2 Characterization.............. ................. ...............26
3.2.2.1 Atomic Force Microscopy (AFM) ................................................26
3.2.2.2 X -Ray Reflectivity (XRR) .................................... ............... 26
3 .3 R results and D iscu ssion ........................................ ...........................................26
3.4 C conclusion .......................................................... ........ ............ 39









4 RUPTURE OF POLYMER THIN FILMS DURING SPIN-COATING PROCESS.40

4 .1 In tro d u ctio n ..................................................................................................... 4 0
4.2 Experiment.............................................................40
4.2.1 Sam ple Preparation .......................................................... ............... 40
4.2.2 Characterization............. ...... ....................... 41
4 .3 R esu lts an d D iscu ssion ........................................ ...........................................4 1
4.4 C conclusion ..................................................................... .......... 53

5 SOLVENT, CONCENTRATION, AND ANNEALING EFFECTS ON
PHOTOLUMINESCENCE OF POLYMER SOLUTIONS AND THIN FILMS......55

5 .1 Introdu action ......... .......... ... .........................................................55
5.2 Experiment........................................... 56
5.2.1 Sam ple Preparation .......................................................... ............... 56
5.2.2 PL Measurement .............. ........... ...................... 56
5 .3 R esu lts an d D iscu ssion ............................................ ....................................... 57
5.3.1 PL of Polym er Solutions ........................................ ........................ 57
5.3.1.1 Solvent effects .................................. ........ .... ................. 57
5.3.1.2 Concentration effects............................. ..... .................. 64
5.3.1.3 Water effect on PFO in THF solutions ............................. ......66
5.3.2 PL of Polymer Thin films .............................................67
5.3.2.1 Solvent effects .................................................................. .......... .... 67
5.3.2.2 Concentration and spin-speed effects....................................68
5.3.3.3 A nnealing effects......................................... .......... ............... 72
5.4 Conclusion ............... ......... ..................... ........................ 73

6 DEGRAGATION OF METAL CONTACT IN POLYMER LIGHT-EMITTING
DIODE DEVICES .................................... .. .......... .. ............75

6.1 Introduction........................................................................ ....... ...... 75
6.2 Experiment.......................................... 75
6.2.1 D evice Fabrication........................................................... ............... 75
6.2.2 Characterization...... ...................................... ............ ........ 76
6.3 R results and D discussion .................................................................................. 77
6.3.1 Bubble Formation Phenomenon and Mechanism............... .......... 77
6.3.2 Surface Com position (AES) .....................................................................83
6.4 C conclusion ..................................................................... .......... 87

7 EFFECT OF THERMO-INDUCED OXIDATION ON POLYMER STRUCTURE
AND OPTICAL PROPERTIES OF POLYFLUORENE................ .....................88

7.1 Introduction ..................................................................... 88
7.2 Experiment........................................... 89
7.2.1 Sam ple Preparation .......................................................... ............... 89
7.2.2 Characterization...... ............ .......................... ...... ...... ........ 89
7 .3 R results and D iscu ssion ........................................ ...........................................90









7.3.1 Chain Conformation in PFO Thick Films ................................................90
7.3.1.1 M molecular w eight effect ....................................... ............... 90
7.3.1.2 Therm al-oxidation effect...................................... ......... ............... 95
7.3.2 Molecular Effect on PFO Thin Film Structure.............. ..................98
7 .4 C o n c lu sio n ............. ......... .. .. .............. ...................... ................ 10 1

8 CONCLUSIONS AND RECOMMENDATIONS FOR FUTURE WORK ............103

L IST O F R E F E R E N C E S ............................................... ....................... ..................... 108

BIOGRAPHICAL SKETCH ...................................... ..................................115
















LIST OF TABLES


Table page

3-1. Characteristic data of solvents at room temperature. .............................................25

3-2. The Mark-Houwink parameters for PVK at 25C .................................................25

3-3. Characteristics of PVK films obtained from XRR measurements ......................35

3-4. Characteristics of MEH-PPV films obtained from XRR measurements..................38

4-1. Roughness of MEH-PPV polymer films prepared from the dried THF and
THF/w after solutions .............................................. .. ...... ................. 48

4-2. PL of MEH-PPV polymer films prepared from the dried THF and THF/water
so lu tio n s .......................................................................... 5 2

5-1. Solubility parameter and polymer solubility. ................................ .................58















LIST OF FIGURES


Figure page

2-1. Structure of conjugated polymers: a) cis-polyacetylene; b) trans-polyacetylene; c)
polythiophene; d) poly(phenylene vinylene); e) poly(pyridyl vinylene); f)
p oly (flu oren e)........................................................................................... 4

2-2. G general structure of PLED s ............................................................................ 7

2-3. The mechanism of EL in PLEDs ............. ............ ............... 7

2-4. Absorption and PL of MEH-PPV in CB and THF solutions ............... .................

2-5. a) PL of CB solution and thin films; b) PL of thin films coated from CB and THF
so lu tio n s. ........................................................... ................ 9

2-6. Characterization of PLEDs: a) I-V curve and b) B-V curve for DCB and THF
devices having sim ilar thickness. ....................................................................... 10

2-7. PL of MEH-PPV films annealed at different temperatures................................. 11

2-8. EL vs. current of the devices annealed at different temperature after Al deposition. 12

2-9. Evolution of black spot formation in air: a) t = 2hrs; b) t = 10 hrs; c) t = 20 hrs; d) t
= 30 hrs; e) t = 40hrs. ....................................... ............... .... ....... 14

2-10. Schematic diagram showing dark spot induced by pinholes ..............................15

2-11. The linear growth of dark spot with various particles......................... ..........15

2-12. Proposed electrochemical half-cell reactions showing the oxidation of metal (M)
and the reduction of doped PEDOT. .......................................................... 16

2-13. Polystyrene coated from THF solution................................. ..............17

2-14. Polystyrene coated from cyclohexane and acetone mixed solution .......................18

2-15. The M arangoni convection ......... ................. ......... .................... ............... 18

2-16. Micrograph of different stages of polystyrene dewetting on Si wafer: a) 15 mins
and b) Ihr. ...........................................................................2 1









2-17. The droplets left on the Si w afer. ........................................ ......................... 21

2-18. Film structure of dewetted thin films: a) A dewetting pattern centered at a defect
(50 x 50 [tm) and b) pattern at high magnification (5 x 5 [tm).............................22

3-1. Chemical structure of a) PVK and b) M EH-PPV.................................................. 24

3-2. Film structure of polymer thin-film observed with optical microscopy: (a) prepared
from THF solution, center area; (b) prepared from THF solution, away from center
area; (c) prepared from chloroform solution, center area; (d) prepared from
chloroform solution, away from center area. Length of scale bar 200tm...............27

3-3. 3D tapping mode AFM image of PVK film surface prepared from (a) chloroform
solution, cell structure near center area; (b) chloroform solution, striation structure
aw ay from center area. ................................. ............ ....................... 28

3-4. The Marangoni convection model (including composition-gradient-driven and
temperature-gradient-driven processes) for the spin-coating process. T1 and C2
represent the temperature and solution concentration near the surface of the
solution layer; T2 and Ci represent the temperature and solution concentration near
the interface between the solution layer and the ITO substrate. ...........................29

3-5. Film structure of PVK thin film observed with optical microscopy prepared from
TCE solution, center area. .............................................. .............................. 30

3-6. 2D tapping mode AFM image of PVK surface prepared from toluene solution......31

3-7. RMS Roughness plotted as a function of evaporation rate of four solvents. ...........32

3-8. MEH-PPV Film structure of polymer thin-film observed with optical microscopy:
(a) prepared from THF solution, center area; (b) prepared from THF solution, away
from center area; (c) prepared from chloroform solution, center area; (d) prepared
from chloroform solution, away from center area; e) prepared from CB solution;
prepared from T C E solution .............................................................. ..... ........33

3-9. 2D tapping mode AFM image of MEH-PPV film surface prepared from (a)
chloroform solution, cell structure near center area; (b) TCE solution, smooth
surface near center area. .......................... .................... ......................... 34

3-10. X-ray reflectivity spectra acquired from PVK thin films prepared from chloroform,
benzene, toluene, and TCE solutions. Experimentally measured data are shown as a
solid line; fitted results are shown as a dotted line......... ......... .......... ............... 35

3-11. X-ray reflectivity spectra acquired from MEH-PPV thin films prepared from
chloroform, chlorobenzene, and TCE solutions. Experimentally measured data are
shown as a solid line of light color, and fitted results are shown as a solid line with
d a rk c o lo r .......................................... ....... ................. ................ 3 7









4-1. SEM image of polymer prepared from THF solution: (a) center area with a low
magnification, length of scale bar 200[Jm; (b) away from center area with a low
magnification, length of scale bar 10m; (c) center area with a high magnification,
length of scale bar 200am; (d) away from center area with a high magnification,
length of scale bar 10 m .............................................................. ....................4 1

4-2. SEM image of polymer thin-film prepared from THF solution with different H20
contents: (a and b) THF dried by molecular sieve for 1 day with a low and high
magnification, length of scale bar 200 and 10 am, respectively; (c), (d) 10tL of
H20 added in 5 mL of THF with a low and high magnification, length of scale bar
200 and 10 am respectively; (e), (f) 100al H20 added in 5mL of THF with a low
and high magnification, length of scale bar 200am and 10 am, respectively..........43

4-3. PEDOT-PSS film structure on ITO substrate observed with optical microscopy: (a)
low magnification, length of scale bar = 200am; (b) high magnification, showing
presence of a hole, length of scale bar = 20m.................................................. 44

4-4. Structure of ruptured thin films: a) Optical microscopy image of MEH-PPV thin
film prepared from 5% H20 in THF, striation structure away from center; b) SEM
image of same film with higher magnification. ............................ ...................46

4-5. Height contrast of polymer thin-films from (a) dried THF solution with polymer
concentration of 5 mg/ml; (b) 5% water in THF solution; (c) height profile of film
along the line indicated in (a); (d) height profile of film along the line indicated in
(b ) ......................................................... .................................4 7

4-6. PL spectra of MEH-PPV polymer thin-films. ............... .... .................49

4-7. PL spectra of MEH-PPV in THF solutions (7 mg/ml) with different water
con centration s............................................................. .. ................. 50

4-8. PLED based on ruptured thin films: (a) The layer structure of PLED device. (b) The
light-emitting output after the device is turned on. ............. .................................. 53

5-1. PL of 0.5% M EH -PPV solutions ........................................ ......................... 57

5-2. PFO dissolved in toluene, THF, chloroform, CB, TCE, and cyclohexanone (from
left to right) ..................................................... ........ ...... 58

5-3. PL of PFO in THF, TCE, Chloroform and CB.................................................... 59

5-4. The measurement of kinematic viscosities of MEH-PPV in CB, chloroform, and
TCE at different concentrations. ........................................ .......................... 61

5-5. rsp/c plotted as a function of c for MEH-PPV in CB, chloroform, and TCE. .........62

5-6. PL of MEH-PPV in chloroform in concentrations of 0.1, 1, 5, and 10 mg/mL. ......64









5-7. PL of MEH-PPV in CB and TCE. .............................. ... ..... ................ 65

5-8. PL of PFO in chloroform ......... .................. ................................... ............... 65

5-9. PL of PFO in the THF solution with different water concentrations ..................67

5-10. Comparison of PL of MEH-PPV in polymer solutions and thin films.....................68

5-11. PL of MEH-PPV thin films from chloroform solutions with concentrations of 2.5,
5, and 7 m g/mL. (Spin-speed = 1000 rpm) ................................... .................69

5-12. PL of MEH-PPV thin films coated from chloroform solution (7 mg/mL) at 1000,
3000, and 5000 rpm ........... .................................. .................... 70

5-13. PL of MEH-PPV thin films from CB solutions with concentrations of 2.5, 5, and 7
mg/mL. (Spin-speed = 1000 rpm) .......... ...... .......................... ............... 71

5-14. PL of MEH-PPV thin films coated from chloroform solutions (7 mg/mL) at 1000,
3000, and 5000 rpm ........... .................................. .................... 71

5-15. The PL of MEH-PPV coated from chloroform solution before and after annealing at
120 C for 3 hrs in a vacuum .................................. ............... ............... 72

5-16. PL of MEH-PPV coated from CB solution before and after annealing at 120 C for
3 hrs in a vacuum .....................................................................73

6-1. Schematic of the PLEDs used in this study.............. .............................................76

6-2. Bubble formation observed with optical microscopy: (a) biased with 7 V, 10-5
A-cm2 for 1 min. within one hour after thermal evaporation of Al; (b) biased with 7
V, 10-5 A.cm2 for 1 min after storage in laboratory air for 1 day (-60% relative
humidity, 27 C). Note the lower magnification in (b). .........................................77

6-3. The effects of storage ambient and encapsulation upon bubble formation as a result
of electrical bias (8 V for 2 min.). Encapsulated or non-encapsulated devices were
stored for four days in argon or air as follows: (a) argon with encapsulation; (b)
argon without encapsulation; (c) air with encapsulation; (d) air without
encapsulation. Photos on the left and right are before and after the devices were
biased, respectively. ..................... .... ............ ................. .... ....... 79

6-4. The dependence of bubble formation on the type of conjugated polymer and
temperature in the absence of electrical bias. Devices consisted of the normal
ITO/polymer/Ca/Al structure, where the polymer layer is either PVK or PEDOT-
PSS. Devices were examined in real time by optical microscopy while being
heated on a hot stage either immediately after deposition of the cathode, or 24
hours after this deposition with storage in air. ......................... ............... ......80









6-5. Formation of bubbles by water droplets in the absence of an electrical bias when a
Ca layer is present. The photos are of Al only or normal Ca/Al cathode contacts
with water droplets on the surface. Al-only contact (a) immediately after placing a
water droplet on the surface, and (b) after a few minutes; no bubbles are observed.
Ca/Al contact (c) immediately after placing a water droplet on the surface, and (d)
after a few minutes; large bubbles are obvious in (d). ...........................................82

6-6. Pinhole structure and composition: (a) SEM image of a pinhole in the Al/Ca
cathode dot contact and surrounding bubbles generated by application of an
electrical bias; (b) AES spectra from the cathode surface (i) inside the pinhole, and
(ii) outside the pinhole............ ...................................................... ...... .... .... 85

6-7. AES sputter depth profile of a normal 10 nm Ca/60 nm Al PLED cathode dot
contact (a) before and (b) after an electrical bias was applied. Note the increase of
oxygen at the Al/Ca interface after electrical biasing (8 V for 10 min. in laboratory
air) ........................................................ .................................86

7-1. Proposed mechanism for the generation of keto-defect sites ................................88

7-2. Images of a low Mw PFO under a polarized optical microscope at (a) 1200C, (b)
150C, (c) 210C, and (d) 220 C ............................................................................. 90

7-3. X-ray diffraction of low Mw PFO samples at room temperature, 125, 150, 200, and
250C. Sample was annealed at 2500C for 10 minutes before cooling to the
m easurem ent tem perature. ............................................. .............................. 91

7-4. X-ray diffraction of high Mw PFO samples at room temperature, 125, 150, 200, and
250C. Sample was annealed at 2500C for 15 minutes cooling to the measurement
tem perature..................................... ........................... .... ........... 92

7-5. Characteristic XRD profile for film in N, a, a', and 0 phases [114]........................93

7-6. XRD profile of low Mw and high Mw PFO samples annealed at 1700C for 3 and 15
hours in A r......................................................................... .. ....... ....... 94

7-7. Comparison of X-ray diffraction pattern of high Mw PFO samples annealed at
170C in Ar and in air for 3 and 15 hours. ....................................... .................. 95

7-8. PL of high Mw PFO samples annealed in air at different temperatures: 170C for 1
hour (solid line), 200C for 1 hour (dash dot line), 2000C for 3 hours (dash line),
and 240C for 1 hour. ....................................... ............... .... ....... 96

7-9. The X-ray diffraction pattern of annealed high Mw PFO samples corresponding to
the sam ples show n in Figure 7-8 ....................................................................... 97

7-10. XRR spectra acquired from spin-coated high Mw PFO thin films. Solid line and
dash line represent the experimental data and the fitted results (one layer model)
re sp e ctiv e ly ................................................................................ 9 8









7-11. XRR spectra acquired from spin-coated low Mw PFO thin films. Solid line and
dash line represent the experimental data and the fitted results (one layer model)
respectively..................................... ................................ ........... 99

7-12. XRR spectra acquired from spin-coated low Mw PFO thin films. Solid line and
dash line represent the experimental data and the fitted results (two layer model)
respectively............................................................... ... .... ......... 100

8-1. The proposed model of ruptured polymer film formation during the spin-coating
process due to the surface-tension gradient induced by water ............................105

8-2. The proposed model of ruptured polymer film formation during the spin-coating
process due to phase separation. ........................................ ........................ 106















Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy

THE BEHAVIOR OF LIGHT-EMITTING POLYMERS: FILM FORMATION,
OPTICAL PROPERTIES, AND DEGRADATION

By

Shyh-Chyang Luo

August 2005

Chair: Elliot P. Douglas
Major Department: Materials Science and Engineering

The properties of light-emitting polymer thin films prepared via spin coating from

different organic solvents were studied. The surface analysis as observed by atomic force

microscopy, scanning microscopy, and x-ray reflectivity has shown that the structure of

polymer thin-films is controlled by the properties of organic solvents. The

photoluminescence (PL) spectrum of polymer thin-films and polymer solutions has

shown the correlation between the chain conformation and light-emitting properties. A

combination of viscosity measurements is used to illustrate the solvent effects on the

optical properties of light-emitting polymers in solutions and thin films.

The solvent-induced film structure of poly(n-vinyl carbazole) (PVK) and poly(2-

methoxy-5-(2'-ethyl-hexyloxy)-1,4-phenyl vinylene) (MEH-PPV) thin films on indium

tin oxide coated glass was examined. The spin-coated polymer thin films are not in

thermodynamic equilibrium; rather, the film properties are affected by the dynamics of

the spin-coating process. The water in tetrahydrofuran induces rupturing of polymer









films during the spin-coating process. Solvents with a high-evaporation rate lead to high-

surface roughness due to Benard-Marangoni convection. Results show that the surface

roughness and structure of the films are dominated by the dynamics of the film-formation

process rather than the thermodynamic interactions between the polymer and solvents.

Two degradation phenomena are studied and discussed in this work. The first

manifestation of degradation is the bubble formation on the polymer light-emitting

diodes. Water can diffuse through pinholes to reach polymer layers. Heat and electrical

bias can induce the reaction between water and metal contacts, which leads to the bubble

formation. The second manifestation of degradation is the thermal-induced oxidation of

poly[9,9-dioctylfluorene] (PFO). The oxidation of PFO can lead to an extra red-shifted

emission and the crosslinking of polymer chains.














CHAPTER 1
INTRODUCTION

The first report of light-emitting diode devices with conjugated polymers as the

active layer by Burroughes et al. in 1990 [1] resulted in extensive investigation of

polymer light-emitting diodes (PLEDs). The demonstration of PLEDs fabricated by

casting the polymer film from solutions [2] and their successful application onto flexible

substrates both suggest how PLEDs possess a great potential for display technology [3].

These devices show many advantages including low cost of processing, low-power

consumption, and a flexible display. So far, many studies have been done to improve the

device's efficiency and lifetime [4-12].

The optical properties of light-emitting polymers are determined by their chemical

and physical structures, which are the configuration and conformation respectively of

polymer chains. The light-emitting color can be tuned by changing the chemical structure

of conjugated polymers. Typical examples are poly (p-phenylene vinylene) (PPV) [1,2]

and poly (fluorene) (PF) [13,14]. The photoluminescence (PL) and electroluminescence

(EL) of PPV-related polymers show a red-orange range and PF-related polymers show a

blue-green range. White light PLEDs can be achieved by synthesizing copolymers [15] or

by using polymer-polymer, or polymer-phosphor blend systems [16-18]. More recently,

polysilane-related polymers show EL in the UV and NUV range [19-21]. Different from

PPV or PF, EL from polysilane is due to the a-conjugation instead of x-conjugation on

the main chain. Additionally, a few non-conjugated polymers, such as poly (9-vinyl









carbazole) (PVK) can also generate EL [22]. PVK can emit blue light via radiative decay

from carbazole chromophores.

Emission wavelength is also affected by the conformation of the polymer chains.

Steady red-shifted absorption and emission of oligomers have been shown to occur as the

conjugation length increases [23-25]. Change of the polymer chain conformation can

also affect the conjugation length. An extended chain conformation results in a longer

conjugation length. Studies have already shown the emission wavelength as being

affected by solvents [26-29], blending with other polymers [30], and the evaporation

process [31] due to the change of the chain conformation. The quantum efficiency, life

time, and turn-on voltage were also seen as being affected by the morphology of

polymers[28, 32-34].

In this research, three different light-emitting polymers, including two conjugated

polymers, poly [2-methoxy-5-(2'-ethylhexyloxy)-1,4-phenylenevinylene] (MEH-PPV)

and poly[9,9-dioctylfluorenyl-2,7-diyl] (PFO), and one non-conjugated polymer, PVK,

were used. The research topics include film formation, optical properties, and the

degradation mechanisms of polymer and PLEDs. Chapter 2 will begin by providing the

general background of light-emitting polymers, including the physics of light generation

as well as the mechanism and fabrication of PLEDs. Second, previous research about the

relationship between polymer morphology and the PLED performance will be illustrated.

Spin-coating is the most general method for thin-film fabrication. During spin-

coating processes, Marangoni convection can lead to high surface roughness. Details of

the Marangoni effect is additionally delineated in chapter 2, and results of this

phenomenon on light-emitting polymer systems are shown in chapter 3.









In chapter 4, the rupture of polymer thin-films during spin-coating processes is

illustrated. Previous studies have shown the dewetting of polymer thin-film after the

annealing process due to the high interface tension between polymers and substrates.

Unlike these studies, we found the formation of rupture of polymer thin-films during

spin-coating processes. Thus, an alternate mechanism of this phenomenon is proposed.

Chapter 5 primarily shows the solvent effect on the optical properties of light-

emitting polymers. Due to the interaction between polymers and solvent molecules, the

chain conformation can be changed by dissolving in different solvents. The change of a

polymer conformation leads to the shift of PL of that polymer solution. According to

previous studies, the conformation of polymer chains after the spin-coating process is

affected by the polymer conformation in solutions, which means the PL of the polymer

thin film is also affected by the solvents. We also show the concentration and annealing

effects on PL in this chapter.

Chapters 6 and 7 present the influence of degradation of the electrode contacts in

PLEDs and the oxidation of light-emitting polymers respectively. The cause of contact

degradation is mainly due to the reaction between metal electrodes, like Ca, and water or

oxygen. The oxidation of PFO leads to an extra emission peak.

Chapter 8 is a summary and some suggestions for future work.














CHAPTER 2
REVIEW OF LITERATURE

2.1 Fundamentals of n-Conjugated Polymers

2.1.1 A Brief History of n-Conjugated Polymers

The first report about doped polyacetylene showing high conductivity in 1977 [35]

opened a new field in polymer applications. Electronic polymers show the same

feature-a backbone consisting of alternating single and double bonds. Polymers with

this bonding structure are called tr-conjugated polymers.

Before the 1990s, the main focus of this field had been studies of the mechanism of

charge transport and the synthesis of new conjugated polymers with higher conductivity

and stability. Since the first PLEDs were successfully fabricated in 1990 [1], the

application of semiconducting (undoped) forms of these polymers attracted lots of

attention. These applications include PLEDs, solar cells [36], sensors [37], and thin-film

transistors [38]. Figure 1 shows the structure of the conjugated polymers studied to

examine their conducting and semiconducting properties.

a) .,b) c) :,


d) e) f)






Figure 2-1. Structure of conjugated polymers: a) cis-polyacetylene; b) trans-
polyacetylene; c) polythiophene; d) poly(phenylene vinylene); e) poly(pyridyl
vinylene); f) poly(fluorene).









2.1.2 Photophysics

According to molecular orbital theory, a conjugated system contains 7t and 7t*

orbitals. By linking these conjugated molecules together, the disparate 7t and 7* orbitals

will merge into continuous bands, which is similar to the valence band and conducting

band in the semiconductor. The energy spacing between the conducting and valence

band is called the bandgap. For conjugated polymers, band gaps are determined by their

chemical structures (configurations) and physical structures (conformation).

When a photon is absorbed by conjugated polymers, an electron is excited from the

valence band to the conducting band and a hole is generated in the valence band. Several

possible excited states can generate after that. Generally, an electron-hole pair forms a

singlet exciton, which means the electron and hole are bound on the sample chain by their

electrostatic attraction [39]. Excitons migrate to sites with lower energy (i.e. segment

with longer conjugated length) before decaying and lead to red-shifted emission [40].

An excimer (short for "excited dimer") is a complex between the excited state of a

molecule and another molecule of the same type that is in its ground state [41]. The ideal

separation is in the range of 0.3 to 0.4 nm. The formation of excimers is accompanied by

a geometry distortion and leads to lower energy [41]. Because of the lower energy in the

excimer state, the emission is strongly red-shift. Different from excimers, aggregate

states mean that the electron is delocalized over two chains [42]. The delocalization then

leads to a lower energy state and a reduced PL efficiency. A lot of examples of

aggregates in conjugated polymers have been shown [27, 28, 43, 44]. Film samples show

strongly red-shifted emissions with much lower quantum efficiency compared to solution

samples.









Another related state is polarons. Formation of a polaron requires sufficient

distance between the electron and the hole such that they do not interact to form excitons

[45]. The electron and hole then separate onto different chains. Alternately, the electron

and hole may be captured by an impurity that provides a lower energy state. This process

is fundamental for the application of solar cells.

Since polaron pairs are non-radiative, the formation of these states results in

reduction of quantum efficiency of luminescence. Even if the excimers and aggregates

are radiative, the quantum efficiency also degrades because of a longer lifetime, which

allows for non-radiative decay. Moreover, the emission is strongly red-shifted. To

increase the quantum efficiency of luminescence, the formations of polaron pairs,

excimers, and aggregates have to be reduced. This goal can be achieved by changing the

chemical structures of polymers, including inserting another segment and shortening the

conjugation length [46], to reduce the interchain interaction. The interchain interaction

can also be reduced by using a polymer blend system [47] or by varying the solvents and

thermal treatments [48]. More details will be illustrated in the following section.

2.2 Polymer Light-Emitting Diodes (PLEDs)

Figure 2 shows the general structure of PLEDs. The light-emitting polymer is

sandwiched between an anode and a cathode. Indium-tin-oxide is the most common of

materials for anodes because of its high-work function and its transparency. Several low-

work function metals are used as cathodes, including Mg, Al, or Ca. After PLEDs are

biased, the electrons are injected from the cathode to the conducting band of the polymer,

and the holes are injected from the anode to the valence band of the polymer, as shown in

Figure 2-3.

















Cathode
Polymer.

Anode
Substrate

Figure 2-2. General structure of PLEDs.
Afterwards, the electrons and holes recombine in the polymer to form excitons.

The excitons then decay and emit light. Since the cathodes are not transparent, light can

only emit from the anode side of PLEDs.


Electron Injection


Hole


Figure 2-3. The mechanism of EL in PLEDs.









2.2.1 Solvent Effects

As noted in the previous section, the interchain interactions of conjugated

polymers are of great interests because the PL and EL strongly depend on the chain

conformation of these polymers.

Nguyen et al. [28] measured the PL of MEH-PPV dissolved in two different

solvents, chlorobenzene (CB) and tetrahydrofuran (THF), as shown in Figure 2-4. The

PL of MEH-PPV in the CB solution is red-shifted compared to the THF solution. From

dynamic light scattering, the study showed the MEH-PPV chains in the CB have more of

an extended conformation than in the THF. This result revealed that the degree of

polymer interchain interaction increases when the polymer is dissolved in a better solvent

and is in more of an extended chain structure.


S, I -


-0.6 /' I
MEH-PPV

0.4 CB

<0.2 %


425 500 575 650 725
Wavelength (nm)

Figure 2-4. Absorption and PL of MEH-PPV in CB and THF solutions.

They further proved [49] the morphology of the cast polymer film is related to the

polymer morphology in solution. In figure 2-5, the film coated from CB shows a red-

shifted emission compared to the film coated from THF at the same concentration and

spin-speed.



































550 600 650 700 750
Wavelength (nm)

Figure 2-5. a) PL of CB solution and thin films; b) PL of thin films coated from CB and
THF solutions.

That means the degree of interchain interaction in the cast film can be affected by

solvents [50].

Liu et al. [33, 34] also showed the relationship between polymer morphology and

device performance including turn-on voltage, quantum efficiency, and current injection

efficiency, as shown in Figure 2-6.














l T
MUH-PPV -*
r. f 6CBr m /CS -- '"
128Am ltI /
MeH-PPV
S- 130 nrm









0 1 2 3 4
BIor (v)


S
P




10"


1 15 2 2.5
Bks (v)


Figure 2-6. Characterization of PLEDs: a) I-V curve and b) B-V curve for DCB and THF
devices having similar thickness.

PLEDs prepared from the good solvent, dichlorobenzene (DCB), show a higher


current density and a lower turn-on voltage compared to devices from the poor solvent,


THF. This variance is due to the different chain conformation in the polymer thin-films.


Said result indicates that contacts between anodes and polymers are sensitive to the chain


conformation. Polymer chains with more extended conformation can improve the contact


between polymers and substrates, which lead to lower turn-on voltage and higher


efficiency of charge injection from electrodes.









2.2.2 Annealing Effects


550 600 650 700
Wavelenglh (nm)


3 xlO




' 2x10f


C
-J
0tixa


0x10 -
500


550 600 650 700
Wavelength (nm)


Figure 2-7. PL of MEH-PPV films annealed at different temperatures.

Thermal annealing also plays an important role in controlling the chain

conformation and the film morphology. Since the polymer chains in spin-coated thin









films are not in thermodynamic equilibrium, thermal annealing can assist in the

movement of polymer chains toward a thermodynamic equilibrium and enhance the

packing of polymer chains, which leads to the increase of interchain interaction.

Therefore, the PL of annealed thin films shows a red-shifted emission due to a generation

of more interchain interaction species such as aggregates and excimers, as shown in

Figure 2-7(a) [51].

In Figure 2-7(b), relative PL quantum efficiency decreases with an increase of the

annealing temperature due to the generation of interchain interaction. However, instead

of showing decreasing EL efficiency, Lee and Park [52] and Liu et al. [51] found the EL

efficiency is much enhanced when devices are annealed as shown in Figure 2-8.




A11CA.




106W' N




a Asc 1 pota nme
1r 3postmnaeled


0 5 10 15 20 25
Current (mA)

Figure 2-8. EL vs. current of the devices annealed at different temperature after Al
deposition.

The annealing at higher temperatures may improve the interfaces between

electrodes and polymers and improve the electron- or hole-injection efficiency, which









may lead to higher EL efficiency despite the generation of more chain-chain interactions.

Furthermore, Kim et al. [53] also showed that thermal annealing can improve the life-

time of PLEDs. The annealing of polymer films can improve the packing of polymer

chains and reduce the defects, which leads to higher thermal stability when an electric

field is applied.

2.2.3 Degradation of PLED Devices

The short lifetime and poor durability of PLEDs are critical issues for their

commercial applications. Some degradation mechanisms have been proposed for PLEDs,

especially for devices made using poly(phenylene vinylene) [54-57] and polyfluorene

[58]. The device lifetime may be limited by photo-degradation during its operation

[59,60]. Several other factors have been reported to increase the rate of degradation of

polymer materials, including oxygen exposure [61], photo-irradiation [62], and electrical-

field induced aging. [55,58,63]. In addition to the polymer itself, degradation of its

metallic cathode causes degradation of the device [63-67]. PLED devices have been

reported to exhibit "black spots," as shown in Figure 2-9.

When PLED devices are operated for two hours, a lot of small black spots emerge

and distribute randomly on the device as shown in Figure (a). These black spots then

expand when the operation of PLED devices continue. Figure 2-9 (e) shows that most of

this PLED device is covered by black spots and can not emit light after 40 hour

operation.








































)

Figure 2-9. Evolution of black spot formation in air: a) t = 2hrs; b) t = 10 hrs; c) t = 20
hrs; d) t = 30 hrs; e) t = 40hrs.

Black spots are non-emitting areas that severely reduce the light output. Several

mechanisms have been proposed to explain the formation of black spots, including metal

from the electrode migrating through the polymer layer under electrical stress. A metallic

short is created between the electrodes leading to large currents that carbonize the

polymer, thus forming a black spot [63-65]. Local heating due to large currents was

reported to also stimulate bubble formation due to reactions between metal electrodes and










polymers. Pinhole defects in the cathode are commonly founds at the center of black

spots. Indeed, pinholes could form either from particulate contamination during thermal

evaporation or from voids in the granular structure of the metal cathode thin films [64], as

shown in Figure 2-10.


Dark Spot Center
S|Ag
a 2 or Ca

b 4--- ** b EL
C(arhm I I Cabon IrTO

SII Substrate


I'I Non-ermssihe area
A radiant nng formed at optimllzai polymr Thckness
H Weakly emissive area -

Figure 2-10. Schematic diagram showing dark spot induced by pinholes.

By using uniformly sized silica micro particles Lim et al. made devices with size-

controllable pinholes [66]. The relationship between the growth rates of dark spots and

the pinhole size is shown in Figure 2-11.




12500- 7 6


g 2 0-
0 756 0040 r 4e

g d, j) 0____3_ ,0 00 ]
t 0 o00
o ; Ioo _..,,.."""*
2500 19Z000 -2**0**SS
F5W( o '3 10~~cr


500 1000 1500
t (sec.)


Figure 2-11. The linear growth of dark spot with various particles.


2000









The growth rates of dark spots are linearly correlated to the pinhole sizes, which

indicate the dark-spot formation is due to the pinholes on the protective layer, which

create a path for water or oxygen.

Kim et al. [67] also pointed out that the transport of water and oxygen through the

pinholes to reach the cathode/polymer interface can lead to oxidation of the metal

cathode and formation of a higher resistance poly(3,4-ethylenedioxythiophene)-poly(4-

styrenesulfonate) (PEDOT-PSS) layer, as shown in Figure 2-12.

Oxidation half-cell:

M M "I + ne



Reduction half-cell:





0 0

doped PEDOT dedoped PEDOT
(y 0.3)

Figure 2-12. Proposed electrochemical half-cell reactions showing the oxidation of metal
(M) and the reduction of doped PEDOT.

The oxidation of active metal and reduction of PEDOT generate non-conducting

materials near pinholes, which increase the local resistance dramatically and cut off the

current density. These non-emissive areas, specifically the dark spots, then cause a

reduction of device active areas and luminescence outputs.










2.3 Defects of Spin-Coated Polymer Thin-Film

The spin-coating process has been studied for many years and many models have

been proposed [70-73] to explain the mechanism of this process. Due to the complexity

of the spin-coating process itself, simplified assumptions have been made in these

models. Most studies have focused on the relationship between film thickness and

various parameters, including the solution concentration, viscosity, evaporation rate, and

spinning speed [70-75]. Some important factors, like the interaction between the

polymers and solvents during evaporation, have been ignored. Recently, the influence of

the type of solvents used on the structure of polystyrene thin films on Si wafer has been

reported [76,77]. Figure 2-13 and 2-14 show the surface structure of polymer thin-films

coated from solvents with a high-evaporation rate.











Figure 2-13. Polystyrene coated from T.F solution.
*F A





;_




A ,JS






Figure 2-13. Polystyrene coated from THF solution.

























Figure 2-14. Polystyrene coated from cyclohexane and acetone mixed solution.

These studies showed that polymer films which were coated from high-

evaporation-rate solvents exhibited a higher roughness surface due to the Marangoni

instability [78].

2.3.1 Benard-Marangoni Convection

In the classical thermocapillary Marangoni convection, rapid evaporation causes

local cooling at the surface. The temperature difference between the surface (Ti) and the

interface (T2) leads to a surface tension gradient and drives convective flows, as shown in

Figure 2-15.

T, TI

Air Ao High Y


Polymer Solution

ITO substrate T.


Figure 2-15. The Marangoni convection.









The formation of this instability is governed by the competition among surface

tension, thermal diffusion, and viscosity, which can be expressed by the Marangoni

number Ma:


Ma = / T)HAT Equation 2-1
pa

where (au / T) is the temperature derivative of the surface tension; AT is the

temperature difference between the bottom and surface of the liquid; H is the film

thickness; and ~ and a are the viscosity and the thermal diffusivity of the system,

respectively. The critical value of Ma which can trigger instability is around 80 [79].

However, the classical Marangoni convection is proposed on the assumption that the

temperature gradient is linear through the solution layer. During the spin-coating

process, the rapid evaporation near the solution surface leads to non-uniform temperature

gradients. The thermocapillary effects upon this development are expressed as:


Ma = ( / T)H 2VT Equation 2-2
fa

More recently Haas et al. [80] have determined the conditions that would lead to

thermocapillary convection during the spin-coating process. Due to the squared

dependence of H, the thermocapillary driven convection can be induced within a

sufficiently thick solution layer of high-evaporation rate. For thin-solution films, Ma

would be too small and thermocapillary-induced-convection thus becomes unstable.

Birnie [81] has pointed out that the concentration effects (solutocapillary) are more

important than the temperature gradient (thermocapillary) during the spin-coating

process. The driving force is a composition gradient rather than a temperature gradient

and is expressed as:









(Ber/ D8T)H 2
Ma = / T)H2VC Equation 2-3
Da

where C is the relevant composition variable and D is the diffusion rate of the component

driving the composition dependent surface-tension change. Evaporation of solvents

during the spin-coating process generates the composition gradient at the solution surface

and leads to convection. All of these instabilities induced by the differences of surface

tension are called the Marangoni effect.

2.3.2 Dewetting Phenomenon

Dewetting of polymer thin-films is another potential limitation for their use in

PLEDs. Although polymer thin-films can be deposited on non-wetting substrates via the

spin-coating process, these polymer films are usually unstable or metastable.

Perturbations, like heat loading, lead to a relaxation towards thermodynamic equilibrium.

Many studies have been done, especially for the system of polystyrene (PS) on silicon

[82-86]. In the beginning, the polymer films tear up and holes form, as shown in Figure

2-16. These holes then expand gradually. The dark part around the holes in these images

is the rim. Finally, the polymer films are totally dewetted to form polymer droplets,

which are in thermodynamic stability on the Si wafer, as shown in Figure 2-17 [86].

The general mechanisms [84,86,87] to describe the dewetting process due to heat

loading include spinodal dewetting (by the amplification of capillary wave fluctuation),

and nucleation (from defects) and growth.












































Figure 2-16. Micrograph of different stages of polystyrene dewetting on Si wafer: a) 15
mins and b) Ihr.





S.. W



S.. .


i U e .

*. -a ,


Figure 2-17. The droplets left on the Si wafer.










Recent research [89] also has shown dewetting when aqueous polymer solutions

were coated on glassy hydrophobic substrates. In these cases, the film dewetting

occurred during evaporation of the solvent. The unstable aqueous films broke up through

the nucleation of holes, and the growth of these holes led to a fingering pattern, as shown

in Figure 2-18.


(A)






C13)


m I m "i

Figure 2-18. Film structure of dewetted thin films: a) A dewetting pattern centered at a
defect (50 x 50 tm) and b) pattern at high magnification (5 x 5 tm).


-iT-
-T




Jim~














CHAPTER 3
MARANGONI EFFECTS ON POLYMER THIN FILM STRUCTURE

3.1 Introduction

So far, most studies of polymer thin-film morphology and its instability phenomena

have been limited to polystyrene deposited on silicon. No studies about the instability

during thin-film formation of light-emitting polymers have been reported. Since the

quality of polymer thin-films is a critical factor in the performance of polymer electronic

devices, it is necessary to understand the instability phenomenon in light-emitting

polymer systems. In this chapter, the focus is on the relationship between solvents and

the film structure of poly(n-vinyl carbazole) (PVK) on indium tin oxide (ITO) glass.

Containing the positive charge transporting carbazole pendant groups [90], PVK is now

extensively used as the hole-transport layer in low-molecule weight organic light-

emitting diodes (OLEDs) [91]. Recent studies have also shown that PVK is a candidate

for blue polymer light-emitting diodes (PLEDs) [92,93]. These similar results can also

be observed by using poly(2-methoxy-5-(2'-ethyl-hexyloxy)-l,4-phenyl vinylene)

(MEH-PPV). MEH-PPV, which is a common light-emitting polymer, has attracted much

attention recently due to its high solubility in many organic solvents.

In this chapter, the instability phenomenon induced by Marangoni effects on the

light-emitting polymer thin films is developed. The other instability phenomenon,

rupture of polymer thin-films during the spin-coating process, will be illustrated in

chapter 4. The chemical structures of PVK and MEH-PPV are shown in Figure 3-1.









a) )-













Figure 3-1. Chemical structure of a) PVK and b) MEH-PPV.



3.2 Experiment

3.2.1 Sample Preparation

ITO coated glass (CG-51IN, ITO layer = 120-140 nm, Delta Technologies,

Limited) was cleaned in ultrasonic baths of deionized water, ethanol, and acetone

respectively for 15 minutes each before being used. PVK (Mw = 90,000, Polydispersity

= 2.40, Scientific Polymer Product) and MEH-PPV (Mn = 40,000-70,000, Aldrich) were

used as received. PVK solutions with fixed concentration (10 mg/mL) were prepared.

MEH-PPV solutions with lower concentrations (2-7 mg/mL) were prepared due to its

lower solubility. All solutions were filtered (0.2 [tm PTFE filter, Fisher) before being

used. The physical properties and Mark-Houwink parameters for the five solvents used

are shown in Tables 1 and 2. Spin coating (time = 30 seconds) was conducted in ambient

laboratory air.
















Table 3-1. Characteristic data of solvents at room temperature.
Solvent Vp 1) rI a p 0

(mmHg) (Debye) (mPa s) (mN/m) (g/cm3)

THF 129 8 1.75 0.456 26.40 0.889 *M

Chloroform 160 11.6 1.04 0.537 26.67 1.492 0.8

Benzene 75 5.1 0 0.604 28.88 0.874 0.18

Toluene 22 2.24 0.37 0.560 27.93 0.865 0.05

TCE 8 0.65 1.32 1.437 35.58 1.586 0.3

Vp: vapor pressure; u: relative evaporation rate (BuAc=l); [a: electric dipole moment; fl:
viscosity; y: surface tension; p: density; 0: solubility (g/100g water), *M: miscible





Table 3-2. The Mark-Houwink parameters for PVK at 250C.
Solvent K a

THF 1.55 0.65

Chloroform 1.36 0.67

Benzene 3.35 0.58

Toluene 7.72 0.50

TCE 1.29 0.68









3.2.2 Characterization

3.2.2.1 Atomic Force Microscopy (AFM)

The film structure and root mean square (RMS) roughness were measured using

tapping mode AFM (Digital Instruments Nanoscope III). A single silicon probe with a

tip radius of 5-10 nm was used. The cantilever was oscillated at or near its resonance

frequency between 300 to 350 kHz. The set point for the feed-back control was between

0.7 and 4V. Height images, which represent the topography of the sample surface, were

recorded with a scan range from 10 to 100 tm. All AFM images were obtained in

ambient air.

3.2.2.2 X-Ray Reflectivity (XRR)

XRR measurements were performed with a Panalytical X'Pert MRD System. A

ceramic filament tube with a Cu-target was used as the X-ray source (k = 1.54A). The

scan rate was 0.02 degree/sec. A parallel plate collimator was attached to the detector to

improve the peak-to-noise ratio. The reflectivity curves provide information about the

film structure perpendicular to the surface [97,98]. The acquired XRR spectra were fitted

using the WinGixa software package from Panalytical to obtain the film thickness,

density profile, surface and interface roughness.

3.3 Results and Discussion

3.3.1 High-Evaporation-Rate Solvents. Table 3-1 shows that THF, chloroform, and

benzene have a relatively higher evaporation rate than toluene and TCE. For PVK, THF

and chloroform are good solvents (a z 2/3), whereas benzene is a poorer solvent, as

shown in Table 3-2.

Figure 3-2 shows the film structure of PVK thin films prepared from THF and

chloroform solutions.











-I
2001pm 200pm

















(c) (c>

Figure 3-2. Film structure of polymer thin-film observed with optical microscopy: (a)
prepared from THF solution, center area; (b) prepared from THF solution,
away from center area; (c) prepared from chloroform solution, center area; (d)
prepared from chloroform solution, away from center area. Length of scale
bar 200am.

The film structure from benzene is similar to the film from chloroform solution.

The film structure near the sample's center (Fig. 3-2(a) and 3-2(c)) shows a cellular

structure, which is similar to Benard cells [99] caused by Marangoni convection. The

film structure away from the center shows striations (Fig. 3-2(b) and 3-2(d)) which

commonly occur in spin-on thin films prepared with high-evaporation-rate solvents.

Comparing Fig. 3-2(a) and 3-2(b) or 3-2(c) and 3-2(d), the film prepared from the THF

solution shows a much higher contrast than from the chloroform solution, which indicates

that the polymer film from the THF solution is rougher than from the chloroform solution

Figure 3-3 shows AFM images of thin film surfaces prepared from chloroform.






















(a)















(b)

Figure 3-3. 3D tapping mode AFM image of PVK film surface prepared from (a)
chloroform solution, cell structure near center area; (b) chloroform solution,
striation structure away from center area.
These images show a clear cell structure in the center part of the polymer film

(Fig. 3-3(a)) and striations away from the center (Fig. 3-3(b)) although the polymer films

are locally smooth. These results show that Marangoni convection has changed the

topography of these polymer films. The structure of polymer thin-films from the benzene









solution is similar to the films from chloroform. Different from the film structure of

polymer films prepared from the THF solution, the polymer films prepared from the

chloroform and benzene solutions exhibit continuous film surfaces instead of the

formation of holes or polymer droplets. These results show that polymer films from

chloroform and benzene, which are immiscible with water, do not show polymer

dewetting during the spin-coating process.

We can understand these results by considering the effect of Marangoni convection.

During the spin-coating process, the evaporation of the organic solvent induces the

composition and temperature gradients at the same time in the thin solution film, as

shown in Fig. 3-4.

T, T,

Air A o High o



Polymer Solution C
ITO substrate T,


Figure 3-4. The Marangoni convection model (including composition-gradient-driven
and temperature-gradient-driven processes) for the spin-coating process. T1
and C2 represent the temperature and solution concentration near the surface
of the solution layer; T2 and Ci represent the temperature and solution
concentration near the interface between the solution layer and the ITO
substrate.

The evaporation of the solvent leaves the solution layer with a higher polymer

concentration near the surface. In addition, the evaporation can also induce cooling near

the surface. Because surface tension is a function of solution concentration and









temperature, the concentration and temperature gradients will change the surface tension

during the spin-coating process, inducing flow near the surface of the thin solution film.

The high-surface-tension area pulls the solution from the low-surface-tension area,

leading to Marangoni convection. According to the reviews in Chapter 2, the

concentration effects will be more important for this work due to the film thickness that

results from spin coating. The thickness in our system is smaller than 1 [am, which leads

to a small Ma in equation 2-2 and unstable thermocapillary-induced convection. The

composition-gradient-driven convection from equation 2-3 dominates in our system.

Figure 3-2 shows that this is because the topography of polymer films from THF is

not only affected by the Marangoni convection but also, as illustrated further in chapter 4,

the rupture of the polymer thin film during the spin-coating process.

3.3.2 Low-Evaporation-Rate Solvents.

From Table 1, toluene and TCE show relatively low-evaporation rates. The center part of

the film coated from TCE under the optical microscope is shown in Figure 3-5.


200Up












Figure 3-5. Film structure of PVK thin film observed with optical microscopy prepared
from TCE solution, center area.

The image shows the film coated from TCE is very smooth without cell structure

and striations. Films coated from toluene also show the same film structure as those










coated from TCE. The image from tapping-mode AFM (Figure 3-6) also shows smooth

surfaces.










0.0 r,









0 2.5 5.0 7.5 LO.1,



Figure 3-6. 2D tapping mode AFM image of PVK surface prepared from toluene
solution.

These results show that Marangoni convection is minimal in low-evaporation-rate

solutions including both a good solvent (TCE) and a poor solvent (toluene). The low-

evaporation rate decreases the composition gradient near the surface of the solution film,

which leads to a lower Ma value based on equation 2-3, and therefore no Marangoni

convection occurs during the spin-coating process. Figure 3-7 shows the RMS roughness

measured by AFM.











7
S6
SChloroform
5- 5
a 4 Benzene
3
2
TCE
S1 Toluene
0
0 2 4 6 8 10 12

Evaporation rate (BuAc=l)


Figure 3-7. RMS Roughness plotted as a function of evaporation rate of four solvents.

The roughness of polymer films from the chloroform and benzene solutions shown

here was measured near the cell-structure area. The roughness of the striations is slightly

higher than the cell-structure area. These results clearly show the dependence of surface

roughness on evaporation rate. During spin coating, the solutions with a high

evaporation rate lead to high surface roughness due to the Marangoni effect.

For making MEH-PPV polymer thin-films, chlorobenzene (CB) is substituted due

to the low solubility of MEH-PPV in toluene. The boiling point of CB (132C) is higher

than toluene, which indicates CB is also a low-evaporation-rate solvent. The structures

of MEH-PPV thin films coated from THF, chloroform, CB and TCE as imaged by optical

microscopy are shown in Figure 3-8. The surface topography of MEH-PPV thin films

from chloroform and TCE are shown in Figure 3-9.

From these results, the structures of MEH-PPV thin films show the same trend as

PVK thin films, which are controlled by the Marangoni effect and dewetting during the

spin-coating process.



















a) b)


c) f)

Figure 3-8. MEH-PPV Film structure of polymer thin-film observed with optical
microscopy: (a) prepared from THF solution, center area; (b) prepared from
THF solution, away from center area; (c) prepared from chloroform solution,
center area; (d) prepared from chloroform solution, away from center area; e)
prepared from CB solution; prepared from TCE solution.








100 mn11111


I
0

50O~uu


a)


10 mu1
[7--


b)


1 Opum


Figure 3-9. 2D tapping mode AFM image of MEH-PPV film surface prepared from (a)
chloroform solution, cell structure near center area; (b) TCE solution, smooth
surface near center area.
3.3.3 XRR Measurement. Typical XRR spectra acquired from the polymer films are
shown in Fig. 3-9, while results from the simulations, including RMS roughness,


thickness, and density, are shown in Table 3-3.




















TCE
4-'






0) \ ', Toluene
0
_J
I \ It^'^ "^ '" I

w ft!^ *~'^ k Benz ne


Chloroform

1 2 3 4 5
0 degree



Figure 3-10. X-ray reflectivity spectra acquired from PVK thin films prepared from
chloroform, benzene, toluene, and TCE solutions. Experimentally measured
data are shown as a solid line; fitted results are shown as a dotted line.


Table 3-3. Characteristics of PVK films obtained from XRR measurements.
Solvent RMS Roughness Density Thickness

(A) (g/cm3) (A)

Chloroform 24 0.88 151

Benzene 18 0.96 852

Toluene 9 1.23 811


9 1.29


TCE









The fitting results (dotted line) are in favorable agreement with the measured

spectra (solid line). The surface roughness values displayed in Table 3-3 show the same

trend as the values from AFM measurements.

The polymer films deposited using high-evaporation-rate solvents, such as

chloroform and benzene, lead to significantly higher surface roughness than using low-

evaporation-rate solvents, like toluene and TCE. The densities of polymer films

deposited using the low-evaporation-rate solvents, toluene and TCE, showed values

similar to solid bulk PVK (1.2g/cm3). One possible explanation for a lower density of

polymer films from the chloroform and benzene solutions is the high surface roughness.

The other possibility is due to the trapping of defects or free volume in polymer films

during the fast evaporation of the solvent and the presence of some gas trapped near the

surface of the polymer films due to the Marangoni convection. The film densities did not

show any obvious differences between a good solvent (TCE) and a poor solvent

(toluene).

Comparing the thickness of the polymer films, the films from the chloroform and

TCE solutions show lower thickness compared to other solvents. According to previous

studies [73], the spin-coating process can be divided into three phases. During the first

phase, most of the solution is flung from the substrate. Only a much thinner and uniform

film is left. At this time, the spin speed is high and shear thinning may occur. The film

structure is dominated by the inertia of the solution. The second phase begins when the

inertial effects become negligible and the film thickness is determined by the balance of

viscosity and centrifugal forces. In the final phase, further fluid is lost due to solvent

evaporation. From this model, solutions with higher viscosity will lead to thicker









polymer films. From Table 3-2, PVK dissolved in ideal solvents exhibits higher viscosity

than in poor solvents ([r]benzene ~ [rlitoluene < [Tllchloroform < [T]TCE). However, the higher

density of chloroform and TCE lead to higher centrifugal forces and cause much more

solution to be flung from the substrate in the first and second phases during spin coating.

Our experimental results show that films from benzene and toluene are thicker than from

chloroform and TCE. The lower viscosity of the benzene and toluene solutions would

lead to thinner films, while the lower density would lead to thicker films; comparison of

these predictions to the experimental results indicates that the density of solvents

(centrifugal force) dominates the polymer film thickness in this system.

The XRR results from MEH-PPV films are shown in Figure 3-10 and simulated

values are shown in Table 3-4.










S\7 mg/mL, TE
C


Io \ 7 mg/mL, Chlorobenzene
0)

7 mg/mL, Chloroform

5 mg/mL in Chloroform

2.5 mg/mL, Chlorfor
1 2 3 4

Figure 3-11. X-ray reflectivity spectra acquired from MEH-PPV thin films prepared from
chloroform, chlorobenzene, and TCE solutions. Experimentally measured data
are shown as a solid line of light color, and fitted results are shown as a solid
line with dark color.









Table 3-4. Characteristics of MEH-PPV films obtained from XRR measurements.
Solvent Concentration Thickness Density RMS Roughness

(mg/mL) (A) (g/cm3) (A)

TCE 7 227 1.32 13

Chlorobenzene 7 391 1.26 15

Chloroform 7 379 0.72 40

Chloroform 5 216 0.70 23

Chloroform 2.5 149 1.28 16



From the results shown in Table 3-4, the MEH-PPV films coated from TCE and

CB show lower roughness and higher density than the films coated from chloroform with

the same polymer concentrations (7mg/mL). Compared to the film coated from

chloroform solutions with different concentrations, films coated from chloroform

solutions with lower concentrations have lower thickness. The thinnest film, which was

coated from chloroform solution with 2.5 mg/mL concentration, has a similar roughness

and density as the films coated from TCE and CB.

We can see these results are coincident to the results of PVK films when MEH-

PPV films are coated from different solvents with the same concentration. The thinnest

film coated from chloroform solution has a similar roughness as films coated from TCE

and CB, which indicates Marangoni convection is not induced in this system. From

equation 2-3, the Ma number is proportional to the height squared of the thin film. It

indicates that when films are too thin, Marangoni convection could not be induced even if

solvents with high-evaporation-rate are used.









3.4 Conclusion

The polymer film structure is determined by the dynamics of the spin-coating

process, which is controlled to a large effect by the solvent properties. The chloroform

and benzene solutions, which have a high-evaporation rate, lead to Benard cell and

striation morphology on polymer films and result in high-surface roughness due to

composition-gradient-driven Marangoni convection. The film structure of polymer films

from good solvents and poor solvents did not exhibit obvious differences, which shows

the interaction between polymer chains and solvents is not an important factor in the

dynamic instabilities. During the spin-coating process, the solvents with higher density

have higher centrifugal forces and lead to more solvent being flung from the substrate

and therefore thinner polymer films.

We expect these differences in film structure to have an important effect on PLED

device performance, depending on the solvent used. In films from THF solution, the

large-scale inhomogeneities in the film will lead to inhomogeneous light emission (More

details will be illustrated in Chapter 4). In addition, according to previous studies

[100,101], one of the intrinsic degradation mechanisms of OLED and PLED devices is

due to the diffusion of the metal from the electrode into the organic layers and the

quenching of luminescence in the emission zone. In our studies, the polymer films from

the chloroform and benzene solutions showed higher roughness and lower density due to

Marangoni convection. We believe the inhomogeneous polymer surface could affect the

durability of the PLED devices by enhancing the possibility of metal diffusion into the

emissive layer.














CHAPTER 4
RUPTURE OF POLYMER THIN FILMS DURING SPIN-COATING PROCESS

4.1 Introduction

As noted in chapter 3, this chapter focuses on another instability phenomenon-the

rupture of polymer thin-films. Although ruptured thin films show a similar film structure

as dewetted thin films, the mechanism is different from the dewetting phenomenon

induced by heat loading after the spin-coating process, because the rupture of polymer

thin films described in this chapter occurs during the spin-coating process. In this

chapter, results show that the water in THF is an important factor for the rupture of

polymer films. Two polymer systems, PVK and MEH-PPV, are used in this chapter. As

noted in chapter 2, MEH-PPV has attracted much attention recently due to its high

solubility in many organic solvents, which is helpful for investigating the relationship

between chain conformation and optical properties. In this chapter, MEH-PPV is used to

study the influence of polymer conformation when ruptured polymer films are formed.

The effect of film rupturing on the performance of devices is also illustrated.

4.2 Experiment

4.2.1 Sample Preparation

Cleaned indium tin oxide (ITO) coated glass (Delta Technologies) was used as the

substrate. PVK solutions with fixed concentration (10 mg/mL) were prepared. MEH-

PPV was dissolved in THF (Aldrich) at three concentrations (2.5, 5, and 7 mg/ml). These

solutions were filtered (0.2 atm PTFE filter) and spin-coated on ITO glass at 1000, 3000,

and 5000 rpm. "Dried" THF means that THF was dried by molecular sieves (4A, 8-12









mesh, Aldrich) overnight before being used. To induce dewetting, 2-5% (v/v) water was

added to the polymer solutions. Poly(3,4-ethylenedioxythiophene)-poly(4-

styrenesulfonate) (PEDOT-PSS, Bayer) water solution was filtered before being used.

4.2.2 Characterization

The topology and structure of polymer films were observed by tapping-mode AFM

(Digital Instruments Nanoscope III). The surface root-mean-square (RMS) roughness

was calculated by the software package included with the AFM. PL spectra were

obtained in ambient air with an excitation energy of 490 nm. The emission was collected

by a monochromator (Oriel instrument, Electo-optical systems).

4.3 Results and Discussion

Figure 4-1 shows SEM images of a polymer thin film prepared from THF.



















(c) (d)


Figure 4-1. SEM image of polymer prepared from THF solution: (a) center area with a
low magnification, length of scale bar 200[am; (b) away from center area with
a low magnification, length of scale bar 10tam; (c) center area with a high
magnification, length of scale bar 200[am; (d) away from center area with a
high magnification, length of scale bar 10tam.









The images with high magnifications (Fig. 4-1(c), 4-1(d)) show obvious rupturing

of the film. Near the center of the film (Fig. 4-1(c)), many holes formed and few polymer

droplets formed inside these holes. Away from the center area, striations formed

(running from top-left to bottom-right in Fig. 4-1(b)), and polymer droplets formed along

with the striations (Fig. 4-1(d)). Compared with the film structure in chapter 3, only

polymer films prepared from THF show the ruptured thin film structure. Polymer films

from other solutions showed smooth and continuous surfaces. To investigate the basis of

the rupture of polymer films during the spin-coating process from the THF solutions, we

compared the properties of these five solvents and found only THF shows good

miscibility with water (Table 3-1). Since the entire process of solution preparation and

spin coating were done in ambient air, it was believed that a small amount of water was

dissolved in the THF solution during the spin-coating process even when THF was dried

previously. To test the influence of water on the thin-film structure, several THF

solutions with different concentrations of water were prepared for film deposition. The

SEM images of the film structures are shown in Figure 4-2. Figure 4-2(b) shows the

structure of polymer films prepared from dried THF. It shows that many holes formed on

the surface without the presence of polymer droplets. This ruptured film structure is

similar to the beginning of polymer dewetting due to heat loading, although in this case

the rupture occurs due to evaporation of the solvent (discussed further below) rather than

the application of heat. As the water content increased, the structure changed from a

continuous film with holes to isolated polymer droplets (Fig. 4-2(d)). Finally, polymer

droplets replaced the entire polymer film, which is similar to the final stage of polymer









dewetting (Fig. 4-2(f)). These results indicate that the degree of polymer rupturing is

related to the concentration of water.


teJ


Figure 4-2. SEM image of polymer thin-film prepared from THF solution with different
H20 contents: (a and b) THF dried by molecular sieve for 1 day with a low
and high magnification, length of scale bar 200 and 10 am, respectively; (c),
(d) 10L of H20 added in 5 mL of THF with a low and high magnification,
length of scale bar 200 and 10 am respectively; (e), (f) 100al H20 added in
5mL of THF with a low and high magnification, length of scale bar 200am
and 10 am, respectively.

Different from the dewetting phenomenon induced by the heat loading after the

spin-coating processes, this polymer rupturing happened during the spin-coating process.

Two proposed mechanisms, spinodal dewetting or nucleation and growth, have been used

to explain the relaxation processes of metastable or unstable polymer films toward









thermodynamic equilibrium with heat loading. The polymer film rupturing during spin

coating is dynamic in nature and cannot be explained by these two mechanisms. Gu, et

al. [89] have shown the dewetting of aqueous polymers coated on hydrophobic substrates.

Since the ruptured film structure only occurs when some water is added to the THF

solution, we need to test if the effect between water and ITO is the cause for this ruptured

film structure. We conducted contact angle measurements of water on ITO and found a

contact angle around 90, indicating that our ITO substrate is hydrophobic. To test the

effect of water during spin coating of a water-soluble polymer on the ITO substrate, we

studied the surface structure of a PEDOT-PSS thin film spin-coated from water solution

(Fig 4-3).











(a) (b)


Figure 4-3. PEDOT-PSS film structure on ITO substrate observed with optical
microscopy: (a) low magnification, length of scale bar = 200[am; (b) high
magnification, showing presence of a hole, length of scale bar = 20[am.

Although there are some holes on the PEDOT-PSS film, the film structure is totally

different from the PVK film from the THF solution shown in Fig 4-1. Recall, as

described previously, that the PVK thin films spin-coated from other water immiscible

organic solvents, like chloroform or chlorobenzene, show smooth surface. These results









indicate the substrate effect is not the reason for the formation of the ruptured film

structure.

In our system, polymers are dissolved in THF/water solvent. Comparing the

properties of water and THF, water has a higher surface tension (72.8 dynes/cm) than

THF (24.8 dynes/cm). Furthermore, polymers, including PVK and MEH-PPV, show

poor interaction with water and cannot dissolve in water. Therefore, the process of fast

evaporation in this system is very complicated and can be analyzed as follows: in the

very beginning, during the spin-coating process, the evaporation of THF induces

Marangoni convection which will lead to thicker and thinner parts around the whole

polymer film. Due to the higher evaporation rate of THF than water, the solution layer

near the surface has a higher water concentration and leads to a higher surface tension

gradient, which can potentially drive a substantial convection. Additionally, the

precipitation of PVK or MEH-PPV can potentially occur at the same time due to the

incompatibility between the water and polymers. These phenomena could be important

factors for the formation of ruptured thin films, but the details of the rupture mechanism

are still not clear. Two possible mechanisms for the film rupturing during the spin-

coating process are proposed and discussed further in chapter 8. In the last stage the thin

film develops a few holes in the thick area and small polymer droplets in the thinner area

on the ITO substrate.

The water-induced dewetting phenomenon was also studied by Bonaccurso et al.

[102], who found dewetting of ultrathin polystyrene films induced by saturated water

vapor after spin-coating. More recently, Muller-Buschbaum et al. [103] also showed that

diblock copolymers dewetted on silicon during the spin-coating process. These results






46


show the interaction between the polymers, solvents, and substrate plays a significant role

in the dewetting phenomenon and the rupture of polymer thin films.

Figure 4-4 shows the optical microscopy and the SEM images of dewetted MEH-

PPV film structures.

40 Lll


(b)


Figure 4-4. Structure of ruptured thin films: a) Optical microscopy image of MEH-PPV
thin film prepared from 5% H20 in THF, striation structure away from center;
b) SEM image of same film with higher magnification.


20J lull









Figure 4-4(a) shows a similar striation structure as PVK films under an optical

microscope. Figure 4-4(b) shows the polymer films are ruptured and a lot of polymer

droplets formed on the substrate. The AFM image of MEH-PPV thin films prepared

from dried THF solution and 5% water in THF solution lead to different surface

structures, as shown in Figure 4-5.




150 10l






I ti
Slull, .-' ( *
707 c } 500:,d}



-70. -500!

Figure 4-5. Height contrast of polymer thin-films from (a) dried THF solution with
polymer concentration of 5 mg/ml; (b) 5% water in THF solution; (c) height
profile of film along the line indicated in (a); (d) height profile of film along
the line indicated in (b).

The thin films from dried THF (Fig. 4-5(a)) are much smoother than from

THF/water solution (Fig. 4-5(c)). Film structures from THF/water solution consist of

polymer droplets instead of a continuous polymer thin film. The height profile of the

polymer film from the THF/water solution clearly shows the presence of polymer

droplets on the ITO substrate (Fig. 4-5(d)). The dewetting ofMEH-PPV films is due to

the incompatibility of the water and MEH-PPV. During the spin-coating process, THF

evaporates faster than water due to its high-evaporation rate, leading to phase separation









into a water-rich phase and a polymer-rich phase. After the THF and water completely

evaporate, polymer droplets are formed on the ITO substrate (Fig. 4-5(d)).

The RMS roughness of polymer films from dried THF solutions of different

polymer concentrations are similar to each other (<5nm) as shown in Table 4-1.

Table 4-1. Roughness of MEH-PPV polymer films prepared from the dried THF and
THF/water solutions
Concentration Spin Speed Roughness (nm)
(mg/mL) (rpm) Dried THF 2% H20 5% H20
7 1000 2.7 32.2 *
3000 2.2 14.3 *
5000 2.3 12.0 *
5 1000 3.1 15.8 32.1
2.5 1000 2.0 8.1 11.7



For polymer films spin-coated from the THF/water solutions, the film from a high-

polymer concentration solution (7 mg/ml) shows much higher roughness. The roughness

also increases as the water concentration increases in the polymer solutions at the same

polymer concentration (5 mg/ml). In Table 4-1, the roughness of polymer films from the

dried THF solution with different spin-speed is similar. After 2% water is added, the

polymer film from the lowest spin-speed shows the highest roughness. According to the

model of the spin-coating process described previously, higher concentration and low

spin-speed lead to higher thickness. Based on our results, roughness increases

dramatically as the thickness of the polymer film increases.

The PL measurement of these polymer films are shown in Figure 4-7.






49




(a) 7 mg/ml
7 mg/ml + 2% H20
S/- ,/5 mg/ml
-.^ \ 5 mg/ml + 2% H20
S'5 mg/ml + 5% H20
g -----
(W II

2 mg/ml "*..
'. | 2 mg/ml + 50 d'6


520 540 560 580 600 620 640 660 680 700
Wavelength (nm)


(b) 1000rpm; dried THF
S 1000rpm; 2% HO2 added

3 3000rpm; dried THF
C_ -.3000rpm; 2% H20 added

c. '' s,. 5000rpm; dried THF

/ /
000rpm; 2% H2O added -'..


520 540 560 580 600 620 640 660 680 700
Wavelength (nm)



Figure 4-6. PL spectra of MEH-PPV polymer thin-films.

Polymer films with higher thickness caused by higher polymer concentration (Fig.

4-6(a)) or lower spin-speed (Fig. 4-6(b)) show higher PL intensity. Comparing the PL of

non-dewetted and dewetted polymer films, the addition of water not only lowers the PL

intensity, but also leads to red-shifted PL peaks. The red shift is also related to the









thickness of the polymer films. The polymer films with higher thickness (higher

concentration or lower spin-speed) show more red shift (Fig. 4-6(a) and (b)) and the

increase of water concentration also leads to more red shift (Fig. 4-6(a)). From previous

studies, the red shift of PL is due to the enhancement of chain interaction and the

formation of aggregates. To identify the cause for this phenomenon in our system, we

measured the PL of MEH-PPV in THF solution (7 mg/ml) with different concentrations

of water, as shown in Figure 4-7.



Dried THF
1%HO

3% H O

22
c 5% HO

S7.5%H20 10% H,0
12% H
-.'" 15% H20








520 540 560 580 600 620 640 660 680 700
Wavelength (nm)

Figure 4-7. PL spectra of MEH-PPV in THF solutions (7 mg/ml) with different water
concentrations.

The emission (-550 nm) from polymer solutions does not show a red shift when

water concentration is increased up to 7.5%. When the water concentration is higher than

7.5%, an additional lower energy emission at 590 nm appears due to the precipitation of

MEH-PPV in the THF solutions. These two separate emissions indicate two different

polymer-chain conformations. According to the description by Hu et al. [104] stiff-chain









conjugated polymers can exist in either a non-collapsed defect-coil conformation or a

collapsed defect-cylinder conformation. The collapsed conformation leads to a lower

energy emission due to the enhancement of chain-chain interaction and the formation of

aggregates [105]. Therefore, the emission from 550 nm and 590 nm are assigned to the

non-collapsed defect-coil conformation and the collapsed defect-cylinder conformation

respectively [105], and thus the precipitation of IMEH-PPV causes a coil-cylinder

transition. Comparing Figure 4-6 and Figure 4-7, the emission (-570 nm) of the thin

films is located between the emission from solution (550 nm) and precipitate (590 nm).

This indicates that the red-shifted emission from polymer thin films is mainly due to the

enhancement of interchain interaction [106] of polymer chains in the coil conformation

instead of the transition of polymer conformation from the coil to the cylinder. In turn,

the results show that the polymer conformation in the thin films is dominated by the

dynamics of spin-coating rather than thermal equilibrium. Due to the fast evaporation of

solvents during the spin-coating process, most polymer chains remain in the coil

conformation. Some polymer chains collapse to the cylinder conformation. Polymer

chains with the transition conformation between the stable coil and cylinder conformation

could also exist in the polymer thin film. The polymer chains with this complex

conformation lead to a single and broad emission band shown in Figure 4-6. Since the

PL of the polymer solutions does not alter with a change in the water content before

precipitation occurs (Fig. 4-7), the red-shifted emission of the polymer films from the

THF/water polymer solutions (Fig. 4-6) is proposed to occur during the formation of

polymer thin films. The details are discussed as follows:









According to our model, due to the incompatibility of MEH-PPV and water,

precipitation of MEH-PPV occurs during the spin-coating process. Higher

concentrations of water can generate higher content of precipitation of MEH-PPV in thin

films. The precipitate has a cylinder conformation, which has a higher interchain

interaction and leads to a red-shifted emission. Because of the fast evaporation rate

during spin coating, the polymer chains cannot be transferred to the cylinder

conformation in thermodynamic equilibrium like the process of precipitation formation

shown in Figure 4-7. Table 4-2 shows the emission peaks as illustrated in Figure 4-6.



Table 4-2. PL of MEH-PPV polymer films prepared from the dried THF and THF/water
solutions
Concentration Spin Speed Emission peaks (nm)
(mg/mL) (rpm) Dried THF 2% H20 5% H20
7 1000 573 583 *
3000 572 577 *
5000 572 573 *
5 1000 572 573 577
2.5 1000 571 571 572
*: Precipitate generation.



The results show that the red shift of these emission peaks is correlated to the water

content in the THF solutions. With increasing water content, the emission peaks are

more red-shifted and the emission peaks are located between 550 nm and 590 nm. These

results are coincident to the proposed model for the formation of ruptured thin films.

Finally, we demonstrate the influence of this dewetting phenomenon on the

performance of PLED devices. Figure 4-8(a) shows the layer structure of this PLED







device. A polymer thin-film coated from 2% water in the THF solution (dewetted
polymer thin-film) was used as the light-emitting layer. Figure 4-9(b) shows clear
striation structure after this PLED device was turn on.

Ca /Al



SPolymer
ITO
Glass
(a)




.. ... ....








(b)

Figure 4-8. PLED based on ruptured thin films: (a) The layer structure of PLED device.
(b) The light-emitting output after the device is turned on.
4.4 Conclusion
In conclusion, we have found that the rupture of polymer films is related to the
water in THF. The rupture of the polymer solution could be related to the change of the









surface-tension gradient caused by the increased content of water near the solution

surface and the incompatibility between the water and polymers, but the workings of this

mechanism are still not so clear and need further study. Compared to smooth polymer

films, the ruptured polymer films show a more red-shifted emission. We also observe

that the conformation of polymer chains in solution and precipitation are of the un-

collapsed and collapsed conformation respectively. The polymer chain undergoes a coil-

cylinder transition during the process of precipitation. Therefore, polymer films coated in

the THF solution with higher concentrations of water can lead to more polymer chains

that have a cylinder conformation and red-shifted emission. These results point out that

the water effect is a critical factor in the performance of polymer thin-film devices.














CHAPTER 5
SOLVENT, CONCENTRATION, AND ANNEALING EFFECTS ON
PHOTOLUMINESCENCE OF POLYMER SOLUTIONS AND THIN FILMS

5.1 Introduction

As noted in chapter 2, the PL and EL of conjugated polymer thin films is related to

fabrication processes, including solvents, concentrations, spin-speeds and annealing

processes. This chapter focuses on the study of the relationship between the PL and

chain conformation in MEH-PPV and PFO.

Some details of the relationship between the PL and the chain conformation in

MEH-PPV are illustrated in chapter 2. From light-scattering experiments [28], MEH-

PPV shows a more-extended chain conformation in the "good solvent," CB, than in the

"poor solvent," THF. Polymer chains with a more-extended conformation lead to a red-

shifted emission due to longer resonance length. Therefore, the PL of MEH-PPV in CB

shows a red-shifted emission compared to the PL of MEH-PPV in THF solutions.

Grell et al. [13,14] who worked on the PFO system, define "good solvents" and

"poor solvents" according to solubility parameters. Chloroform and THF are "good

solvents" because of their higher solubility in PFO. Toluene is a "moderately good

solvent" because of its slightly lower solubility, while cyclohexane is a "poor solvent"

because of its low solubility. According to the results, PFO in poor solvents shows an

additional red-shifted peak. Their study proposed that this emission is due to a more

compact polymer chain conformation in poor solvents, which leads to the generation of

aggregates and an additional emission peak. Comparing the results from these two









research groups, the controversy on the definition of "good solvents" and "poor solvents"

exists, as well as the explanation of solvent effects on photoluminescence.

In this chapter, the first part focuses on the solvent effect on the PL of MEH-PPV

and PFO solutions. The PL of polymers in more than two solvent systems is measured.

Combined with the results from the viscosity measurements, more details about the

polymer conformation in solutions are given and subsequently are used to explain the

relationship between the PL and solvent properties. The second part focuses on the

influences of the solvent, the polymer concentration, and the annealing process on the PL

of polymer thin-films.

5.2 Experiment

5.2.1 Sample Preparation

MEH-PPV (Mn = 40,000-70,000, Aldrich) and PFO (Mw = 52,000, American Dye

Source, Inc) were used as received. MEH-PPV was dissolved in organic solvents at

several concentrations from 0.1 10 mg/mL. These solutions were filtered (1 [tm PTFE

filter) and spin-coated on ITO glass at 1000 rpm for 30 seconds.

5.2.2 PL Measurement

The PL of polymer solutions was measured in a quartz cell with the excitation

energy of 490 nm for MEH-PPV and 380 nm for PFO. The emission was collected by a

monochromator (Oriel instrument, Electo-optical systems) with a range from 530 nm to

700 nm for MEH-PPV and 400 nm to 650 nm for the PFO solutions and thin films.

5.2.3 Viscosity Measurement

The viscosity was measured with a standard Ubbelohde capillary viscometer in a

temperature-controlled water bath at 28 C. Polymer solutions were filtered through a

1.0 atm PTFE filter to remove dust particles and other impurities. Polymer solutions were









held in the temperature bath for approximately 15 minutes to reach thermal equilibrium.

Typically, the measurement was repeated at least three times.

5.3 Results and Discussion

5.3.1 PL of Polymer Solutions

5.3.1.1 Solvent effects

Figure 5-1 shows the PL of the MEH-PPV solutions in four organic solvents:

chloroform, CB, THF, and Tetrachloroethane (TCE).


0.5% MEH-PPV solutions
Chloroform
CB
Sf --- THF
I ; \\\ ---TCE



i I7/ "

zI II





540 560 580 600 620 640 660 680 700
Wavelength (nm)

Figure 5-1. PL of 0.5% MEH-PPV solutions.

The PL from the CB solutions shows a red-shifted emission compared to the THF

solutions, which is coincident to the results from Nguyen et al. [28]. In addition, the

emission peak of the chloroform solution is located between the peaks of THF and CB.

The PL from the TCE solutions shows the most red-shifted emission.

Figure 5-2 shows results of PFO dissolved in six organic solvents: toluene, THF,

chloroform, CB, TCE, and cyclohexaone (from left to right) with a concentration of 0.5









%. From the figure, chloroform shows the highest solubility and cyclohexanone shows

the poorest solubility, which is coincident to the results from Grell et al. [13].



















Figure 5-2. PFO dissolved in toluene, THF, chloroform, CB, TCE, and cyclohexanone
(from left to right).

Table 5-1. Solubility parameter and polymer solubility.
Solvent Solubility Parameter PFO solubility MEH-PPV solubility
(cal/cm3)1/2 (mg/ml) (mg/ml)
Toluene 8.9 <1 <5
THF 9.1 3-5 7-10
Chloroform 9.3 >5 >15
CB 9.5 3-5 7-10
TCE 9.7 -1 7-10
Cyclohexanone 9.9 <1 <5
Table 5-1 shows the solubility parameters of these organic solvents [107] and the

solubility of PFO and MEH-PPV in these solvents based on our test results. Based on the

solubility of PFO and MEH-PPV, both polymers have similar solubility parameters near

9.3, and chloroform is a better solvent than the other solvents. From these results,










chloroform is a good solvent compared to the other organic solvents. This conclusion is

consistent with the result from Grell et al. [13].

Figure 5-3 shows the PL of the PFO solutions.

0.05% PFO solution
/. THF
I TCE
S\\ Ch
.-I ---CB
fI
-



z
i I







400 420 440 460 480
Wavelength (nm)

Figure 5-3. PL of PFO in THF, TCE, Chloroform and CB.

The figure shows that PL from TCE solution has the most red-shifted emission.

The PL from THF solution shows the most blue shifted emission, which is similar to the

resultsfor MEH-PPV.

From these results, we can see the controversy of the definition of "good" solvents

and "poor" solvents. Assuming that chloroform is the best solvent compared to other

solvents, the chain conformation of MEH-PPV and PFO in chloroform should be more

extended than in the other solvents. Polymer chains with a more-extended conformation

have a longer resonance length and lead to more red-shifted emission, which means the

PL of MEH-PPV and PFO in chloroform should show the most red-shifted emission

compared to other solvents. In Figure 5-1, however, the PL of MEH-PPV in chloroform

solution does not show the most red-shifted emission. In the PFO system shown in









Figure 5-3, the PL from the chloroform solutions does not show the most red-shifted

emission, either. Instead, the PL from the "poor" solvent, TCE, shows a more red-shifted

emission compared to the "moderately good" solvent, CB, and the "good" solvents, THF

and chloroform.

To solve this controversy, we need to figure out the chain conformation of these

polymers in different solvent systems. The viscosity of a polymer solution is related to

the dimension of the polymer chains and can be used to estimate the polymer

conformation in solution.

The measurement of the viscosity of a solution is based on capillary flow. The

fluid flowing through a capillary follows Poiseuille's Law:

)rr4tAP
l r4
8V1

where r is the radius of the capillary; t is the flow time; V is the fluid volume between the

two markers on the viscometer; and AP is the pressure drop. For a capillary viscometer

in the vertical position, AP is generated from gravity:

AP = plgl

where pliq is the density of the liquid and 1 is the length between the two markers. The

kinematic viscosity is defined as follows:


kinematic viscosity = r= t
Phq 8V

4
where is a constant for a given viscometer and can be determined by measuring a
8V

fluid of known viscosity. Therefore, the kinematic viscosity of any polymer solution can

be measured by measuring the flow time, t.







61


Figure 5-4 shows the kinematic viscosities of MEH-PPV in chloroform, CB, and

TCE at different concentrations.


12
m TCE
Chloroform
10- Chlorobenze

Cy
E 8-
E



CU
6
O ..



2 4-
0-

O- -- *
0 1 2 3 4 5 6 7 8
Concentration (mg/mL)


Figure 5-4. The measurement of kinematic viscosities of MEH-PPV in CB, chloroform,
and TCE at different concentrations.

Polymer chains are isolated and do not interact with each other at low-polymer

concentrations. When the polymer concentration is increased, the polymer chains start to

entangle with each other and the slope of the viscosity-concentration curve is changed.

In this figure, the curve for MEH-PPV in TCE changes slope at a lower concentration

compared to CB and chloroform, which indicates that the polymer chains in TCE more

easily interact with each other and form entanglements.

By measuring the viscosity of the pure solvent, ls, the specific viscosity, rsp, can be

calculated:


rsp -l = [rl]c + Kc +...
r)










where [r] is the intrinsic viscosity, K, is a second-order coefficient and c is the

concentration. Therefore, the intrinsic viscosity, [r], can be determined by:


[1] = lim 7
c-O C


The intrinsic viscosity is a quantity characteristic of a polymer. A polymer

molecule with a larger dimension has a larger intrinsic viscosity. Figure 5-5 shows rsp/c

plotted as a function of c for the MEH-PPV solutions.


1200
TCE
Chlorobenzene
1000 Chloroform. -"
1000


800 -
JRB-
E -
o 600 -

____9---
400 -


200

0.05 0.10 0.15 0.20 0.25 0.30
Concentration (mg/mL)

Figure 5-5. rsp/c plotted as a function of c for MEH-PPV in CB, chloroform, and TCE.

This figure shows that MEH-PPV has the highest intrinsic viscosity in TCE and the

lowest intrinsic viscosity in chloroform. This indicates that MEH-PPV has a larger

dimension and a more extended chain conformation in TCE. MEH-PPV in chloroform

shows a smaller dimension and a more compact chain conformation. These results are

coincident to the PL shown in figure 5-1. MEH-PPV in TCE shows the most red-shifted









emission and MEH-PPV in chloroform shows the most blue-shifted emission among the

three solutions.

Therefore, from the PL results, the MEH-PPV and PFO polymer chains do not

show a longer conjugated length when they are dissolved in solvents in which they are

more soluble. This phenomenon is different from what is seen in traditional polymers.

For traditional polymers, like PS or PMMA, polymer chains show a random-walk

conformation when dissolved in good solvents, and the polymer chains are extended.

When solvents become poorer, the polymer chains contract due to the unfavorable

interaction between the polymer chains and the solvent. Therefore, the polymer chains in

poor solvents are more contracted. The difference in trend between traditional polymers

and conjugated polymers is quite possibly due to the specific chemical structure of the

conjugated polymers. MEH-PPV and PFO have rigid aromatic groups as main chains

and soft alkyl groups as side chains. The solubility of these polymers is mainly

controlled by the side chain groups, whereas the PL is controlled by the resonance length

on the main chains. So far, there are very few studies that have focused on solvent

effects on the chain conformation of conjugated polymers using computer simulation.

Recently, one study [33] was conducted comparing aromatic and non-aromatic solvent

effects on the chain conformation of MEH-PPV. Although the results could be used to

explain the red-shifted emission of the CB solution compared to the chloroform or THF

solutions, it is still not sufficient to explain the red-shifted emission as seen in the TCE

solution. The effect of this solvent on chain conformation could be very complicated and

may not conform to a simple rule.










5.3.1.2 Concentration effects

Figure 5-6 shows the PL of MEH-PPV in chloroform at four different

concentrations.


MEH-PPV in Chloroform
0.1 mg/mL
1 mg/mL
-5 mg/mL
S---10 mg/mL












550 600 650 700
Wavelength (nm)

Figure 5-6. PL of MEH-PPV in chloroform in concentrations of 0.1, 1, 5, and 10 mg/mL.

Compared to the results from viscosity measurement shown in Figure 5-4, Figure

5-6 shows that the PL of MEH-PPV does not change even when the polymer chains start

to form entanglements (concentration higher than 5 mg/mL) in chloroform solutions.

This indicates that the formation of entanglements and the inter-chain interactions of

MEH-PPV polymer chains in solution does not change the chain conformation and affect

the PL. MEH-PPV in CB and TCE also show similar results, as seen in figure 5-7. The

PL of MEH-PPV in CB and TCE does not change when concentration is changed.






























550 600 650
Wavelength (nm)


Figure 5-7. PL of MEH-PPV in CB and TCE.

Figure 5-8 shows the PL of PFO in chloroform with different polymer

concentrations.


400 420 440 460
Wavelength (nm)


Figure 5-8. PL of PFO in chloroform.


480 500









Unlike the PL of MEH-PPV solutions, the PL of PFO solutions shows an extra

emission peak near 435 nm when the polymer concentration is increased. This suggests

that the PFO can form a special morphology in solutions when the concentration is

increased, and that interchain interaction exists in said morphology, which can lead to an

extra red-shifted emission. This phenomenon is also observed in other solvent systems.

Previous studies have shown that PFO exists as a P-phase [14] via suitable

physicochemical treatment in the solid state. This P-phase has a high degree of interchain

order and a longer conjugation length, which leads to a red-shifted emission. Our results

show that gelation of PFO in the polymer solution at a higher concentration also leads to

an extended chain conformation and can generate an extra emission peak as shown in the

PL measurement.

5.3.1.3 Water effect on PFO in THF solutions

In the previous chapter, the effect of water on MEH-PPV was explored. After

adding water to the THF solutions, the MEH-PPV precipitates and the polymer chains

form a collapsed conformation.

Figure 5-9 shows the PL of PFO in THF after water is added. In contrast to the PL

of MEH-PPV in THF solutions (Figure 4-8), PFO shows three emission peaks during the

process of precipitation formation. Based on the previous discussion, the emission peak

near 440 nm is due to the formation of aggregates in the THF solution. The emission

peak near 470 nm is due to the formation of precipitates.






67



PFO in THF solution
Dried THF
S- 1% H20
S. --- 2%
\ ---3%
/----7.5%
U/. ......... *..10%












Figure 5-9. PL of PFO in the THF solution with different water concentrations.

Therefore, these results show that the water can also induce aggregation of PFO in

THF solutions. After adding more water, PFO can also undergo a transformation of

chain conformation from coil to cylinder.
5.3.2 PL of Polymer Thin films










As noted in chapter 2, the polymer chains have a much stronger interchain

interaction in polymer thin films compared to polymer solutions. Therefore, the PL of

polymer thin films shows lower quantum efficiency and a red-shifted emission due to the

generation of aggregates.
.%


400 420 440 460 480 500












Wavelength5.3.2.1 Solvent effects(nm)

Figure 5-10 shows the PL of MEH-PPVPFO in the THF solution within four different water concents: chloroform, CB,
Therefore, these results show that the water can also induce aggregation of PFO in

THF solutions. After adding more water, PFO can also undergo a transformation of

chain conformation from coil to cylinder.

5.3.2 PL of Polymer Thin films

As noted in chapter 2, the polymer chains have a much stronger interchain

interaction in polymer thin films compared to polymer solutions. Therefore, the PL of

polymer thin films shows lower quantum efficiency and a red-shifted emission due to the

generation of aggregates.

5.3.2.1 Solvent effects

Figure 5-10 shows the PL of MEH-PPV in four different solvents: chloroform, CB,

THF, and TCE, as well as the PL of thin films coated from these solutions under the same

polymer concentration and same spin-speed.













\\ --TCE Sol
ID -- -- Chloroform Films
S\ -A CB Films
S\ -- -- THF Films
-- TCE Films


00





550 600 650 700
Wavelength (nm)

Figure 5-10. Comparison of PL of MEH-PPV in polymer solutions and thin films.

This figure illustrates how the PL from films shows an emission with the longest

wavelength when the film is cast from TCE and the shortest wavelength when it is cast

from THF. This result shows that the degree of red-shift occurs in the same order as the

PL from the polymer solutions, which indicates that the chain conformation of polymers

can be maintained during the spin-coating process [28]. The red-shifted emission from

polymer films compared to polymer solutions is due to the formation of aggregates.

5.3.2.2 Concentration and spin-speed effects

Based on the classic model of the spin-coating process [73], the film thickness is

determined by the concentration of the polymer solution and the spin-speed. Recent

research also has shown that the conformation of polymer chains and the interchain

interaction can be altered by changing the concentration and the spin-speed [33]. Figure

5-11 shows the PL of MEH-PPV thin films cast from chloroform solutions with three

different concentrations.







69



PL of MEH-PPV films
from chloroform
7 mg/mL
S- 5 mg/mL
S- 2.5 mg/mL

S/ \
D- I' \



SI I





550 600 650 700
Wavelength (nm)

Figure 5-11. PL of VMEH-PPV thin films from chloroform solutions with concentrations
of 2.5, 5, and 7 mg/mL. (Spin-speed = 1000 rpm)

The polymer film spin-coated from the solution with a lower polymer concentration

shows a lower thickness, which leads to a lower PL intensity. The films coated from

solutions of lower concentration (2.5 mg/mL and 5 mg/mL) show slightly blue-shifted

emissions compared to the films coated from solutions of higher concentration (7

mg/mL). This indicates that the films coated from solutions with higher concentrations

have a better interchain interaction compared to the films coated from solutions with

lower concentrations. Therefore, the increase of polymer concentration can increase the

film thickness as well as the chain interaction, which leads to a red-shifted emission.

Figure 5-12 shows the PL from MEH-PPV thin films coated from the chloroform

solutions with the same concentration at three different spin-speeds.







70



PL of MEH-PPV films
from chloroform
S--1000 rpm
S/ \ 3000 rpm
d --- 5000rpm


I.


I,.

I I I



550 600 650 700
Wavelength (nm)

Figure 5-12. PL of MEH-PPV thin films coated from chloroform solution (7 mg/mL) at
1000, 3000, and 5000 rpm.

An increase of spin-speed will lead to a thinner film and will result in a lower PL

intensity, as shown in Figure 5-12. Moreover, the PL of the thin films from higher spin-

speeds (3000 rpm and 5000 rpm) show slightly blue-shifted emissions compared to the

films from a lower spin-speed (1000 rpm). This indicates that the polymer chains which

cast using higher spin-speeds have less interchain interaction. Therefore, the spin-speed

effect shows the same effect as concentration: i.e. polymer films coated from lower

speeds show higher thickness and more interchain interactions, which lead to a red-

shifted emission.

Figure 5-13 and Figure 5-14 show the effect of concentration and spin-speed on the

PL of MEH-PPV films coated from CB solutions. The results show that the PL of MEH-

PPV films coated from CB solutions also follow the same trend as the PL of MEH-PPV

films coated from chloroform solutions.





























Wavelength (nm)


Figure 5-13. PL of MEH-PPV thin films
and 7 mg/mL. (Spin-speed =


from CB solutions with concentrations of 2.5, 5,
1000 rpm)


550 600 650 700
Wavelength (nm)

Figure 5-14. PL of MEH-PPV thin films coated from chloroform solutions (7 mg/mL) at
1000, 3000, and 5000 rpm.







72


5.3.3.3 Annealing effects

As described in chapter 2, the annealing process can increase the interchain

interaction of polymer chains in the spin-coated thin films, which leads to the generation

of aggregates. The formation of aggregates leads to the reduction of PL efficiency and a

red-shifted emission.

Figure 5-15 shows the annealing effect on the MEH-PPV thin film coated from

chloroform solution.


Before annealed
After annealed




-E









550 600 650 700
Wavelength (nm)

Figure 5-15. The PL of MEH-PPV coated from chloroform solution before and after
annealing at 120 C for 3 hrs in a vacuum.

Comparing the PL of the thin film before and after the annealing process, the

annealed MEH-PPV thin films show lower PL efficiency and a red-shifted emission,

which is consistent with the previous studies. Figure 5-16 shows that the annealing effect

on MEH-PPV thin films coated from CB solutions is similar.













'E



c-2

-j / ,






550 600 650 700
Wavelength (nm)

Figure 5-16. PL of MEH-PPV coated from CB solution before and after annealing at 120
C for 3 hrs in a vacuum.

5.4 Conclusion

From the PL and viscosity measurements of polymer solutions, we conclude that

the emission peak of MEH-PPV and PFO solutions depend on the chain conformation (or

conjugation length). The PL of polymer chains with a large dimension shows a red-

shifted emission compared to polymer chains with a small dimension, due to the increase

of the conjugation length. The PL of MEH-PPV solution is not affected when the

polymer concentration is increased, even once entanglements are formed. Different from

MEH-PPV, the PL of PFO solutions show an extra red-shifted emission when

concentration is increased. This is due to the gelation of PFO in the polymer solution and

the formation of an extended-chain conformation. More details about the chain

conformation of PFO will be discussed in chapter 7.









The precipitation of MEH-PPV by adding water to THF can result in a coil-

cylinder transition, which leads to an extra red-shifted emission. Different than MEH-

PPV, water leads to gelation before the coil-cylinder transition occurs in PFO solutions.

The PL of polymer thin films is related to the chain conformation as well as the

interchain interaction in the thin films. The formation of aggregates, which are caused by

the interchain interaction, leads to lower PL quantum efficiency and a red-shifted

emission. Therefore, those spin-coating conditions which can improve the packing of

polymer chains and lead to formation of aggregates show a red-shifted emission.














CHAPTER 6
DEGRAGATION OF METAL CONTACT IN POLYMER LIGHT-EMITTING DIODE
DEVICES

6.1 Introduction

Bubble formation on calcium/aluminum contacts to bilayer poly(n-vinyl carbazole)

(PVK)/poly(3,4-ethylenedioxythiophene)-poly(4-styrenesulfonate) polymer light-

emitting diode (PLED) devices was studied using optical and electron microscopy and

Auger electron spectroscopy (AES). The formation of bubbles is shown to correlate with

pinholes in the metal contact thin film and with excess absorbed water in the presence of

the calcium layer. AES data show oxidation at the Ca/Al interface during this process.

Gas evolution and relaxation of compressive stress when voltage is applied to the PLED

devices are postulated to cause the formation of bubbles.



6.2 Experiment

6.2.1 Device Fabrication

The multilayer PLED device consisted of an indium-tin-oxide (ITO) coated glass

substrate coated with layers of PEDOT-PSS/PVK/Ca/Al, as shown schematically in

Figure 6-1. The 50 nm PEDOT-PSS (Bayer) layer was spin-coated (3000 rpm) on the

cleaned ITO coated glass, then dried at 1200C for 20 minutes. The PVK (Aldrich)

solution was prepared by dissolving 50 mg of the polymer in 5 ml of toluene and spin-

coated (700 rpm) to form a 70 nm layer on the PEDOT-PSS. The 10 nm Ca layer and








200 nm Al layer were deposited by thermal evaporation in a glass bell jar, an oil-

diffusion-pumped vacuum system with a base pressure of 4x10-6 Torr.



/
Al

Ca // --

PVK

PEDOT-PSS

ITO

Figure 6-1. Schematic of the PLEDs used in this study.

To study the contributions of the various layers to bubble formation in the cathode

contact, multilayer devices were prepared as described below without selected layers, e.g.

without the Ca or PEDOT-PSS layers. Encapsulated devices means that the cathode

contact would be covered by epoxy (QuicksetTM, Loctite), sealing a glass slide (Micro

slide 2947, Corning) over the surface in an atmosphere of argon.

6.2.2 Characterization

Bubble formation was observed by optical microscopy after applying a current to

the devices in ambient laboratory air, or heating in various atmospheres. The chemical

composition of the cathode surface of the PLED device was determined by using a

Perkin-Elmer PHI 660 Auger electron spectrometer (AES). A primary electron beam of

5 keV was used. The Auger electron peaks at energies around 271, 294, 511, and 1398

eV are chosen for the elements C, Ca, 0, and Al respectively. The composition through

the cathode and at the polymer/cathode interface was determined by using 1 keV argon

sputter depth profiling on the AES.








6.3 Results and Discussion

6.3.1 Bubble Formation Phenomenon and Mechanism

Figure 6-2 shows bubbles that formed in the Ca/Al cathode film on a standard

PEDOT-PSS/PVK PLED upon the application of an electrical bias within one hour after

depositing the Ca/Al cathode (Fig. 6-2(a)), or after depositing the Ca/Al cathode and

storing the device for 24 hours in lab air (Fig. 6-2(b)).


25 pin


* *.


0 (b)


50FC)II


O '



0


Figure 6-2. Bubble formation observed with optical microscopy: (a) biased with 7 V, 10-5
A-cm2 for 1 min. within one hour after thermal evaporation of Al; (b) biased
with 7 V, 10-5 A-cm-2 for 1 min after storage in laboratory air for 1 day (-60%
relative humidity, 27 C). Note the lower magnification in (b).









Prior to the application of an electrical bias, no bubbles were observed in either

sample. Both devices were biased at 7 V, which caused a current density of 10-5 Acm-2

to flow through the devices. The small bubbles in Fig. 6-2(a) developed after 15 minutes

of bias, while large bubbles developed after only 1 minute of bias on the device stored in

lab air for 24 hours. Note that in both cases, the bubbles formed around a defect

(pinhole) in the Ca/Al cathode thin film.

Do et al. [69] and McElvain et al. [68] reported that bubbles formed in Mg/Al

cathode thin films on Alq3 emitter layers due to heat, gas evolution and oxidation of the

Al as a result of electrical biasing. Savvate'ev et al. [65] reported that gas evolution

caused delamination of the Al cathode thin films on the PPV layers, resulting in electrical

noise and ultimately a current spike that resulted in carbonized polymer areas. Do et al.

[69] showed that an inert atmosphere inhibited the formation of bubbles. The effects of

the environment and encapsulation on the formation of cathode bubbles in our PEDOT-

PSS/PVK conjugated polymer devices were tested by storing devices for four days under

argon or air with and without epoxy/glass encapsulation, followed by electrical biasing at

8 V for 2 minutes (Figure 6-3). The cathode surfaces of these devices before electrical

biasing, shown on the left, are free of bubbles even when stored without encapsulation.

After electrical biasing, the surface of encapsulated devices stored in argon did not

bubble, but bubbles were observed on the surfaces of the other devices stored in argon or

air without encapsulation, or in air even with encapsulation. The bubbles were larger and

more numerous for devices stored in air (Figs. 6-3(c) and 6-3(d)) versus argon (Figs. 6-

3(a) and 6-3(b)). Devices stored in air without encapsulation formed the greatest amount

of bubbles after biasing (Fig. 6-3(d)).



















































Figure 6-3. The effects of storage ambient and encapsulation upon bubble formation as a
result of electrical bias (8 V for 2 min.). Encapsulated or non-encapsulated
devices were stored for four days in argon or air as follows: (a) argon with
encapsulation; (b) argon without encapsulation; (c) air with encapsulation; (d)
air without encapsulation. Photos on the left and right are before and after the
devices were biased, respectively.









These data are consistent with the results of Do et al. [69] and show that argon

reduced the formation of bubbles, and with their general conclusion that moisture

accelerates their formation. It is interesting to note that bubbles were denser near the

edges of the dot contacts, suggesting that moisture could more easily penetrate into the

devices at the step edges where there was no protective Al layer.


21.io.'pm


(a) 2 '.1 1LtL


(d)


(b) 2' I IIm


2i1.1.11'1


2'.It)( lL


(0
0 S
0


Figure 6-4. The dependence of bubble formation on the type of conjugated polymer and
temperature in the absence of electrical bias. Devices consisted of the normal
ITO/polymer/Ca/Al structure, where the polymer layer is either PVK or
PEDOT-PSS. Devices were examined in real time by optical microscopy
while being heated on a hot stage either immediately after deposition of the
cathode, or 24 hours after this deposition with storage in air. The samples and
temperatures were as follows: (a) PVK, fresh, 120C; (b) PVK, fresh, 250C;
(c) PVK, 24 hour storage, 120C; : (d) PEDOT-PSS, fresh, 120C; (e)
PEDOT-PSS, fresh, 250C; (f) PEDOT-PSS, storage, 120C. Note that
bubbles are observed only when PEDOT-PSS is present. Note also that
bubbles were formed by heating alone without an electrical bias.









To determine whether PEDOT-PSS or PVK was contributing to the formation of

bubbles, devices were prepared with only a PVK (Figs. 6-4(a) 6-4(c)) or a PEDOT-PSS

(Figs. 6-4(d) 6-4(e)) polymer layer, respectively. In addition, to determine if simple

heating without electrical currents would cause bubbles to form, the devices were heated

to either 120C or 250C on a hot plate in air. Finally, devices were heated immediately

after the deposition of the cathode (fresh devices) and monitored while hot using the

optical microscope. Neither fresh PVK nor fresh PEDOT-PSS devices showed bubbles

when heated to 1200C (Figs. 6-4(a) and (d)). When the temperature was raised to 2500C,

only the device with a PEDOT-PSS layer formed bubbles (compare Figs. 6-4(b) and (e)).

While storage in air for 24 hours did not cause bubbles to form at 1200C on PVK devices,

large bubbles were found on PEDOT-PSS devices that were treated in a similar fashion.

These data demonstrate that water is absorbed by and evolved from PEDOT-PSS layers

to form bubbles. They also demonstrate that simple heating without an electrical bias can

result in the formation of bubbles on these devices. This is consistent with the postulate

that local heating can cause gas evolution from some polymers [65]. By real time

observations in the optical microscope, the fresh PEDOT-PSS devices started to evolve

gas and form bubbles at T -220C. While storage of PVK devices in air for 24 hours did

not cause bubbles to emerge (Fig. 6-4(c)), storage accelerated their formation in PEDOT-

PSS devices. After 24 hours in air, PEDOT-PSS devices started to form bubbles at

temperatures of -90C, and they were well-formed by 120C (Fig. 6-4(f)). Observation

of the bubbles at 900C after a 24-hour exposure to air supports the conclusion that

PEDOT-PSS layers absorb and evolve water to create bubbles. Because of the low water

content and experience with other hydrophobic polymers, it is believed that the bubbles









formed in fresh PEDOT-PSS samples heated to 2500C result from decomposition of the

polymer. Since PEDOT-PSS is soluble in water, it is reasonable that these layers would

absorb water under ambient storage conditions, and evolve water when heated to T-900C.

Do et al. [69] also suggested that water can lead to formation of A1203 with

evolution of H2, which can cause bubbles to form. In our devices, water in the PEDOT-

PSS layer could react either with Ca or Al to form H2, e.g.:

Ca + H20 CaO + H2

2A1 + 3H20 A1203 + 3H2

To evaluate the importance of these reactions, devices were made with either the

normal Ca/Al or with only an Al cathode dot contact to the normal PVK/PEDOT-PSS

bilayer structure. A small water-droplet was placed on the surface of these cathodes,

without an electrical bias being applied. The results are shown in Figure 6-5.

M ,- 7100 iu (e)
(a)





..i. (d)
(b)






Figure 6-5. Formation of bubbles by water droplets in the absence of an electrical bias
when a Ca layer is present. The photos are of Al only or normal Ca/Al cathode
contacts with water droplets on the surface. Al-only contact (a) immediately
after placing a water droplet on the surface, and (b) after a few minutes; no
bubbles are observed. Ca/Al contact (c) immediately after placing a water
droplet on the surface, and (d) after a few minutes; large bubbles are obvious
in (d).









No bubbles are observed immediately after placing the water droplet on either

contact (Figs. 6-5(a) and 6-5(c)). For the cathode consisting of only an Al layer, no

bubbles were observed upon contact after 5 min (Fig. 6-5(b)). By contrast, large bubbles

are observed after <5 minutes on devices with a normal Ca/Al cathode (Fig. 6-5(d)).

Pinholes certainly existed in both the Al and the Ca/Al cathodes, allowing water to

quickly reach the metal/polymer interface. Formation of bubbles in only a few minutes

when the Ca layer was present demonstrates its importance to the formation of bubbles.

Perhaps the Ca reacts rapidly to evolve gas quickly and thereby form a bubble. This is

consistent with the fact that CaO should not form a passivating oxide layer. Conversely,

Al forms a protective A1203 layer and would not result in rapid H2 evolution from the

above chemical reaction. A second factor may be the mechanical strength of the

interfacial oxide formed by such reactions. An interfacial layer of CaO may fracture

more easily than A1203 and thus allow for a gas pocket to form a bubble.

6.3.2 Surface Composition (AES)

A scanning electron microscope (SEM) in the AES mode was used to examine the

surface for pinholes surrounded by bubbles that were produced by electrical biasing at 7

V for 10 minutes on a device stored in air for 1 day (see Fig. 6-6(a)). The AES spectra

collected from the pinhole and adjacent cathode surface are shown in Fig. 6-6(b). The

spectra from the pinhole (i) contain Auger peaks from only C and 0, presumably from

the polymer thin film structure. No Al was detected in the pinhole. The spectrum from

the adjacent cathode surface (ii) contains peaks from C, O and Al. The strong oxygen

signal in (ii) is due to A1203 on the Al surface. The C peak is from normal adventitious

contamination due to adsorption of organic and hydrocarbons from laboratory air. Since

pinhole areas lack the protection of a continuous Al layer, they offer paths for transport of




Full Text

PAGE 1

BEHAVIOR OF LIGHT-EMITTING POLYME RS: FILM FORMATION, OPTICAL PROPERTIES, AND DEGRADATION By SHYH-CHYANG LUO A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2005

PAGE 2

Copyright 2005 by Shyh-Chyang Luo

PAGE 3

This dissertation is dedicated to my parents, my brothers, and Inga for their love, support and encouragement.

PAGE 4

ACKNOWLEDGMENTS First of all, I would like to thank my advisor and my committee chair, Dr. Elliot Douglas, for lending an unwavering supportive presence to my studies in the United States throughout these years. Without a doubt, I feel very fortunate to have received the benefits of his guidance. His welcomed openness and confidence in my ideas have made working in his group a pleasant and enriching experience. I would like to express sincere gratitude to my committee members, Dr. Paul Holloway, Dr. Ronald Baney, Dr. David Norton, and Dr. Kirk Schanze, for their invaluable help and support of my research. I would like to thank Dr. Valentin Craciun, Dr. Eric Lambers, Andrew Gerger, and Juhyun Woo in the Major Analytical and Instrumentation Center for their assistance in the measurements of X-ray reflectivity, Auger electron spectroscopy, and atomic force microscopy. I also wish to thank Dr. Mark Davidson in Microfabritech for helping with the photoluminescence measurement. Without their aid and input, this work could not have been done. I would like to extend a special thank you to all of the members in Dr. Douglas and Dr. Holloways group and my classmates, especially Dr. Wenling Jia, Lewei Bu, Brendan Collins, Hsiu-Hsin Chung, Phil Chung, and JiaHau Yen, for their help and support. I am so lucky to have worked with all of them. Many thanks go to Jennifer Wrighton for her attention to all the details that made my working in this group all the more enlightening. iv

PAGE 5

I especially thank Ying-Chih Wang. Her support and encouragement enabled me to overcome the challenges of this project. Finally, I would like to thank my parents and my younger brothers. They are always there for me. Without them, I could not have gotten this far. v

PAGE 6

TABLE OF CONTENTS Page ACKNOWLEDGMENTS .................................................................................................iv TABLE OF CONTENTS ...................................................................................................vi LIST OF TABLES .............................................................................................................ix LIST OF FIGURES .............................................................................................................x ABSTRACT .....................................................................................................................xvi CHAPTER 1 INTRODUCTION........................................................................................................1 2 REVIEW OF LITERATURE.......................................................................................4 2.1 Fundamentals of -Conjugated Polymers...............................................................4 2.1.1 A Brief History of -Conjugated Polymers..................................................4 2.1.2 Photophysics.................................................................................................5 2.2 Polymer Light-Emitting Diodes (PLEDs)..............................................................6 2.2.1 Solvent Effects..............................................................................................8 2.2.2 Annealing Effects.......................................................................................11 2.2.3 Degradation of PLED Devices...................................................................13 2.3 Defects of Spin-Coated Polymer Thin-Film.........................................................17 2.3.1 Bnard-Marangoni Convection..................................................................18 2.3.2 Dewetting Phenomenon..............................................................................20 3 MARANGONI EFFECTS ON POLYMER THIN FILM STRUCTURE.................23 3.1 Introduction...........................................................................................................23 3.2 Experiment............................................................................................................24 3.2.1 Sample Preparation.....................................................................................24 3.2.2 Characterization..........................................................................................26 3.2.2.1 Atomic Force Microscopy (AFM)...................................................26 3.2.2.2 X-Ray Reflectivity (XRR)...............................................................26 3.3 Results and Discussion.........................................................................................26 3.4 Conclusion............................................................................................................39 vi

PAGE 7

4 RUPTURE OF POLYMER THIN FILMS DURING SPIN-COATING PROCESS.40 4.1 Introduction...........................................................................................................40 4.2 Experiment............................................................................................................40 4.2.1 Sample Preparation.....................................................................................40 4.2.2 Characterization..........................................................................................41 4.3 Results and Discussion.........................................................................................41 4.4 Conclusion............................................................................................................53 5 SOLVENT, CONCENTRATION, AND ANNEALING EFFECTS ON PHOTOLUMINESCENCE OF POLYMER SOLUTIONS AND THIN FILMS......55 5.1 Introduction...........................................................................................................55 5.2 Experiment............................................................................................................56 5.2.1 Sample Preparation.....................................................................................56 5.2.2 PL Measurement.........................................................................................56 5.3 Results and Discussion.........................................................................................57 5.3.1 PL of Polymer Solutions............................................................................57 5.3.1.1 Solvent effects..................................................................................57 5.3.1.2 Concentration effects........................................................................64 5.3.1.3 Water effect on PFO in THF solutions............................................66 5.3.2 PL of Polymer Thin films...........................................................................67 5.3.2.1 Solvent effects..................................................................................67 5.3.2.2 Concentration and spin-speed effects...............................................68 5.3.3.3 Annealing effects..............................................................................72 5.4 Conclusion............................................................................................................73 6 DEGRAGATION OF METAL CONTACT IN POLYMER LIGHT-EMITTING DIODE DEVICES......................................................................................................75 6.1 Introduction...........................................................................................................75 6.2 Experiment............................................................................................................75 6.2.1 Device Fabrication......................................................................................75 6.2.2 Characterization..........................................................................................76 6.3 Results and Discussion.........................................................................................77 6.3.1 Bubble Formation Phenomenon and Mechanism.......................................77 6.3.2 Surface Composition (AES).......................................................................83 6.4 Conclusion............................................................................................................87 7 EFFECT OF THERMO-INDUCED OXIDATION ON POLYMER STRUCTURE AND OPTICAL PROPERTIES OF POLYFLUORENE...........................................88 7.1 Introduction...........................................................................................................88 7.2 Experiment............................................................................................................89 7.2.1 Sample Preparation.....................................................................................89 7.2.2 Characterization..........................................................................................89 7.3 Results and Discussion.........................................................................................90 vii

PAGE 8

7.3.1 Chain Conformation in PFO Thick Films..................................................90 7.3.1.1 Molecular weight effect...................................................................90 7.3.1.2 Thermal-oxidation effect..................................................................95 7.3.2 Molecular Effect on PFO Thin Film Structure...........................................98 7.4 Conclusion..........................................................................................................101 8 CONCLUSIONS AND RECOMMENDATIONS FOR FUTURE WORK............103 LIST OF REFERENCES.................................................................................................108 BIOGRAPHICAL SKETCH...........................................................................................115 viii

PAGE 9

LIST OF TABLES Table page 3-1. Characteristic data of solvents at room temperature................................................25 3-2. The Mark-Houwink parameters for PVK at 25 o C....................................................25 3-3. Characteristics of PVK films obtained from XRR measurements...........................35 3-4. Characteristics of MEH-PPV films obtained from XRR measurements..................38 4-1. Roughness of MEH-PPV polymer films prepared from the dried THF and THF/water solutions.................................................................................................48 4-2. PL of MEH-PPV polymer films prepared from the dried THF and THF/water solutions...................................................................................................................52 5-1. Solubility parameter and polymer solubility............................................................58 ix

PAGE 10

LIST OF FIGURES Figure page 2-1. Structure of conjugated polymers: a) cis-polyacetylene; b) trans-polyacetylene; c) polythiophene; d) poly(phenylene vinylene); e) poly(pyridyl vinylene); f) poly(fluorene).............................................................................................................4 2-2. General structure of PLEDs........................................................................................7 2-3. The mechanism of EL in PLEDs................................................................................7 2-4. Absorption and PL of MEH-PPV in CB and THF solutions......................................8 2-5. a) PL of CB solution and thin films; b) PL of thin films coated from CB and THF solutions.....................................................................................................................9 2-6. Characterization of PLEDs: a) I-V curve and b) B-V curve for DCB and THF devices having similar thickness..............................................................................10 2-7. PL of MEH-PPV films annealed at different temperatures......................................11 2-8. EL vs. current of the devices annealed at different temperature after Al deposition.12 2-9. Evolution of black spot formation in air: a) t = 2hrs; b) t = 10 hrs; c) t = 20 hrs; d) t = 30 hrs; e) t = 40hrs................................................................................................14 2-10. Schematic diagram showing dark spot induced by pinholes....................................15 2-11. The linear growth of dark spot with various particles..............................................15 2-12. Proposed electrochemical half-cell reactions showing the oxidation of metal (M) and the reduction of doped PEDOT.........................................................................16 2-13. Polystyrene coated from THF solution.....................................................................17 2-14. Polystyrene coated from cyclohexane and acetone mixed solution.........................18 2-15. The Marangoni convection.......................................................................................18 2-16. Micrograph of different stages of polystyrene dewetting on Si wafer: a) 15 mins and b) 1hr.................................................................................................................21 x

PAGE 11

2-17. The droplets left on the Si wafer..............................................................................21 2-18. Film structure of dewetted thin films: a) A dewetting pattern centered at a defect (50 x 50 m) and b) pattern at high magnification (5 x 5 m).................................22 3-1. Chemical structure of a) PVK and b) MEH-PPV.....................................................24 3-2. Film structure of polymer thin-film observed with optical microscopy: (a) prepared from THF solution, center area; (b) prepared from THF solution, away from center area; (c) prepared from chloroform solution, center area; (d) prepared from chloroform solution, away from center area. Length of scale bar 200m...............27 3-3. 3D tapping mode AFM image of PVK film surface prepared from (a) chloroform solution, cell structure near center area; (b) chloroform solution, striation structure away from center area..............................................................................................28 3-4. The Marangoni convection model (including composition-gradient-driven and temperature-gradient-driven processes) for the spin-coating process. T 1 and C 2 represent the temperature and solution concentration near the surface of the solution layer; T 2 and C 1 represent the temperature and solution concentration near the interface between the solution layer and the ITO substrate...............................29 3-5. Film structure of PVK thin film observed with optical microscopy prepared from TCE solution, center area.........................................................................................30 3-6. 2D tapping mode AFM image of PVK surface prepared from toluene solution......31 3-7. RMS Roughness plotted as a function of evaporation rate of four solvents............32 3-8. MEH-PPV Film structure of polymer thin-film observed with optical microscopy: (a) prepared from THF solution, center area; (b) prepared from THF solution, away from center area; (c) prepared from chloroform solution, center area; (d) prepared from chloroform solution, away from center area; e) prepared from CB solution; prepared from TCE solution.....................................................................................33 3-9. 2D tapping mode AFM image of MEH-PPV film surface prepared from (a) chloroform solution, cell structure near center area; (b) TCE solution, smooth surface near center area............................................................................................34 3-10. X-ray reflectivity spectra acquired from PVK thin films prepared from chloroform, benzene, toluene, and TCE solutions. Experimentally measured data are shown as a solid line; fitted results are shown as a dotted line...................................................35 3-11. X-ray reflectivity spectra acquired from MEH-PPV thin films prepared from chloroform, chlorobenzene, and TCE solutions. Experimentally measured data are shown as a solid line of light color, and fitted results are shown as a solid line with dark color..................................................................................................................37 xi

PAGE 12

4-1. SEM image of polymer prepared from THF solution: (a) center area with a low magnification, length of scale bar 200m; (b) away from center area with a low magnification, length of scale bar 10m; (c) center area with a high magnification, length of scale bar 200m; (d) away from center area with a high magnification, length of scale bar 10m..........................................................................................41 4-2. SEM image of polymer thin-film prepared from THF solution with different H 2 O contents: (a and b) THF dried by molecular sieve for 1 day with a low and high magnification, length of scale bar 200 and 10 m, respectively; (c), (d) 10L of H 2 O added in 5 mL of THF with a low and high magnification, length of scale bar 200 and 10 m respectively; (e), (f) 100l H 2 O added in 5mL of THF with a low and high magnification, length of scale bar 200m and 10 m, respectively..........43 4-3. PEDOT-PSS film structure on ITO substrate observed with optical microscopy: (a) low magnification, length of scale bar = 200m; (b) high magnification, showing presence of a hole, length of scale bar = 20m........................................................44 4-4. Structure of ruptured thin films: a) Optical microscopy image of MEH-PPV thin film prepared from 5% H 2 O in THF, striation structure away from center; b) SEM image of same film with higher magnification........................................................46 4-5. Height contrast of polymer thin-films from (a) dried THF solution with polymer concentration of 5 mg/ml; (b) 5% water in THF solution; (c) height profile of film along the line indicated in (a); (d) height profile of film along the line indicated in (b).............................................................................................................................47 4-6. PL spectra of MEH-PPV polymer thin-films...........................................................49 4-7. PL spectra of MEH-PPV in THF solutions (7 mg/ml) with different water concentrations...........................................................................................................50 4-8. PLED based on ruptured thin films: (a) The layer structure of PLED device. (b) The light-emitting output after the device is turned on...................................................53 5-1. PL of 0.5% MEH-PPV solutions..............................................................................57 5-2. PFO dissolved in toluene, THF, chloroform, CB, TCE, and cyclohexanone (from left to right)...............................................................................................................58 5-3. PL of PFO in THF, TCE, Chloroform and CB.........................................................59 5-4. The measurement of kinematic viscosities of MEH-PPV in CB, chloroform, and TCE at different concentrations...............................................................................61 5-5. sp /c plotted as a function of c for MEH-PPV in CB, chloroform, and TCE...........62 5-6. PL of MEH-PPV in chloroform in concentrations of 0.1, 1, 5, and 10 mg/mL.......64 xii

PAGE 13

5-7. PL of MEH-PPV in CB and TCE.............................................................................65 5-8. PL of PFO in chloroform..........................................................................................65 5-9. PL of PFO in the THF solution with different water concentrations.......................67 5-10. Comparison of PL of MEH-PPV in polymer solutions and thin films.....................68 5-11. PL of MEH-PPV thin films from chloroform solutions with concentrations of 2.5, 5, and 7 mg/mL. (Spin-speed = 1000 rpm)..............................................................69 5-12. PL of MEH-PPV thin films coated from chloroform solution (7 mg/mL) at 1000, 3000, and 5000 rpm..................................................................................................70 5-13. PL of MEH-PPV thin films from CB solutions with concentrations of 2.5, 5, and 7 mg/mL. (Spin-speed = 1000 rpm)............................................................................71 5-14. PL of MEH-PPV thin films coated from chloroform solutions (7 mg/mL) at 1000, 3000, and 5000 rpm..................................................................................................71 5-15. The PL of MEH-PPV coated from chloroform solution before and after annealing at 120 o C for 3 hrs in a vacuum....................................................................................72 5-16. PL of MEH-PPV coated from CB solution before and after annealing at 120 o C for 3 hrs in a vacuum.....................................................................................................73 6-1. Schematic of the PLEDs used in this study..............................................................76 6-2. Bubble formation observed with optical microscopy: (a) biased with 7 V, 10 -5 Acm -2 for 1 min. within one hour after thermal evaporation of Al; (b) biased with 7 V, 10 -5 Acm -2 for 1 min after storage in laboratory air for 1 day (60% relative humidity, 27 o C). Note the lower magnification in (b)............................................77 6-3. The effects of storage ambient and encapsulation upon bubble formation as a result of electrical bias (8 V for 2 min.). Encapsulated or non-encapsulated devices were stored for four days in argon or air as follows: (a) argon with encapsulation; (b) argon without encapsulation; (c) air with encapsulation; (d) air without encapsulation. Photos on the left and right are before and after the devices were biased, respectively..................................................................................................79 6-4. The dependence of bubble formation on the type of conjugated polymer and temperature in the absence of electrical bias. Devices consisted of the normal ITO/polymer/Ca/Al structure, where the polymer layer is either PVK or PEDOT-PSS. Devices were examined in real time by optical microscopy while being heated on a hot stage either immediately after deposition of the cathode, or 24 hours after this deposition with storage in air..........................................................80 xiii

PAGE 14

6-5. Formation of bubbles by water droplets in the absence of an electrical bias when a Ca layer is present. The photos are of Al only or normal Ca/Al cathode contacts with water droplets on the surface. Al-only contact (a) immediately after placing a water droplet on the surface, and (b) after a few minutes; no bubbles are observed. Ca/Al contact (c) immediately after placing a water droplet on the surface, and (d) after a few minutes; large bubbles are obvious in (d)..............................................82 6-6. Pinhole structure and composition: (a) SEM image of a pinhole in the Al/Ca cathode dot contact and surrounding bubbles generated by application of an electrical bias; (b) AES spectra from the cathode surface (i) inside the pinhole, and (ii) outside the pinhole..............................................................................................85 6-7. AES sputter depth profile of a normal 10 nm Ca/60 nm Al PLED cathode dot contact (a) before and (b) after an electrical bias was applied. Note the increase of oxygen at the Al/Ca interface after electrical biasing (8 V for 10 min. in laboratory air)............................................................................................................................86 7-1. Proposed mechanism for the generation of keto-defect sites...................................88 7-2. Images of a low Mw PFO under a polarized optical microscope at (a) 120 o C, (b) 150 o C, (c) 210 o C, and (d) 220 o C..............................................................................90 7-3. X-ray diffraction of low Mw PFO samples at room temperature, 125, 150, 200, and 250 o C. Sample was annealed at 250 o C for 10 minutes before cooling to the measurement temperature........................................................................................91 7-4. X-ray diffraction of high Mw PFO samples at room temperature, 125, 150, 200, and 250 o C. Sample was annealed at 250 o C for 15 minutes cooling to the measurement temperature...............................................................................................................92 7-5. Characteristic XRD profile for film in N, , and phases [114]........................93 7-6. XRD profile of low Mw and high Mw PFO samples annealed at 170 o C for 3 and 15 hours in Ar................................................................................................................94 7-7. Comparison of X-ray diffraction pattern of high Mw PFO samples annealed at 170 o C in Ar and in air for 3 and 15 hours................................................................95 7-8. PL of high Mw PFO samples annealed in air at different temperatures: 170 o C for 1 hour (solid line), 200 o C for 1 hour (dash dot line), 200 o C for 3 hours (dash line), and 240 o C for 1 hour................................................................................................96 7-9. The X-ray diffraction pattern of annealed high Mw PFO samples corresponding to the samples shown in Figure 7-8..............................................................................97 7-10. XRR spectra acquired from spin-coated high Mw PFO thin films. Solid line and dash line represent the experimental data and the fitted results (one layer model) respectively...............................................................................................................98 xiv

PAGE 15

7-11. XRR spectra acquired from spin-coated low Mw PFO thin films. Solid line and dash line represent the experimental data and the fitted results (one layer model) respectively...............................................................................................................99 7-12. XRR spectra acquired from spin-coated low Mw PFO thin films. Solid line and dash line represent the experimental data and the fitted results (two layer model) respectively.............................................................................................................100 8-1. The proposed model of ruptured polymer film formation during the spin-coating process due to the surface-tension gradient induced by water...............................105 8-2. The proposed model of ruptured polymer film formation during the spin-coating process due to phase separation.............................................................................106 xv

PAGE 16

Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy THE BEHAVIOR OF LIGHT-EMITTING POLYMERS: FILM FORMATION, OPTICAL PROPERTIES, AND DEGRADATION By Shyh-Chyang Luo August 2005 Chair: Elliot P. Douglas Major Department: Materials Science and Engineering The properties of light-emitting polymer thin films prepared via spin coating from different organic solvents were studied. The surface analysis as observed by atomic force microscopy, scanning microscopy, and x-ray reflectivity has shown that the structure of polymer thin-films is controlled by the properties of organic solvents. The photoluminescence (PL) spectrum of polymer thin-films and polymer solutions has shown the correlation between the chain conformation and light-emitting properties. A combination of viscosity measurements is used to illustrate the solvent effects on the optical properties of light-emitting polymers in solutions and thin films. The solvent-induced film structure of poly(n-vinyl carbazole) (PVK) and poly(2-methoxy-5-(2-ethyl-hexyloxy)-1,4-phenyl vinylene) (MEH-PPV) thin films on indium tin oxide coated glass was examined. The spin-coated polymer thin films are not in thermodynamic equilibrium; rather, the film properties are affected by the dynamics of the spin-coating process. The water in tetrahydrofuran induces rupturing of polymer xvi

PAGE 17

films during the spin-coating process. Solvents with a high-evaporation rate lead to high-surface roughness due to Bnard-Marangoni convection. Results show that the surface roughness and structure of the films are dominated by the dynamics of the film-formation process rather than the thermodynamic interactions between the polymer and solvents. Two degradation phenomena are studied and discussed in this work. The first manifestation of degradation is the bubble formation on the polymer light-emitting diodes. Water can diffuse through pinholes to reach polymer layers. Heat and electrical bias can induce the reaction between water and metal contacts, which leads to the bubble formation. The second manifestation of degradation is the thermal-induced oxidation of poly[9,9-dioctylfluorene] (PFO). The oxidation of PFO can lead to an extra red-shifted emission and the crosslinking of polymer chains. xvii

PAGE 18

CHAPTER 1 INTRODUCTION The first report of light-emitting diode devices with conjugated polymers as the active layer by Burroughes et al. in 1990 [1] resulted in extensive investigation of polymer light-emitting diodes (PLEDs). The demonstration of PLEDs fabricated by casting the polymer film from solutions [2] and their successful application onto flexible substrates both suggest how PLEDs possess a great potential for display technology [3]. These devices show many advantages including low cost of processing, low-power consumption, and a flexible display. So far, many studies have been done to improve the devices efficiency and lifetime [4-12]. The optical properties of light-emitting polymers are determined by their chemical and physical structures, which are the configuration and conformation respectively of polymer chains. The light-emitting color can be tuned by changing the chemical structure of conjugated polymers. Typical examples are poly (p-phenylene vinylene) (PPV) [1,2] and poly (fluorene) (PF) [13,14]. The photoluminescence (PL) and electroluminescence (EL) of PPV-related polymers show a red-orange range and PF-related polymers show a blue-green range. White light PLEDs can be achieved by synthesizing copolymers [15] or by using polymer-polymer, or polymer-phosphor blend systems [16-18]. More recently, polysilane-related polymers show EL in the UV and NUV range [19-21]. Different from PPV or PF, EL from polysilane is due to the -conjugation instead of -conjugation on the main chain. Additionally, a few non-conjugated polymers, such as poly (9-vinyl 1

PAGE 19

2 carbazole) (PVK) can also generate EL [22]. PVK can emit blue light via radiative decay from carbazole chromophores. Emission wavelength is also affected by the conformation of the polymer chains. Steady red-shifted absorption and emission of oligomers have been shown to occur as the conjugation length increases [23-25]. Change of the polymer chain conformation can also affect the conjugation length. An extended chain conformation results in a longer conjugation length. Studies have already shown the emission wavelength as being affected by solvents [26-29], blending with other polymers [30], and the evaporation process [31] due to the change of the chain conformation. The quantum efficiency, life time, and turn-on voltage were also seen as being affected by the morphology of polymers[28, 32-34]. In this research, three different light-emitting polymers, including two conjugated polymers, poly [2-methoxy-5-(2-ethylhexyloxy)-1,4-phenylenevinylene] (MEH-PPV) and poly[9,9-dioctylfluorenyl-2,7-diyl] (PFO), and one non-conjugated polymer, PVK, were used. The research topics include film formation, optical properties, and the degradation mechanisms of polymer and PLEDs. Chapter 2 will begin by providing the general background of light-emitting polymers, including the physics of light generation as well as the mechanism and fabrication of PLEDs. Second, previous research about the relationship between polymer morphology and the PLED performance will be illustrated. Spin-coating is the most general method for thin-film fabrication. During spin-coating processes, Marangoni convection can lead to high surface roughness. Details of the Marangoni effect is additionally delineated in chapter 2, and results of this phenomenon on light-emitting polymer systems are shown in chapter 3.

PAGE 20

3 In chapter 4, the rupture of polymer thin-films during spin-coating processes is illustrated. Previous studies have shown the dewetting of polymer thin-film after the annealing process due to the high interface tension between polymers and substrates. Unlike these studies, we found the formation of rupture of polymer thin-films during spin-coating processes. Thus, an alternate mechanism of this phenomenon is proposed. Chapter 5 primarily shows the solvent effect on the optical properties of light-emitting polymers. Due to the interaction between polymers and solvent molecules, the chain conformation can be changed by dissolving in different solvents. The change of a polymer conformation leads to the shift of PL of that polymer solution. According to previous studies, the conformation of polymer chains after the spin-coating process is affected by the polymer conformation in solutions, which means the PL of the polymer thin film is also affected by the solvents. We also show the concentration and annealing effects on PL in this chapter. Chapters 6 and 7 present the influence of degradation of the electrode contacts in PLEDs and the oxidation of light-emitting polymers respectively. The cause of contact degradation is mainly due to the reaction between metal electrodes, like Ca, and water or oxygen. The oxidation of PFO leads to an extra emission peak. Chapter 8 is a summary and some suggestions for future work.

PAGE 21

CHAPTER 2 REVIEW OF LITERATURE 2.1 Fundamentals of -Conjugated Polymers 2.1.1 A Brief History of -Conjugated Polymers The first report about doped polyacetylene showing high conductivity in 1977 [35] opened a new field in polymer applications. Electronic polymers show the same featurea backbone consisting of alternating single and double bonds. Polymers with this bonding structure are called -conjugated polymers. Before the 1990s, the main focus of this field had been studies of the mechanism of charge transport and the synthesis of new conjugated polymers with higher conductivity and stability. Since the first PLEDs were successfully fabricated in 1990 [1], the application of semiconducting (undoped) forms of these polymers attracted lots of attention. These applications include PLEDs, solar cells [36], sensors [37], and thin-film transistors [38]. Figure 1 shows the structure of the conjugated polymers studied to examine their conducting and semiconducting properties. Figure 2-1. Structure of conjugated polymers: a) cis-polyacetylene; b) trans-polyacetylene; c) polythiophene; d) poly(phenylene vinylene); e) poly(pyridyl vinylene); f) poly(fluorene). 4

PAGE 22

5 2.1.2 Photophysics According to molecular orbital theory, a conjugated system contains and orbitals. By linking these conjugated molecules together, the disparate and orbitals will merge into continuous bands, which is similar to the valence band and conducting band in the semiconductor. The energy spacing between the conducting and valence band is called the band gap. For conjugated polymers, band gaps are determined by their chemical structures (configurations) and physical structures (conformation). When a photon is absorbed by conjugated polymers, an electron is excited from the valence band to the conducting band and a hole is generated in the valence band. Several possible excited states can generate after that. Generally, an electron-hole pair forms a singlet exciton, which means the electron and hole are bound on the sample chain by their electrostatic attraction [39]. Excitons migrate to sites with lower energy (i.e. segment with longer conjugated length) before decaying and lead to red-shifted emission [40]. An excimer (short for excited dimer) is a complex between the excited state of a molecule and another molecule of the same type that is in its ground state [41]. The ideal separation is in the range of 0.3 to 0.4 nm. The formation of excimers is accompanied by a geometry distortion and leads to lower energy [41]. Because of the lower energy in the excimer state, the emission is strongly red-shift. Different from excimers, aggregate states mean that the electron is delocalized over two chains [42]. The delocalization then leads to a lower energy state and a reduced PL efficiency. A lot of examples of aggregates in conjugated polymers have been shown [27, 28, 43, 44]. Film samples show strongly red-shifted emissions with much lower quantum efficiency compared to solution samples.

PAGE 23

6 Another related state is polarons. Formation of a polaron requires sufficient distance between the electron and the hole such that they do not interact to form excitons [45]. The electron and hole then separate onto different chains. Alternately, the electron and hole may be captured by an impurity that provides a lower energy state. This process is fundamental for the application of solar cells. Since polaron pairs are non-radiative, the formation of these states results in reduction of quantum efficiency of luminescence. Even if the excimers and aggregates are radiative, the quantum efficiency also degrades because of a longer lifetime, which allows for non-radiative decay. Moreover, the emission is strongly red-shifted. To increase the quantum efficiency of luminescence, the formations of polaron pairs, excimers, and aggregates have to be reduced. This goal can be achieved by changing the chemical structures of polymers, including inserting another segment and shortening the conjugation length [46], to reduce the interchain interaction. The interchain interaction can also be reduced by using a polymer blend system [47] or by varying the solvents and thermal treatments [48]. More details will be illustrated in the following section. 2.2 Polymer Light-Emitting Diodes (PLEDs) Figure 2 shows the general structure of PLEDs. The light-emitting polymer is sandwiched between an anode and a cathode. Indium-tin-oxide is the most common of materials for anodes because of its high-work function and its transparency. Several low-work function metals are used as cathodes, including Mg, Al, or Ca. After PLEDs are biased, the electrons are injected from the cathode to the conducting band of the polymer, and the holes are injected from the anode to the valence band of the polymer, as shown in Figure 2-3.

PAGE 24

7 Cathode Anode Substr at e Polymer Figure 2-2. General structure of PLEDs. Afterwards, the electrons and holes recom b ine in the polym er to for m excitons. The excitons then decay and em it light. Sin ce the cathodes are not tr ansparent, light can only em it from the anode side of PLEDs. Electron Injection Figure 2-3. The m echanism of EL in PLEDs. Cathode Polymer Recombination Anode + Hole

PAGE 25

8 2.2.1 Solvent Effects As noted in the previous section, the interchain interactions of conjugated polymers are of great interests because the PL and EL strongly depend on the chain conformation of these polymers. Nguyen et al. [28] measured the PL of MEH-PPV dissolved in two different solvents, chlorobenzene (CB) and tetrahydrofuran (THF), as shown in Figure 2-4. The PL of MEH-PPV in the CB solution is red-shifted compared to the THF solution. From dynamic light scattering, the study showed the MEH-PPV chains in the CB have more of an extended conformation than in the THF. This result revealed that the degree of polymer interchain interaction increases when the polymer is dissolved in a better solvent and is in more of an extended chain structure. Figure 2-4. Absorption and PL of MEH-PPV in CB and THF solutions. They further proved [49] the morphology of the cast polymer film is related to the polymer morphology in solution. In figure 2-5, the film coated from CB shows a red-shifted emission compared to the film coated from THF at the same concentration and spin-speed.

PAGE 26

9 Figure 2-5. a) PL of CB solution and thin films; b) PL of thin films coated from CB and THF solutions. That means the degree of interchain interaction in the cast film can be affected by solvents [50]. Liu et al. [33, 34] also showed the relationship between polymer morphology and device performance including turn-on voltage, quantum efficiency, and current injection efficiency, as shown in Figure 2-6.

PAGE 27

10 Figure 2-6. Characterization of PLEDs: a) I-V curve and b) B-V curve for DCB and THF devices having similar thickness. PLEDs prepared from the good solvent, dichlorobenzene (DCB), show a higher current density and a lower turn-on voltage compared to devices from the poor solvent, THF. This variance is due to the different chain conformation in the polymer thin-films. Said result indicates that contacts between anodes and polymers are sensitive to the chain conformation. Polymer chains with more extended conformation can improve the contact between polymers and substrates, which lead to lower turn-on voltage and higher efficiency of charge injection from electrodes.

PAGE 28

11 2.2.2 Annealing Effects Figure 2-7. PL of MEH-PPV films annealed at different temperatures. Thermal annealing also plays an important role in controlling the chain conformation and the film morphology. Since the polymer chains in spin-coated thin

PAGE 29

12 films are not in thermodynamic equilibrium, thermal annealing can assist in the movement of polymer chains toward a thermodynamic equilibrium and enhance the packing of polymer chains, which leads to the increase of interchain interaction. Therefore, the PL of annealed thin films shows a red-shifted emission due to a generation of more interchain interaction species such as aggregates and excimers, as shown in Figure 2-7(a) [51]. In Figure 2-7(b), relative PL quantum efficiency decreases with an increase of the annealing temperature due to the generation of interchain interaction. However, instead of showing decreasing EL efficiency, Lee and Park [52] and Liu et al. [51] found the EL efficiency is much enhanced when devices are annealed as shown in Figure 2-8. Figure 2-8. EL vs. current of the devices annealed at different temperature after Al deposition. The annealing at higher temperatures may improve the interfaces between electrodes and polymers and improve the electronor hole-injection efficiency, which

PAGE 30

13 may lead to higher EL efficiency despite the generation of more chain-chain interactions. Furthermore, Kim et al. [53] also showed that thermal annealing can improve the life-time of PLEDs. The annealing of polymer films can improve the packing of polymer chains and reduce the defects, which leads to higher thermal stability when an electric field is applied. 2.2.3 Degradation of PLED Devices The short lifetime and poor durability of PLEDs are critical issues for their commercial applications. Some degradation mechanisms have been proposed for PLEDs, especially for devices made using poly(phenylene vinylene) [54-57] and polyfluorene [58]. The device lifetime may be limited by photo-degradation during its operation [59,60]. Several other factors have been reported to increase the rate of degradation of polymer materials, including oxygen exposure [61], photo-irradiation [62], and electrical-field induced aging. [55,58,63]. In addition to the polymer itself, degradation of its metallic cathode causes degradation of the device [63-67]. PLED devices have been reported to exhibit black spots, as shown in Figure 2-9. When PLED devices are operated for two hours, a lot of small black spots emerge and distribute randomly on the device as shown in Figure (a). These black spots then expand when the operation of PLED devices continue. Figure 2-9 (e) shows that most of this PLED device is covered by black spots and can not emit light after 40 hour operation.

PAGE 31

14 Figure 2-9. Evolution of black spot formation in air: a) t = 2hrs; b) t = 10 hrs; c) t = 20 hrs; d) t = 30 hrs; e) t = 40hrs. Black spots are non-emitting areas that severely reduce the light output. Several mechanisms have been proposed to explain the formation of black spots, including metal from the electrode migrating through the polymer layer under electrical stress. A metallic short is created between the electrodes leading to large currents that carbonize the polymer, thus forming a black spot [63-65]. Local heating due to large currents was reported to also stimulate bubble formation due to reactions between metal electrodes and

PAGE 32

15 polymers. Pinhole defects in the cathode are commonly founds at the center of black spots. Indeed, pinholes could form either from particulate contamination during thermal evaporation or from voids in the granular structure of the metal cathode thin films [64], as shown in Figure 2-10. Figure 2-10. Schematic diagram showing dark spot induced by pinholes. By using uniformly sized silica micro particles Lim et al. made devices with size-controllable pinholes [66]. The relationship between the growth rates of dark spots and the pinhole size is shown in Figure 2-11. Figure 2-11. The linear growth of dark spot with various particles.

PAGE 33

16 The growth rates of dark spots are linearly correlated to the pinhole sizes, which indicate the dark-spot formation is due to the pinholes on the protective layer, which create a path for water or oxygen. Kim et al. [67] also pointed out that the transport of water and oxygen through the pinholes to reach the cathode/polymer interface can lead to oxidation of the metal cathode and formation of a higher resistance poly(3,4-ethylenedioxythiophene)-poly(4-styrenesulfonate) (PEDOT-PSS) layer, as shown in Figure 2-12. Figure 2-12. Proposed electrochemical half-cell reactions showing the oxidation of metal (M) and the reduction of doped PEDOT. The oxidation of active metal and reduction of PEDOT generate non-conducting materials near pinholes, which increase the local resistance dramatically and cut off the current density. These non-emissive areas, specifically the dark spots, then cause a reduction of device active areas and luminescence outputs.

PAGE 34

17 2.3 Defects of Spin-Coated Polymer Thin-Film The spin-coating process has been studied for many years and many models have been proposed [70-73] to explain the mechanism of this process. Due to the complexity of the spin-coating process itself, simplified assumptions have been made in these models. Most studies have focused on the relationship between film thickness and various parameters, including the solution concentration, viscosity, evaporation rate, and spinning speed [70-75]. Some important factors, like the interaction between the polymers and solvents during evaporation, have been ignored. Recently, the influence of the type of solvents used on the structure of polystyrene thin films on Si wafer has been reported [76,77]. Figure 2-13 and 2-14 show the surface structure of polymer thin-films coated from solvents with a high-evaporation rate. Figure 2-13. Polystyrene coated from THF solution.

PAGE 35

18 Figure 2-14. Polystyrene coated from cyclohexane and acetone mixed solution. These studies showed that polymer films which were coated from high-evaporation-rate solvents exhibited a higher roughness surface due to the Marangoni instability [78]. 2.3.1 Bnard-Marangoni Convection In the classical thermocapillary Marangoni convection, rapid evaporation causes local cooling at the surface. The temperature difference between the surface (T 1 ) and the interface (T 2 ) leads to a surface tension gradient and drives convective flows, as shown in Figure 2-15. Figure 2-15. The Marangoni convection.

PAGE 36

19 The formation of this instability is g overned by the competition among surface tensio n, thermal diffusion, and viscosity, which can be expressed by the Marangoni number Ma: aTHTMa )/( Equation 2-1 where )/(T is the temperature derivative of the surface tension; T is the is temperature difference between the bottom and surface of the liquid; H the filmthickness; and and a are the viscosity and the thermal diffusivity of the system, respectively. The critical value of Ma which can trigger instability is around 80 [79]However, the classical Marangoni convection is proposed on the assumption that the temperature gradient is linear through the solution layer. During the spin-coating process, the rapid evaporation near the solution surface leads to non-uniform tempgradients. The thermocapillary effects upon this development are expressed as: erature a THT2)/( Ma Equation 2-2 More recently Haas et al. [80] have determined the conditions that would lead to thermMa ient and is expressed as: ocapillary convection during the spin-coating process. Due to the squared dependence of H, the thermocapillary driven convection can be induced within a sufficiently thick solution layer of high-evaporation rate. For thin-solution films, would be too small and thermocapillary-induced-convection thus becomes unstable. Birnie [81] has pointed out that the concentration effects (solutocapillary) are more important than the temperature gradient (thermocapillary) during the spin-coating process. The driving force is a composition gradient rather than a temperature grad

PAGE 37

20 Da CHTMa2)/( Equation 2-3 where C is the relevant composition variable and D is the diffusion rate of the component driving the composition dependent se another potential limitation for their use in -films can be deposited on non-wetting substrates via the spin-c silicon es at ), and n urface-tension change. Evaporation of solvents during the spin-coating process generates the composition gradient at the solution surface and leads to convection. All of these instabilities induced by the differences of surfactension are called the Marangoni effect. 2.3.2 Dewetting Phenomenon Dewetting of polymer thin-films is PLEDs. Although polymer thin oating process, these polymer films are usually unstable or metastable. Perturbations, like heat loading, lead to a relaxation towards thermodynamic equilibrium. Many studies have been done, especially for the system of polystyrene (PS) on[82-86]. In the beginning, the polymer films tear up and holes form, as shown in Figure 2-16. These holes then expand gradually. The dark part around the holes in these imagis the rim. Finally, the polymer films are totally dewetted to form polymer droplets, which are in thermodynamic stability on the Si wafer, as shown in Figure 2-17 [86]. The general mechanisms [84,86,87] to describe the dewetting process due to heloading include spinodal dewetting (by the amplification of capillary wave fluctuation ucleation (from defects) and growth.

PAGE 38

21 Figure 2-16. Micrograph of different stages of polystyrene dewetting on Si wafer: a) 15 mins and b) 1hr. Figure 2-17. The droplets left on the Si wafer.

PAGE 39

22 Recent research [89] also has shown dewetting when aqueous polymer solutions were coated on glassy hydrophobic substrates. In these cases, the film dewetting occurred during evaporation of the solvent. The unstable aqueous films broke up through the nucleation of holes, and the growth of these holes led to a fingering pattern, as shown in Figure 2-18. Figure 2-18. Film structure of dewetted thin films: a) A dewetting pattern centered at a defect (50 x 50 m) and b) pattern at high magnification (5 x 5 m).

PAGE 40

CHAPTER 3 MARANGONI EFFECTS ON POLYMER THIN FILM STRUCTURE 3.1 Introduction So far, most studies of polymer thin-film morphology and its instability phenomena have been limited to polystyrene deposited on silicon. No studies about the instability during thin-film formation of light-emitting polymers have been reported. Since the quality of polymer thin-films is a critical factor in the performance of polymer electronic devices, it is necessary to understand the instability phenomenon in light-emitting polymer systems. In this chapter, the focus is on the relationship between solvents and the film structure of poly(n-vinyl carbazole) (PVK) on indium tin oxide (ITO) glass. Containing the positive charge transporting carbazole pendant groups [90], PVK is now extensively used as the hole-transport layer in low-molecule weight organic light-emitting diodes (OLEDs) [91]. Recent studies have also shown that PVK is a candidate for blue polymer light-emitting diodes (PLEDs) [92,93]. These similar results can also be observed by using poly(2-methoxy-5-(2-ethyl-hexyloxy)-1,4-phenyl vinylene) (MEH-PPV). MEH-PPV, which is a common light-emitting polymer, has attracted much attention recently due to its high solubility in many organic solvents. In this chapter, the instability phenomenon induced by Marangoni effects on the light-emitting polymer thin films is developed. The other instability phenomenon, rupture of polymer thin-films during the spin-coating process, will be illustrated in chapter 4. The chemical structures of PVK and MEH-PPV are shown in Figure 3-1. 23

PAGE 41

24 Figure 3-1. Chemical structure of a) PVK and b) MEH-PPV. 3.2 Experiment 3.2.1 Sample Preparation ITO coated glass (CG-51IN, ITO layer = 120~140 nm, Delta Technologies, Limited) was cleaned in ultrasonic baths of deionized water, ethanol, and acetone respectively for 15 minutes each before being used. PVK (Mw = 90,000, Polydispersity = 2.40, Scientific Polymer Product) and MEH-PPV (Mn = 40,000-70,000, Aldrich) were used as received. PVK solutions with fixed concentration (10 mg/mL) were prepared. MEH-PPV solutions with lower concentrations (2-7 mg/mL) were prepared due to its lower solubility. All solutions were filtered (0.2 m PTFE filter, Fisher) before being used. The physical properties and Mark-Houwink parameters for the five solvents used are shown in Tables 1 and 2. Spin coating (time = 30 seconds) was conducted in ambient laboratory air.

PAGE 42

25 Table 3-1. Characteristic data of solvents at room temperature. Solvent p (mmHg) (Debye) (mPa s) (mN/m) (g/cm 3 ) THF 129 8 1.75 0.456 26.40 0.889 *M Chloroform 160 11.6 1.04 0.537 26.67 1.492 0.8 Benzene 75 5.1 0 0.604 28.88 0.874 0.18 Toluene 22 2.24 0.37 0.560 27.93 0.865 0.05 TCE 8 0.65 1.32 1.437 35.58 1.586 0.3 p : vapor pressure; : relative evaporation rate (BuAc=1); : electric dipole moment; : viscosity; : surface tension; : density; : solubility (g/100g water), M: miscible Table 3-2. The Mark-Houwink parameters for PVK at 25 o C. Solvent K a THF 1.55 0.65 Chloroform 1.36 0.67 Benzene 3.35 0.58 Toluene 7.72 0.50 TCE 1.29 0.68

PAGE 43

26 3.2.2 Characterization 3.2.2.1 Atomic Force Microscopy (AFM) The film structure and root mean square (RMS) roughness were measured using tapping mode AFM (Digital Instruments Nanoscope III). A single silicon probe with a tip radius of 5-10 nm was used. The cantilever was oscillated at or near its resonance frequency between 300 to 350 kHz. The set point for the feed-back control was between 0.7 and 4V. Height images, which represent the topography of the sample surface, were recorded with a scan range from 10 to 100 m. All AFM images were obtained in ambient air. 3.2.2.2 X-Ray Reflectivity (XRR) XRR measurements were performed with a Panalytical XPert MRD System. A ceramic filament tube with a Cu-target was used as the X-ray source ( = 1.54). The scan rate was 0.02 degree/sec. A parallel plate collimator was attached to the detector to improve the peak-to-noise ratio. The reflectivity curves provide information about the film structure perpendicular to the surface [97,98]. The acquired XRR spectra were fitted using the WinGixa software package from Panalytical to obtain the film thickness, density profile, surface and interface roughness. 3.3 Results and Discussion 3.3.1 High-Evaporation-Rate Solvents. Table 3-1 shows that THF, chloroform, and benzene have a relatively higher evaporation rate than toluene and TCE. For PVK, THF and chloroform are good solvents (a 2/3), whereas benzene is a poorer solvent, as shown in Table 3-2. Figure 3-2 shows the film structure of PVK thin films prepared from THF and chloroform solutions.

PAGE 44

27 Figure 3-2. Film structure of polymer thin-film observed with optical microscopy: (a) prepared from THF solution, center area; (b) prepared from THF solution, away from center area; (c) prepared from chloroform solution, center area; (d) prepared from chloroform solution, away from center area. Length of scale bar 200m. The film structure from benzene is similar to the film from chloroform solution. The film structure near the samples center (Fig. 3-2(a) and 3-2(c)) shows a cellular structure, which is similar to Bnard cells [99] caused by Marangoni convection. The film structure away from the center shows striations (Fig. 3-2(b) and 3-2(d)) which commonly occur in spin-on thin films prepared with high-evaporation-rate solvents. Comparing Fig. 3-2(a) and 3-2(b) or 3-2(c) and 3-2(d), the film prepared from the THF solution shows a much higher contrast than from the chloroform solution, which indicates that the polymer film from the THF solution is rougher than from the chloroform solution Figure 3-3 shows AFM images of thin film surfaces prepared from chloroform.

PAGE 45

28 Figure 3-3. 3D tapping mode AFM image of PVK film surface prepared from (a) chloroform solution, cell structure near center area; (b) chloroform solution, striation structure away from center area. These images show a clear cell structure in the center part of the polymer film (Fig. 3-3(a)) and striations away from the center (Fig. 3-3(b)) although the polymer films are locally smooth. These results show that Marangoni convection has changed the topography of these polymer films. The structure of polymer thin-films from the benzene

PAGE 46

29 solution is similar to the films from chloroform. Different from the film structure of polymer films prepared from the THF solution, the polymer films prepared from the chloroform and benzene solutions exhibit continuous film surfaces instead of the formation of holes or polymer droplets. These results show that polymer films from chloroform and benzene, which are immiscible with water, do not show polymer dewetting during the spin-coating process. We can understand these results by considering the effect of Marangoni convection. During the spin-coating process, the evaporation of the organic solvent induces the composition and temperature gradients at the same time in the thin solution film, as shown in Fig. 3-4. Figure 3-4. The Marangoni convection model (including composition-gradient-driven and temperature-gradient-driven processes) for the spin-coating process. T 1 and C 2 represent the temperature and solution concentration near the surface of the solution layer; T 2 and C 1 represent the temperature and solution concentration near the interface between the solution layer and the ITO substrate. The evaporation of the solvent leaves the solution layer with a higher polymer concentration near the surface. In addition, the evaporation can also induce cooling near the surface. Because surface tension is a function of solution concentration and

PAGE 47

30 temperature, the concentration and temperature gradients will change the surface tension during the spin-coating process, inducing flow near the surface of the thin solution film. The high-surface-tension area pulls the solution from the low-surface-tension area, leading to Marangoni convection. According to the reviews in Chapter 2, the concentration effects will be more important for this work due to the film thickness that results from spin coating. The thickness in our system is smaller than 1 m, which leads to a small Ma in equation 2-2 and unstable thermocapillary-induced convection. The composition-gradient-driven convection from equation 2-3 dominates in our system. Figure 3-2 shows that this is because the topography of polymer films from THF is not only affected by the Marangoni convection but also, as illustrated further in chapter 4, the rupture of the polymer thin film during the spin-coating process. 3.3.2 Low-Evaporation-Rate Solvents. From Table 1, toluene and TCE show relatively low-evaporation rates. The center part of the film coated from TCE under the optical microscope is shown in Figure 3-5. Figure 3-5. Film structure of PVK thin film observed with optical microscopy prepared from TCE solution, center area. The image shows the film coated from TCE is very smooth without cell structure and striations. Films coated from toluene also show the same film structure as those

PAGE 48

31 coated from TCE. The image from tapping-mode AFM (Figure 3-6) also shows smooth surfaces. Figure 3-6. 2D tapping mode AFM image of PVK surface prepared from toluene solution. These results show that Marangoni convection is minimal in low-evaporation-rate solutions including both a good solvent (TCE) and a poor solvent (toluene). The low-evaporation rate decreases the composition gradient near the surface of the solution film, which leads to a lower Ma value based on equation 2-3, and therefore no Marangoni convection occurs during the spin-coating process. Figure 3-7 shows the RMS roughness measured by AFM.

PAGE 49

32 01234567024681012Evaporation rate (BuAc=1)RMS roughness (nm)ChloroformBenzeneTolueneTCE Figure 3-7. RMS Roughness plotted as a function of evaporation rate of four solvents. The roughness of polymer films from the chloroform and benzene solutions shown here was measured near the cell-structure area. The roughness of the striations is slightly higher than the cell-structure area. These results clearly show the dependence of surface roughness on evaporation rate. During spin coating, the solutions with a high evaporation rate lead to high surface roughness due to the Marangoni effect. For making MEH-PPV polymer thin-films, chlorobenzene (CB) is substituted due to the low solubility of MEH-PPV in toluene. The boiling point of CB (132 o C) is higher than toluene, which indicates CB is also a low-evaporation-rate solvent. The structures of MEH-PPV thin films coated from THF, chloroform, CB and TCE as imaged by optical microscopy are shown in Figure 3-8. The surface topography of MEH-PPV thin films from chloroform and TCE are shown in Figure 3-9. From these results, the structures of MEH-PPV thin films show the same trend as PVK thin films, which are controlled by the Marangoni effect and dewetting during the spin-coating process.

PAGE 50

33 Figure 3-8. MEH-PPV Film structure of polymer thin-film observed with optical microscopy: (a) prepared from THF solution, center area; (b) prepared from THF solution, away from center area; (c) prepared from chloroform solution, center area; (d) prepared from chloroform solution, away from center area; e) prepared from CB solution; prepared from TCE solution.

PAGE 51

34 Figure 3-9. 2D tapping mode AFM image of MEH-PPV film surface prepared from (a) chloroform solution, cell structure near center area; (b) TCE solution, smooth surface near center area. 3.3.3 XRR Measurement. Typical XRR spectra acquired from the polymer films are shown in Fig. 3-9, while results from the simulations, including RMS roughness, thickness, and density, are shown in Table 3-3.

PAGE 52

35 1234 5 TCETolueneBenzeneChloroformLog (Intensity) (arb. unit) degree Figure 3-10. X-ray reflectivity spectra acquired from PVK thin films prepared from chloroform, benzene, toluene, and TCE solutions. Experimentally measured data are shown as a solid line; fitted results are shown as a dotted line. Table 3-3. Characteristics of PVK films obtained from XRR measurements. Solvent RMS Roughness () Density (g/cm 3 ) Thickness () Chloroform 24 0.88 151 Benzene 18 0.96 852 Toluene 9 1.23 811 TCE 9 1.29 320

PAGE 53

36 The fitting results (dotted line) are in favorable agreement with the measured spectra (solid line). The surface roughness values displayed in Table 3-3 show the same trend as the values from AFM measurements. The polymer films deposited using high-evaporation-rate solvents, such as chloroform and benzene, lead to significantly higher surface roughness than using low-evaporation-rate solvents, like toluene and TCE. The densities of polymer films deposited using the low-evaporation-rate solvents, toluene and TCE, showed values similar to solid bulk PVK (1.2g/cm 3 ). One possible explanation for a lower density of polymer films from the chloroform and benzene solutions is the high surface roughness. The other possibility is due to the trapping of defects or free volume in polymer films during the fast evaporation of the solvent and the presence of some gas trapped near the surface of the polymer films due to the Marangoni convection. The film densities did not show any obvious differences between a good solvent (TCE) and a poor solvent (toluene). Comparing the thickness of the polymer films, the films from the chloroform and TCE solutions show lower thickness compared to other solvents. According to previous studies [73], the spin-coating process can be divided into three phases. During the first phase, most of the solution is flung from the substrate. Only a much thinner and uniform film is left. At this time, the spin speed is high and shear thinning may occur. The film structure is dominated by the inertia of the solution. The second phase begins when the inertial effects become negligible and the film thickness is determined by the balance of viscosity and centrifugal forces. In the final phase, further fluid is lost due to solvent evaporation. From this model, solutions with higher viscosity will lead to thicker

PAGE 54

37 polymer films. From Table 3-2, PVK dissolved in ideal solvents exhibits higher viscosity than in poor solvents ([] benzene ~ [] toluene < [] chloroform < [] TCE ). However, the higher density of chloroform and TCE lead to higher centrifugal forces and cause much more solution to be flung from the substrate in the first and second phases during spin coating. Our experimental results show that films from benzene and toluene are thicker than from chloroform and TCE. The lower viscosity of the benzene and toluene solutions would lead to thinner films, while the lower density would lead to thicker films; comparison of these predictions to the experimental results indicates that the density of solvents (centrifugal force) dominates the polymer film thickness in this system. The XRR results from MEH-PPV films are shown in Figure 3-10 and simulated values are shown in Table 3-4. 1234 2.5 mg/mL, Chloroform5 mg/mL in Chloroform7 mg/mL, Chloroform7 mg/mL, Chlorobenzene7 mg/mL, TCELog Intensity (arb. unit) Figure 3-11. X-ray reflectivity spectra acquired from MEH-PPV thin films prepared from chloroform, chlorobenzene, and TCE solutions. Experimentally measured data are shown as a solid line of light color, and fitted results are shown as a solid line with dark color.

PAGE 55

38 Table 3-4. Characteristics of MEH-PPV films obtained from XRR measurements. Solvent Concentration (mg/mL) Thickness () Density (g/cm 3 ) RMS Roughness () TCE 7 227 1.32 13 Chlorobenzene 7 391 1.26 15 Chloroform 7 379 0.72 40 Chloroform 5 216 0.70 23 Chloroform 2.5 149 1.28 16 From the results shown in Table 3-4, the MEH-PPV films coated from TCE and CB show lower roughness and higher density than the films coated from chloroform with the same polymer concentrations (7mg/mL). Compared to the film coated from chloroform solutions with different concentrations, films coated from chloroform solutions with lower concentrations have lower thickness. The thinnest film, which was coated from chloroform solution with 2.5 mg/mL concentration, has a similar roughness and density as the films coated from TCE and CB. We can see these results are coincident to the results of PVK films when MEH-PPV films are coated from different solvents with the same concentration. The thinnest film coated from chloroform solution has a similar roughness as films coated from TCE and CB, which indicates Marangoni convection is not induced in this system. From equation 2-3, the Ma number is proportional to the height squared of the thin film. It indicates that when films are too thin, Marangoni convection could not be induced even if solvents with high-evaporation-rate are used.

PAGE 56

39 3.4 Conclusion The polymer film structure is determined by the dynamics of the spin-coating process, which is controlled to a large effect by the solvent properties. The chloroform and benzene solutions, which have a high-evaporation rate, lead to Bnard cell and striation morphology on polymer films and result in high-surface roughness due to composition-gradient-driven Marangoni convection. The film structure of polymer films from good solvents and poor solvents did not exhibit obvious differences, which shows the interaction between polymer chains and solvents is not an important factor in the dynamic instabilities. During the spin-coating process, the solvents with higher density have higher centrifugal forces and lead to more solvent being flung from the substrate and therefore thinner polymer films. We expect these differences in film structure to have an important effect on PLED device performance, depending on the solvent used. In films from THF solution, the large-scale inhomogeneities in the film will lead to inhomogeneous light emission (More details will be illustrated in Chapter 4). In addition, according to previous studies [100,101], one of the intrinsic degradation mechanisms of OLED and PLED devices is due to the diffusion of the metal from the electrode into the organic layers and the quenching of luminescence in the emission zone. In our studies, the polymer films from the chloroform and benzene solutions showed higher roughness and lower density due to Marangoni convection. We believe the inhomogeneous polymer surface could affect the durability of the PLED devices by enhancing the possibility of metal diffusion into the emissive layer.

PAGE 57

CHAPTER 4 RUPTURE OF POLYMER THIN FILMS DURING SPIN-COATING PROCESS 4.1 Introduction As noted in chapter 3, this chapter focuses on another instability phenomenonthe rupture of polymer thin-films. Although ruptured thin films show a similar film structure as dewetted thin films, the mechanism is different from the dewetting phenomenon induced by heat loading after the spin-coating process, because the rupture of polymer thin films described in this chapter occurs during the spin-coating process. In this chapter, results show that the water in THF is an important factor for the rupture of polymer films. Two polymer systems, PVK and MEH-PPV, are used in this chapter. As noted in chapter 2, MEH-PPV has attracted much attention recently due to its high solubility in many organic solvents, which is helpful for investigating the relationship between chain conformation and optical properties. In this chapter, MEH-PPV is used to study the influence of polymer conformation when ruptured polymer films are formed. The effect of film rupturing on the performance of devices is also illustrated. 4.2 Experiment 4.2.1 Sample Preparation Cleaned indium tin oxide (ITO) coated glass (Delta Technologies) was used as the substrate. PVK solutions with fixed concentration (10 mg/mL) were prepared. MEH-PPV was dissolved in THF (Aldrich) at three concentrations (2.5, 5, and 7 mg/ml). These solutions were filtered (0.2 m PTFE filter) and spin-coated on ITO glass at 1000, 3000, and 5000 rpm. Dried THF means that THF was dried by molecular sieves (4, 8-12 40

PAGE 58

41 mesh, Aldrich) overnight before being used. To induce dewetting, 2-5% (v/v) water was added to the polymer solutions. Poly(3,4-ethylenedioxythiophene)-poly(4-styrenesulfonate) (PEDOT-PSS, Bayer) water solution was filtered before being used. 4.2.2 Characterization The topology and structure of polymer films were observed by tapping-mode AFM (Digital Instruments Nanoscope III). The surface root-mean-square (RMS) roughness was calculated by the software package included with the AFM. PL spectra were obtained in ambient air with an excitation energy of 490 nm. The emission was collected by a monochromator (Oriel instrument, Electo-optical systems). 4.3 Results and Discussion Figure 4-1 shows SEM images of a polymer thin film prepared from THF. Figure 4-1. SEM image of polymer prepared from THF solution: (a) center area with a low magnification, length of scale bar 200m; (b) away from center area with a low magnification, length of scale bar 10m; (c) center area with a high magnification, length of scale bar 200m; (d) away from center area with a high magnification, length of scale bar 10m.

PAGE 59

42 The images with high magnifications (Fig. 4-1(c), 4-1(d)) show obvious rupturing of the film. Near the center of the film (Fig. 4-1(c)), many holes formed and few polymer droplets formed inside these holes. Away from the center area, striations formed (running from top-left to bottom-right in Fig. 4-1(b)), and polymer droplets formed along with the striations (Fig. 4-1(d)). Compared with the film structure in chapter 3, only polymer films prepared from THF show the ruptured thin film structure. Polymer films from other solutions showed smooth and continuous surfaces. To investigate the basis of the rupture of polymer films during the spin-coating process from the THF solutions, we compared the properties of these five solvents and found only THF shows good miscibility with water (Table 3-1). Since the entire process of solution preparation and spin coating were done in ambient air, it was believed that a small amount of water was dissolved in the THF solution during the spin-coating process even when THF was dried previously. To test the influence of water on the thin-film structure, several THF solutions with different concentrations of water were prepared for film deposition. The SEM images of the film structures are shown in Figure 4-2. Figure 4-2(b) shows the structure of polymer films prepared from dried THF. It shows that many holes formed on the surface without the presence of polymer droplets. This ruptured film structure is similar to the beginning of polymer dewetting due to heat loading, although in this case the rupture occurs due to evaporation of the solvent (discussed further below) rather than the application of heat. As the water content increased, the structure changed from a continuous film with holes to isolated polymer droplets (Fig. 4-2(d)). Finally, polymer droplets replaced the entire polymer film, which is similar to the final stage of polymer

PAGE 60

43 dewetting (Fig. 4-2(f)). These results indicate that the degree of polymer rupturing is related to the concentration of water. Figure 4-2. SEM image of polymer thin-film prepared from THF solution with different H 2 O contents: (a and b) THF dried by molecular sieve for 1 day with a low and high magnification, length of scale bar 200 and 10 m, respectively; (c), (d) 10L of H 2 O added in 5 mL of THF with a low and high magnification, length of scale bar 200 and 10 m respectively; (e), (f) 100l H 2 O added in 5mL of THF with a low and high magnification, length of scale bar 200m and 10 m, respectively. Different from the dewetting phenomenon induced by the heat loading after the spin-coating processes, this polymer rupturing happened during the spin-coating process. Two proposed mechanisms, spinodal dewetting or nucleation and growth, have been used to explain the relaxation processes of metastable or unstable polymer films toward

PAGE 61

44 thermodynamic equilibrium with heat loading. The polymer film rupturing during spin coating is dynamic in nature and cannot be explained by these two mechanisms. Gu, et al. [89] have shown the dewetting of aqueous polymers coated on hydrophobic substrates. Since the ruptured film structure only occurs when some water is added to the THF solution, we need to test if the effect between water and ITO is the cause for this ruptured film structure. We conducted contact angle measurements of water on ITO and found a contact angle around 90 o indicating that our ITO substrate is hydrophobic. To test the effect of water during spin coating of a water-soluble polymer on the ITO substrate, we studied the surface structure of a PEDOT-PSS thin film spin-coated from water solution (Fig 4-3). Figure 4-3. PEDOT-PSS film structure on ITO substrate observed with optical microscopy: (a) low magnification, length of scale bar = 200m; (b) high magnification, showing presence of a hole, length of scale bar = 20m. Although there are some holes on the PEDOT-PSS film, the film structure is totally different from the PVK film from the THF solution shown in Fig 4-1. Recall, as described previously, that the PVK thin films spin-coated from other water immiscible organic solvents, like chloroform or chlorobenzene, show smooth surface. These results

PAGE 62

45 indicate the substrate effect is not the reason for the formation of the ruptured film structure. In our system, polymers are dissolved in THF/water solvent. Comparing the properties of water and THF, water has a higher surface tension (72.8 dynes/cm) than THF (24.8 dynes/cm). Furthermore, polymers, including PVK and MEH-PPV, show poor interaction with water and cannot dissolve in water. Therefore, the process of fast evaporation in this system is very complicated and can be analyzed as follows: in the very beginning, during the spin-coating process, the evaporation of THF induces Marangoni convection which will lead to thicker and thinner parts around the whole polymer film. Due to the higher evaporation rate of THF than water, the solution layer near the surface has a higher water concentration and leads to a higher surface tension gradient, which can potentially drive a substantial convection. Additionally, the precipitation of PVK or MEH-PPV can potentially occur at the same time due to the incompatibility between the water and polymers. These phenomena could be important factors for the formation of ruptured thin films, but the details of the rupture mechanism are still not clear. Two possible mechanisms for the film rupturing during the spin-coating process are proposed and discussed further in chapter 8. In the last stage the thin film develops a few holes in the thick area and small polymer droplets in the thinner area on the ITO substrate. The water-induced dewetting phenomenon was also studied by Bonaccurso et al. [102], who found dewetting of ultrathin polystyrene films induced by saturated water vapor after spin-coating. More recently, Muller-Buschbaum et al. [103] also showed that diblock copolymers dewetted on silicon during the spin-coating process. These results

PAGE 63

46 show the interaction between the polymers, solvents, and substrate plays a significant role in the dewetting phenomenon and the rupture of polymer thin films. Figure 4-4 shows the optical microscopy and the SEM images of dewetted MEH-PPV film structures. Figure 4-4. Structure of ruptured thin films: a) Optical microscopy image of MEH-PPV thin film prepared from 5% H 2 O in THF, striation structure away from center; b) SEM image of same film with higher magnification.

PAGE 64

47 Figure 4-4(a) shows a similar striation structure as PVK films under an optical microscope. Figure 4-4(b) shows the polymer films are ruptured and a lot of polymer droplets formed on the substrate. The AFM image of MEH-PPV thin films prepared from dried THF solution and 5% water in THF solution lead to different surface structures, as shown in Figure 4-5. Figure 4-5. Height contrast of polymer thin-films from (a) dried THF solution with polymer concentration of 5 mg/ml; (b) 5% water in THF solution; (c) height profile of film along the line indicated in (a); (d) height profile of film along the line indicated in (b). The thin films from dried THF (Fig. 4-5(a)) are much smoother than from THF/water solution (Fig. 4-5(c)). Film structures from THF/water solution consist of polymer droplets instead of a continuous polymer thin film. The height profile of the polymer film from the THF/water solution clearly shows the presence of polymer droplets on the ITO substrate (Fig. 4-5(d)). The dewetting of MEH-PPV films is due to the incompatibility of the water and MEH-PPV. During the spin-coating process, THF evaporates faster than water due to its high-evaporation rate, leading to phase separation

PAGE 65

48 into a water-rich phase and a polymer-rich phase. After the THF and water completely evaporate, polymer droplets are formed on the ITO substrate (Fig. 4-5(d)). The RMS roughness of polymer films from dried THF solutions of different polymer concentrations are similar to each other (<5nm) as shown in Table 4-1. Table 4-1. Roughness of MEH-PPV polymer films prepared from the dried THF and THF/water solutions Roughness (nm) Concentration (mg/mL) Spin Speed (rpm) Dried THF 2% H 2 O 5% H 2 O 1000 2.7 32.2 3000 2.2 14.3 7 5000 2.3 12.0 5 1000 3.1 15.8 32.1 2.5 1000 2.0 8.1 11.7 For polymer films spin-coated from the THF/water solutions, the film from a high-polymer concentration solution (7 mg/ml) shows much higher roughness. The roughness also increases as the water concentration increases in the polymer solutions at the same polymer concentration (5 mg/ml). In Table 4-1, the roughness of polymer films from the dried THF solution with different spin-speed is similar. After 2% water is added, the polymer film from the lowest spin-speed shows the highest roughness. According to the model of the spin-coating process described previously, higher concentration and low spin-speed lead to higher thickness. Based on our results, roughness increases dramatically as the thickness of the polymer film increases. The PL measurement of these polymer films are shown in Figure 4-7.

PAGE 66

49 520540560580600620640660680700 2 mg/ml + 5% H2O 5 mg/ml + 5% H2O2 mg/ml 5 mg/ml + 2% H2O 5 mg/ml 7 mg/ml + 2% H2O7 mg/ml Intensity (Arb. Unit)Wavelength (nm)520540560580600620640660680700 5000rpm; 2% H2O added 5000rpm; dried THF 3000rpm; 2% H2O added 3000rpm; dried THF1000rpm; 2% H2O added 1000rpm; dried THF Intensity (Arb. unit)Wavelength (nm) (a) (b) Figure 4-6. PL spectra of MEH-PPV polymer thin-films. Polymer films with higher thickness caused by higher polymer concentration (Fig. 4-6(a)) or lower spin-speed (Fig. 4-6(b)) show higher PL intensity. Comparing the PL of non-dewetted and dewetted polymer films, the addition of water not only lowers the PL intensity, but also leads to red-shifted PL peaks. The red shift is also related to the

PAGE 67

50 thickness of the polymer films. The polymer films with higher thickness (higher concentration or lower spin-speed) show more red shift (Fig. 4-6(a) and (b)) and the increase of water concentration also leads to more red shift (Fig. 4-6(a)). From previous studies, the red shift of PL is due to the enhancement of chain interaction and the formation of aggregates. To identify the cause for this phenomenon in our system, we measured the PL of MEH-PPV in THF solution (7 mg/ml) with different concentrations of water, as shown in Figure 4-7. 520540560580600620640660680700 15% H2O12% H2O10% H2O 7.5% H2O 5% H2O 3% H2O2% H2O 1% H2O Dried THF Intensity (Arb. unit)Wavelength (nm) Figure 4-7. PL spectra of MEH-PPV in THF solutions (7 mg/ml) with different water concentrations. The emission (~550 nm) from polymer solutions does not show a red shift when water concentration is increased up to 7.5%. When the water concentration is higher than 7.5%, an additional lower energy emission at 590 nm appears due to the precipitation of MEH-PPV in the THF solutions. These two separate emissions indicate two different polymer-chain conformations. According to the description by Hu et al. [104] stiff-chain

PAGE 68

51 conjugated polymers can exist in either a non-collapsed defect-coil conformation or a collapsed defect-cylinder conformation. The collapsed conformation leads to a lower energy emission due to the enhancement of chain-chain interaction and the formation of aggregates [105]. Therefore, the emission from 550 nm and 590 nm are assigned to the non-collapsed defect-coil conformation and the collapsed defect-cylinder conformation respectively [105], and thus the precipitation of MEH-PPV causes a coil-cylinder transition. Comparing Figure 4-6 and Figure 4-7, the emission (~570 nm) of the thin films is located between the emission from solution (550 nm) and precipitate (590 nm). This indicates that the red-shifted emission from polymer thin films is mainly due to the enhancement of interchain interaction [106] of polymer chains in the coil conformation instead of the transition of polymer conformation from the coil to the cylinder. In turn, the results show that the polymer conformation in the thin films is dominated by the dynamics of spin-coating rather than thermal equilibrium. Due to the fast evaporation of solvents during the spin-coating process, most polymer chains remain in the coil conformation. Some polymer chains collapse to the cylinder conformation. Polymer chains with the transition conformation between the stable coil and cylinder conformation could also exist in the polymer thin film. The polymer chains with this complex conformation lead to a single and broad emission band shown in Figure 4-6. Since the PL of the polymer solutions does not alter with a change in the water content before precipitation occurs (Fig. 4-7), the red-shifted emission of the polymer films from the THF/water polymer solutions (Fig. 4-6) is proposed to occur during the formation of polymer thin films. The details are discussed as follows:

PAGE 69

52 According to our model, due to the incompatibility of MEH-PPV and water, precipitation of MEH-PPV occurs during the spin-coating process. Higher concentrations of water can generate higher content of precipitation of MEH-PPV in thin films. The precipitate has a cylinder conformation, which has a higher interchain interaction and leads to a red-shifted emission. Because of the fast evaporation rate during spin coating, the polymer chains cannot be transferred to the cylinder conformation in thermodynamic equilibrium like the process of precipitation formation shown in Figure 4-7. Table 4-2 shows the emission peaks as illustrated in Figure 4-6. Table 4-2. PL of MEH-PPV polymer films prepared from the dried THF and THF/water solutions Emission peaks (nm) Concentration (mg/mL) Spin Speed (rpm) Dried THF 2% H 2 O 5% H 2 O 1000 573 583 3000 572 577 7 5000 572 573 5 1000 572 573 577 2.5 1000 571 571 572 *: Precipitate generation. The results show that the red shift of these emission peaks is correlated to the water content in the THF solutions. With increasing water content, the emission peaks are more red-shifted and the emission peaks are located between 550 nm and 590 nm. These results are coincident to the proposed model for the formation of ruptured thin films. Finally, we demonstrate the influence of this dewetting phenomenon on the performance of PLED devices. Figure 4-8(a) shows the layer structure of this PLED

PAGE 70

53 device. A polymer thin-film coated from 2% water in the THF solution (dewetted polymer thin-film) was used as the light-emitting layer. Figure 4-9(b) shows clear striation structure after this PLED device was turn on. Figure 4-8. PLED based on ruptured thin films: (a) The layer structure of PLED device. (b) The light-emitting output after the device is turned on. 4.4 Conclusion In conclusion, we have found that the rupture of polymer films is related to the water in THF. The rupture of the polymer solution could be related to the change of the

PAGE 71

54 surface-tension gradient caused by the increased content of water near the solution surface and the incompatibility between the water and polymers, but the workings of this mechanism are still not so clear and need further study. Compared to smooth polymer films, the ruptured polymer films show a more red-shifted emission. We also observe that the conformation of polymer chains in solution and precipitation are of the un-collapsed and collapsed conformation respectively. The polymer chain undergoes a coil-cylinder transition during the process of precipitation. Therefore, polymer films coated in the THF solution with higher concentrations of water can lead to more polymer chains that have a cylinder conformation and red-shifted emission. These results point out that the water effect is a critical factor in the performance of polymer thin-film devices.

PAGE 72

CHAPTER 5 SOLVENT, CONCENTRATION, AND ANNEALING EFFECTS ON PHOTOLUMINESCENCE OF POLYMER SOLUTIONS AND THIN FILMS 5.1 Introduction As noted in chapter 2, the PL and EL of conjugated polymer thin films is related to fabrication processes, including solvents, concentrations, spin-speeds and annealing processes. This chapter focuses on the study of the relationship between the PL and chain conformation in MEH-PPV and PFO. Some details of the relationship between the PL and the chain conformation in MEH-PPV are illustrated in chapter 2. From light-scattering experiments [28], MEH-PPV shows a more-extended chain conformation in the good solvent, CB, than in the poor solvent, THF. Polymer chains with a more-extended conformation lead to a red-shifted emission due to longer resonance length. Therefore, the PL of MEH-PPV in CB shows a red-shifted emission compared to the PL of MEH-PPV in THF solutions. Grell et al. [13,14] who worked on the PFO system, define good solvents and poor solvents according to solubility parameters. Chloroform and THF are good solvents because of their higher solubility in PFO. Toluene is a moderately good solvent because of its slightly lower solubility, while cyclohexane is a poor solvent because of its low solubility. According to the results, PFO in poor solvents shows an additional red-shifted peak. Their study proposed that this emission is due to a more compact polymer chain conformation in poor solvents, which leads to the generation of aggregates and an additional emission peak. Comparing the results from these two 55

PAGE 73

56 research groups, the controversy on the definition of good solvents and poor solvents exists, as well as the explanation of solvent effects on photoluminescence. In this chapter, the first part focuses on the solvent effect on the PL of MEH-PPV and PFO solutions. The PL of polymers in more than two solvent systems is measured. Combined with the results from the viscosity measurements, more details about the polymer conformation in solutions are given and subsequently are used to explain the relationship between the PL and solvent properties. The second part focuses on the influences of the solvent, the polymer concentration, and the annealing process on the PL of polymer thin-films. 5.2 Experiment 5.2.1 Sample Preparation MEH-PPV (Mn = 40,000-70,000, Aldrich) and PFO (Mw = 52,000, American Dye Source, Inc) were used as received. MEH-PPV was dissolved in organic solvents at several concentrations from 0.1 ~ 10 mg/mL. These solutions were filtered (1 m PTFE filter) and spin-coated on ITO glass at 1000 rpm for 30 seconds. 5.2.2 PL Measurement The PL of polymer solutions was measured in a quartz cell with the excitation energy of 490 nm for MEH-PPV and 380 nm for PFO. The emission was collected by a monochromator (Oriel instrument, Electo-optical systems) with a range from 530 nm to 700 nm for MEH-PPV and 400 nm to 650 nm for the PFO solutions and thin films. 5.2.3 Viscosity Measurement The viscosity was measured with a standard Ubbelohde capillary viscometer in a temperature-controlled water bath at ~ 28 C. Polymer solutions were filtered through a 1.0 m PTFE filter to remove dust particles and other impurities. Polymer solutions were

PAGE 74

57 held in the temperature bath for approximately 15 minutes to reach thermal equilibrium. Typically, the measurement was repeated at least three times. 5.3 Results and Discussion 5.3.1 PL of Polymer Solutions 5.3.1.1 Solvent effects Figure 5-1 shows the PL of the MEH-PPV solutions in four organic solvents: chloroform, CB, THF, and Tetrachloroethane (TCE). 540560580600620640660680700 Normalized PL IntensityWavelength (nm) 0.5% MEH-PPV solutions Chloroform CB THF TCE Figure 5-1. PL of 0.5% MEH-PPV solutions. The PL from the CB solutions shows a red-shifted emission compared to the THF solutions, which is coincident to the results from Nguyen et al. [28]. In addition, the emission peak of the chloroform solution is located between the peaks of THF and CB. The PL from the TCE solutions shows the most red-shifted emission. Figure 5-2 shows results of PFO dissolved in six organic solvents: toluene, THF, chloroform, CB, TCE, and cyclohexaone (from left to right) with a concentration of 0.5

PAGE 75

58 %. From the figure, chloroform shows the highest solubility and cyclohexanone shows the poorest solubility, which is coincident to the results from Grell et al. [13]. Figure 5-2. PFO dissolved in toluene, THF, chloroform, CB, TCE, and cyclohexanone (from left to right). Table 5-1. Solubility parameter and polymer solubility. Solvent Solubility Parameter (cal/cm 3 ) 1/2 PFO solubility (mg/ml) MEH-PPV solubility (mg/ml) Toluene 8.9 <1 <5 THF 9.1 3~5 7~10 Chloroform 9.3 >5 >15 CB 9.5 3~5 7~10 TCE 9.7 ~1 7~10 Cyclohexanone 9.9 <1 <5 Table 5-1 shows the solubility parameters of these organic solvents [107] and the solubility of PFO and MEH-PPV in these solvents based on our test results. Based on the solubility of PFO and MEH-PPV, both polymers have similar solubility parameters near 9.3, and chloroform is a better solvent than the other solvents. From these results,

PAGE 76

59 chloroform is a good solvent compared to the other organic solvents. This conclusion is consistent with the result from Grell et al. [13]. Figure 5-3 shows the PL of the PFO solutions. 400420440460480 Normalized PL IntensityWavelen g th ( nm ) 0.05% PFO solution THF TCE Ch CB Figure 5-3. PL of PFO in THF, TCE, Chloroform and CB. The figure shows that PL from TCE solution has the most red-shifted emission. The PL from THF solution shows the most blue shifted emission, which is similar to the resultsfor MEH-PPV. From these results, we can see the controversy of the definition of good solvents and poor solvents. Assuming that chloroform is the best solvent compared to other solvents, the chain conformation of MEH-PPV and PFO in chloroform should be more extended than in the other solvents. Polymer chains with a more-extended conformation have a longer resonance length and lead to more red-shifted emission, which means the PL of MEH-PPV and PFO in chloroform should show the most red-shifted emission compared to other solvents. In Figure 5-1, however, the PL of MEH-PPV in chloroform solution does not show the most red-shifted emission. In the PFO system shown in

PAGE 77

60 Figure 5-3, the PL from the chloroform solutions does not show the most red-shifted emission, either. Instead, the PL from the poor solvent, TCE, shows a more red-shifted emission compared to the moderately good solvent, CB, and the good solvents, THF and chloroform. To solve this controversy, we need to figure out the chain conformation of these polymers in different solvent systems. The viscosity of a polymer solution is related to the dimension of the polymer chains and can be used to estimate the polymer conformation in solution. The measurement of the viscosity of a solution is based on capillary flow. The fluid flowing through a capillary follows Poiseuilles Law: VlPtr8 4 where r is the radius of the capillary; t is the flow time; V is the fluid volume between the two markers on the viscometer; and P is the pressure drop. For a capillary viscometer in the vertical position, P is generated from gravity: glPlig where liq is the density of the liquid and l is the length between the two markers. The kinematic viscosity is defined as follows: kinematic viscosity tVgrliq8 4 where Vgr8 4 is a constant for a given viscometer and can be determined by measuring a fluid of known viscosity. Therefore, the kinematic viscosity of any polymer solution can be measured by measuring the flow time, t.

PAGE 78

61 Figure 5-4 shows the kinematic viscosities of MEH-PPV in chloroform, CB, and TCE at different concentrations. 012345678024681012 Kinematic viscosity (mm2/s)Concentration (mg/mL) TCE Chloroform Chlorobenze Figure 5-4. The measurement of kinematic viscosities of MEH-PPV in CB, chloroform, and TCE at different concentrations. Polymer chains are isolated and do not interact with each other at low-polymer concentrations. When the polymer concentration is increased, the polymer chains start to entangle with each other and the slope of the viscosity-concentration curve is changed. In this figure, the curve for MEH-PPV in TCE changes slope at a lower concentration compared to CB and chloroform, which indicates that the polymer chains in TCE more easily interact with each other and form entanglements. By measuring the viscosity of the pure solvent, s the specific viscosity, sp can be calculated: 2][cKcvsssp

PAGE 79

62 where [] is the intrinsic viscosity, K v is a second-order coefficient and c is the concentration. Therefore, the intrinsic viscosity, [], can be determined by: cspc 0lim][ The intrinsic viscosity is a quantity characteristic of a polymer. A polymer molecule with a larger dimension has a larger intrinsic viscosity. Figure 5-5 shows sp /c plotted as a function of c for the MEH-PPV solutions. 0.050.100.150.200.250.3020040060080010001200 sp/c (mL/g)Concentration (mg/mL) TCE Chlorobenzene Chloroform Figure 5-5. sp /c plotted as a function of c for MEH-PPV in CB, chloroform, and TCE. This figure shows that MEH-PPV has the highest intrinsic viscosity in TCE and the lowest intrinsic viscosity in chloroform. This indicates that MEH-PPV has a larger dimension and a more extended chain conformation in TCE. MEH-PPV in chloroform shows a smaller dimension and a more compact chain conformation. These results are coincident to the PL shown in figure 5-1. MEH-PPV in TCE shows the most red-shifted

PAGE 80

63 emission and MEH-PPV in chloroform shows the most blue-shifted emission among the three solutions. Therefore, from the PL results, the MEH-PPV and PFO polymer chains do not show a longer conjugated length when they are dissolved in solvents in which they are more soluble. This phenomenon is different from what is seen in traditional polymers. For traditional polymers, like PS or PMMA, polymer chains show a random-walk conformation when dissolved in good solvents, and the polymer chains are extended. When solvents become poorer, the polymer chains contract due to the unfavorable interaction between the polymer chains and the solvent. Therefore, the polymer chains in poor solvents are more contracted. The difference in trend between traditional polymers and conjugated polymers is quite possibly due to the specific chemical structure of the conjugated polymers. MEH-PPV and PFO have rigid aromatic groups as main chains and soft alkyl groups as side chains. The solubility of these polymers is mainly controlled by the side chain groups, whereas the PL is controlled by the resonance length on the main chains. So far, there are very few studies that have focused on solvent effects on the chain conformation of conjugated polymers using computer simulation. Recently, one study [33] was conducted comparing aromatic and non-aromatic solvent effects on the chain conformation of MEH-PPV. Although the results could be used to explain the red-shifted emission of the CB solution compared to the chloroform or THF solutions, it is still not sufficient to explain the red-shifted emission as seen in the TCE solution. The effect of this solvent on chain conformation could be very complicated and may not conform to a simple rule.

PAGE 81

64 5.3.1.2 Concentration effects Figure 5-6 shows the PL of MEH-PPV in chloroform at four different concentrations. 550600650700 PL intensityWavelength (nm) MEH-PPV in Chloroform 0.1 mg/mL 1 mg/mL 5 mg/mL 10 mg/mL Figure 5-6. PL of MEH-PPV in chloroform in concentrations of 0.1, 1, 5, and 10 mg/mL. Compared to the results from viscosity measurement shown in Figure 5-4, Figure 5-6 shows that the PL of MEH-PPV does not change even when the polymer chains start to form entanglements (concentration higher than 5 mg/mL) in chloroform solutions. This indicates that the formation of entanglements and the inter-chain interactions of MEH-PPV polymer chains in solution does not change the chain conformation and affect the PL. MEH-PPV in CB and TCE also show similar results, as seen in figure 5-7. The PL of MEH-PPV in CB and TCE does not change when concentration is changed.

PAGE 82

65 550600650700 PL IntensityWavelength (nm) PL of MEH-PPV CB 0.5 mg/mL CB 2.5 mg/mL CB 7 mg/mL TCE 0.5 mg/mL TCE 2.5 mg/mL TCE 7 mg/mL Figure 5-7. PL of MEH-PPV in CB and TCE. Figure 5-8 shows the PL of PFO in chloroform with different polymer concentrations. 400420440460480500 PL Intensity (Arb. unit)Wavelength (nm) PFO in Chloroform 10 mg/mL 7.5 mg/mL 5 mg/mL 1 mg/mL 0.1 mg/mL Figure 5-8. PL of PFO in chloroform.

PAGE 83

66 Unlike the PL of MEH-PPV solutions, the PL of PFO solutions shows an extra emission peak near 435 nm when the polymer concentration is increased. This suggests that the PFO can form a special morphology in solutions when the concentration is increased, and that interchain interaction exists in said morphology, which can lead to an extra red-shifted emission. This phenomenon is also observed in other solvent systems. Previous studies have shown that PFO exists as a -phase [14] via suitable physicochemical treatment in the solid state. This -phase has a high degree of interchain order and a longer conjugation length, which leads to a red-shifted emission. Our results show that gelation of PFO in the polymer solution at a higher concentration also leads to an extended chain conformation and can generate an extra emission peak as shown in the PL measurement. 5.3.1.3 Water effect on PFO in THF solutions In the previous chapter, the effect of water on MEH-PPV was explored. After adding water to the THF solutions, the MEH-PPV precipitates and the polymer chains form a collapsed conformation. Figure 5-9 shows the PL of PFO in THF after water is added. In contrast to the PL of MEH-PPV in THF solutions (Figure 4-8), PFO shows three emission peaks during the process of precipitation formation. Based on the previous discussion, the emission peak near 440 nm is due to the formation of aggregates in the THF solution. The emission peak near 470 nm is due to the formation of precipitates.

PAGE 84

67 400420440460480500 PL Intensity (arb. unit)Wavelength (nm) PFO in THF solution Dried THF 1% H2O 2% 3% 5% 7.5% 10% Figure 5-9. PL of PFO in the THF solution with different water concentrations. Therefore, these results show that the water can also induce aggregation of PFO in THF solutions. After adding more water, PFO can also undergo a transformation of chain conformation from coil to cylinder. 5.3.2 PL of Polymer Thin films As noted in chapter 2, the polymer chains have a much stronger interchain interaction in polymer thin films compared to polymer solutions. Therefore, the PL of polymer thin films shows lower quantum efficiency and a red-shifted emission due to the generation of aggregates. 5.3.2.1 Solvent effects Figure 5-10 shows the PL of MEH-PPV in four different solvents: chloroform, CB, THF, and TCE, as well as the PL of thin films coated from these solutions under the same polymer concentration and same spin-speed.

PAGE 85

68 550600650700 PL Intensity (Arb. Unit)Wavelength (nm) Chloroform Sol CB Sol THF Sol TCE Sol Chloroform Films CB Films THF Films TCE Films Figure 5-10. Comparison of PL of MEH-PPV in polymer solutions and thin films. This figure illustrates how the PL from films shows an emission with the longest wavelength when the film is cast from TCE and the shortest wavelength when it is cast from THF. This result shows that the degree of red-shift occurs in the same order as the PL from the polymer solutions, which indicates that the chain conformation of polymers can be maintained during the spin-coating process [28]. The red-shifted emission from polymer films compared to polymer solutions is due to the formation of aggregates. 5.3.2.2 Concentration and spin-speed effects Based on the classic model of the spin-coating process [73], the film thickness is determined by the concentration of the polymer solution and the spin-speed. Recent research also has shown that the conformation of polymer chains and the interchain interaction can be altered by changing the concentration and the spin-speed [33]. Figure 5-11 shows the PL of MEH-PPV thin films cast from chloroform solutions with three different concentrations.

PAGE 86

69 550600650700 PL Intensity (Arb. Unit)Wavelength (nm) PL of MEH-PPV films from chloroform 7 mg/mL 5 mg/mL 2.5 mg/mL Figure 5-11. PL of MEH-PPV thin films from chloroform solutions with concentrations of 2.5, 5, and 7 mg/mL. (Spin-speed = 1000 rpm) The polymer film spin-coated from the solution with a lower polymer concentration shows a lower thickness, which leads to a lower PL intensity. The films coated from solutions of lower concentration (2.5 mg/mL and 5 mg/mL) show slightly blue-shifted emissions compared to the films coated from solutions of higher concentration (7 mg/mL). This indicates that the films coated from solutions with higher concentrations have a better interchain interaction compared to the films coated from solutions with lower concentrations. Therefore, the increase of polymer concentration can increase the film thickness as well as the chain interaction, which leads to a red-shifted emission. Figure 5-12 shows the PL from MEH-PPV thin films coated from the chloroform solutions with the same concentration at three different spin-speeds.

PAGE 87

70 550600650700 PL Intensity (Arb. Unit)Wavelength (nm) PL of MEH-PPV films from chloroform 1000 rpm 3000 rpm 5000 rpm Figure 5-12. PL of MEH-PPV thin films coated from chloroform solution (7 mg/mL) at 1000, 3000, and 5000 rpm. An increase of spin-speed will lead to a thinner film and will result in a lower PL intensity, as shown in Figure 5-12. Moreover, the PL of the thin films from higher spin-speeds (3000 rpm and 5000 rpm) show slightly blue-shifted emissions compared to the films from a lower spin-speed (1000 rpm). This indicates that the polymer chains which cast using higher spin-speeds have less interchain interaction. Therefore, the spin-speed effect shows the same effect as concentration: i.e. polymer films coated from lower speeds show higher thickness and more interchain interactions, which lead to a red-shifted emission. Figure 5-13 and Figure 5-14 show the effect of concentration and spin-speed on the PL of MEH-PPV films coated from CB solutions. The results show that the PL of MEH-PPV films coated from CB solutions also follow the same trend as the PL of MEH-PPV films coated from chloroform solutions.

PAGE 88

71 550600650700 PL Intensity (Arb. Unit)Wavelength (nm) PL of MEH-PPV filmsfrom CB 7 mg/mL 5 mg/mL 2.5 mg/mL Figure 5-13. PL of MEH-PPV thin films from CB solutions with concentrations of 2.5, 5, and 7 mg/mL. (Spin-speed = 1000 rpm) 550600650700 PL Intensity (Arb. Unit)Wavelength (nm) PL of MEH-PPV in CB solutions 1000 rpm 3000 rpm 5000 rpm Figure 5-14. PL of MEH-PPV thin films coated from chloroform solutions (7 mg/mL) at 1000, 3000, and 5000 rpm.

PAGE 89

72 5.3.3.3 Annealing effects As described in chapter 2, the annealing process can increase the interchain interaction of polymer chains in the spin-coated thin films, which leads to the generation of aggregates. The formation of aggregates leads to the reduction of PL efficiency and a red-shifted emission. Figure 5-15 shows the annealing effect on the MEH-PPV thin film coated from chloroform solution. 550600650700 PL Intensity (Arb. Unit)Wavelength (nm) Before annealed After annealed Figure 5-15. The PL of MEH-PPV coated from chloroform solution before and after annealing at 120 o C for 3 hrs in a vacuum. Comparing the PL of the thin film before and after the annealing process, the annealed MEH-PPV thin films show lower PL efficiency and a red-shifted emission, which is consistent with the previous studies. Figure 5-16 shows that the annealing effect on MEH-PPV thin films coated from CB solutions is similar.

PAGE 90

73 550600650700 PL Intensity (Arb. Unit)Wavelength (nm) Before annealing After annealing Figure 5-16. PL of MEH-PPV coated from CB solution before and after annealing at 120 o C for 3 hrs in a vacuum. 5.4 Conclusion From the PL and viscosity measurements of polymer solutions, we conclude that the emission peak of MEH-PPV and PFO solutions depend on the chain conformation (or conjugation length). The PL of polymer chains with a large dimension shows a red-shifted emission compared to polymer chains with a small dimension, due to the increase of the conjugation length. The PL of MEH-PPV solution is not affected when the polymer concentration is increased, even once entanglements are formed. Different from MEH-PPV, the PL of PFO solutions show an extra red-shifted emission when concentration is increased. This is due to the gelation of PFO in the polymer solution and the formation of an extended-chain conformation. More details about the chain conformation of PFO will be discussed in chapter 7.

PAGE 91

74 The precipitation of MEH-PPV by adding water to THF can result in a coil-cylinder transition, which leads to an extra red-shifted emission. Different than MEH-PPV, water leads to gelation before the coil-cylinder transition occurs in PFO solutions. The PL of polymer thin films is related to the chain conformation as well as the interchain interaction in the thin films. The formation of aggregates, which are caused by the interchain interaction, leads to lower PL quantum efficiency and a red-shifted emission. Therefore, those spin-coating conditions which can improve the packing of polymer chains and lead to formation of aggregates show a red-shifted emission.

PAGE 92

CHAPTER 6 DEGRAGATION OF METAL CONTACT IN POLYMER LIGHT-EMITTING DIODE DEVICES 6.1 Introduction Bubble formation on calcium/aluminum contacts to bilayer poly(n-vinyl carbazole) (PVK)/poly(3,4-ethylenedioxythiophene)-poly(4-styrenesulfonate) polymer light-emitting diode (PLED) devices was studied using optical and electron microscopy and Auger electron spectroscopy (AES). The formation of bubbles is shown to correlate with pinholes in the metal contact thin film and with excess absorbed water in the presence of the calcium layer. AES data show oxidation at the Ca/Al interface during this process. Gas evolution and relaxation of compressive stress when voltage is applied to the PLED devices are postulated to cause the formation of bubbles. 6.2 Experiment 6.2.1 Device Fabrication The multilayer PLED device consisted of an indium-tin-oxide (ITO) coated glass substrate coated with layers of PEDOT-PSS/PVK/Ca/Al, as shown schematically in Figure 6-1. The 50 nm PEDOT-PSS (Bayer) layer was spin-coated (3000 rpm) on the cleaned ITO coated glass, then dried at 120 o C for 20 minutes. The PVK (Aldrich) solution was prepared by dissolving 50 mg of the polymer in 5 ml of toluene and spin-coated (700 rpm) to form a 70 nm layer on the PEDOT-PSS. The 10 nm Ca layer and 75

PAGE 93

76 200 nm Al layer were deposited by thermal evaporation in a glass bell jar, an oil-diffusion-pumped vacuum system with a base pressure of 4x10 -6 Torr. Figure 6-1. Schematic of the PLEDs used in this study. To study the contributions of the various layers to bubble formation in the cathode contact, multilayer devices were prepared as described below without selected layers, e.g. without the Ca or PEDOT-PSS layers. Encapsulated devices means that the cathode contact would be covered by epoxy (Quickset, Loctite), sealing a glass slide (Micro slide 2947, Corning) over the surface in an atmosphere of argon. 6.2.2 Characterization Bubble formation was observed by optical microscopy after applying a current to the devices in ambient laboratory air, or heating in various atmospheres. The chemical composition of the cathode surface of the PLED device was determined by using a Perkin-Elmer PHI 660 Auger electron spectrometer (AES). A primary electron beam of 5 keV was used. The Auger electron peaks at energies around 271, 294, 511, and 1398 eV are chosen for the elements C, Ca, O, and Al respectively. The composition through the cathode and at the polymer/cathode interface was determined by using 1 keV argon sputter depth profiling on the AES.

PAGE 94

77 6.3 Results and Discussion 6.3.1 Bubble Formation Phenomenon and Mechanism Figure 6-2 shows bubbles that formed in the Ca/Al cathode film on a standard PEDOT-PSS/PVK PLED upon the application of an electrical bias within one hour after depositing the Ca/Al cathode (Fig. 6-2(a)), or after depositing the Ca/Al cathode and storing the device for 24 hours in lab air (Fig. 6-2(b)). Figure 6-2. Bubble formation observed with optical microscopy: (a) biased with 7 V, 10 -5 Acm -2 for 1 min. within one hour after thermal evaporation of Al; (b) biased with 7 V, 10 -5 Acm -2 for 1 min after storage in laboratory air for 1 day (60% relative humidity, 27 o C). Note the lower magnification in (b).

PAGE 95

78 Prior to the application of an electrical bias, no bubbles were observed in either sample. Both devices were biased at 7 V, which caused a current density of 10 -5 Acm -2 to flow through the devices. The small bubbles in Fig. 6-2(a) developed after 15 minutes of bias, while large bubbles developed after only 1 minute of bias on the device stored in lab air for 24 hours. Note that in both cases, the bubbles formed around a defect (pinhole) in the Ca/Al cathode thin film. Do et al. [69] and McElvain et al. [68] reported that bubbles formed in Mg/Al cathode thin films on Alq 3 emitter layers due to heat, gas evolution and oxidation of the Al as a result of electrical biasing. Savvateev et al. [65] reported that gas evolution caused delamination of the Al cathode thin films on the PPV layers, resulting in electrical noise and ultimately a current spike that resulted in carbonized polymer areas. Do et al. [69] showed that an inert atmosphere inhibited the formation of bubbles. The effects of the environment and encapsulation on the formation of cathode bubbles in our PEDOT-PSS/PVK conjugated polymer devices were tested by storing devices for four days under argon or air with and without epoxy/glass encapsulation, followed by electrical biasing at 8 V for 2 minutes (Figure 6-3). The cathode surfaces of these devices before electrical biasing, shown on the left, are free of bubbles even when stored without encapsulation. After electrical biasing, the surface of encapsulated devices stored in argon did not bubble, but bubbles were observed on the surfaces of the other devices stored in argon or air without encapsulation, or in air even with encapsulation. The bubbles were larger and more numerous for devices stored in air (Figs. 6-3(c) and 6-3(d)) versus argon (Figs. 6-3(a) and 6-3(b)). Devices stored in air without encapsulation formed the greatest amount of bubbles after biasing (Fig. 6-3(d)).

PAGE 96

79 Figure 6-3. The effects of storage ambient and encapsulation upon bubble formation as a result of electrical bias (8 V for 2 min.). Encapsulated or non-encapsulated devices were stored for four days in argon or air as follows: (a) argon with encapsulation; (b) argon without encapsulation; (c) air with encapsulation; (d) air without encapsulation. Photos on the left and right are before and after the devices were biased, respectively.

PAGE 97

80 These data are consistent with the results of Do et al. [69] and show that argon reduced the formation of bubbles, and with their general conclusion that moisture accelerates their formation. It is interesting to note that bubbles were denser near the edges of the dot contacts, suggesting that moisture could more easily penetrate into the devices at the step edges where there was no protective Al layer. Figure 6-4. The dependence of bubble formation on the type of conjugated polymer and temperature in the absence of electrical bias. Devices consisted of the normal ITO/polymer/Ca/Al structure, where the polymer layer is either PVK or PEDOT-PSS. Devices were examined in real time by optical microscopy while being heated on a hot stage either immediately after deposition of the cathode, or 24 hours after this deposition with storage in air. The samples and temperatures were as follows: (a) PVK, fresh, 120 o C; (b) PVK, fresh, 250 o C; (c) PVK, 24 hour storage, 120 o C; : (d) PEDOT-PSS, fresh, 120 o C; (e) PEDOT-PSS, fresh, 250 o C; (f) PEDOT-PSS, storage, 120 o C. Note that bubbles are observed only when PEDOT-PSS is present. Note also that bubbles were formed by heating alone without an electrical bias.

PAGE 98

81 To determine whether PEDOT-PSS or PVK was contributing to the formation of bubbles, devices were prepared with only a PVK (Figs. 6-4(a) 6-4(c)) or a PEDOT-PSS (Figs. 6-4(d) 6-4(e)) polymer layer, respectively. In addition, to determine if simple heating without electrical currents would cause bubbles to form, the devices were heated to either 120 o C or 250 o C on a hot plate in air. Finally, devices were heated immediately after the deposition of the cathode (fresh devices) and monitored while hot using the optical microscope. Neither fresh PVK nor fresh PEDOT-PSS devices showed bubbles when heated to 120 o C (Figs. 6-4(a) and (d)). When the temperature was raised to 250 o C, only the device with a PEDOT-PSS layer formed bubbles (compare Figs. 6-4(b) and (e)). While storage in air for 24 hours did not cause bubbles to form at 120 o C on PVK devices, large bubbles were found on PEDOT-PSS devices that were treated in a similar fashion. These data demonstrate that water is absorbed by and evolved from PEDOT-PSS layers to form bubbles. They also demonstrate that simple heating without an electrical bias can result in the formation of bubbles on these devices. This is consistent with the postulate that local heating can cause gas evolution from some polymers [65]. By real time observations in the optical microscope, the fresh PEDOT-PSS devices started to evolve gas and form bubbles at T 220 o C. While storage of PVK devices in air for 24 hours did not cause bubbles to emerge (Fig. 6-4(c)), storage accelerated their formation in PEDOT-PSS devices. After 24 hours in air, PEDOT-PSS devices started to form bubbles at temperatures of 90 o C, and they were well-formed by 120 o C (Fig. 6-4(f)). Observation of the bubbles at 90 o C after a 24-hour exposure to air supports the conclusion that PEDOT-PSS layers absorb and evolve water to create bubbles. Because of the low water content and experience with other hydrophobic polymers, it is believed that the bubbles

PAGE 99

82 formed in fresh PEDOT-PSS samples heated to 250 o C result from decomposition of the polymer. Since PEDOT-PSS is soluble in water, it is reasonable that these layers would absorb water under ambient storage conditions, and evolve water when heated to T~90 o C. Do et al. [69] also suggested that water can lead to formation of Al 2 O 3 with evolution of H 2 which can cause bubbles to form. In our devices, water in the PEDOT-PSS layer could react either with Ca or Al to form H 2 e.g.: Ca + H 2 O CaO + H 2 2Al + 3H 2 O Al 2 O 3 + 3H 2 To evaluate the importance of these reactions, devices were made with either the normal Ca/Al or with only an Al cathode dot contact to the normal PVK/PEDOT-PSS bilayer structure. A small water-droplet was placed on the surface of these cathodes, without an electrical bias being applied. The results are shown in Figure 6-5. Figure 6-5. Formation of bubbles by water droplets in the absence of an electrical bias when a Ca layer is present. The photos are of Al only or normal Ca/Al cathode contacts with water droplets on the surface. Al-only contact (a) immediately after placing a water droplet on the surface, and (b) after a few minutes; no bubbles are observed. Ca/Al contact (c) immediately after placing a water droplet on the surface, and (d) after a few minutes; large bubbles are obvious in (d).

PAGE 100

83 No bubbles are observed immediately after placing the water droplet on either contact (Figs. 6-5(a) and 6-5(c)). For the cathode consisting of only an Al layer, no bubbles were observed upon contact after 5 min (Fig. 6-5(b)). By contrast, large bubbles are observed after <5 minutes on devices with a normal Ca/Al cathode (Fig. 6-5(d)). Pinholes certainly existed in both the Al and the Ca/Al cathodes, allowing water to quickly reach the metal/polymer interface. Formation of bubbles in only a few minutes when the Ca layer was present demonstrates its importance to the formation of bubbles. Perhaps the Ca reacts rapidly to evolve gas quickly and thereby form a bubble. This is consistent with the fact that CaO should not form a passivating oxide layer. Conversely, Al forms a protective Al 2 O 3 layer and would not result in rapid H 2 evolution from the above chemical reaction. A second factor may be the mechanical strength of the interfacial oxide formed by such reactions. An interfacial layer of CaO may fracture more easily than Al 2 O 3 and thus allow for a gas pocket to form a bubble. 6.3.2 Surface Composition (AES) A scanning electron microscope (SEM) in the AES mode was used to examine the surface for pinholes surrounded by bubbles that were produced by electrical biasing at 7 V for 10 minutes on a device stored in air for 1 day (see Fig. 6-6(a)). The AES spectra collected from the pinhole and adjacent cathode surface are shown in Fig. 6-6(b). The spectra from the pinhole (i) contain Auger peaks from only C and O, presumably from the polymer thin film structure. No Al was detected in the pinhole. The spectrum from the adjacent cathode surface (ii) contains peaks from C, O and Al. The strong oxygen signal in (ii) is due to Al 2 O 3 on the Al surface. The C peak is from normal adventitious contamination due to adsorption of organics and hydrocarbons from laboratory air. Since pinhole areas lack the protection of a continuous Al layer, they offer paths for transport of

PAGE 101

84 water and oxygen to the Ca and polymer layers. The composition versus depth in the PLED before and after electrical biasing was determined by sputter depth profiling shown in Figs. 6-7(a) and 6-7(b), respectively. Both devices showed high oxygen signals at the surface due to the formation of a few nanometer-thick native oxide films by exposure of freshly deposited Al to air. For the unbiased sample in Fig. 6-7(a), the oxygen concentration shows a second increase after sputtering through the Al films and reaching the Ca underlayer. The Ca peak-intensity rises simultaneously with that of oxygen, suggesting that the Ca was at least partially oxidized during the thermal-evaporation process or during the delay in switching the power supply and heating the Al boat for deposition of the Al overlayer. However, the Ca layer was not completely oxidized to CaO since the device turn-on voltages were lower (4 V) when the Ca was present at the cathode interface. By contrast, when Al contacted the polymer layers directly, the turn-on voltage increased to 12 V.

PAGE 102

85 (a) 50 m (i) (ii) 500100015002000 (b) (ii) Figure 6-6. Pinhole structure and composition: (a) SEM image of a pinhole in the Al/Ca cathode dot contact and surrounding bubbles generated by application of an electrical bias; (b) AES spectra from the cathode surface (i) inside the pinhole, and (ii) outside the pinhole. Kinetic Energy (i) 139 Al C O 51 27 51 O 27 C

PAGE 103

86 Cycle dN(E)/dE (a) Figure 6-7. AES sputter depth profile of a normal 10 nm Ca/60 nm Al PLED cathode dot contact (a) before and (b) after an electrical bias was applied. Note the increase of oxygen at the Al/Ca interface after electrical biasing (8 V for 10 min. in laboratory air). As a result of biasing (Fig. 6-7(b)), the oxygen Auger peak intensity increased before the Ca Auger peak, suggesting that the electrical bias led to oxidation of Al at the Al/Ca interface. Based on the data reported above that the Ca layer led to the formation of bubbles, water that evolved from the PEDOT-PSS layer may have reacted with the Ca layer to form H 2 and bubbles when the devices were biased. Once the Ca layer was dN(E)/dE Al O Ca (b) O Al Ca Cycle

PAGE 104

87 completely converted to CaO, the water would attack the Al layer to form Al 2 O 3 or Al(OH) 3 This interfacial oxide/hydroxide would be expected to have poor mechanical properties and could lead to loss of interfacial adhesion and the generation of a bubble. 6.4 Conclusion The formation of bubbles in cathode contacts to Al/Ca/PVK/PEDOT-PSS/ITO-glass PLEDs was studied using microscopy and Auger electron spectroscopy. The data showed that bubbles could be generated three ways: 1. by electrical biasing, 2. by simply heating to T>90 o C without an electrical bias when moisture and PEDOT-PSS were present, and 3. by the presence of a water droplet without an electrical bias when a Ca layer (under an Al overlayer film) and PEDOT-PSS were present. Bubbles tended to form around pinhole defects in the cathode thin films or around the edge of the dot contact. The mechanism of bubble formation was suggested to be an absorption of moisture by the PEDOT-PSS layer where the Ca/polymer interface was not protected (i.e. at pinholes and the edge of the dot contact), and subsequent gas evolution due to interfacial chemical reactions and/or heating. While the presence of a Ca layer in the cathode led to significantly lower PLED turn-on voltages, it also led to interfacial reactions and bubble formation in the presence of water droplets at room temperature without an electrical bias. The Auger electron sputter depth profile data are consistent with this mechanism.

PAGE 105

CHAPTER 7 EFFECT OF THERMO-INDUCED OXIDATION ON POLYMER STRUCTURE AND OPTICAL PROPERTIES OF POLYFLUORENE 7.1 Introduction .Polyfluorenes (PF), which shows a blue light emission of high quantum efficiency, can be used as the light-emitting layer in blue PLEDs [108]. The formation of a liquid-crystalline phase further makes PF a potential material for use in polarized polymer light-emitting devices [14]. Upon device operation, the formation of an additional emission peak around 535 nm is usually observed. Recently, it has been shown that this emission is due to the photoor electro-induced oxidation of polyfluorene chains and the formation of ketonic defects [11,109-111], as shown in Figure 7-1. Figure 7-1. Proposed mechanism for the generation of keto-defect sites. In first part of this chapter, the relationship between the optical properties and the polymer morphology of poly[9,9-dioctylfluorenyl-2,7-diyl] (PFO) will be illustrated. Then the thermo-induced oxidation of PFO and its effects on optical properties and polymer morphology will be discussed. 88

PAGE 106

89 7.2 Experiment 7.2.1 Sample Preparation PFO (Mw = 52,000 and 341,000, American Dye Source, Inc) is used as received. Polymer solutions were prepared by dissolving PFO in chloroform at concentrations between 1 to 5mg/mL. These solutions were filtered (1 m PTFE filter) and spin-coated on ITO glass and Si wafer to make thin film samples. Thick film samples were prepared by using a dip-coating process. 7.2.2 Characterization Polarized Optical Microscopy. Polymer solutions were dropped on a glass slide and dried in the air. Samples were examined by using an Olympus BX 60 polarized optical microscopy in the transmission mode. A hot stage was used to control the temperature. Images were captured by a CCD camera (3CDD, MTI). X-Ray Diffraction (XRD). XRD was used to identify the crystallinity of the polymer and orientation of the polymer crystals. A Phillips APD 3720 diffractometer with a Cu K source was used to measure the diffraction pattern from 4 to 27 degrees. The sample holder was filled with Ar when the measurements were conducted at temperatures higher than room temperature. Samples were annealed at each temperature for 10 minutes before measurement. The details of PL and XRR measurements were already described in chapters 5 and 4.

PAGE 107

90 7.3 Results and Discussion 7.3.1 Chain Conformation in PFO Thick Films 7.3.1.1 Molecular weight effect Figure 7-2 shows the images of low Mw PFO samples under a polarized optical microscope at several different temperatures. Figure 7-2. Images of a low Mw PFO under a polarized optical microscope at (a) 120 o C, (b) 150 o C, (c) 210 o C, and (d) 220 o C. Figure 7-2(a) shows the crystalline phase still exists in a polymer sample when the temperature was increased to 120 o C. When the temperature was raised to 150 o C (Figure 7-2(b)), the polymer crystal melted and a liquid crystalline phase was revealed. After the temperature reached 210 o C, the liquid crystalline started to transform and the polymer showed a totally isotropic and amorphous phase at temperatures higher than 210 o C (Figure 7-2(c) and 7-2(d)). This result is coincident to previous studies [14]. Indeed, the

PAGE 108

91 PFO crystals melt when the temperature is higher than 150 o C; in addition, the liquid crystalline phase exists when T is between ~150 and 200 o C. The phase transformation can also be observed by X-ray diffraction, as shown in Figure 7-3 (low Mw PFO samples) and Figure 7-4 (high Mw PFO samples). 510152025 (250)(530)(330)(310)(200)Cooling 250 oC200 oC150 oC125 oCRoom Temp.Counts/sec (Arb. unit)2 (degree) Figure 7-3. X-ray diffraction of low Mw PFO samples at room temperature, 125, 150, 200, and 250 o C. Sample was annealed at 250 o C for 10 minutes before cooling to the measurement temperature.

PAGE 109

92 510152025 (250)(530)(520)(330)(310)(210)(200)(110)27oC120oC150oC200oC250oCCounts/sec (Arb. unit)2 (degree) Figure 7-4. X-ray diffraction of high Mw PFO samples at room temperature, 125, 150, 200, and 250 o C. Sample was annealed at 250 o C for 15 minutes cooling to the measurement temperature. The polymer samples were first annealed at 250 o C in the sample holder filled with Ar for 15 minutes. Figure 7-3 and Figure 7-4 both show the formation of diffraction peaks when the temperature is lower than 150 o C, which indicates that PFO starts to crystallize when the temperature is lower than 150 o C. Comparing the XRD profiles measured at room temperature in these two figures, the profile from low Mw PFO shows a much stronger (200) diffraction peak than other peaks, which is different from the profile from high Mw PFO. From Chen et als work [113], the XRD profile from high Mw PFO is mostly composed of the phase crystalline structure with an orthorhombic unit cell (a = 2.56 nm, b = 2.34 nm, c = 3.32 nm). More recently, their research showed the formation of a modified form, which is kinetically favored at lower crystallization temperatures, as shown in Figure 7-5 [114].

PAGE 110

93 Figure 7-5. Characteristic XRD profile for film in N, , and phases [114]. The change involves a slightly longer b-axis and a preferred orientation of the a-axis along the film normal. Our XRD profile from low Mw PFO shows the same results as their profile from the form, which indicates that the low Mw PFO is dominantly composed of form crystals. Chen et als results show that the same PFO sample can form an or phase depending on the crystallization temperature. In our experiments, the PFO of a different Mw shows a different crystalline structure even at the same crystallization temperature. These results indicate that the molecular weight is a critical factor for the formation of the crystalline phase. The PFO of low Mw generates more phase crystals compared to the PFO of higher Mw, which tends to form phase crystals. As shown in both figures, the positions of the diffraction peaks that were measured at 125 o C occur at slightly lower angles than the peaks that were measured at room temperature. From Braggs Law, nd sin2 a smaller value of indicates a larger value of d. The increase in size is due to the thermal expansion of PFO unit cells. The diffraction patterns which were measured at 150, 200, and 250 o C, do not show much difference. From the images observed using polarized optical microscopy (Figure 7-2),

PAGE 111

94 PFO forms a liquid-crystalline phase between 150 and 210 o C. The same diffraction patterns observed in the liquid-crystalline and amorphous phases indicate the PFO has a nematic liquid-crystalline phase. Figure 7-6 shows the effect of annealing on the crystallization of PFO. 510152025 Low Mw; 3 hrsLow Mw; 15 hrsHigh Mw; 15 hrsHigh Mw; 3 hrsCounts/sec (Arb. unit)2 (degree) Figure 7-6. XRD profile of low Mw and high Mw PFO samples annealed at 170 o C for 3 and 15 hours in Ar. Both low and high Mw PFO samples were annealed at 170 o C for 3 to 15 hours and cooled to room temperature. Based on the results shown in Figure 7-2, 7-3 and 7-4, PFO forms a liquid-crystalline phase at 170 o C. For low Mw PFO samples, the diffraction patterns for samples with different annealing times do not show much difference. For high Mw PFO samples, the intensities of the diffraction peaks are much stronger for the sample with the longer annealing time. These results indicate that annealing at 170 o C can increase the crystallinity due to the formation of the liquid-crystalline phase, which can improve the order of the polymer chains. The low Mw PFO samples, which have a higher mobility and diffusion rate, can

PAGE 112

95 form a liquid-crystalline phase faster than high Mw PFO samples. Therefore the diffraction patterns after 3 and 15 hours of annealing for low MW samples do not show an obvious difference. For high Mw PFO samples, the chain diffusion rate is much lower. The high Mw PFO sample annealed for15 hours forms a more ordered liquid crystalline phase than the one annealed for 3 hrs, which leads to higher crystallinity as shown in the X-ray diffraction pattern. 7.3.1.2 Thermal-oxidation effect Figure 7-10 shows the XRD diffraction pattern of PFO samples annealed in Ar and air. 510152025 15 hrs; Ar3 hrs; Ar15 hrs; Air3 hrs; AirDipcoatingCounts/sec (Arb. unit)2 (degree) Figure 7-7. Comparison of X-ray diffraction pattern of high Mw PFO samples annealed at 170 o C in Ar and in air for 3 and 15 hours. From Figure 7-7, the samples annealed in Ar show clear diffraction peaks, which means these samples can crystallize. Furthermore, the crystallinity becomes higher as the annealing time increases. For samples that were annealed in air, the sample annealed for

PAGE 113

96 3 hours shows diffraction peaks whereas the sample annealed for 15 hours does not. Moreover, we also found that the sample which was annealed for 15 hours can not dissolve in chloroform. This indicates oxidation of polymer chains can induce crosslinking, which inhibits the formation of polymer crystals. The PFO samples of low Mw show similar results. The relationships between the PL and PFO chain conformation have been well studied by Grell et al. [13,14]. Here we focus on the thermal-induced oxidation effect on PL and polymer morphology. Figure 7-8 and Figure 7-9 show the PL spectra and XRD profiles of high Mw PFO samples which were annealed at temperatures higher than 150 o C for 1 or 3 hours in the air and cooled down to room temperature. The PFO sample of low Mw shows similar results. 400450500550600650 Annealing in air Normalized PL Intensity (Arb. unit)Wavelength (nm) 170 oC; 1hr 200 oC; 1hr 200 oC; 3hr 240 oC; 1hr Figure 7-8. PL of high Mw PFO samples annealed in air at different temperatures: 170 o C for 1 hour (solid line), 200 o C for 1 hour (dash dot line), 200 o C for 3 hours (dash line), and 240 o C for 1 hour.

PAGE 114

97 From Figure 7-8, the sample annealed at 170 o C for 1 hour does not show a clear emission peak near 550 nm, which indicates only a few ketonic defects exist on the polymer chains. The sample annealed at 200 o C for 1 hour shows emission peaks at both 450nm and 550nm, which indicates the formation of many of ketonic defects on the polymer chains. The sample annealed at 200 o C for 3 hours shows primarily one emission near 550nm, which indicates most of the polymer repeat units were oxidized. The sample annealed at 240 o C for 1 hour shows similar results as the sample annealed at 200 o C for 1 hour. 510152025 240 oC; 1hr200 oC; 3hr200 oC; 1hr170 oC; 1hrCounts/sec (Arb. unit)2 (degree) Figure 7-9. The X-ray diffraction pattern of annealed high Mw PFO samples corresponding to the samples shown in Figure 7-8. As shown in Figure 7-9, the samples annealed at 170 o C for 1 hour and at 200 o C for 1 hour show clear diffraction peaks, which indicate that these two samples can crystallize. The sample annealed at 200 o C for 1 hour show higher crystallinity than the sample annealed at 170 o C because a more ordered liquid-crystalline phase can form at

PAGE 115

98 higher temperatures for the high Mw PFO sample, which is helpful for the formation of polymer crystallinity as described in the previous section. The samples annealed at 200 o C for 3 hours and 240 o C for 1 hour do not show diffraction peaks, which indicates no formation of polymer crystals in these two samples. These results show that the thermal-induced oxidation of PFO can lead to the crosslinking of polymer chains and inhibit the formation of polymer crystals. 7.3.2 Molecular Effect on PFO Thin Film Structure In this research, we used XRR to measure the electron density, roughness, and thickness of polymer thin films. Figure 7-10 shows the XRR spectra and the fitting result from spin-coated high Mw PFO thin films on ITO glass. 1234 Log (Reflectivity)2 (degree)7075808590 PFO layerITO Electron density (arb. unit)Thickness (nm) Figure 7-10. XRR spectra acquired from spin-coated high Mw PFO thin films. Solid line and dash line represent the experimental data and the fitted results (one layer model) respectively.

PAGE 116

99 The fitting result is from one homogenous polymer film of ~ 90 The roughness is less than 10 The experimental results can be fit well using a one layer model. Figure 7-11 shows the XRR spectra and the fitting result from the low Mw PFO thin films on ITO glass. The fitting curve is also from one homogeneous polymer film, in this case of ~ 130 There is a major difference between the model and experiment in the region 2 ~ 1 degree. Figure 7-12 shows the fitting results using a two layer model. An interface layer of lower density exists between the ITO and PFO thin film of higher density. 1234 Log (Reflectivity)2 (degree)707580859095 PFO layerITO Electron density (Arb. unit)Thickness (nm) Figure 7-11. XRR spectra acquired from spin-coated low Mw PFO thin films. Solid line and dash line represent the experimental data and the fitted results (one layer model) respectively.

PAGE 117

100 1234 Log (Reflectivity)2 (degree)707580859095 PFO layerITO Electron density (Arb. unit)Thickness (nm) Figure 7-12. XRR spectra acquired from spin-coated low Mw PFO thin films. Solid line and dash line represent the experimental data and the fitted results (two layer model) respectively. From Figure 7-12, it can be seen that this model gives a good fit to the data. The poor fit in Figure 7-11 near 1 degree is corrected, which indicates that an interface layer exists between the low Mw PFO thin films and ITO glass. In contrast, no interface layer is observed between the high Mw PFO thin films and ITO glass. The properties of polymers near solid substrates are critical for numerous commercial applications and have been studied extensively. For the application of polymer thin film technologies like PLEDs, the interface behavior is an especially important factor for the performance of devices. Many studies have shown changes in Tg for supported thin films due to the interaction of polymers with substrates [115-119]. Several models have been used to interpret the effects of polymer film interfaces on the change of Tg, including change in film density [120], reduction in the extent of entanglements [121], and enrichment of chain end-groups within the interface [122,123].

PAGE 118

101 In our system, we use PFO of two different molecular weights. From the XRR results, the high Mw PFO thin-film shows a simple homogeneous polymer layer, whereas the low Mw PFO shows an interface layer with lower density. Comparing PFO of different Mws, PFO of low Mw has a higher diffusion rate and more chain ends than PFO of a higher Mw. Mayer [122] has pointed out the chain end effect on the glass transition of polymers near an interface. A decrease of density near the surface or interface is caused by chain end segregation. Therefore, the interface layer, which is observed only in the low Mw PFO system, could be related to the increase in number of chain ends and the segregation of chain ends near the interface. On the other hand, Bollinne et al. [124] also examined the interface layer of PS and PMMA on an SiO 2 substrate by using XRR, but they did not find a molecular weight effect in their system. This controversy could be due to the specific polymer chain conformation of PFO compared to traditional polymers, including PS and PMMA. Compared to PS and PMMA, PFO is a stiff main-chain polymer and tends to form a gel when the polymer concentration is increased. In fact, as illustrated previously, PFO of a different Mw tends to form into a different crystalline-phase under the thermal conditions, which indicates that Mw is a critical factor in chain conformation. The chain conformation near the interface in the PFO system needs further study to understand the effects that molecular weight has on the chain morphology of spin-coated thin films. 7.4 Conclusion From the XRD results, we found that molecular weight is an important factor in the crystallization process. High Mw PFO tends to form an phase crystalline whereas low Mw PFO tends to form an phase crystalline at the same crystallization

PAGE 119

102 temperature. Annealing at 170 o C, which induces the formation of a liquid-crystalline phase, can increase the crystallinity of PFO. The thermal-induced oxidation of PFO causes the formation of ketonic defects and leads to an extra red-shifted emission. XRD also shows that oxidized PFO samples cannot crystallize due to crosslinking. From the fitting of XRR results, the high Mw PFO thin film shows one homogeneous polymer layer whereas the low Mw PFO thin film shows the existence of an interface layer. This interface layer is proposed as being related to the enrichment of chain ends near the interface and needs further study.

PAGE 120

CHAPTER 8 CONCLUSIONS AND RECOMMENDATIONS FOR FUTURE WORK For polymer-based thin film devices, the performance and efficiency are influenced by the quality of the films, polymer/substrate interfaces, and even the chain conformation in the polymer thin films. Therefore, the control of the polymer chain conformation as well as the film quality will be very important. In this work, the physical properties of three light-emitting polymers, PVK, MEH-PPV, and PFO, were investigated for their application in polymer thin film devices. The spin-coating technique is the most common method to make polymer thin films. The solvents evaporate so fast that the polymers are quenched and smooth thin films are formed. During spin-coating, evaporation of solvents of a high evaporation rate leads to temperature and concentration gradients, which can induce Marangoni convection. In turn, Marangoni convection can then lead to cell and striation structures near and away from the center area, respectively, of the polymer thin films. Thin films prepared from THF/water solutions show result in the formation of ruptured films and polymer droplets on ITO substrates. Due to the incompatibility of water and MEH-PPV, water in polymer solutions can induce a coil-cylinder transition of the polymer chains and lead to a red-shifted emission. Due to fast evaporation, the polymer chains in thin films are not in a state of thermodynamic equilibrium. Therefore, the conformation of polymer chains in thin films is determined by a variety of critical factors involved in the spin-coating process, including solvents used, concentrations, and spin-speeds, or post-processes such as 103

PAGE 121

104 annealing. For conjugated polymers, the band gap is related to the conjugation length: the longer the conjugation length, the smaller the band gap. Polymer chains that have a more extended conformation show a longer conjugation length and lead to a red-shifted emission compared to polymer chains that have a more compact chain conformation. Therefore, solvents which can affect the chain conformation can also influence the optical properties of conjugated polymers. Likewise, for conjugated polymers the formation of aggregates in polymer thin films can lower the PL quantum efficiency and can lead to a red-shifted emission. Therefore, the critical factors for the fabrication of polymer thin films can also influence the optical properties of those films. For PLEDs, the short life-time is due to the degradation of both the metal contacts and the polymer layer. Pinholes give pathways for the diffusion of moisture through the protection layer. The metal contact, such as Ca, can react with water from air or from PEDOT-PSS that then generates H 2 which leads to the formation of bubbles around pinholes. The thermal-induced oxidation of PFO can lead to an extra red-shifted peak due to the generation of ketonic defects. In addition, the thermal-induced oxidation also causes crosslinking of PFO. In this study, the details of some interesting phenomenon are still not clear and need further study. First, the effects of solvents on the chain conformation of conjugated polymers in solution are still not clear. From chapter 5, although the viscosity results are coincident with the PL (the polymers with a larger dimension show a more red-shifted emission), the polymer dimension cannot be explained from the view of polymer solubility. Generally, solvents that dissolve a polymer well are called good solvents and those which do not dissolve a polymer are called poor solvents. Because contact

PAGE 122

105 between polymers and good solvents is favorable, the polymer chain tends to adopt a more expanded random-coil conformation. When the solvent quality becomes poor, contacts between those polymers and ill-suited solvents become unfavorable, and thus the polymer chain tends to contract. Therefore, polymers in good solvents (from the view of solubility) should have a red-shifted emission compared to polymers in poor solvents. Nonetheless, the PL results did not show this coincidence. MEH-PPV and PFO in the good solvent chloroform do not show a red-shifted emission compared to what is seen in other solvents. We believe this is due to the complex chemical structure of conjugated polymers, which contain rigid aromatic groups in the main chains and soft alkyl groups as side chains. We suggest that computer simulation is needed to illustrate the solvent effects on this complex system. Second, the mechanism of how a ruptured polymer thin film forms from THF/water solution is not clear. Two speculated mechanisms are proposed, which are based on the results from this work, as shown in Figure 8-1 and Figure 8-2. Thicker area Figure 8-1. The proposed model of ruptured polymer film formation during the spin-coating process due to the surface-tension gradient induced by water. ITO substrate Pol y mer in THF and Trace H 2 O solution Thinner area THF evaporates faster then water H 2 O and THF evaporate High surface tension gradient due to highontents of water er c Ru p tured Pol y mer thin fil m Polymer droplets

PAGE 123

106 In Figure 8-1, since water has a lower evaporation rate than THF, the polymer solution thin films have a higher water concentration near the solution surface during the evaporation of the solvents. Due to the high surface tension of water, the surface tension gradient then increases and drives much faster convection near the surface. This convection then leads to rupture of the polymer solution from the thinner area of the solution thin films. For thicker area THF evaporate faster Polymer-rich phase H 2 O-rich p hase H 2 O and THF evaporate Figure 8-2. The proposed model of ruptured polymer film formation during the spin-coating process due to phase separation. On the other hand, due to the incompatibility of water and MEH-PPV, phase separation may generate water-rich and polymer-rich phases in the solution. The phase H 2 O-rich p hase Polymer-rich phase H 2 O-rich p hase For thinner area THF evaporate faster H 2 O-rich p hase Polymer-rich phase Polymer droplet H 2 O and THF evaporate Polymer droplet Ruptured Polymer film Hole Ruptured Polymer film

PAGE 124

107 separation then leads to rupture of the thin film and the formation of polymer droplets, as shown in Figure 8-2. Again, these two models are highly speculative and need more data to prove. Third, the molecular weight effect of PFO on chain conformation and thin film structure is not clear. Although a few models, like the chain end effect model, can be used to explain this phenomenon, more details are still needed to understand the mechanism. Ultimately, the effect of the interface layer is an important factor on the performance of thin film devices and is of enormous interest.

PAGE 125

LIST OF REFERENCES 1. Burroughes, J. H.; Bradley, D. D. C.; Brown, A. R.; Marks, R. N.; Mackey, K.; Friend, R. H.; Burns, P. L.; and Holmes, A. B. Nature 1990, 347, 539. 2. Braun, D.; Heeger, A. J., Appl. Phys. Lett. 1991, 58, 1982. 3. Gustafsson, G.; Cao, Y.; Treacy, G. M.; Klavetter, F.; Colaneri, N.; Hegger, A. J., Nature 1992, 357, 477. 4. Friend, R. H.; R. W. Gymer, A. B. Holmes, J. H. Burroughes, T. N. Marks, C. Taliani, D. D. C. Bradley, D. A. Dos Santos, J. L. Bredas, M. Lgdluned, and W. R. Salaneck, Nature 1999, 397, 121. 5. Patel, N. K.; Cin, S.; Burroughes J. H. IEEE J. Select. Top. Quant. Elect. 2002, 8, 346. 6. Redecker, M.; Bradley, D. D. C.; Bernius, M. T.; Inbasekaran, M.; Wu, W. W.; Woo, E. P. Appl. Phys. Lett. 1998, 73, 1565. 7. Yu, L. S.; Chen, S. A. Synth. Met. 2002, 132, 81. 8. He, G. F.; Liu, J.; Li, Y. F.; Yang, Y. Appl. Phys. Lett. 2002, 80, 1891. 9. Leising, G.; Tasch, S.; Brandstatter, C.; Graupner, W.; Hampel, S.; List, E. J. W.; Meghdadi, F.; Zenz, C.; Schlichting, P.; Rohr, U.; Geerts, Y.; Scherf, U.; Mullen, K. Synth. Met. 1997, 91, 41. 10. Jin, Y. D.; Chen, H. Z.; Heremans, P. L.; Aleksandrzak, K.; Geise, H. J.; Borghs, G.; Van der Auweraer, M. Synth. Met. 2002, 127, 155. 11. Sims, M.; Bradley, D. D. C.; Ariu, M.; Koeberg, M.; Asimakis, A.; Grell, M.; Lidzey, D. G. Adv. Funct. Mater. 2004, 14, 765. 12. Popovic, Z. D.; Aziz, H. IEEE J. Select. Top. Quant. Elect. 2002, 8, 362. 13. Grell, M.; Bradley, D. D. C.; Long, X.; Chamberlain, T.; Inbasekaran, M.; Woo, E. P.; Soliman, M. Acta Poly. 1998, 49, 439. 14. Grell, M.; Bradley, D. D. C.; Ungar, G.; Hill, J.; Whitehead, K. S. Macromolecules 1999, 32, 5810. 15. Shim, H. K.; Ahn, T.; Song, S. Y. Thin Solid Films, 2002, 417, 7. 108

PAGE 126

109 16. Leising, G.; Tasch, S.; Brandstatter, C.; Graupner, W.; Hampel, S.; List, E. J. W.; Meghdadi, F.; Zenz, C.; Schlichting, P.; Rohr, U.; Geerts, Y.; Scherf, U.; Mullen, K. Synth. Met. 1997, 91, 47. 17. Gao, M. Y.; Richter, B.; Kirstein, S. Adv. Mater. 1997, 9, 802. 18. Thompson, J.; Blyth, R. I. R.; Mazzeo, M.; Anni, M.; Gigli, G.; Cingolani, R. Appl. Phys. Lett. 2001, 79, 560. 19. Suzuki, H.; Hoshino, S.; Yuan, C. H.; Fujiki, M.; Toyoda, S.; Matsumoto, N. Thin Solid Film 1998, 331, 64. 20. Suzuki, H.; Hoshino, S.; Yuan, C. H.; Fujiki, M.; Toyoda, S.; Matsumoto, N. H. IEEE J. Select. Top. Quant. Elect. 1998, 4, 129. 21. Suzuki, H. J. Lumin. 1997, 72-4, 1005. 22. Partridge, R. H. Polymer 1982, 24, 733. 23. Kersting, R.; Lemmer, U.; Mahrt, R. F.; Leo, K.; Kurz, H.; Bassler, H.; Gobel, E. O. Phys. Rev. Lett. 1993, 70, 3820. 24. Hayes, G. R.; Samuel, I. D. W.; Phillips, R. T. Phys. Rev. B 1995, 52, 11569. 25. Kersting, R.; Mollay, B.; Rusch, M.; Wenisch, J.; Leising, G.; Kauffmann, H. F. J. Chem. Phys. 1997, 106, 2850. 26. Zheng, M.; Bai, G. L.; Zhu, D. B. J. Photochem. Photobiol. A 1998, 116, 143. 27. Schwartz, B. J. Ann. Rev. Phys. Chem. 2003, 54, 141. 28. Nguyen, T. Q.; Doan, V.; Schwartz, B. J. J. Chem. Phys. 1999, 110, 4068. 29. Kumar, P.; Mehta, A.; Mahurin, S. M.; Dai, S.; Dadmun, M. D.; Sumpter, B. G.; Barnes, M. D. Macromolecules 2004 37, 6132. 30. Fritz, K. P.; Scholes, G. D. J. Phys. Chem. B. 2003; 107(37); 10141. 31. Amautov, S. A.; Nechvolodova, E. M.; Bakulin, A. A.; Elizarov, S. G.; Khodarev, A.; Martyanov, D. S.; Paraschuk, D. Y. Synth. Met. 2004, 147, 287. 32. He, G. F.; Li, Y. F.; Liu, J.; Yang, Y. Appl. Phys. 2002, 80, 4247. 33. Shi, Y.; Liu, J.; Yang, Y. J. Appl. Phys. 2000, 87, 4254. 34. Liu, J.; Shi, Y.J.; Ma, L.P.; Yang, Y. J. Appl. Phys. 2000, 88, 605. 35. Chiang, C. K.; Fincher, C. R.; Park, Y. W.; Heeger, A. J.; Chirakawa, H.; Louis, E. J.; Gua, S. C.; MacDiarmid, A. G. Phys. Rev. Lett. 1977, 39, 1098

PAGE 127

110 36. Antoniadis, H.; Hsieh, B. R.; Abkowitz, M. A.; Jenekhe, S. A.; Stolka, M. Synth. Met. 1994, 62, 265. 37. McQuade, D. T.; Pullen, A. C.; Swager, T. M. Chem. Rev. 2000, 100, 2537. 38. Torsi, L.; Dodabalapur, A.; Rothberg, L. J.; Fung, A. W. P.; Katz, H. E. Science 1996, 272, 1462. 39. Rauscher, U.; Schutz, L.; Greiner, A.; Bassler, H. J. Phys. Condens. Matt. 1989, 1, 9751. 40. Kersting, R.; Lemmer, U.; Mahrt, R. F.; Leo, K.; Kurz, H.; Bassler, H.; Gobel, E. O. Phys. Rev. Lett. 1993, 70, 3820. 41. Birks, J. B. Photophysics of Aromatic Molecules, Wiley Interscience, London, 1970. 42. Pope, M.; Swenberg, C. E. Electronic Processes in Organic Crystals, Oxford University Press, New York, 1999. 43. Blatchford, J. W.; Gustafson, T. L.; Epstein, A. J. J. Chem. Phys. 1996, 105, 9214. 44. Blatchford, J. W.; Jessen, S. W.; Lin, L. -B., Gustafson, T. L.; Fu, D. -K.; Wang, H. L.; Swager, T. M.; MacDiarmid, A. G.; Epstein, A. J. Synth. Met. 1997, 84, 501. 45. Yan, M.; Rothberg, L. J.; Kwock, E. W.l Miller, T. M., Phys. Rev. Lett. 1995, 75, 1992. 46. Hagler, T. W.; Pakbaz, K.; Voss. K. F.; Heeger, A. J. Phys. Rev. B 1991, 44, 8651. 47. Berggren, M.; Ingans, O.; Gustafsson, G.; Andersson, M. R.; Hjertberg, T.; Wennerstrm, O. Nature, 1994, 372, 444. 48. Bradley, D. D. C.; Grell, M.; Long, X.; Mellor, H.; Grice, A. SPIE 1997, 254, 3145. 49. Nguyen, T. -Q.; Martini, I. B.; Liu, J.; Schwartz, N. J. J. Phys. Chem. B, 2000, 104, 237. 50. Nguyen, T. -Q.; Schwartz, N. J. J. Chem. Phys, 2002, 116, 8198. 51. Liu, J.; Guo, T. F.; Yang, Y. J. Appl. Phys. 2002, 91, 1595. 52. Lee. T; Park, O. O. Adv. Mater. 2002. 12. 801. 53. Kim, J.; Lee, J.; Han, C. W.; Lee, N. Y.; Chung, I. -J. Appl Phys. Lett. 2003, 82, 4238. 54. Silvestre, G. C. M.; Johnson, M. T.; Giraldo, A.; Shannon, J. M. Appl. Phys. Lett. 2001, 78, 1619.

PAGE 128

111 55. Aziz, H.; Xu, G. J. Phys. Chem. B 1997, 101, 4009. 56. Sutherland, D.G. J.; Carlisle, J. A.; Elliker,P.; Fox, G.; Hagler, T. W.; Jimenez, I.; Lee, H. W.; Pakaz, K.; Terminello, L. J.; Williams, S. C. Appl. Phys. Lett. 1996, 68, 2046. 57. Nguyen, T. P.; Speiesser, M.; Garnier, A.; de Kok, M.; Tran, V. H. Mater. Sci. Eng. 1999, B60, 76. 58. Gamerith, S.; Gaal, M.; Romaner, L.; Nothofer, H. -G.; Guntner, R.; Scandiucci de Freitas, P.; Scherf, U.; List, E. J. W. Synth Met. 2003, 139, 855. 59. Cumpston, B. H.; Parker, I. D.; Jensen, K. F. J. Appl. Phys. 1997, 81, 3716. 60. Scherf, U.; List, E. J. W. Adv. Mater. 2002, 14, 477. 61. Andersson, G. G.; de Jong, M. P.; Winands, G. J. J.; Denier van der Gon, A. W.; van IJzendoorn, L. J.; Brongersma, H. H.; de Voigt, M. J. A. J. Phys. D: Appl. Phys. 2002, 35, 1103. 62. Tada, K.; Onoda, M. Appl. Phys. Lett. 2000, 77, 2539. 63. Ke, L.; Zhang, K.; Yakovlev, N.; Chua, S. -J.; Chen, P. Mater. Sci. Eng. 2003, B97, 1. 64. Ke, L.; Chua, S. -J.; Zuang, K.; Yakovlev, N. Appl. Phys. Lett. 2002, 80, 2195. 65. Savvateev, V. N.; Yakimov, A. V.; Davidov, D. Appl. Phys. Lett. 1997, 71, 3344. 66. Lim, S. L.; Ke, L.; Wang, W.; Chua, S. J.; Appl. Phys. Lett. 2001, 78, 2116. 67. Kim, J. -S.; Ho, P. K. H.; Murphy, C. E.; Baynes, N.; Friend, R. H. Adv. Mater. 2002, 14, 206. 68. McElvain, J.; Antomiadis, H.; Hueschen, M. R.; Miller, J. N.; Roitman, D. M.; Sheets, J. R.; Moon, R. L. J. Appl. Phys. 1996, 80, 6002. 69. Do, L. -M.; Oyamada, M.; Koike, A.; Han, E. -M.; Yamamoto, N.; Fujihira, M. Thin Solid Films 1996, 273, 209. 70. Emslie, A. G.; Bonner, F. T.; Peck, L. G. J. Appl. Phys. 1958, 29, 858. 71. Meyerhofer, D. J. Appl. Phys. 1978, 49, 3993. 72. Flack, W. W.; Soong, D. S.; Bell, A. T.; Hess, L. E. J. Appl. Phys. 1984, 56, 1199 73. Lawrence, C. J. Phys. Fluids 1988, 31, 2786. 74. Spangler, L. L.; Torkelson, J. M.; Poyal, J. S. Polym. Eng. Sci. 1990, 30, 644.

PAGE 129

112 75. Schubert, D. W. Polym. Bull. 1997, 37, 177. 76. Strawhecker, K. E.; Kumar, S. K.; Douglas, J. F.; Karim, A. Macromolecules 2001, 34, 4669. 77. Mller-Buschbaum, P.; Gutmann, J. S.; Wolkenhauer, M.; Kraus, J.; Stamm, M.; Smiligies, D.; Petry, W. Macromolecules 2001, 34, 1369. 78. Scriven, L. E.; Sternling, C. V. Nature 1960, 187, 186. 79. Pearson, J. R. A. Fluid Mech. 1958, 4, 489. 80. Haas, D. E.; Birnie, III, D. P. J. Mater. Sci. 2002, 39, 2100. 81. Birnie, III, D. P. J. Mater. Sci. 2001, 16, 1145. 82. Mller-Buschbaum, P.; Stamm, M. Macromolecules 1998, 31, 3686. 83. Mller-Buschbaum, P.; Gutmann, J. S.; Lorenz, C.; Schmitt, T.; Stamm, M. Macromolecules 2001, 34, 1369. 84. Bollinne, C.; Cuenot, S.; Nysten, B.; Jonas, A. M. Eur. Phys. J. E 2003, 12, 389. 85. Mller-Buschbaum, P. J. Phys.: Condens. Matter 2003, 15, R1549. 86. Bollinne, C.; Stone, V. W.; Carlier, V.; Jonas, A. M. Macromolecules 1999, 32, 4719. 87. Jacobs, K.; Herminghaus, S. Langmuir 1998, 14, 965. 88. Renger, C.; Mller-Buschbaum, P.; Stamm, M.; Hinrichsen, G. Macromolecules 2000, 33, 8388. 89. Gu, X.; Raghavan, D.; Douglas, J. F., Karim, A. J. Polm. Sci. Part B: Polym. Phys. 2002, 40. 2825. 90. Mort, J.; Pfister, G. Electronic Properties of Polymers, Wiley-Interscience, New York, 1982. 91. Mitschke, U.; Bauerle, P. J. Mater. Chem. 2000, 10, 1471. 92. Kido, J.; Hongawa, K.; Okuyama, K.; Nagai, K. Appl. Phys. Lett. 1993, 63, 2627. 93. Peng, J. B.; Yu, B. Y.; Pyun C. H.; Kim, C. H.; Jin, J. I. Jpn. J. Appl. Phys., Part 1 1996, 35, 4379. 94. Linde, D. R. Ed CRC Handbook of Chemistry and Physics, 78th ed. CRC Press Inc., New York, 1999.

PAGE 130

113 95. Sitaramaiah, G.; Jacobs, D. Polymer 1970, 11, 165. 96. Kuwahara, N.; Higashida, S.; Nakato, M.; Kaneko, M. J. Poly. Sci,. Part A: Polym. Chem. 1969, 7, 285. 97. Yoneda, Y. Phys. Rev. 1963, 131, 2010. 98. Parrat. L. G. Phys. Rev. 1954, 55, 359. 99. Bnard, H. Rev. Gen. Sci. Pures Appl. 1900, 11, 1261. 100. Lee, S. T.; Gao, Z. Q.; Hung, L. S. Appl. Phys. Lett. 1999, 75, 1900. 101. Ke, L.; Chua, S.-J.; Zuang, K.; Yakovlev, N. Appl. Phys. Lett. 2002, 80, 2195. 102. Bonaccurso, E.; Butt, H. J.; Franz, V.; Graf. K.; Kappl M.; Loi, S.; Niesenhaus, B.; Chemnitz, S.; Bohm, M.; Petrova, B.; Jonas, U.; Spiess, H. W. Langmuir 2002, 18, 8056. 103. Mller-Buschbaum, P.; Hermsdorf, N.; Gutmann, J. S.; Stamm, M.; Cunis, S.; Gehrke, R.; Petry, W. J. Macromol. Sci., Phys. 2004, 43, 29. 104. Hu, D.; Yu, J.; Wong, K.; Bagchi, B.; Rossky, P. J.; Barbara, P. F. Nature 2000, 405, 1030. 105. Yu, J.; Hu, D.; Barbara, P. F. Science 2000, 289, 1327. 106. Jakubiak, R.; Collision, C. J.; Wen, W. C.; Rothberg, L. J.; Hsieh, B. R. J Phys Chem A 1999, 103, 2394. 107. Brandrup, J.; Immergut, E. H.; Grulke, E. A. Ed. Polymer Handbook (4 th Ed), Wiley: New York 1999. 108. Grice, A. W.; Bradley, D. D. C; Bernius, M. T.; Inbasekaran, M.; Wu, W. W.; Woo, E. P. Appl. Phys. Lett. 1998, 73, 629. 109. Zojer, E.; Pogantsch, A.; Hennebicq, E.; Beljonne, D.; Bredas, J. L.; de Freitas, P. S.; Scherf, U.; List, E. J. W. J. Chem. Phys. 2002, 117, 6794. 110. Gamerith, S.; Gaal, M.; Romaner, L.; Nothofer, H. G.; Guntner, R.; de Freitas, P. S.; Scherf, U.; List, E. J. W. Synth. Met. 2003, 139, 855. 111. List, E. J. W.; Guentner, R.; de Freitas, P. S.; Scherf, U. Synth. Met. 2003, 139, 759. 112. Yang, X. H.; Jaiser, F.; Neher, D.; Lawson, P. V.; Bredas, J. L.; Zojer, E.; Guentner, R.; de Freitas, P. S.; Forster, M.; Scherf, U. Adv. Funct. Mater. 2004, 14, 1097. 113. Chen, S. H.; Chou, H. L.; Su, A. C.; Chen, S. A. Macromolecules 2004, 37, 6833.

PAGE 131

114 114. Chen, S. H.; Su, A. C.; Su. C. H.; Chen, S. A. Macromolecules 2005, 38, 379. 115. Keddie, J. L.; Jones, R. A. L.; R. A. Cory, Faraday Discuss. Chem. Soc. 1994, 98, 219. 116. Keddie, J. L.; Jones, R. A. L.; R. A. Cory, Europhy. Lett 1994, 27, 59. 117. Forrest, J. A.; Dalnoki-Veress, K.; Dutcher J. R. Phys. Rev. E 1997, 56, 5705. 118. van Zanten, J. H.; Wallace, W. E.; Wu, W. -L. Phys. Rev. E 1996, 53, R2053. 119. Tsui, O. K. C.; Russell, T. P.; Hawker, C. J. Macromolecules 2001, 34, 5535. 120. Reiter, G. Macromolecules 1994, 27, 3046. 121. Brown H.; Russell, T. P. Macromolecules 1996, 29, 798. 122. Mayers, A. M. Macromolecules 1994, 27, 3114. 123. Tanaka, K.; Takahara, A.; Kaijiyama, T. Macromolecules 1997, 30, 6626. 124. Bollinne, C; Stone, V. W.; Carlier, V.; Jonas, A. M. Macromolecules 1999, 32, 4719.

PAGE 132

BIOGRAPHICAL SKETCH Shyh-Chyang Luo was born in Kauhsung, a port city in south Taiwan. He entered the National Taiwan University (NTU) in 1992 and majored in chemistry. He was awarded his bachelors degree in 1996 and continued his study in the Department of Materials Science and Engineering. He joined Dr. Wen-Bin Liaus group and started his research in polymer science. His investigation was related to the relaxation behavior of C 60 -based star polyether urethane. After graduation in 1998, he enlisted in the Army for two years, serving as an artillery lieutenant during his tour of duty. Thereafter, he returned to university as a research assistant for one year and then was granted admission to the University of Florida in the doctoral program for the Materials Science and Engineering Department. With Dr. Elliot P. Douglas as his mentor, Mr. Luo honed his topic of study to specialize in light-emitting polymers. 115