University of Kentucky University of Kentucky

UKnowledge UKnowledge

Theses and Dissertations--Physics and Astronomy Physics and Astronomy

2017

THERMAL CONDUCTIVITIES OF ORGANIC SEMICONDUCTORS THERMAL CONDUCTIVITIES OF ORGANIC SEMICONDUCTORS

Yulong Yao University of Kentucky, yya233@gmail.com Digital Object Identifier: https://doi.org/10.13023/ETD.2017.337

Right click to open a feedback form in a new tab to let us know how this document benefits you. Right click to open a feedback form in a new tab to let us know how this document benefits you.

Recommended Citation Recommended Citation Yao, Yulong, "THERMAL CONDUCTIVITIES OF ORGANIC SEMICONDUCTORS" (2017). Theses and Dissertations--Physics and Astronomy. 48. https://uknowledge.uky.edu/physastron_etds/48

This Doctoral Dissertation is brought to you for free and open access by the Physics and Astronomy at UKnowledge. It has been accepted for inclusion in Theses and Dissertations--Physics and Astronomy by an authorized administrator of UKnowledge. For more information, please contact UKnowledge@lsv.uky.edu.

STUDENT AGREEMENT: STUDENT AGREEMENT:

I represent that my thesis or dissertation and abstract are my original work. Proper attribution

has been given to all outside sources. I understand that I am solely responsible for obtaining

any needed copyright permissions. I have obtained needed written permission statement(s)

from the owner(s) of each third-party copyrighted matter to be included in my work, allowing

electronic distribution (if such use is not permitted by the fair use doctrine) which will be

submitted to UKnowledge as Additional File.

I hereby grant to The University of Kentucky and its agents the irrevocable, non-exclusive, and

royalty-free license to archive and make accessible my work in whole or in part in all forms of

media, now or hereafter known. I agree that the document mentioned above may be made

available immediately for worldwide access unless an embargo applies.

I retain all other ownership rights to the copyright of my work. I also retain the right to use in

future works (such as articles or books) all or part of my work. I understand that I am free to

register the copyright to my work.

REVIEW, APPROVAL AND ACCEPTANCE REVIEW, APPROVAL AND ACCEPTANCE

The document mentioned above has been reviewed and accepted by the student’s advisor, on

behalf of the advisory committee, and by the Director of Graduate Studies (DGS), on behalf of

the program; we verify that this is the final, approved version of the student’s thesis including all

changes required by the advisory committee. The undersigned agree to abide by the statements

above.

Yulong Yao, Student

Dr. Joseph W. Brill, Major Professor

Dr. Christopher Crawford, Director of Graduate Studies

THERMAL CONDUCTIVITIES OF ORGANIC SEMICONDUCTORS

________________________________________

DISSERTATION ________________________________________

A dissertation submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy in the College of Arts and Sciences

at the University of Kentucky

By Yulong Yao

Lexington, Kentucky

Director: Dr. Joseph W. Brill, Professor of Physics and Astronomy Lexington, Kentucky 2017

Copyright © Yulong Yao 2017

ABSTRACT OF DISSERTATION

THERMAL CONDUCTIVITIES OF ORGANIC SEMICONDUCTORS

Organic semiconductors have gained a lot of interest due to their ease of processing, low-cost and inherent mechanical flexibility. Although most of the research has been on their electronic and optical properties, knowledge of the thermal properties is important in the design of electronic devices as well. Our group has used ac-calorimetric techniques to measure both in-plane and transverse thermal conductivities of a variety of organic semiconductors including small-molecule crystals and polymer blends. For layered crystals composed of molecules with planar backbones and silylethynyl (or germylethynyl) sidegroups projecting between the layers, very high interplanar thermal conductivities have been observed, presumably implying that heat flows between layers mostly via interactions between librations on these sidegoups.

Since most organic semiconducting devices require materials in thin film rather than bulk crystal form, I have focused on using the “3ω- technique” to measure the thermal resistances of thin films of this class of organic semiconductors, including bis(triisopropylsilylethynyl) pentacene (TIPS-pn), bis(triethylsilylethynyl) anthradithiophene (TES-ADT), and difluoro bis(triethylsilylethynyl) anthradithiophene (diF-TES-ADT). For each material, several films of different thicknesses have been measured to separate the effects of intrinsic thermal conductivity from interface thermal resistance. For sublimed films of TIPS-pn and diF-TES-ADT, with thicknesses ranging from less than 100 nm to greater than 4 μm, the thermal conductivities are similar to those of polymers and over an order of magnitude smaller than those of single crystals, presumably reflecting the large reduction in phonon mean-free path due to disorder in the films. On the other hand, the thermal resistances of thin (≤ 205 nm) crystalline films of TES-ADT, prepared by vapor-annealing of spin-cast films, are dominated by their interface resistances, possibly due to dewetting of the film from the substrate during the annealing process. KEYWORDS: organic semiconductors, ac-calorimetry, thermal conductivity, thin film, 3ω-technique, interface thermal resistance

Author’s Signature: Yulong Yao

Date: July 5, 2017

THERMAL CONDUCTIVITIES OF ORGANIC SEMICONDUCTORS

By Yulong Yao

Joseph W. Brill Director of thesis

Christopher Crawford

Director of graduate studies

July 5, 2017

iii

Acknowledgments

This work would not have been possible without the help, encouragement and

guidance of many people.

First of all, I would like to express my deepest gratitude to my advisor, Dr. Joseph

Brill, for his intelligent guidance, inspiration, encouragement throughout this study and

providing me with an excellent environment for doing research. Also, I would like to thank

the other members of my advisory committee, Dr. Kwok-Wai Ng, Dr. Ambrose Seo and

Dr. John Anthony for their guidance and support to my project.

In addition, I give my special thanks to my former colleague Dr. Hao Zhang for his

caring and help at the beginning of my research, John Connell and Mohsen Nasseri for

their help with AFM measurements and E-beam evaporation. Also, I wish to thank Dr.

Emily Bittle, Maryam Shahi, Shaoqian Wang and Gen Wang for their help and friendship.

I also share the credit of my work with Dr. Marcia Payne for her small molecule

organic crystals, and the Crispin group from Linköping University for the polymer

samples.

I would like to thank the entire department staff for all the help they provided,

especially Greg Porter for the support he provided in electronics, Jim Morris, Charles

Tipton and Steve Maynard for machining many of the pieces required for my work, and

Gene Baber for his help maintaining our apparatuses.

I would also like to thank Dr. Todd Hastings from Electrical and Computer

Engineering at University of Kentucky for helping us with photomasks and Dr. Ahmet

Alatas at Argonne National Lab for working with us on the inelastic X-ray scattering.

Finally, I am grateful to my parents in China who always supported me with deepest

love, blessing and understanding while I was pursuing this doctoral degree.

I really appreciate and will always remember my experience and everybody's help

during my graduation education.

My research was supported with the following grant from the U.S. National Science

Foundation DMR-1262261.

iv

TABLE OF CONTENTS

Acknowledgments.............................................................................................................. iii

List of Figures .................................................................................................................... vi

Chapter 1 Background and Theory ..................................................................................... 1

1.1 Introduction ........................................................................................................... 1

1.2 AC-calorimetry Techniques .................................................................................. 5

1.3 3ω-Technique ........................................................................................................ 9

Chapter 2 Experimental Setup and Sample Preparation ................................................... 12

2.1 In-Plane Thermal Conductivity Measurements .................................................. 12

2.2 3ω signal measurements ..................................................................................... 14

2.3 Temperature coefficient of resistance measurements ......................................... 16

2.4 Optical measurements ......................................................................................... 18

2.5 Sample Fabrication ............................................................................................. 20

2.5.1 Substrate Surface Cleaning .............................................................................. 20

2.5.2 Thin Films Deposition ..................................................................................... 21

2.5.3 Heater Deposition ............................................................................................ 25

Chapter 3 In-Plane Thermal Conductivity Measurements of Organic Semiconductors ... 28

3.1 NFC-PEDOT Paper ............................................................................................ 28

3.2 FS-PEDOT:PSS .................................................................................................. 33

3.3 TIPS-pn Functionalized Pentacene Semiconductors .......................................... 36

3.3.1 In-Plane Measurements .................................................................................... 37

3.3.2 Interlayer Measurements .................................................................................. 38

3.3.3 Discussion ........................................................................................................ 39

3.3.4 Inelastic X-ray Measurements ......................................................................... 43

v

3.3.5 Summary .......................................................................................................... 45

Chapter 4 Thermal Resistances of Thin-Films of Small Molecule Organic Semiconductors ................................................................................................ 47

4.1 Introduction ......................................................................................................... 47

4.2 Results and Discussion ....................................................................................... 48

4.3 Conclusion .......................................................................................................... 53

Chapter 5 Conclusions ...................................................................................................... 54

References ......................................................................................................................... 57

VITA ................................................................................................................................. 66

vi

List of Figures

FIGURE 1.1 MOLECULAR STRUCTURE OF (A) BIS(TRIISOPROPYLSILYLETHYNYL)

PENTACENE (TIPS-PN); (B) BIS(CYCLOPROPYL-DIISOPROPYLSILYLETHYNYL)

PENTACENE (CP-DIPS-PN [24]); (C) BIS(TRIETHYLSILYLETHYNYL)

ANTHRADITHIOPHENE (TES-ADT); (D) DIFLUORO

BIS(TRIETHYLSILYLETHYNYL) ANTHRADITHIOPHENE (DIF-TES-ADT); (E)

BIS(TRIISOPROPYLGERMYLETHYNYL) PENTACENE (TIPGE-PN [24]); (F)

BIS(TRIISOPROPYLSILYLETHYNYL) OCTAFLUOROPENTACENE (F8-TIPS-PN

[25]); (G) BIS(TRIISOPROPYLSILYLETHYNYL) PERIFLUOROPENTACENE

(F2-TIPS-PN[24]). (H)

TETRAETHYL-BIS(TRIISOPROPYLSILYLETHYNYL)-TETRAOXADICYCLOPENTA[

B, M] PENTACENE (ETTP-5 [26]) ........................................................................ 3

FIGURE 1.2 TIPS-PN: (A) AB-PLANE BRICKWORK STRUCTURE (THE SOLID BARS

REPRESENT THE PENTACENE BACKBONES); (B) BC-PLANE STRUCTURE. (THE

GREEN CIRCLES REPRESENT THE SILICON ATOMS.) [29] ..................................... 4

FIGURE 1.3 SCHEMATIC (NOT TO SCALE) OF THE SAMPLE ARRANGEMENT FOR

PHOTOTHERMAL MEASUREMENTS OF THE TRANSVERSE THERMAL

DIFFUSIVITY. (THE LPF IS LONG-WAVE PASS FILTER NEEDED FOR

TIPS-PENTACENE SAMPLES) ............................................................................ 7

FIGURE 1.4 SCHEMATIC DIAGRAM FOR THE LONGITUDINAL THERMAL DIFFUSIVITY

MEASUREMENTS. ............................................................................................... 8

FIGURE 1.5 3Ω-TECHNIQUE: MEASUREMENTS OF (A) YIELD THE THERMAL CONDUCTIVITY

OF THE SUBSTRATE; OFFSET OF (B) COMPARED TO (A) YIELDS THE THERMAL

RESISTANCE OF THE ORGANIC THIN FILMS ........................................................ 11

FIGURE 2.1 INSIDE (A) AND OUTSIDE (B) PICTURES OF THE VACUUM CHAMBER BUILT AT

UNIVERSITY OF KENTUCKY FOR AC CALORIMETRY MEASUREMENTS. .............. 13

FIGURE 2.2 CIRCUIT DIAGRAM FOR 3Ω MEASUREMENTS. ................................................... 14

vii

FIGURE 2.3 NULL BRIDGE SCHEMATIC WITH THREE TUNING POT SETTING DESIGNED BY

GREG PORTER AT UNIVERSITY OF KENTUCKY. ................................................ 15

FIGURE 2.4 3Ω SIGNAL MEASURED AT SEVERAL PRESET FREQUENCIES AND CONVERTED

TO ΔT/P FOR A VAPOR ANNEALED TES ADT FILM ON A SAPPHIRE

SUBSTRATE. THE LINE SHOWS THE RESPONSE FOR THE BARE SAPPHIRE

SUBSTRATE. ..................................................................................................... 16

FIGURE 2.5 CIRCUIT DIAGRAM FOR TEMPERATURE COEFFICIENT OF RESISTANCE

MEASUREMENTS. ............................................................................................. 17

FIGURE 2.6 INSIDE PICTURE OF THE VACUUM CHAMBER (AN OXFOR MICROSTATN

LIQUID NITROGEN CRYOSTAT) FOR 3Ω MEASUREMENTS. .................................. 17

FIGURE 2.7 RESISTANCE-TEMPERATURE CHARACTERIZATION OF A COPPER STRIP

DEPOSITED ON DIF-TES-ADT THIN FILMS BEFORE (RED) AND AFTER (BLUE)

3Ω MEASUREMENTS. ........................................................................................ 18

FIGURE 2.8 TWO DIFFERENT MODES OF OPERATION OF THE IR MICROSCOPE: THE

REFLECTION MODE AND THE TRANSMISSION MODE. ......................................... 19

FIGURE 2.9 SINGLE SCAN SPECTRUM OF TRANSMISSION THROUGH A EMPTY HOLE ON

MICROSCOPE STAGE, THE OVERALL SHAPE OF THE SPECTRUM IS DUE TO THE

SENSITIVITY OF THE DETECTOR, TRANSMISSION AND REFLECTIVE PROPERTIES

OF THE MIRRORS, AND SOURCE SPECTRUM. COMMON FEATURES AROUND

3500 AND 1630 CM-1 ARE DUE TO ATMOSPHERIC WATER VAPOR, AND THE

BANDS AT 2350 AND 667 ARE DUE TO ATMOSPHERIC CARBON DIOXIDE. .......... 20

FIGURE 2.10 DROP CAST FILMS FROM 10% TIPS-PN CHLOROBENZENE SOLUTION BY

SLOW DRY (A) AND 3% TIPS-PN CHLOROBENZENE SOLUTION BY FAST DRY

(B).................................................................................................................... 21

FIGURE 2.11 DROP CASTING WITH A TEFLON CONTAINER ................................................ 22

FIGURE 2.12 A SPIN-COATED TES-ADT THIN FILMS ON SILICON WAFER BEFORE (A)

AND AFTER (B) VAPOR ANNEALING UNDER MICROSCOPE. (C) DIAGRAM OF

VAPOR ANNEALING PROCESS. ........................................................................... 23

viii

FIGURE 2.13 UV-VISIBLE ABSORPTION SPECTRUM OF THE SUBLIMED TIPS-PN FILMS

(DASH RED), TIPS-PN SOLUTION (GREEN), SUBLIMED TES-ADT FILMS (DASH

BLUE), AND TES-ADT SOLUTION (BLACK). ..................................................... 24

FIGURE 2.14 POLARIZED INFRARED TRANSMISSION SPECTRA OF SMALL AREAS OF

DIF-TES-ADT FILMS EVAPORATED ON A KRS5 SUBSTRATE, TAKEN IN (50

µM)2 SPOTS. ...................................................................................................... 25

FIGURE 2.15 SHADOW MASK SAMPLES WITH TWO DIFFERENT FOUR PROBES FEATURE

MADE AT UNIVERSITY OF LOUISVILLE BY A DEEP REACTIVE-ION ETCHING

PROCESS. .......................................................................................................... 26

FIGURE 2.16 POLARIZED INFRARED TRANSMISSION SPECTRA OF A SINGLE SPHERULITE

IN A SPIN-CAST, VAPOR ANNEALED TES-ADT FILM ON A KRS5 SUBSTRATE,

TAKEN IN (100 µM)2 SPOTS ADJACENT TO THE COPPER HEATER LINE AND 1

MM AND 2 MM AWAY FROM THE LINE. (X AND Y REFER TO PERPENDICULAR

MICROSCOPE AXES.) THE CURVES FOR EACH POSITION ARE VERTICALLY

OFFSET. ............................................................................................................ 27

FIGURE 3.1 SURVEY IONIC AND/OR ELECTRONIC CONDUCTORS. WITH THE EXCEPTION OF

IONIC LIQUIDS, ONLY SOLID CONDUCTORS ARE INCLUDED. THE POINTS IN THE

GRAPH REPRESENTS THE FOLLOWING MATERIALS: A: NAFION [62]; B:

POLY(DIALLYLDIMETHYL AMMONIUM

CHLORIDE)/POLY(2,6-DIMETHYL1,4-PHENYLENE OXIDE) [62]; C:

POLY(4-STYRENESULFONIC ACID) (105); D: POLY(ETHYLENE

OXIDE)/POLY(ACRYLIC) ACID/POLY(ETHYLENE OXIDE)/(POLY(ACRYLIC)

ACID/MULTI WALLED CARBON NANOTUBES) [63]; E: POLYVINYLIDENE

FLUORIDE/POLYETHYLENE OXID/PROPYLENE CARBONATE/ LICLO4 [64]; F:

(LITHIUM BIS(OXLATE)BORATE AND LITHIUM

TETRAFLUOROBORATE)/1-ETHYL-3-METHYL-IMIDAZOLIUM

TETRAFLUOROBORATE [65]; G: LICF3SO3/POLY(METHYL METHACRYLATE),

LICLO4/POLY(METHYL METHACRYLATE) AND LICLO4/PROPYLENE

ix

CARBONATE/ETHYLENE CARBONATE/

DIMETHYLFORMAMIDE/POLY(ACRYLONITRILE) [66]; H: LI10GEP22S12 [67]; I:

AG2HFS3 [68]; J: AG2S [69]; K: LI3.5V0.5G0.5O4 [70]; L:

CE0.8GD0.2O2-D-COFE2O4 [45]; M:

POLY(3,4-ETHYLENEDIOXYTHIOPHENE):POLYSTYRENE SULFONATE AND :

POLY(3,4-ETHYLENEDIOXYTHIOPHENE):POLYSTYRENE SULFONATE/SODIUM

POLYSTYRENE SULFONATE [59]; N:

POLY-[1-METHYL-3-(PYRROL-L-YLMETHYL)PYRIDINIUM PERCHLORATE]

[71]; O: POLYANILINE [46, 47]; P: POLYPYRROLE [72, 73]; Q:

POLY(3,4-ETHYLENEDIOXYTHIOPHENE):POLYSTYRENE

SULFONATE/NANOFIBRILLATED CELLULOSE/DIMETHYL

SULFOXIDE/POLYETHYLENE GLYCOL (THIS WORK); R/S: GAAS [74]; T:

NICHROME [75]; U: AG [75]. ............................................................................ 31

FIGURE 3.2 POSITION DEPENDENCE OF THE OSCILLATING THERMOCOUPLE SIGNAL (VAC)

FOR TWO DIFFERENT CHOPPING FREQUENCIES FOR A 30 M THICK SAMPLE;

X = DISTANCE BETWEEN THE EDGE OF THE SCREEN AND THE THERMOCOUPLE,

AND X0 = CONSTANT OFFSET. TOP INSET: SPECIFIC HEAT OF A 14 MG PELLET

MEASURED WITH DIFFERENTIAL SCANNING CALORIMETRY AT A SCANNING

RATE OF 18 K/MINUTE. .................................................................................... 33

FIGURE 3.3 EXPERIMENTAL DATA USED TO DETERMINE THE IN-PLANE DIFFUSIVITY:

SPATIAL DEPENDENCE OF THE THERMOCOUPLE SIGNAL AT A FEW

FREQUENCIES. DLONG IS DETERMINED FROM THE SLOPES BETWEEN THE

DOTTED LINES, AS DISCUSSED IN THE TEXT. A SOLID LINE WITH SLOPE =

2.8/MM IS SHOWN FOR REFERENCE. .................................................................. 36

FIGURE 3.4 RESULTS FOR AN 8 MM LONG TIPS-PN NEEDLE. THE DOTTED LINES SHOW

THE REGION OF THE EXPECTED LINEAR VARIATION OF F-1/2LN(VAC) FROM

WHICH THE DIFFUSIVITY IS CALCULATED. THE INSET SHOWS THE SPECIFIC

HEAT OF A PELLET MEASURED BY DSC. ........................................................... 38

x

FIGURE 3.5 FREQUENCY DEPENDENCE OF FVΩ FOR REPRESENTATIVE CRYSTALS; THE

CURVES SHOW FITS TO EQN. 1.4. (A) TIPS-PN (SIGNAL × 3), D = 610 ΜM,

τMEAS = 2.8 MS → DC > 14 MM2/S; (B) F2-TIPS-PN, D ≈ 300 ΜM, τMEAS = 1.26

MS → DC > 7 MM2/S; (C) TIPGE-PN (SIGNAL × 2), D = 460 ΜM, τMEAS = 1.03

MS → DC > 20 MM2/S; (D) ETTP-5, D ≈ 300 µM, τMEAS = 0.68 MS → DC > 14

MM2/S. INSET: SPECIFIC HEAT OF TIPS-PN. ...................................................... 39

FIGURE 3.6 SPATIAL DEPENDENCE OF VΩ FOR CRYSTALS OF CP-DIPS-PN (LENGTH = 6

MM, -F-1/2 DLNVΩ/DX = 2.7 ± 0.3 (HZ1/2·MM)-1, DLONG = 0.43 ± 0.10 MM2/S),

F2-TIPS-PN (LENGTH = 10 MM, -F-1/2 DLNVω/DX = 1.89 ± 0.07 (HZ1/2·MM)-1,

DLONG = 0.88 ± 0.09 MM2/S), AND F8-TIPS-PN (LENGTH = 11 MM, -F-1/2

DLNVω/DX = 1.98 ± 0.10 (HZ1/2·MM)-1, DLONG = 0.80 ± 0.10 MM2/S) AT

SELECTED FREQUENCIES IN THEIR LINEAR REGIONS. ........................................ 43

FIGURE 3.7 INELASTIC X-RAY SPECTRA TAKEN AT THE ADVANCED PHOTON SOURCE AT

ARGONNE NATIONAL LAB. (A)ENERGY RESOLUTION; (B)FOUR C* WAVE

VECTORS Q = [0, 0, -(2+Q)] ENERGY SCANS; (C)FOUR B* WAVE VECTORS Q =

[0, 1+Q, 0] ENERGY SCANS; (D)FOUR A* WAVE VECTORS Q = [-(2+Q), 0, 0]

ENERGY SCANS. ............................................................................................... 45

FIGURE 4.1 MEASURED FREQUENCY DEPENDENCE OF THE ∆T/P FOR SPIN-CAST

TES-ADT FILMS ON SAPPHIRE SUBSTRATES. THE OPEN TRIANGLES SHOW

THE RESULTS ON BARE SAPPHIRE AND THE SOLID LINE IS A FIT TO EQN. 1.5.

THE DASHED LINE SHOWS THE EXPECTED RESULTS FOR 500 NM THICK

TES-ADT, ASSUMING Κ = ΚC(CRYSTAL) = 5W/M⋅K. THE SOLID SYMBOLS

SHOW RESULTS FOR VAPOR-ANNEALED FILMS OF DIFFERENT THICKNESSES,

AS SHOWN, AND THE OPEN CIRCLES AND SQUARES SHOW RESULTS FOR

NON-ANNEALED FILMS. .................................................................................... 49

FIGURE 4.2 MEASURED FREQUENCY DEPENDENCE OF ∆T/P FOR ≈ 100 NM THICK

VAPOR-ANNEALED (SOLID BLUE CIRCLES) AND NON-ANNEALED (OPEN BLUE

CIRCLES) SPIN-CAST TES-ADT FILMS ON THERMALLY OXIDIZED SILICON.

xi

THE RED TRIANGLES SHOW THE RESULTS ON THE BARE OXIDIZED SILICON

AND THE CALCULATED SILICON BASELINE [FROM EQN. 1.5] IS SHOWN BY THE

SOLID LINE. ...................................................................................................... 50

FIGURE 4.3 FREQUENCY DEPENDENCE OF SUBLIMED FILMS OF DIF-TES-ADT (LEFT

PANELS) AND TIPS-PN (RIGHT PANELS) OF THE INDICATED THICKNESSES ON

SAPPHIRE SUBSTRATES. THE REFERENCE BARE SAPPHIRE LINE (FROM FIGURE

4.1) IS SHOWN IN THE LOWER LEFT PANEL. ...................................................... 51

FIGURE 4.4 THICKNESS DEPENDENCE OF FILM THERMAL RESISTANCE (∆TFILM/P) FOR

SUBLIMED FILMS OF DIF-TES-ADT (TRIANGLES) AND TIPS-PN (CIRCLES).

THE SOLID SYMBOLS SHOW THE PROFILOMETER MEASUREMENTS OF THE

FILM THICKNESSES WHILE THE OPEN SYMBOLS SHOW THE THICKNESSES AS

DETERMINED BY THE QUARTZ CRYSTAL MONITOR DURING SUBLIMATION.

ALSO SHOWN (OPEN INVERTED RED TRIANGLES) ARE THE RESULTS FOR THE

NON-ANNEALED SPIN-CAST TES-ADT FILMS. THE INSET SHOWS A BLOW-UP

OF THE RESULTS WITH T < 1 ΜM. ...................................................................... 52

1

Chapter 1 Background and Theory

1.1 Introduction

In recent years, there has been extensive research on the properties and possible

applications of organic semiconductors due to their ease of processing, low-cost and

inherent mechanical flexibility. Most of the interest is on electronic applications, e.g.

displays (i.e. organic light emitting diodes: “OLEDs”), electronic paper (and paper

electronic [1]), solar cells and other photovoltaic devices, and thermoelectric generators.

Although most of the research has been on characterizing and understanding

electronic and optical properties of this material class [2-10], knowledge of the thermal

properties is, of course, important in the design of electronic devices. For example, for

submicron thin-film transistors, one would desire a thermal conductivity κ > κ0 ~ 1

W/m·K [11] to prevent device failure due to thermal fatigue after prolonged usage. On

the other hand, one desires low thermal conductivity, κ < κ0, for thermoelectric

applications; the dimensionless thermoelectric figure of merit is

κσ /2 TSZT = (1.1)

where S is the Seebeck coefficient, σ the electrical conductivity, T the temperature, and κ

= κel + κph, the sum of the electron and phonon thermal conductivties, and for lightly

doped semiconductors, the thermal conductivity is predominantly due to phonons.

Therefore organic semiconductors have been used to make thermoelectric generators.

Lightly doped semiconducting polymers can have Seebeck coefficients S ~ 100 μV/K but

very low phonon thermal conductivitieds (κ ~ 0.3 W/m·K), reflecting the strong

scattering of phonons in disordered polymers. Thus, values of ZT > 0.4 have been

obtained in suitably doped polymer films [6, 9].

Organic semiconductors can be broadly classified in two groups based on their

molecular weight: heterocyclic polymers with molecular weight greater than 1000, and

crystals of conjugated polycyclic molecules with molecular weight less than 1000 [12].

While most work on organic semiconductors has concentrated on polymers, there is

growing interest in using small-molecule crystals due to their higher electronic mobilities.

For electronic applications, small-molecule crystals can be used to make both faster and

2

higher current devices. For example, an electronic mobility of ~ 40 cm2/V·s, greater than

that of amorphous silicon, has been obtained in crystals of rubrene [13]. Because of their

higher charge mobilities, small-molecule crystals may also have higher thermoelectric

efficiency than polymers; crystals don’t have to be doped to the same extent as polymers

to obtain sufficient conductivity, increasing their Seebeck coefficients [14]. Of course, to

obtain high values of ZT, the material must also have small thermal conductivity, but

rubrene was observed to have thermal conductivities similar to that of polymers, i.e.

in-plane and interlayer values κin-plane ~ 0.4 W/m·K [2] and κc ~ 0.07 W/m·K [3].

Assuming that heat is carried predominantly by acoustic phonons, these values

correspond to mean-free paths of ~10 and 1 molecular spacings, respectively; i.e.

interlayer phonon transport is borderline between propagating and diffusive behavior [3].

While small-molecule materials like rubrene and pentacene are widely studied in

organic semiconductors research, they have the disadvantage of being relatively insoluble

so that thin films, desired for most applications, could only be prepared by sublimation

which is not suitable for low-cost mass production. This limitation was overcome over a

decade ago by the Anthony group by adding side groups to pentacenes and

anthradithiophenes. The substituted materials, such as 6,13 bis (triisopropylsilylethynyl)

pentacene (TIPS-pn) and bis(triethylsilylethynyl) anthradithiophene (TES-ADT) (Figure

1.1) [15-20], become soluble and are able to self-assemble into thin films when deposited

from common solvents including acetone and toluene. The deposition could be done by a

variety of techniques, such as drop casting, spin coating, dip coating [21], ink-jet printing

[22], and spray coating [23].

The molecular structures of small molecule organic semiconductors used in this

study are shown in Figure 1.1.

3

Figure 1.1 Molecular structure of (a) bis(triisopropylsilylethynyl) pentacene (TIPS-pn); (b) bis(cyclopropyl-diisopropylsilylethynyl) pentacene (CP-DIPS-pn [24]); (c) bis(triethylsilylethynyl) anthradithiophene (TES-ADT); (d) difluoro bis(triethylsilylethynyl) anthradithiophene (diF-TES-ADT); (e) bis(triisopropylgermylethynyl) pentacene (TIPGe-pn [24]); (f) bis(triisopropylsilylethynyl) octafluoropentacene (F8-TIPS-pn [25]); (g) bis(triisopropylsilylethynyl) perifluoropentacene (F2-TIPS-pn[24]). (h) tetraethyl-bis(triisopropylsilylethynyl)-tetraoxadicyclopenta[b, m] pentacene (EtTP-5 [26])

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

(e) (e) (g) (f)

(h)

4

All of these materials have layered crystal structures. The pentacene backbones of

TIPS-pn molecules form a “brick-layer” pattern in the ab-plane, with side groups

projecting between layers along c (Figure 1.2). The TIPS-pn crystals usually have a

needle shape with typical size 10 × 1 × 0.1 mm where the needle direction is along [2, 1,

0]. The needle-axis is the direction of the best π-orbital overlap [26, 27, 28], i.e. the

direction of crystal growth and highest electrical conductivity.

Figure 1.2 TIPS-pn: (a) ab-plane brickwork structure (the solid bars represent the pentacene backbones); (b) bc-plane structure. (The green circles represent the silicon atoms.) [29]

In our initial work, the thermal conductivities of layered cyrstals of several small

molecule organic semiconductors were reported [3, 4, 5]. For molecules with planar

backbones and silylethynyl (or germanylethynyl) sidegroups projecting between planes,

very high interplanar thermal conductivities were observed [4, 5]. For example, while we

found the in-plane, needle-axis thermal conductivity of TIPS-pn κneedle ≈ 1.6 W/m·K, the

inter-plane (c-axis) thermal conductivity was measured, using an ac-photothermal

technique [5], to be κc ≈ 21 W/m·K, a value close to that of sapphire [30]. As discussed

later, this large value and inverted anisotropy (i.e. κc > κneedle) has tentatively been

associated with heat flowing between layers via interactions between librations of alkyl

c

b

(a) (b)

5

groups terminating the silylethynyl sidegroups on the molecules [5]. In contrast, rubrene,

with tetracene backbones and much more rigid phenyl sidegroups, has an interlayer

thermal conductivity of only κc ≈ 0.07 W/m·K ≈ κneedle/6 [3].

While such large thermal conductivities would provide efficient dissipation of Joule

heat and therefore bode well for electronic applications of these materials, most organic

semiconducting devices require materials in thin film rather than bulk crystal form. Thin

film thermal resistances, even for crystalline films, can be much larger than the values

deduced from bulk crystalline conductivities, either because of reduced mean-free paths

in the material due to increased disorder or because of interfacial thermal resistance with

the substrate. Therefore, it was desirable to measure the thin film thermal resistance of

TIPS-pn and other small molecule materials being considered for electronic applications.

In our thin film studies, the samples we measured are “vapor annealed” [31-33] and

non-annealed spin-cast films of TES-ADT, and sublimed films of TIPS-pn and

diF-TES-ADT. The vapor annealed films are crystalline, with mm sized crystallites

oriented with c in the through-plane direction [31-33], while the other films are thought

to be ab-plane disordered, but with the molecular sidegroups also largely oriented

through-plane [33-35].

1.2 AC-calorimetry Techniques

Because of the layered crystal structures and the silyl or germyl side groups

extending between layers, the interlayer and in-plane phonon thermal conductivities of

small molecule organic semiconductors like TIPS-pn are expected to be different, and we

measured them separately. Because the crystals are small (in-plane dimensions typically

0.5-10 mm and interlayer direction less than 0.6 mm), we used ac-calorimetry techniques

[3, 36, 37], which yeild the thermal diffusivity, D ≡ κ/cρ, where c is the specific heat and

ρ is the mass density.

The interlayer measurement techniques are described in detail in References [3, 5]. If

a thin sample is heated uniformly on its “front” surface with light chopped at frequency ω

= 2πF, the magnitude and phase of the temperature oscillations on its “back” surface will

6

depend on its external (τ1) and internal (τ2) thermal time constants. Here τ1 ≡ C/Λ [37], is

the time constant with which the sample comes to equilibrium with the bath, where C is

the heat capacity of the sample and Λ is the net thermal conductance of the sample to its

thermal bath. The interlayer measurements determine τ2, the time for heat to diffuse

through the sample:

cc cdDd κρτ 90/90/ 222 ≡≡ (1.2)

where d is the thickness of the sample and Dc is the transverse (i.e. c-axis) thermal

diffusivity. In the limit ωτ1 >> 1, the complex oscillating temperature is given by [37, 38]

]sincosh)1(cossinh)1[(/4)( χχχχωπχω iiCPT inac +−−= , (1.3a)

with 2/12/12

2/1 )2/()2/90( cDd ωωτχ =≡ . (1.3b)

where Pin is the absorbed incident power and the phase of the oscillations are measured

with respect to that of the incident chopped light. The magnitude of Tac is [37]

])(1[/2|| 2/122ωτπω +≈ CPT inac (1.4)

For ωτ2 << 1, measurement of |Tac| is commonly used to measure the sample heat

capacity [37]. We used measurements at larger values of ω to determine τ2 and hence Dc

and κc for rubrene [3].

An important limitation in using Tac measurements to determine Dc, however, is that

one must also be concerned about the time constant of the thermometer used to measure

Tac. For example, we have found that for fine (e.g. 25 μm diameter) thermocouple wires

glued to the sample with silver paint, the response time of the thermometer is mostly

determined by the thermal resistivity of the paint/sample interface, which generally varies

between 0.1 and 1.0 cm2⋅K/W, giving a thermometer time constant between 1 and 10 ms

[3]. While this did not significantly affect the results for rubrene (as we determined from

the measured temperature dependence of τ2 [3]), we found that it was a problem for many

other materials, i.e. for these materials, the measured time constant, τ2,effective > intrinsic τ2

of the sample.

7

Figure 1.3 Schematic (not to scale) of the sample arrangement for photothermal measurements of the transverse thermal diffusivity. (The LPF is long-wave pass filter needed for TIPS-pentacene samples)

One can avoid this problem by measuring the oscillating thermal radiation, e.g. with

a liquid nitrogen cooled mercury cadmium telluride (MCT) detector, from the back

surface of the sample to find the frequency dependence of Tac. (This is the Fourier

transform of the “Laser-Flash” technique of measuring the thermal conductivity.

Working in the frequency domain makes it possible to measure much smaller samples

and use less intense light to minimize heating or damaging the sample.) In References [38,

39], this was done using chopped light from a laser to heat the front of the sample and

focusing the radiation from the back of the sample on the detector with off-axis parabolic

mirrors; samples had areas > 20 mm2. Because our samples were typically much smaller

than this, we chose to mount the sample in the MCT detector directly in front (within 1

cm) of the sample, thus avoiding the need for vacuum windows between the sample and

detector and losses of light signal due to alignment and aberrations in the optics (see

Figure 1.3). Instead of a laser, we used the intense quartz-halogen light source we’ve

used for ac-calorimetry, but the final incident intensity, after passing through an optical

fiber with lenses was ~10 mW/cm2.

For our measurements, although the optical fiber, lenses, and dewar window strongly

attenuate the incident infrared light, there is usually a fairly large signal from the light

8

that leaks through or around (e.g. through mounting glue) the sample. In those cases, we

only fit the response that is in quadrature with the incident light:

)sincoshcos/(sinh)sincoshcos(sinh)sin()( 22220 χχχχχχχχχθθ ++=+ RFFVac .

(1.5)

Here θ is the phase shift between the measured signal and that of the chopped light.

Fitting parameters are θ0, the error in setting the lock-in amplifier phase (typically a few

degrees), the magnitude of the signal R, and τ2, from which we determine the transverse

diffusivity by Eqn. 1.2. Experimental details, including the approximations made in using

Eqn. 1.5 are discussed in Ref. [5].

We use the technique of Hatta, et al, [4, 36] to measure the needle-axis

(“longitudinal”) thermal diffusivities of bulk (e.g. crystalline) organic semiconductors. A

schematic of the apparatus is shown in Figure 1.4; chopped light illuminates the front of

the sample uniformly, which is partly blocked by a movable screen, and the temperature

oscillations on the back surface measured with a thermocouple which is glued to the

sample with silver paint.

Figure 1.4 Schematic diagram for the longitudinal thermal diffusivity measurements.

Measurements are made at frequencies 1/τ1 < ω < 1/τ2, effective. A preliminary

transverse Tac measurement will be performed to determine τ2, effective [3]. In this limit, the

position dependence of the oscillating temperature is given by [36]

2/1)2/(/)(ln longDdxxTd ωω −= (1.6)

Vac x

Sample Thermocouple

Light

Movable screen

9

where Dlong = κlong/cρ and x = the distance between the thermocouple and edge of the

screen. For these measurements, the frequency is fixed so that the thermal response of the

thermometer does not affect results.

1.3 3ω-Technique

The “3ω” technique is a widely used method for measuring the thermal resistance of

thin films [40]. This technique for thermal conductivity measurements uses a single metal

strip (length = L and width = W = 2b) deposited on the film or base substrate to act as

both a resistive heater and thermometer (Figure 1.5(a)) [41]. An ac driving current (at

frequency ω) is applied to the heater, so that its temperature (and the temperature of the

nearby substrate) will oscillate at frequency 2ω. Consequently, its electrical resistance

will oscillate at 2ω, producing a third harmonic (V3ω) in the voltage drop across the metal

strip, with V3ω∝ΔT, the magnitude of its (in-phase) temperature oscillation (see Eqn. 1.6

below). We use a sensitive bridge circuit and a lock-in amplifier (with differential input)

to remove the drive voltage (including its third harmonic distortion), so that the measured

3ω signal can be used to infer the magnitude of the temperature oscillations. For

measurements on the base substrate, the thermal response is given by: [40,41]

].79.2)W/ln()ln()[2/(T 2 ++−=∆ subsubsub cLP κωπκ (1.5)

Here P is the applied power and csub and κsub are the specific heat per unit volume and

thermal conductivity of the substrate. ΔT was determined from V3ω using [41]

,/2T 03 VV αω=∆ (1.6)

where V0 is the voltage across the heater (measured in a 4-probe configuration) and α ≡

(1/R)dR/dT the temperature coefficient of resistance for the copper strip. We found α by

measuring the dc resistance of the heater as its temperature was slowly heated and cooled

(~30 oC) in the same vacuum cryostat in which the thermal conductivity is measured, as

discussed later in Section 2.3

Lee and Cahill [40] have shown that the 3ω-technique could be used to measure the

transverse thermal conductivity of a thin film of thickness t, deposited between the

10

substrate and heater (Figure 1.5(b)), once the substrate thermal conductivity was known;

the (transverse) film thermal resistance just added a frequency independent offset to the

ln(ω) dependence of the substrate:

film

app

PRtWLPLPtsubstrateTsubstratefilmTfilm

=+=

=∆−+∆≡∆

)/)(/()W/()()()(T

intρκ

κ (1.7)

where the effective area in Eqn. 1.7, S = WL, and κapp is the apparent thermal

conductivity of the film. The basic assumption is that the heat flow in the film is

one-dimensional, i.e. W >> t, while, as for bulk measurements, W is much smaller than

the thermal diffusion length in the substrate [40, 41]. In addition, one needs t << the

thermal wavelength in the film [40] and κapp < κsub. [42]

11

Figure 1.5 3ω-technique: measurements of (a) yield the thermal conductivity of the substrate; offset of (b) compared to (a) yields the thermal resistance of the organic thin films

Thin films of organic semiconductor

d

(a)

(b)

12

Chapter 2 Experimental Setup and Sample Preparation

2.1 In-Plane Thermal Conductivity Measurements

In our experiments, the visible/NIR light source we used is a quartz-halogen lamp.

The light was chopped by a voltage controlled chopper and then sent into the vacuum

chamber through a 1.6 mm diameter fiber optic cable. A silvered glass tube and a

cylindrical convex lens was used to converge and cast light uniformly upon the sample. A

25 μm diameter chromel-constantan thermocouple was prepared by spot-welding it in a

cross and glued to the back of the sample with a thin layer (< 10 μm) of butyl acetate

based silver paint, as shown in Figure 2.1(a). The vacuum chamber (Figure 2.1) was built

with the help of the machine shop at University of Kentucky based on my design. In our

setup, the screen is attached to a micrometer (precision ±3 μm) which measures the

distance x0-x, where the offset x0 depends on the relative positions of the sample and

screen. Frequency dependent (i.e. transverse) measurements were preformed first to

determine τ2, effective, i.e. including the response of the thermometer, while the screen was

fully opened. Then the position dependent (i.e. longitudinal) measurements were made at

a frequency 1/τ1 < ω < 1/τ2, effective. For one position dependent measurement, the

frequency is fixed so that the thermal response of the thermometer does not affect results.

To be sure that we are in the correct frequency limit and the edge of the screen is not too

close to the thermocouple (i.e. overlapping with the silver paint spot), we check that we

obtain the same values for the “frequency normalized” slopes, f-1/2dlnTac/dx at a few

different frequencies, where f ≡ ω/2π.

13

Figure 2.1 Inside (a) and outside (b) pictures of the vacuum chamber built at University of Kentucky for ac calorimetry measurements.

Fiber optic cable

Vacuum valve

Micrometer Electrical contacts

Sample supported by thermocouple wires

Screen

(a)

(b)

14

2.2 3ω signal measurements

Figure 2.2 shows the circuit diagram used to measure the 3ω component of the

voltage drop across the metal strip. We use a Hewlett Packard 33120A function generator

to produce a sine wave as the ac signal with a harmonic distortion less than 0.1% of the

fundamental. We use a sensitive bridge circuit and a Stanford Research SR830 DSP

lock-in amplifier (with differential input) to remove the in-phase drive voltage (including

its third harmonic distortion). The gain of the multiplying DAC can be varied by three

tuning pots manually until the residual of ω signal read from lock-in is below a hundred

thousandth of the driving signal. The reference channel of the lock-in is connected to the

voltage output of the function generator directly instead of the SYNC output to avoid

introducing any phase shift. A frequency tripler mentioned in Ref. 41 is not necessary in

our experiment because our lock-in amplifier can measure the third harmonic signal by

itself.

Figure 2.2 Circuit diagram for 3ω measurements.

REF IN Frequency Synthesizer

Multiplying DAC

Lock-In Amplifier

Null Setting

Diff Amp

Diff Amp

Series Resistor

Heater/ Thermometer

Null Box

A

B

15

The null bridge circuit (Figure 2.3) updates the MDAC at a 4 Hz rate, reading the

three tuning pot setting at each update. Each pot is set to yield 1 part in 64 resolution

(about 5 degrees of a turn per step) for a stable final state. The medium tuning pot spans a

range of about ±5 degrees (out of 360 degrees) of motion of the coarse tuning pot.

Similarly, the full range of the fine pot is ±5 degrees of the medium pot. [43]

Figure 2.3 Null bridge schematic with three tuning pot setting designed by Greg Porter at University of Kentucky.

The lock-in amplifier and function generator are connected to the computer through a

GPIB (general purpose interface bus) to USB (universal serial bus) interface. The

program to collect data are written in QuickBasic. The frequency dependence of the 3ω

signal is measured using an entirely automated program where the function generator

frequencies are preset and embedded in the program.

An example of a 3ω-measurements for a vapor annealed TES ADT film with preset

frequencies from 22 Hz to 803 Hz is shown in Figure 2.4.

16

Figure 2.4 3ω signal measured at several preset frequencies and converted to ΔT/P for a vapor annealed TES ADT film on a sapphire substrate. The line shows the response for the bare sapphire substrate.

2.3 Temperature coefficient of resistance measurements

The temperature coefficient of the resistance, α = (1/R) dR/dT is measured for each

copper strip before and after 3ω measurements to correct for changes of its resistance due

to oxidation. The circuit diagram in Figure 2.5 is used to take the measurements. A 100

μA DC current is applied to the copper strip using a Keithley 224 Programmable Current

Source. The copper strip is heated up to 330 K from room temperature in 2 hours using

the SWEEP mode of an OXFORD ITC601 Intelligent Temperature Controller connected

to the heater and a platinum resistance thermometer in an OXFORD MicrostatN cryostat

and cooled down to room temperature naturally within another 4-5 hours. The voltage

across the copper strip and the voltage across the diode thermometer inside the

temperature controller are measured using two Hewlett Packard multimeters as V1 and V2.

Both multimeters are connected with the computer through a GPIB to USB interface. V2

is converted to the temperature of the copper strip after calibration and the

Vapor annealed TES ADT films

Calculated sapphire response

ln(F(Hz))3 4 5 6 7

T/

P (K

/W)

2

3

4

5

6Vapor annealed TES ADT filmsSapphire fit

17

temperature-resistance data is collected at an adjustable time interval. Example

temperature resistance data are shown in Figure 2.7 for a copper strip on a dif-TES-ADT

film.

Figure 2.5 Circuit diagram for temperature coefficient of resistance measurements.

Figure 2.6 Inside picture of the vacuum chamber (an OXFOR MicrostatN liquid nitrogen cryostat) for 3ω measurements.

Electrical contacts with nulling box

Heater of temperature controller underneath

Electrical contacts with temperature controller

Substrate glued down with silver paint on sample holder

V1

Sample Header

Temperature Controller V2

Computer

Current Source

18

In the temperature coefficient of resistance measurements, it is often the case that the

resistance of the copper strip decreases a few percent as it is being heated up. These

effects are often attributed to oxidation and annealing processes of the copper film. We

therefore study the temperature coefficient from the change of resistance while the copper

strip is cooling down. From linear regression of resistance-temperature data in Figure 2.7,

α of the copper film is (7.231±0.0004)×10-4 K-1and (7.233±0.003)×10-4 K-1 at 300K

before and after the 3ω measurements, respectively. The preparation of the copper strip

will be discussed in Section 2.5.3.

Temperature (K)290 300 310 320 330 340

Res

ista

nce

of C

oppe

r stri

p (

)

950

955

960

965

970

975

980

985

Before 3measurementsAfter 3 measurements

Figure 2.7 Resistance-temperature characterization of a copper strip deposited on dif-TES-ADT thin films before (red) and after (blue) 3ω measurements.

2.4 Optical measurements

The Thermo Fisher Scientific infrared microscope has facilitated our measurements

to a great extent. It has two photoconductive mercury-cadmium-telluride (MCT)

detectors with a high sensitivity which operate at liquid nitrogen temperature (77K).

Detector 1 covers the spectral range from 400 cm-1 to 1300 cm-1 with a peak sensitivity of

19

d* = 6.6 × 109 cm√Hz/W at 465 cm-1. Detector 2 covers the spectral range from 600 cm-1

to 4000 cm-1 with a peak sensitivity of d* = 4.4×1010 cm√Hz/W at 740 cm-1 and has better

signal-to-noise. The microscope has two modes of operations: one for reflection

measurements and the other for transmission measurements. It has a measurement area

between 10μm × 10μm and 200μm × 200μm. A diagram of operation in the two different

modes is shown in Figure 2.8.

Figure 2.8 Two different modes of operation of the IR microscope: the reflection mode and the transmission mode.

The Fourier-transform infrared spectrometer used in our experiments is the Thermo

Fisher Scientific Nicolet 6700 which is compatible with our infrared microscope. We

take measurements with the microscope sample stage and detectors instead of the internal

sample holder and detector of the FTIR due to sensitivity and ability to select area of

illumination. The broad band light from the FTIR is sent to the microscope through a

KRS-5 window on the side of the spectrometer. The beam is then transmitted through the

sample and into the microscope detector. OMNIC software that came with the FTIR then

Fourier transforms the resulting interferogram into a spectrum. OMNIC also allows

normalization of the spectra measurements to the background. Background spectra were

either transmission through an empty hole on the microscope stage or reflection on a gold

20

film. An example background single scan of transmission through a hole is shown in

Figure 2.9.

(cm-1)

1000 1500 2000 2500 3000 3500 4000

Inte

nsity

(arb

itrar

y un

its)

0

1

2

3

CO2

H2O H2O

Figure 2.9 Single scan spectrum of transmission through a empty hole on microscope stage, the overall shape of the spectrum is due to the sensitivity of the detector, transmission and reflective properties of the mirrors, and source spectrum. Common features around 3500 and 1630 cm-1 are due to atmospheric water vapor, and the bands at 2350 and 667 are due to atmospheric carbon dioxide.

2.5 Sample Fabrication

2.5.1 Substrate Surface Cleaning

In order to ensure the uniformity of the thin films and minimize the interface thermal

resistance, a clean surface free of particulate matter and solvents is required before

deposition of thin films on our substrates (glass, sapphire and thermally oxidized silicon).

The substrates were first immersed in an acetone bath and sonicated at room temperature

for 5 minutes. To remove the residue of acetone, a similar treatment was applied to the

substrate with isopropanol or mixed alcohols followed by deionised water. Finally, the

substrate surface was blown with compressed nitrogen until dry. In some cases, the

surface was then treated with UV-ozone as a final step, but we found that the latter had

insignificant effect on the interface thermal resistances for our samples.

21

2.5.2 Thin Films Deposition

Thin films of organic semiconductors can be applied by different methods such as

drop casting, spin coating and thermal evaporation on different substrates. Smooth and

uniform thin films are necessary to perform 3-ω measurements. Several films of different

thicknesses for each material should be measured to separate the effects of intrinsic

thermal conductivity from interface thermal resistance.

Drop Casting

We have prepared TIPS-pn and CP-DIPS-pn films with drop casting followed by a

slow or fast dry. For both samples, a 2-10% by weight solution in organic solvent

(toluene, chlorobenzene or mesitylene) is applied with a dropper and dried afterwards by

different methods. For slow dry, the sample is dried under a cover which will allow extra

time for large area crystals to grow. Unfortunately, the geometry of the needle-shape

crystalline films formed by this way is not suitable for application of the heater (Figure

2.10a). On the other hand, when the sample is dried too quickly with no cover,

amorphous or polymorphous films with large holes form as shown in Figure 2.10b.

(a) (b)

Figure 2.10 Drop cast films from 10% TIPS-pn Chlorobenzene solution by slow dry (a) and 3% TIPS-pn Chlorobenzene solution by fast dry (b).

Another drop casting process was tried with a specially made Teflon container as

shown in Figure 2.11. The solution is dropped through a narrow slit opened on top of the

container and fills it. This will allow extra amount of solution to dry out on the substrate.

22

The films coming out of this method do become thicker but the uniformity is still not

improved.

Figure 2.11 Drop casting with a Teflon container

Spin Coating

TES-ADT crystallizes very slowly from solution and non-crystalline films with

uniform thicknesses form from spin-cast solutions [31, 32, 33]. To prepare these,

TES-ADT was dissolved in toluene to form a 2 wt% solution, which was spin-coat on the

substrate at 1000 rpm. Then the sample was heated in air at 80 oC to remove residual

solvent, yielding a ~ uniform thickness film. As described in Ref. 29, molecules in these

films are oriented with their c-axes approximately normal to the film but are mostly

disoriented in the ab-plane, although small monoclinic crystallites may be present.

It was reported that a vapor annealing process is able to induce structural

rearrangement and crystallization of TES-ADT films [31]. To fabricate the crystallized

TES-ADT films, the spin-cast sample was then attached to the dish lid and placed over a

pool of dichloroethane (DCE) solvent for 1-10 minutes. Exposing these films to DCE

vapors promotes spherulite growth into triclinic crystallites, typically ~ 1mm in size [31,

32, 33] as in Figure 2.12, while the thickness stays uniform. The films made by this

process are typically 100 nm thick, as measured with an AFM (after the thermal

resistance measurement).

Teflon Container

Dropper

Substrate

23

Figure 2.12 A spin-coated TES-ADT thin films on silicon wafer before (a) and after (b) vapor annealing under microscope. (c) Diagram of vapor annealing process.

Thermal Evaporation

Unfortunately, much thicker uniform TES-ADT films cannot be grown from solution

and TES-ADT degrades when evaporated. On the other hand, while diF-TES-ADT and

TIPS-pn precipitate quickly from solution to form non-uniform, granular films [17, 35,

44], uniform and thicker films [34] can be prepared by vacuum evaporation [34],

although the uv-visible absorption spectra of the TIPS-pn films shown in Figure 2.13

indicates that there may be small amounts of degradation products present. The

thicknesses of the sublimed films (100 nm to 4 µm), were determined approximately

during sublimation by a quartz crystal thickness monitor and more accurately (± 10 nm)

with a Dektak 6M profilometer after the thermal measurement. (The two thickness

measurements typically agreed within 10%.) We assume that the molecules in the

sublimed films have their sidegroups mostly aligned transversely, although the films are

presumably microcrystalline and/or disordered in the ab-plane, as discussed for

(b) (a)

(c)

24

evaporated TIPS-pn films in Ref. 10. Similarly, spin-cast diF-TES-ADT films on

thermally oxidized silicon were observed to be microcrystalline but with the c-axes ~

80% aligned perpendicular to the substrate [35].

Figure 2.13 Uv-visible absorption spectrum of the sublimed TIPS-pn films (dash red), TIPS-pn solution (green), sublimed TES-ADT films (dash blue), and TES-ADT solution (black).

Supporting our assumption of ab-plane disorder is that no grains were visibly present

in these films and infrared transmission spectra for small areas [~ (50 µm)2] of

dif-TES-ADT films evaporated on (infrared transparent) KRS-5 substrates measured with

an infrared microscope showed no polarization dependence as shown in Figure 2.14. (In

fact, the diF-TES-ADT absorption spectra were identical to that of Figure 4a in Reference

15 for diF-TES-ADT dispersed in KBr.)

Abs

orpt

ion

25

Wavenumber (cm-1)800 1000 1200 1400 1600

Tran

smis

sion

T=10%

E//0o

E//45o 10% shiftedE//90o 20% shifted

Figure 2.14 Polarized infrared transmission spectra of small areas of dif-TES-ADT films evaporated on a KRS5 substrate, taken in (50 µm)2 spots.

2.5.3 Heater Deposition

The shadow masks (Figure 2.15) used to deposit heaters on organic films are made from

300 μm thick silicon wafers at University of Louisville by a deep reactive-ion etching

process. Copper strips (~50 nm thick, L = 6 mm long, w = 50 μm wide, resistances ~ few

hundred ohms) as heaters/thermometers were evaporated through the shadow mask on the

samples (bare substrates or thin films). The metal strip requires a fairly large temperature

coefficient of resistance to generate a measurable resistance change as a function of

temperature. We chose copper for its large temperature coefficient of resistance, ease of

evaporation, and low cost. (While the resistance of these films changed slowly and slightly

due to oxidation, we corrected for these changes by measuring the temperature dependence

of the resistance before and after thermal conductivity measurements, as discussed in Section

2.3)

26

Figure 2.15 Shadow mask samples with two different four probes feature made at University of Louisville by a deep reactive-ion etching process.

Two checks were made to insure that evaporation of the copper heater/thermometer

film did not damage or otherwise affect the organic film. 1) We exposed organic films to

heated tungsten evaporation sources for prolonged periods and checked their thicknesses

with the AFM before and after; thicknesses changed less than 10 nm, indicating that

exposure to the hot source did not cause significant evaporation of the organic film. 2)

Polarized infrared transmission spectra (Figure 2.16) were taken on vapor annealed

TES-ADT films which were deposited on KRS5 substrates (KRS5 was used because it is

transparent in the infrared), as functions of distance from the copper heater using an IR

microscope. The spectra were taken in 100 μm square areas adjacent to the copper line

and 1 mm and 2 mm away, but all on the same crystallite. (Unlike single crystals in

which molecules are uniquely oriented, spherulites comprise crystallites having changing

orientations in order to maintain a radially-symmetric, space-filling growth habit [33]).

Absorption lines at ~728, 857 and 878 cm-1 are strongly polarization dependent. The

spectra were independent of distance, indicating that evaporation through the shadow

mask did not cause degradation or melting (and realignment) of the spherulite-crystallites

in the TES-ADT.

8 mm 100 μm 6 mm 50 μm

27

Figure 2.16 Polarized infrared transmission spectra of a single spherulite in a spin-cast, vapor annealed TES-ADT film on a KRS5 substrate, taken in (100 µm)2 spots adjacent to the copper heater line and 1 mm and 2 mm away from the line. (x and y refer to perpendicular microscope axes.) The curves for each position are vertically offset.

28

Chapter 3 In-Plane Thermal Conductivity Measurements of Organic

Semiconductors

3.1 NFC-PEDOT Paper

In addition to my work on small molecule organic semiconductors, I also measured

the longitudinal thermal conductivity of two bulk polymeric samples provided by Xavier

Crispin on Linköping University. The first was the copolymer poly(3,4-ethylene-

dioxythiophene):poly(styrene-sulfonate) (PEDOT:PSS) coating cellulose nanofibrils.

The material was denoted NFC-PEDOT and has a paper morphology; for example, it can

be folded into origami structures without affecting its electronic properties.

PEDOT:PSS itself has been promoted as a possible thermoelectric material, motivating

our thermal conductivity measurements, but NFC-PEDOT turned out not to have a very

large Seebeck coefficient. Because of its processing and structure, however, it turned

out to be a mixed ionic-electronic conductor (MIEC), attractive for possible

electrochemical applications.

The current civilizational challenges (and opportunities) related to the conversion

and storage of energy urge for the development of eco-friendly bulk MIEC systems.

These MIECs would preferably be based on sustainable, light-weight and abundant

materials that can easily be processed into large (even giant) volumes. Such a “green”

MIEC would enable the mass adoption of fuel cells, batteries and supercapacitors.

Furthermore, this development may also help organic electronics venture into the domain

of high power electronics.

The state-of-the-art in electronic, ionic and mixed conductors is summarized in

Figure 3.1. Besides the standard electronic and ionic conductors, MIECs belong to two

distinct families: ceramics and conducting polymers. There exists a clear trade-off

between the ionic and electronic conductivities, with an unoccupied niche in the upper

right corner of the graph. The latter corner is particularly interesting for electrochemical

conversion and storage of charges, as well as for power electronics. Ceramic materials

with high ionic conductivity (points l in Figure 3.1) [45] have been reported but are far

from reaching the electronic conductivities of the best organic conducting polymers

29

(point “o”) [46, 47], although the ionic conductivity of the latter is two orders of

magnitude lower. The low temperature processability (relative to ceramics) and the ease

with which their wet synthesis may be scaled up make conducting polymers attractive for

mass production and implementation into giant scales [48]. The development of organic

conducting polymers, such as trans-polyacetylene [49], was mainly focused on reaching

high and air-stable electronic conductivity [50, 51] in more or less bulky samples [52].

High electronic conductivity (1000 to 4000 S/cm [53, 54]) has been achieved in organic

polymer thin films (10nm to 10 μm) often with metallic charge transport characteristics.

But there have been no reports of high conductivity for thicker films and bulky

geometries (10 μm to 10 cm). Indeed, most of the applications are currently focused on

ultra-thin transparent electrodes for the replacement of expensive transparent metal oxide

electrodes in solar cells and light-emitting diodes.

In parallel to these developments, (semi)conducting polymers have been investigated

for their reversible electrochemical activity due to the fact that they are intrinsic MIECs.

Hence, (semi)conducting polymers have been explored in electrochromic displays,

electrochemical transistors, diodes and sensors [55], as well as in supercapacitor, batteries

[56] and fuel cells [57]. But there exists a lack of mixed ionic-electronic large-bulk

electrodes combining record-high electronic with record-high ionic conductivity values.

One strategy to improve the ionic conductivity and the aqueous processability has been to

composite a polyelectrolyte with a conjugated polymer [58]. PEDOT:PSS is the most

studied and used conducting polymer (point “m”) [59]. In those blends, the electronic

conductivity is strongly correlated with the phase separation. The latter can be controlled

and suppressed with the addition of high boiling point solvents. This effect is called

“secondary doping” to distinguish it from the “primary doping” which controls the

charge carrier density in the PEDOT phase. Importantly, despite the addition of PSSH in

those blends, the ionic conductivity is limited to 1.5 mS/cm at 80%RH [60]; which is still

well below the values reported for ceramics (Figure 3.1 [7]).

Our samples comprise of PEDOT:PSS, nanofibrillated cellulose (NFC), glycerol and

dimethylsulfoxide (DMSO) blended from solutions. This

NFC-PEDOT:PSS-glycerol-DMSO composite (NFC-PEDOT paper), combines high

30

electronic and ionic conductivity with facile manufacturing, mechanical properties

compatible with paper-maker machine and ease-of-handling. In NFC-PEDOT paper, PSS

acts as a polyanionic counterion to neutralize the highly oxidized PEDOT chains [76].

DMSO is a high-boiling point solvent and is commonly used as a conductivity

enhancement agent for PEDOT:PSS [61]. This secondary dopant impacts the nano- and

micro-morphology of PEDOT:PSS by improving crystallinity (π-stacking) and promote

phase separation of the PSS in excess to promote the creation of highly conductive

percolation paths. NFC is a nanofiber scaffold system with “remarkably high toughness

with a large strain-to-failure” [77]. NFC is here utilized as a 3D-scaffold to improve the

PEDOT:PSS’s micro/mesoscopic organization in 3D (i.e. it acts as a tertiary dopant to

favor high conductivity also in bulky dimensions). Adding glycerol improves the

composite’s plasticity while increasing its hygroscopicity, which allows ions to move

much easier. As shown in Figure 3.1, this material has a higher ionic conductivity than

simple polymers and higher electronic conductivity than ceramics, so may form an

appropriate base for applications such as supercapacitors.

31

Figure 3.1 Survey ionic and/or electronic conductors. With the exception of ionic liquids, only solid conductors are included. The points in the graph represents the following materials: a: Nafion [62]; b: poly(diallyldimethyl ammonium chloride)/poly(2,6-dimethyl1,4-phenylene oxide) [62]; c: poly(4-styrenesulfonic acid) (105); d: poly(ethylene oxide)/poly(acrylic) acid/poly(ethylene oxide)/(poly(acrylic) acid/multi walled carbon nanotubes) [63]; e: polyvinylidene fluoride/polyethylene oxid/propylene carbonate/ LiClO4 [64]; f: (lithium bis(oxlate)borate and lithium tetrafluoroborate)/1-Ethyl-3-methyl-imidazolium tetrafluoroborate [65]; g: LiCF3SO3/poly(methyl methacrylate), LiClO4/poly(methyl methacrylate) and LiClO4/propylene carbonate/ethylene carbonate/ dimethylformamide/poly(acrylonitrile) [66]; h: Li10GeP22S12 [67]; i: Ag2HfS3 [68]; j: Ag2S [69]; k: Li3.5V0.5G0.5O4 [70]; l: Ce0.8Gd0.2O2-d-CoFe2O4 [45]; m: poly(3,4-ethylenedioxythiophene):polystyrene sulfonate and : poly(3,4-ethylenedioxythiophene):polystyrene sulfonate/sodium polystyrene sulfonate [59]; n: poly-[1-methyl-3-(pyrrol-l-ylmethyl)pyridinium perchlorate] [71]; o: Polyaniline [46, 47]; p: Polypyrrole [72, 73]; q: poly(3,4-ethylenedioxythiophene):polystyrene sulfonate/nanofibrillated cellulose/dimethyl sulfoxide/polyethylene glycol (this work); r/s: GaAs [74]; t: Nichrome [75]; u: Ag [75].

Because an initial interest in this material was actually for possible thermoelectric

applications, our group used both the position dependent [3] and photothermal [5]

techniques to measure its in-plane and transverse thermal conductivities. We found the

in-plane thermal diffusivity, Dlong = (7.0 ± 0.2) x 10-3 cm2/s for several sheets of paper of

32

different thicknesses. Experimental results for a 30 μm thick sample are shown in Figure

3.2. For small values of x, the signal saturates as the edge of the screen overlaps the glue

attaching the thermocouple, while for large x the signal becomes comparable to the noise

and offset voltage of the measuring electronics. The value of D is determined from the

measured slope at intermediate values of x: f-1/2 dlnVac/dx = (2.12 ± 0.03) /mm·Hz1/2.

From D ≡ κ/cρ, where κ is the thermal conductivity and ρ = (1.26 ± 0.4) g/cm3 is the

mass density, we found κlong = (11.6 ± 1.0) mW/cm·K. The specific heat, c = 1.32 J/g⋅K

with precision 3%, was measured using differential scanning calorimetry on a pellet

prepared from pressing several sheets of paper together, with results shown in the inset to

Figure 3.2. Note the transition at T ~ -50 oC presumably due to freezing of the glycerol in

the matrix. The values of c, κ, and ρ are very similar to other coated-NFC materials, as

expected

Later, measurements by our group of the transverse diffusivity using the

photothermal technique indicate that it is a few times smaller than the in-plane value,

which is consistent with our assumptions that the cellulose fibers largely aligning in the

plane makes the in-plane value of Dlong larger than the value in the transverse direction.

33

Figure 3.2 Position dependence of the oscillating thermocouple signal (Vac) for two different chopping frequencies for a 30 µm thick sample; x = distance between the edge of the screen and the thermocouple, and x0 = constant offset. Top inset: specific heat of a 14 mg pellet measured with differential scanning calorimetry at a scanning rate of 18 K/minute.

3.2 FS-PEDOT:PSS

The second polymeric material on which I worked was very thin (~ 10 µm) papers of

freed-standing PEDOT:PSS.

To enhance ZT of thermoelectric materials, many effective methods were employed

for improving the electrical conductivity σ of PEDOT thin films (~100 nm) [78-81]. For

example, high σ over 3000 S/cm has been achieved by sulfuric acid post-treatment on the

film [82, 83]. Another effective way to achieve high thermoelectric performance is to

increase the Seebeck coefficient S. For this goal, n-type doping strategy has been

developed by Crispin and co-authors to tune the oxidation level and charge carrier

concentration [84-86]. A record-high power factor of 324 μW/m·K2 was achieved from

PEDOT-Tos materials at the oxidation level of 22% due to its very high S of 780 μV/K.

x0 - x (mm)8.5 9.0 9.5 10.0 10.5 11.0 11.5

ln[V

ac(n

V)] /

f1/2

(H

z-1/2)

0

1

2

3

4

5

6

7

T (oC)-80 -60 -40 -20 0 20

c (J

/gK)

0.8

1.0

1.2

1.4

2.6 Hz

4.0 Hz

34

Recently, the highest ZT of 0.42 was reported based on PEDOT:PSS by K. P. Pipe et al.

[87]. These works strongly stimulate the intensive research on the development of

conducting polymer based thermoelectric devices.

Despite the encouraging thermoelectric effect achieved in PEDOT based materials,

thin films of this materials supported by a substrate are not desirable for a thermoelectric

generator (TEG) since there is large heat leakage through the substrate [7, 88]. Moreover,

to achieve considerable power output, from the perspective of practical TEGs, a large

number of organic thermoelectric films need to be connected in series. Employing thick

films could be more stable and achieve higher power than thin films because they involve

less supporting substrates and interlayer space. Therefore, thick free-standing

PEDOT:PSS films with high electrical conductivity σ are desired for achieving high power

density TEGs. However, it is very challenging to fabricate a free standing PEDOT:PSS

film via normal film-fabrication methods (such as spin-coating, printing) from a

commercialized PEDOT:PSS solution due to its low concentration about 1.3% [89].

Lately, some strategies were proposed for preparing free standing PEDOT:PSS films, but

it was hard to realize high thermoelectric performance because of their moderate values

of σ [89-91]. A high power density TEG is realized by the Crispin group at Linköping

University recently with a highly conductive free-standing PEDOT:PSS (FS-PEDOT:PSS)

film, which possesses a Seebeck coefficient S of 20.6 µV/K and a high power factor of 107

μW/m·K2 at room temperature [92]. Under a heat gradient of 29 K, the FS-PEDOT:PSS

TEG exhibits a power density of 99 μW/cm2, which is the highest reported value for

PEDOT based TEGs. As discussed below, we found that the films had moderate values

of thermal conductivity of 0.64 W/m·K in plane and 0.27 W/m·K through plane. The

striking performance of the TEG is attributed to its high σ (2500 S/cm) and the free

standing property of the FS-PEDOT:PSS film. In addition, in-situ temperature-dependent

x-ray diffraction (XRD) measurement proved good thermal stability of the

FS-PEDOT:PSS since there were neither crystalline nor chemical structure changes even

up to 200 oC. The high power density free-standing flexible TEG indicates its great

potential for low power consumption wearable electronics and biomedical implants in the

future.

35

The longitudinal in-plane thermal diffusivity Dlong of the FS-PEDOT:PSS was

measured using a ~1 cm long strip of d = 10 μm sample. The sample was suspended, in

vacuum, on 12 μm chromel-constantan thermocouple wires which were glued to the

bottom center of the sample. Results at three frequencies are shown in Figure 3.3. The

average slope is (2.8 ± 0.3) /mm, giving Dlong = (0.40 ± 0.08) mm2/s. Assuming values of

ρ = 1.37 g/cm3 and c = 1.17 J/g⋅K from Ref. 59, we find κlong = (0.64 ± 0.13) W/m·K. The

demonstrated κlong of FS-PEDOT:PSS is lower than ethylene glycol (EG) (κlong =

0.94W/m·K) and dimethyl sulfoxide (DMSO) (κlong = 1.0 W/m·K) treated-PEDOT:PSS

[93, 94]. The reduced thermal conductivity of FS-PEDOT:PSS may be caused by its

porosity [95, 96, 97]. The value of κlong gives a thermoelectric coefficient of performance

ZT ~ 0.05.

From photothermal measurements, our group also found κtrans = (0.27 ± 0.06) W/m·K

[96]. The thermal transport anisotropy is κlong/κtrans = 2.37, which is lower than previously

reported (3.33 for DMSO treatment and 5.6 for EG treatment) [93, 98].

36

Figure 3.3 Experimental data used to determine the in-plane diffusivity: spatial dependence of the thermocouple signal at a few frequencies. Dlong is determined from the slopes between the dotted lines, as discussed in the text. A solid line with slope = 2.8/mm is shown for reference.

3.3 TIPS-pn Functionalized Pentacene Semiconductors

In this Section, I discuss measurements of the room temperature thermal

conductivities of TIPS-pn and several other pentacene based materials with related

structures shown in Figure 1.1. Because these crystals are undoped, the electronic density

is negligible and κ is overwhelmingly due to phonons [15]. All of these materials form

layered crystals, with the pentacene (or substituted pentacene) backbones lying

approximately in the ab-plane (where one hopes for good π-orbital overlap between

molecules [99]), and the silyl or germyl sidegroups extending between layers [15, 24, 26,

27, 44, 100, 101] The interlayer (c-axis) and in-plane phonon thermal conductivities are

therefore expected to be different and we measured them separately. Because the crystals

are small (in-plane dimensions typically 0.5-10 mm and interlayer direction less than 0.6

mm), we used ac-calorimetry techniques [3,36, 37] discussed above.

x0 - x (mm)4.4 4.6 4.8 5.0 5.2 5.4 5.6 5.8

ln[V

(nV)

) / F

1/2

(H

z-1/2)

0

1

2

3

4

5

6

noise saturation

2 Hz

3.5 Hz

5 Hz

37

3.3.1 In-Plane Measurements

We used the technique of Hatta, et al, [4, 36] to measure the needle-axis thermal

diffusivities of TIPS-pn. Because we assume that both the illuminated and blocked

portions of the sample are much greater than the longitudinal diffusion length, (Dlong/ω)1/2,

which is typically ~0.5 mm for our samples and frequencies, data were taken on needle

shaped samples ~1 cm long. In practice, this technique works well for needles with

thickness d < 200 μm, lengths ≈ 1 cm, and Dneedle between 0.3 and 10 mm2/s.

Results for an 8 mm long, d ≈ 100 μm crystal of TIPS-pn are shown in Figure 3.4.

We plot f-1/2lnVac as a function of position behind a screen for a TIPS-pn crystal along its

[2,1,0] needle axis at a few frequencies. From Eqn. 1.4, the slope is inversely

proportional to Dlong1/2. For large x0-x, the edge of the screen overlaps the silver paint

holding the thermocouple, and the signal begins to saturate, whereas for small x0-x, the

edge of the screen is far from the thermocouple, so Vac is small and approaches the

thermocouple offset voltage and noise level. For intermediate distances, the same slope,

f-1/2dlnVac/dx, is measured for different frequencies, yielding Dlong = (1.0±0.1) mm2/s.

(Similar values were found for other crystals.) The specific heat of TIPS-pn, measured on

a pellet with differential scanning calorimetry with a precision ~3% [102], is shown in

the Figure 3.4 inset. Using c = 1.48 J/g·K and density ρ = 1.1 g/cm3 [103], we find κlong =

(16±2) mW/cm·K.

38

Figure 3.4 Results for an 8 mm long TIPS-pn needle. The dotted lines show the region of the expected linear variation of f-1/2ln(Vac) from which the diffusivity is calculated. The inset shows the specific heat of a pellet measured by DSC.

This value of κlong is several times larger than the room temperature value for rubrene

[2], but comparable to the phonon thermal conductivity of quasi-one dimensional organic

conductors containing segregated molecular stacks with excellent π-orbital overlap

[104-106].

3.3.2 Interlayer Measurements

Figure 3.5 shows the frequency dependence of fVω of a d=610 μm TIPS-pn crystal,

with τmeas = 2.8 ms measured with a thermocouple thermometer. In Ref. 3, we showed

results for a d = 335 μm crystal, with τmeas = 1.3 ms; much thinner crystals, e.g. d ≈ 100

μm, had similar time constants. These values of τmeas and their lack of correlation with

thickness imply that we are limited by the thermal interface resistance of the silver paint,

and from the 610 μm crystal (the thickest available), we deduce Dc > 14 mm2/s and κc >

225 mW/cm·K. We can avoid this problem by measuring the oscillating thermal radiation,

with a MCT detector, from the backsurface of the sample to find the frequency

dependence of Tac as discussed in Section 1.2. From the photothermal measurements, we

39

found the interlayer thermal diffusivity: Dc = (13 ± 6) mm2/s; using[4, 68] c = 1.48 J/g⋅K

and ρ = 1.1 g/cm3, this corresponds to κc = (210 ± 100) mW/cm⋅K, similar to the lower

limit concluded from Tac measurements.

f (Hz)10 100

f Vω

(µ V

Hz)

0

1

2

3

4

(a)

(b)

(d)(c)

T (K)

240 280

c (

J/gK

)

1.25

1.50

Figure 3.5 Frequency dependence of fVω for representative crystals; the curves show fits to Eqn. 1.4. (a) TIPS-Pn (signal × 3), d = 610 μm, τmeas = 2.8 ms → Dc > 14 mm2/s; (b) F2-TIPS-pn, d ≈ 300 μm, τmeas = 1.26 ms → Dc > 7 mm2/s; (c) TIPGe-Pn (signal × 2), d = 460 μm, τmeas = 1.03 ms → Dc > 20 mm2/s; (d) EtTP-5, d ≈ 300 µm, τmeas = 0.68 ms → Dc > 14 mm2/s. Inset: Specific heat of TIPS-Pn.

3.3.3 Discussion

This interlayer thermal conductivity is much larger than typically found in a

van-der-Waals bonded molecular crystal (e.g. for rubrene we measured Dc ~ 0.05 mm2/s,

giving κc ~0.7 mW/cm·K [3], while for pentacene κc = 5.1 mW/cm·K [8]), and is

comparable to the values for materials with extended bonding (e.g. Al2O3 has D ~ 10

mm2/s and κ ~300 mW/cm·K [30]). It has not been previously observed in a “molecular

40

crystal”, for which the interlayer bonding is generally considered to be due to

van-der-Waals interactions between essentially rigid molecules. It is not observed in

pentacene crystals [8, 107], for which there are no sidegroups projecting between the

planes, nor in rubrene [3, 108], for which tetracene backbones align in the plane with

relatively rigid phenyl sidegroups project between the planes [109]. This suggests that the

large interlayer thermal conductivity of TIPS-pn is associated with the ability of the floppy

TIPS sidegroups to conduct heat.

In fact, its value suggests that we need to reconsider our understanding of phonon

propagation and heat transport in these materials. From kinetic theory, the phonon

thermal conductivity in a solid can be considered as a sum over all modes:

∑= jjjvc λρκ )3/( (3.1)

where ρ is the mass density, cj, vj, and λj are the specific heat, propagation velocity, and

mean-free path associated with phonons of mode j. For a van-der-Waals bonded

molecular crystal, it is usually assumed that the only propagating phonons (i.e. those with

significant velocities and mean free paths) are acoustic modes. At high temperatures, the

acoustic specific heat cacoustic = 3kB/(Ωρ), where kB is Boltzman’s constant and Ω is the

molecular volume, so

><Ω≈ acousticacousticB vk λκ / (3.2)

Using a typical acoustic phonon velocity of ~ 2 km/s gives an unreasonable value of

λacustic ~700 nm ~400 dc, where dc = 1.7 nm is the interlayer spacing. (In contrast, for

rubrene, λacoustic ≈ 1.4 nm [3].) Such a large mean-free path is extremely unlikely in view

of the measured large thermal disorder, e.g. shear motion of the molecules [110].

Therefore, we suggested that most of the heat is carried by optical phonons.

In particular, since the librational modes typically have energies ≤ kBTroom [111], they

can also carry heat at room temperature if they have sufficient dispersion. Furthermore,

because of the large number of terminal methyl groups (12 on each molecule), they can

potentially carry an order of magnitude more heat than the acoustic modes alone. In fact,

the quadrupolar coupling between these groups may give librational phonons sufficient

velocity to contribute. For example, assume a typical quadrupole moment Q ~ 10-39 C⋅m2

41

[112]. Since the distance between isopropyl groups on neighboring layers is r ~ 0.4 nm, the

interaction energy Uquad ~ Q2/(4πε0r5) ~ 5 meV [111]. This bandwidth would give a

librational optical phonon velocity close that of acoustic phonons: vlib ~ Uquad dc/h ~ 2

km/s, where h = Planck’s constant. Direct proof of propagating low-energy optical

phonons in TIPS-pn and related materials would require inelastic neutron or x-ray

measurements of phonon dispersion, which would be discussed later. Indirect proof may

come from measurements of in-plane and interlayer thermal diffusivity in materials with a

variety of interlayer constituents and structures.

Also surprising is the anisotropy, κc > 14 κlong; because the in-plane (π-π) interactions

were assumed to be stronger than the inter-plane (hydrocarbon) interactions, we expected

the longitudinal thermal conductivity to be greater than the transverse, as we observed for

rubrene, κc ~ κlong/6 [3]. For example, the in-plane electrical conductivity of TIPS-pn is

believed to be much greater than the interlayer value; the band structure calculation has

indicated extremely flat-bands (bandwidths << 10 meV) in the interlayer direction but

in-plane electron and hole bandwidths ~300 and 150 meV, respectively [113]. If one

assumes that thermal transport is dominated by intermolecular thermal resistances and

that these were equal for c-axis and in-plane interactions, then the thermal conductivities

in different directions would be proportional to the packing densities in their transverse

planes [114], making κc ~ 2 κlong because a ~ b ~ c/2 [15]. The much larger anisotropy we

measure therefore indicates that the intermolecular thermal resistances along c are

smaller than those in the plane, implying stronger phonon interactions between the TIPS

sidegroups than between the in-plane pentacene backbones.

To check these results for TIPS-Pn, we measured the transverse (Tac method) and

longitudinal diffusivities of the related materials listed in Figure 1.1. Figure 3.3 shows the

results of the frequency dependence of fVω for some crystals, along with the values of d

and τmeas. F2-TIPS-pn crystals are needles with a brick-layer structure [24] similar to that

of TIPS-pn. TIPGe-pn and EtTP-5 crystals, on the other hand, grow as thick (d > 200

µm) flakes (in-plane dimensions < 3 mm). Molecules in EtTP-5 are coplanar but

insulating substituents keep the aromatic faces ~ 10 Å apart along [010] [26]. In

TIPGe-pn, the orientation of the pentacene backbones alternate in the ab-plane so that the

42

aromatic surface of each molecule faces insulating substituents of adjacent molecules

[24]. In both materials, therefore, there is poor π-orbital overlap in the ab-plane, but the

silyl and germyl sidegroups still extend along the interlayer c-axis [24, 26]. For all

materials, a few crystals were measured with representative results shown. In all cases,

the responses are very fast so only lower limits for Dc (given in the caption) were

determined. As for TIPS-pn, the large diffusivities are unusual for van-der-Waals bonded

molecular crystals and suggest strong phonon interactions between the sidegroups, giving

significant dispersion to low-frequency optical phonons which consequently carry most

of the heat.

Figure 3.6 shows representative spatial dependences of f-1/2ln(Vω) for needle

shaped crystals of F2-TIPS-pn, F8-TIPS-pn, and CP-DIPS-pn; for the latter two, all

crystals were very thin (< 100 μm) so frequency-dependent transverse measurements are

not meaningful. All of these have brick-layer structures similar to that of TIPS-pn [24, 26,

101]. Because the cyclopropyl group makes the sidegroups slightly more rigid in

CP-DIPS-pn than in the other crystals, the molecules are more tightly packed; e.g. the

unit cell of CP-DIPS-Pn is 6% smaller than that of TIPS-pn [24]. The measured slopes

and calculated longitudinal diffusivities for the materials are given in the caption. As for

TIPS-pn, slopes were measured at a few frequencies for each crystal and the uncertainties

given in the table reflect the variations in slopes. The fluorinated crystals may have

slightly steeper slopes, and therefore lower thermal diffusivities, than TIPS-pn, but the

differences are less than the uncertainties in the measurement. The slopes for

CP-DIPS-pn had more scatter, probably reflecting the somewhat shorter crystals

available, but were always considerably steeper than for the other compounds. The

resulting lower diffusivity for CP-DIPS-pn is surprising because one would expect the

more rigid sidegroups to reduce scattering of acoustic phonons. It again suggests that

much of the heat is carried by molecular vibrations and that the greater rigidity of the

sidegroups reduces the dispersion and/or reduces the specific heat (by increasing the

energies) of the relevant low-energy modes.

43

x0 - x (mm)5.4 5.6 5.8 6.0

f -1/2 ln

Vω(n

V)

(H

z-1/2)

0.5

1.0

1.5

F8-TIPS-Pn 17 Hz

F2-TIPS-Pn 8.0 Hz

CP-DIPS-Pn 5.5 Hz

Figure 3.6 Spatial dependence of Vω for crystals of CP-DIPS-pn (length = 6 mm, -f-1/2 dlnVω/dx = 2.7 ± 0.3 (Hz1/2·mm)-1, Dlong = 0.43 ± 0.10 mm2/s), F2-TIPS-pn (length = 10 mm, -f-1/2 dlnVω/dx = 1.89 ± 0.07 (Hz1/2·mm)-1, Dlong = 0.88 ± 0.09 mm2/s), and F8-TIPS-pn (length = 11 mm, -f-1/2 dlnVω/dx = 1.98 ± 0.10 (Hz1/2·mm)-1, Dlong = 0.80 ± 0.10 mm2/s) at selected frequencies in their linear regions.

The large thermal diffusivities indicate that, for those of these materials with

brick-layer structures and high electronic mobility, Joule heating of micro-electronic

components should not pose a problem. While the longitudinal diffusivities are slightly

larger than desired for thermoelectric applications, they are not excessive. However,

thermoelectric devices would need to be constructed so that the very high transverse

diffusivities do not create thermal shunts.

3.3.4 Inelastic X-ray Measurements

We used inelastic x-ray scattering at the Advanced Photon Source at Argonne

National Lab to study the dispersion in low-frequency (~ 10 meV) optical phonons

propagating in the interlayer direction in crystals of TIPS-pn.

Figure 3.7 shows inelastic x-ray spectra at four wavevectors each in the c*, b* and a*

direction. For each case, the measurements are for phonons propagating along the

44

q-direction. The results indicate that 11 meV optical phonons have significant energy

dispersion along c*, corresponding to phonon velocities ~2.2 km/s; i.e. the peaks for c*

measurements appear to be slightly shifted by ~ 2 meV at the four wavevectors,

consistent with a possible phonon bandwidth of several meV. Our thermal conductivity

model assumes that some low energy (e.g. E < ~ 30 meV) optical phonons have "large"

dispersion (e.g. 1/h(dE/dq) > ~ 1 km/s) along c*, but not necessarily in the in-plane

directions. Assuming that the peaks we are observing in inelastic x-ray correspond to 2 or

more phonons each that we are not resolving, the phonon “cluster” we observe for c* is

dispersing more than this amount. For the mode propagating along b*, the dispersion is

smaller, corresponding to phonon velocities less than 0.4 km/s. No dispersion outside the

noise is observed for q along a*.

45

Figure 3.7 Inelastic x-ray spectra taken at the Advanced Photon Source at Argonne National Lab. (a)Energy resolution; (b)four c* wave vectors Q = [0, 0, -(2+q)] energy scans; (c)four b* wave vectors Q = [0, 1+q, 0] energy scans; (d)four a* wave vectors Q = [-(2+q), 0, 0] energy scans.

3.3.5 Summary

We have measured the longitudinal (needle-axis) and transverse, interlayer thermal

conductivities of crystals of TIPS-Pn by ac-calorimetry. We have found that the

longitudinal value is higher than that of rubrene [2] and pentacene [1] and comparable to

quasi-one-dimensional conductors with excellent π-orbital overlap [104-106]. The

transverse thermal diffusivity is at least an order of magnitude larger than the

longitudinal. These values and the inverted anisotropy of κ indicate that molecular

vibrations, presumably concentrated on the silyl-containing sidegroups, have sufficient

intermolecular interactions and dispersion to carry most of the heat. Similar values for

Energy (meV)-20 -10 0 10 20

Cou

nts

(2 m

inut

es)

0

50

100

150

200

250

300

q = 0 125

v < 0.4 km/s

energy (meV)-20 -10 0 10 20

Cou

nts

(3 m

inut

es)

0

100

200

300

400

500

600

q = 0.125q = 0.250q = 0.375q = 0.50

v = E/hq ~ 2.2 km/s

Energy (meV)

-6 -4 -2 0 2 4 6

Cou

nts

0

200

400

600

800

1000

1200

1400

FWHM = 2.12 meV

Energy (meV)-20 -10 0 10 20

Inte

nsity

(co

unts

/3 m

in)

0

50

100

150

200

250

q = 0.125q = 0.250q = 0.375q = 0.500 v < 0.4 km/s

(a)

(d) (c)

(b)

46

both the in-plane and interlayer thermal diffusivities were found for several other

materials with related structures.

47

Chapter 4 Thermal Resistances of Thin-Films of Small Molecule Organic

Semiconductors 4.1 Introduction

As discussed above, the thermal conductivities (κ) of layered crystals of several

small molecule organic semiconductors have been reported [1-5, 115]. For molecules

with planar backbones and silylethynyl (or germanylethynyl) sidegroups projecting

between planes, very high interplanar thermal conductivities were observed [4, 5]. For

example, while the in-plane, needle-axis thermal conductivity of TIPS-pn κneedle ≈ 1.6

W/m⋅K [4], the inter-plane (c-axis) thermal conductivity was measured to be κc ≈ 21

W/m⋅K. As discussed below, this large value and surprising inverted anisotropy (i.e. κc >

κneedle) has been tentatively associated with propagating low-frequency optical phonons,

in particular librations of alkyl chains terminating the silylethynyl sidegroups, requiring

relatively strong interactions between these groups on neighboring planes as well as long

mean-free paths [5]. In contrast, rubrene, with tetracene backbones and much more rigid

phenyl sidegroups, has an interlayer thermal conductivity of only κc ≈ 0.07 W/m⋅K ≈

κneedle/6 [3].

While such large thermal conductivities would provide efficient dissipation of Joule

heat and therefore bode well for electronic applications of these materials, most organic

semiconducting devices require materials in thin-film rather than bulk crystal form.

Thin film thermal resistances, even for crystalline films, can be much larger than the

values deduced from bulk crystalline conductivities, either because of reduced mean-free

paths in the material due to increased disorder or because of interfacial thermal resistance

with the substrate. In this chapter, we report on the thin film thermal resistances of films

of TIPS-pn, TES-ADT [30-31, 115] and diF-TES-ADT [19, 32] on sapphire and

thermally oxidized silicon substrates. (Because of small and irregularly shaped crystals,

bulk crystal values of the thermal conductivities of TES-ADT and diF-TES-ADT have

not yet been reported, but our preliminary photothermal measurements on a slightly

irregular TES-ADT crystal gave a value κc ≈ (5 ± 2) W/m⋅K, a few times smaller than the

value for TIPS-pn but still very large.) For these thin-film measurements, we use the

48

3ω-technique [6, 8, 40, 102, 116-119]], which, as described in Section 1.3, yields values

of the film thermal resistance

StRfilm /)/( intρκ += (4.1)

where t and S are the thickness and area of the film, ρint is the interfacial thermal

resistivity, and κ is the through-plane thermal conductivity (i.e. for layered, crystalline

films, κ = κc). Therefore, to separate the interfacial from intrinsic resistivity, it is useful to

measure the thickness dependence of Rfilm, which typically has not been done for organic

films [6, 8, 40, 116-118]. For example, for a 100 nm thick film of a material with κ ≈ 0.3

W/m⋅K (similar to many polymers [6, 117-118]), the two terms in Eqn. 4.1 are

comparable for ρint ≈ 3 × 10-7 Km2/W. This value is only about an order of magnitude

larger than that of evaporated metal [42, 120], oxide [121], or organometallic [122] films

on ceramic or metal substrates, but an order of magnitude smaller than the interface

resistances of the best epoxy resins [123] or thermal greases [123].

The samples measured are “vapor annealed” [6, 31, 32] and non-annealed spin-cast

films of TES-ADT, with thicknesses ranging from 77 nm to 205 nm, and sublimed films

of TIPS-pn and diF-TES-ADT, with thicknesses ranging from 100 nm to 4 μm. As

described in Section 2.5, the vapor annealed films are crystalline, with mm sized

crystallites oriented with c in the through-plane direction [31-33] while the other films

are thought to be ab-plane disordered, but with the molecular sidegroups also largely

oriented through-plane [33-35].

4.2 Results and Discussion

Figure 4.1 shows measured values of ∆T/P as a function of frequency (F = ω/2π) for

several spin-cast TES-ADT films. Also shown (open triangles) are the results for heaters

on four bare sapphire substrates; the solid line shows ∆T/P calculated with κsapphire= 38.9

± 0.8W/m∙K, consistent with the published value [34]. The dashed line shows the

calculated value of ∆T/P for a 500 nm thick film of TES-ADT assuming our measured

bulk crystal thermal conductivity value κc = 5.5 W/m∙K. The solid symbols show our

measurements on vapor annealed TES- but non-annealed (i.e. non-crystalline) films.

49

Thicknesses were determined by atomic-force microscopy. The thermal resistance values

vary from 1.0 to 2.8 K/W, much larger than expected from the bulk crystal measurements

(e.g. as shown by the dashed line), and for the crystalline films don’t scale with film

thickness, suggesting that for these, the thermal resistance is dominated by the interface

thermal resistivity; the resulting values of ρint vary from (3 to 8) × 10-7 Km2/W, which, as

mentioned above, are reasonable values for deposited films. In fact, it has been suggested

that vapor annealing causes dewetting of the films from the substrate [124], so it is not

surprising that the interface resistivity be non-negligible. On the other hand, we will

argue below that for the non-annealed films, the interface resistivity is smaller, i.e. ρint <

2 × 10-7 Km2/W and ρint < t/κ.

Figure 4.1 Measured frequency dependence of the ∆T/P for spin-cast TES-ADT films on sapphire substrates. The open triangles show the results on bare sapphire and the solid line is a fit to Eqn. 1.5. The dashed line shows the expected results for 500 nm thick TES-ADT, assuming κ = κc(crystal) = 5W/m⋅K. The solid symbols show results for vapor-annealed films of different thicknesses, as shown, and the open circles and squares show results for non-annealed films.

ln(F(Hz))3 4 5 6 7

∆T/

P(K

/W)

0

2

4

6

8

saphhire fit ------ 500 nm crystal calculation

~205 nm vapor annealed~100nm vapor annealed

~77nm vapor annealed~90nm vapor annealed

∆ ∆

~90nm non-annealed~80nm non-annealed

∆ ∆ bare sapphire measurements

50

Figure 4.2 Measured frequency dependence of ∆T/P for ≈ 100 nm thick vapor-annealed (solid blue circles) and non-annealed (open blue circles) spin-cast TES-ADT films on thermally oxidized silicon. The red triangles show the results on the bare oxidized silicon and the calculated silicon baseline [from Eqn. 1.5] is shown by the solid line.

Measurements on spin-cast TES-ADT films (~100 nm thick) were also carried out on

doped silicon substrates (with thermally oxidized surfaces), the most common substrate

for organic thin-film transistors. The calculated baseline for silicon, with κSi = 142

W/m⋅K [35], as well as the measured value on the thermally oxidized substrate are shown

in Figure 4.2, with the measured values for vapor-annealed and non-annealed films

prepared at the same time. The thermal resistances are comparable to those measured on

sapphire. (The nonlinearities for F > 30 Hz are caused by capacitive coupling of the

heater to the conducting, grounded doped silicon.)

The results of 3ω measurements on several sublimed diF-TES-ADT films and

TIPS-pn films of different thicknesses on sapphire are shown in Figure 4.3. Note that for

the thicker films, t approaches (D/2ω)1/2, the thermal wave length, explaining the

downward curvature at high frequencies.

Calculated silicon

ln(F(Hz))2.5 3.0 3.5 4.0 4.5 5.0

∆T/

P (K

/W)

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

4.5

calculated silicon base line

silicon wafer with a thin layer of oxide ~300nm

Non-annealed TES-ADT

Annealed TES-ADT

51

Figure 4.3 Frequency dependence of sublimed films of diF-TES-ADT (left panels) and TIPS-pn (right panels) of the indicated thicknesses on sapphire substrates. The reference bare sapphire line (from Figure 4.1) is shown in the lower left panel.

The thickness dependence of the thermal resistances for these sublimed films is

shown in Figure 4.4. Although there is some scatter, presumably reflecting the quality of

the films, both materials exhibit a rough linear dependence of Rfilm ≡ ∆Tfilm/P on

thickness. As shown in the inset, the intercept is very small: Rfilm(t=0) < 0.6 K/W

corresponding to ρint < 2 × 10-7 Km2/W. That is, the sublimed films have lower interface

resistivity than the annealed, spin-cast TES-ADT films.

For TIPS-pn, the average value of Rfilm/t (using the profilometer values of t) gives κ

= (0.104 ± 0.007) W/m⋅K. This value, similar to that of polymers [6, 117-118] is an order

of magnitude smaller than the in-plane (needle axis) thermal conductivity [4] and two

52

orders of magnitude smaller than the c-axis [5] thermal conductivity of crystals. If one

assumes that the TIPS sidegroups are still aligned perpendicular to the substrate in the

sublimed films, the small value of κ in the films seems especially surprising, since the

large value of κc in crystals was associated with interactions between low-energy

librations of these groups in molecules on neighboring layers [5]. However, the disorder

of the sublimed films would reduce the mean-free path of the librations to ~ 1 molecular

layer, preventing propagation of these phonons and removing their contribution to the

thermal conductivity.

Figure 4.4 Thickness dependence of film thermal resistance (∆Tfilm/P) for sublimed films of diF-TES-ADT (triangles) and TIPS-pn (circles). The solid symbols show the profilometer measurements of the film thicknesses while the open symbols show the thicknesses as determined by the quartz crystal monitor during sublimation. Also shown (open inverted red triangles) are the results for the non-annealed spin-cast TES-ADT films. The inset shows a blow-up of the results with t < 1 μm.

For diF-TES-ADT, the average value of Rfilm/t gives κ = (0.13 ± 0.01) W/m⋅K.

Since the structures of diF-TES-ADT and TES-ADT are very similar [31], we expect

similar thermal conductivities for the two materials, and in-fact the thermal resistances of

Film Thickness (nm)0 1000 2000 3000 4000 5000

∆T f

ilm/P

(K

/W)

0

20

40

60

80

100

120

TIPS-pn (profilometer)TIPS-pn (quartz crystal)diF-TES-ADT (profilometer)diF-TES-ADT(quartz crystal)non-annealed TES-ADT(AFM)

20

10

400 800

53

the non-annealed spin-cast TES-ADT films, also shown in Figure 4.4, are consistent with

those of the sublimed films of diF-TES-ADT. We therefore assume that the non-annealed,

spin-cast TES-ADT films also have lower interface resistivity than the annealed films,

presumably reflecting some dewetting and lift-off of the latter. As for TIPS-pn, κ(film)

<< κc(crystal), and since the films presumably have c-axis orientation [33], this

presumably reflects the very short mean-free paths for librational phonons in the films.

4.3 Conclusion

The thermal conductivities of sublimed films of TIPS-pn and diF-TES-ADT are

much smaller than their crystalline values, presumably because the lack of

three-dimensional order in the films severely limits the mean-free path of the conducting

phonons, including the librational optical phonons assumed to carry much of the heat.

On the other hand, the thermal resistances of thin (≤ 205 nm) crystalline films of

TES-ADT, prepared by vapor-annealing of spin-cast films, are dominated by their

interface resistances, possibly due to dewetting of the film from the substrate during the

annealing process. Unfortunately, we have not been able to produce thicker crystalline

films to study their intrinsic conductivities.

54

Chapter 5 Conclusions

We have developed new probes to measure thermal conductivities of small bulk

(crystalline and polymer) samples and thin films of organic semiconductors, both to

identify these materials for possible applications and to gain an understanding of thermal

transport in molecular solids. The emphasis of my work was on the thin-film

measurements, but I have participated with my group in bulk sample measurements as

well and, in particular, I am responsible for the design of the longitudinal bulk sample

apparatus.

Since most of the small molecule organic semiconductor crystals have layered

structures, separate techniques were used for in-plane and transverse thermal

conductivities. For in-plane measurements, we used a position dependent ac-calorimetric

technique because the crystals are typically small (0.5-10 mm in plane and less than 0.6

mm interplane). We found that the needle-axis thermal conductivities of

bis(triisopropylsilylethynyl) pentacene (TIPS-pn) and related crystals were quite high (e.g.

Κneedle ≈ 16 mW/cm⋅K), comparable to that of the quasi-one dimensional conductor

TTF-TCNQ and several times higher than that of rubrene. For the transverse

measurements, we found that the frequency-dependent ac-calorimetric technique did not

work for the TIPS-pn family of materials, because the interlayer thermal transport was

too fast and the response was limited by the thermal resistance of the glue used to attach a

thermocouple to the sample. Therefore, we developed a new ac-photothermal technique

aimed for small (area > 1 mm2) samples, measuring the frequency-dependence of thermal

radiation from the sample by mounting it in front of a mercury-cadmium-telluride

infrared detector inside the detector dewar. These results confirmed TIPS-pn’s high

interlayer (c-axis) thermal conductivity, κc = (210 ± 100) mW/cm⋅K, implying that most

of the heat is carried not by acoustic but by optical phonons, presumably associated with

low energy librations of terminal methyl groups projecting between the layers.

Preliminary inelastic x-ray spectra of TIPS-pn taken at the Advanced Photon Source at

Argonne National Lab has shown a large dispersion of low energy optical phonons in c*

direction which supports our guess.

55

While such large thermal conductivities would provide efficient dissipation of Joule

heat and therefore bode well for electronic applications of these materials, most organic

semiconducting devices require materials in thin-film rather than bulk crystal form. Thin

film thermal resistances, even for crystalline films, can be much larger than the values

deduced from bulk crystalline conductivities, either because of reduced mean-free paths

in the material due to increased disorder or because of interfacial thermal resistance with

the substrate. Therefore, it was desirable to measure the thin film thermal resistance of

TIPS-pn and other small molecule materials, which is important in establishing the heat

dissipation capability in devices such as thin-film transistors.

We used the 3ω-technique of Lee and Cahill to measure the thin film thermal

resistances of films of TIPS-pn and two materials with similar sidegroups and crystal

structures, bis(triethylsilylethynyl) anthradithiophene (TES-ADT) and difluoro

bis(triethylsilylethynyl) anthra-dithiophene (diF-TES-ADT), on sapphire and thermally

oxidized silicon substrates. For each material, several films of different thicknesses have

been measured to separate the effects of intrinsic thermal conductivity from interface

thermal resistance. For sublimed films of TIPS-pn and diF-TES-ADT, with thicknesses

ranging from less than 100 nm to greater than 4 μm, the thermal conductivities are similar

to those of polymers and over an order of magnitude smaller than those of single crystals

presumably because the lack of three-dimensional order in the films severely limits the

mean-free path of the conducting phonons, including the librational optical phonons

proposed to carry much of the heat. On the other hand, the thermal resistances of thin (≤

205 nm) crystalline films of TES-ADT, prepared by vapor-annealing of spin-cast films,

do not depend on thickness and we assumed they are dominated by their interface

resistances, possibly due to dewetting of the film from the substrate during the annealing

process. It is also possible that the molecular packing in crystalline thin films has a

thickness dependence, which will make a big difference in their thermal conductivities

for thin films with different thicknesses. While not excessive, such thermal resistances

might limit the utility of these films in electronic devices. It remains to be determined if

solution-cast, high electronic mobility, crystalline films of TIPS-pn and diF-TES-ADT,

which are too irregular in thickness for our 3ω-measurements, have comparable interface

56

resistances. Other thin films deposition techniques, e.g. inkjet printing, could be tried to

make such films with more uniform thickness.

57

References

1. Daniel Tobjörk and Ronald Österbacka, “Paper Electronics,” Adv. Mater. 23,

1935-1961, 2011. 2. Y. Okada, M. Uno, Y. Nakazawa, K. Sasai, K. Matsukawa, M. Yoshimura, Y.

Kitaoka, Y. Mori, and J. Takeya, “Thermal diffusivity measurement of thin films by mean of an ac calorimetric method,” Phys. Rev. B, 83, 113305, 2011.

3. H. Zhang and J.W. Brill, “Interlayer thermal conductivity of rubrene measured by ac-calorimetry,” J. Appl. Phys., 114, 043508, 2013.

4. H. Zhang, Y. Yao, M.M. Payne, J.E. Anthony, and J.W. Brill, “Thermal diffusivities of functionalized pentacene semiconductors,” Appl. Phys. Lett., 105, 073302, 2014.

5. J.W. Brill, M. Shahi, M.M. Payne, J. Edberg, Y. Yao, X. Crispin, and J.E. Anthony, “Frequency-Dependent Photothermal Measurements of Transverse Thermal Diffusivity of Organic Semiconductors”, J. Appl. Phys., 118, 235501, 2015.

6. G.-H. Kim, L. Shao, K. Zhang, and K.P. Pipe, “Engineered doping of organic semiconductors for enhanced thermoelectric efficiency,” Nature Materials, 12, 719-723, 2013.

7. A. Malti, J. Edberg, H. Granberg, Z.U. Khan, J. W. Andreasen, X. Liu, D. Zao, H. Zhang, Y. Yao, J.W. Brill, I. Engquist, M. Fahlman, L. Wagberg, X. Crispin, and M. Berggren, “An Organic Mixed Ion-Electron Conductor for Power Electronics”, Adv. Sci., 1500305, 2015.

8. Namsu Kim, Benoit Domercq, Seunhyup Yoo, Adam Christensen, Bernard Kippelen, and Samuel Graham, “Thermal transport properties of thin films of small molecule organic semiconductors”, Appl. Phys. Lett. 87, 241908, 2005.

9. Olga Bubnova, Zia Ullah Khan, Abdellah Malti, Slawomir Braun, Mats fahlman, Magnus Berggren, and Xavier Crispin, “Optimization of the thermoelectric figure of merit in the conducting polymer poly(3,4-ethylenedioxythiophene),” Mature Materials, 10, 429-433, 2011.

10. John C. Duda, Patrick E. Hopkins, Yang Shen, and Mool C. Gupta, “Thermal transport in organic semiconducting polymers,” Appl. Phys. Lett., 102, 251912, 2013.

11. L. Maiolo, M. Cuscuna, L. Mariucci, A.Pecora, D. Simeone, A. Valletta, and G. Fortunato, “Analysis if self-heating related instability in n-channel polysilicon thin film transistors fabricated on polyimide”, Thin Solid Films, 517, 6371-6374, 2009.

12. Katz, H. E., Bao, Z., Gilat, S. J., “Synthetic Chemistry for Ultrapure, Processable, and High-Mobility Organic Transistor Semiconductors,” Acc. Chem. Res., 34, 359, 2001.

13. Wei Xie, Kathryn A. McGarry, Feilong Liu, Yanfei Wu, P. Paul Ruden, Christopher J. Douglas, and C. Daniel Frisbie, “High-Mobility Transistors Based

58

on Single Crystals of Isotopically Substituted Rubrene-d28,” J. Phys. Chem. C, 117, 11522-11529, 2013.

14. Dong Wang, Ling Tang, Mengqiu Long, and Zhigang Shuai, “First-principles investigation of organic semiconductors for thermoelectric applications,” J. Chem. Phys., 131, 224704, 2009.

15. John E. Anthony, “Functionalized Acenes and Heteroacenes for Organic Electronics”, Chem.Rev. 106, 5028-5048, 2006.

16. John E. Anthony, “The Larger Acenes: Versatile Organic Semiconductors,” Angew. Chem. Int. Ed., 47, 452-483, 2008.

17. C.E. Sheraw, T.N. Jackson, D.L. Eaton, and J.E. Anthony, “Functionalized Pentacene Active Layer Organic Thin-Film Transistors,” Adv. Mater., 15, 2009-2011, 2003.

18. Katelyn P. Goetz, Zhong Li, Jeremy W. Ward, Cortney Bougher, Jonathan Rivnay, Jeremy Smith, Brad R. Conrad, Sean R. Parkin, Thomas D. Anthopoulos, Alberto Salleo, John E. Anthony, and Oana D. Jurchescu, “Effect of Acne Length on Electronic Properties in 5-,6-, and7-Ringed Heteroacenes”, Adv. Mater., 23, 3698-3073, 2011.

19. O.D. Jurchescu, S. Subramanian, R.J. Kline, S.D. Hudson, J.E. Anthony, T.N. Jackson, and D.J. Gundlach, “Organic Single-Crystal Field-Effect Transistors of a Soluble Anthradithiophene,” Chem. Mater., 2008, 20, 6733-6737.

20. B.R. Conrad, C.K. Chan, M.A. Loth, S.R. Parkin, X. Zhang, D.M. DeLongchamp, J.E. Anthony, and D.J. Gundlach, “Characterization of a soluble anthradithiophene derivative,” App. Phys. Lett., 97, 133306 (3pp) (2010).

21. J. Jang et al., “Highly Creystalline Soluble Acene Crystal Arrays for Organic Transistors: Mechanism of Crystal Growth During Dip-Coating,” Adv. Funct. Mater., vol.22, pp. 1005-1014, 2012.

22. A. C. Arias et al., “All jet-printed polymer thin-film transistor active-matrix backplanes,” Appl. Phys. Lett., vol. 85, n. 15, pp. 3304-3306, 2004.

23. A. Azarova et al., “Fabrication of organic thin-film trasistors by spray-deposition for low-cost, large-area electronic,” Org. Electron., vol.11, pp. 1960-1965, 2010.

24. S. R. Parkin and J. E. Anthony, “private communication.” 25. C. R. Swartz, S. R. Parkin, J. E. Bullock, J. E. Anthony, A. C. Mayer, and G. G.

Malliaras, "Synthesis and characterization of electron-deficient pentacenes," Org. Lett., 7, 3163, 2005.

26. M. A. Wolak, J. Delcamp, C. A. Landis, P. A. Lane, J. Anthony, and Z. H. Kafafi, "High performance organic light-emitting diodes based on dioxolane-substituted pentacene derivatives," Adv. Func. Mater., 16, 1943, 2006.

27. J. E. Anthony, J. S. Brooks, D. L. Eaton, and S. R. Parkin, "Functionalized pentacene: Improved electronic properties from control of solid-state order," J. Am. Chem. Soc., 123, 9482, 2001.

28. J. H. Chen, D. C. Martin, and J. E. Anthony, "Morphology and molecular orientation of thin-film bis(triisopropylsilylethynyl) pentacene," J. Mater. Res., 22, 1701, 2007.

29. J. E. Anthony and Joseph Brill, “private communication.”

59

30. V. Pischik, L.A. Lytvynov, and E.R. Dobrovinskaya, “Sapphire: Material, Manufacturing, Applications,” Springer, New York, 2009.

31. K.C. Dickey, J.E. Anthony, and Y.-L. Loo, “Improving Organic Thin-Film Transistor Performance through Solvent-Vapor Annealing of Solution-Processable Triethylsilylethynyl Anthradithiophene,” Adv. Mater., 2006, 18, 1721-1726.

32. S.S. Lee, C.S. Kim, E.D. Gomez, B. Parushothaman, M.F. Toney, C. Wang, A. Hexemer, .E. Anthony, and Y.-L. Loo, “Controlling Nucleation and Crystallization in Solution-Processed Organic Semiconductors for Thin-Film Transistors,” Adv. Mater., 2009, 21, 3605-3609.

33. S.S. Lee, S.B. Tang, D.-M. Smilgies, A.R. Woll, M.A. Loth, J.M. Mativetsky, J.E. Anthony, and Y.-L. Loo, “Guiding Crystallization around Bends and Sharp Corners,” Adv. Mater., 2012, 24, 2692-2698.

34. S.C.B. Mannsfeld, M.L. Tang, and Z. Bao, “Thin Film Structure of Triisopropylsilylethynyl-Functionalized Pentacene and Tetraceno[2,3-b]thiophene from Grazing Incidence X-Ray Diffraction,” Adv. Mater., 2011, 23, 127-131.

35. J.W. Ward, R. Li, A. Obaid, M.M. Payne, D.- M. Smilgies, J.E. Anthony, A. Amassian, and O.D. Jurchescu, “Rational Design of Organic Semiconductors for Texture Control and Self-Patterning on Halogenated Surfaces,” Adv. Funct. Mater., 2014, 24, 5052-5058.

36. I. Hatta, Y. Sasuga, R. Kato, and A. Maesono, “Thermal diffusivity measurement of thin films by means of an ac calorimetric method,” Rev. Sci. Instrum., 56, 1643, 1985.

37. Paul f. Sullivan and G. Seidel, “Steady-State, ac-Temperature Calorimetry,” Phys. Rev., 173, 679-685, 1968.

38. Raquel Fuente, Estibaliz Apinaniz, Arantza Mendioroz, and Agustin Salazar, “Simultaneous measurement of thermal diffusivity and optical absorption coefficient using photothermal radiometry. I. Homogeneous solids”, J. Appl. Phys., 110, 03315, 2011.

39. P. Martinez-Torres, A. Mandelis, and J.J. Alvarado-Gil, “Photothermal determination of thermal diffusivity and polymerization depth profiles of polymerized dental resins”, J. Appl. Phys., 106, 114906 (7pp), 2009.

40. S.-M. Lee and D.G. Cahill, “Heat transport in thin dielectric films,” J. Appl. Phys., 1997, 81, 2590-2595.

41. D.G. Cahill, “Thermal conductivity measurement from 30 to 750 K: the 3ω method,” Rev. Sci. Inst., 1990, 61, 802-808.

42. A. Jacquot, B. Lenoir, A. Dauscher, M. Stolzer, and J. Meusel, “Numerical simulation of the 3ω method for measuring the thermal conductivity,” J. Appl. Phys., 2002, 91, 4733-4738.

43. Greg Porter and Joseph Brill, “private communication.” 44. J. Chen, D.C. Martin, and J.E. Anthony, “Morphology and molecular orientation

of thin-film bis(triisopropylsilylethynyl) pentacene,” J. Mater. Res., 2007, 22, 1701-1709.

60

45. Y. Lin, S. Fang, D. Su, K. S. Brinkman, F. Chen, “Enhancing grain boundary ionic conductivity in mixed ionic-electronic conductors,” Nature Communications 6, 2015.

46. O. Y. Posudievsky et al., “Metallic Conductivity of Mechanochemically Doped Polyaniline,” Theoretical and Experimental Chemistry 50, 197-203, 2014.

47. N. M. Kocherginsky, Z. Wang, “The role of ionic conductivity and interface in electrical resistance, ion transport and transmembrane redox reactions through polyaniline membranes,” Synthetic Metals 156, 1065-1072, 2006.

48. S. R. Forrest, “The path to ubiquitous and low-cost organic electronic appliances on plastic,” Nature 428, 911-918, 2004.

49. A. J. Heeger, “Semiconducting and Metallic Polymers: The Fourth Generation of Polymeric Materials (Nobel Lecture),” Angewandte Chemie International Edition 40, 2591-2611, 2001.

50. A. G. MacDiarmid, A. J. Heeger, “Organic metals and semiconductors: The chemistry of polyacetylene, (CH)x, and its derivatives,” Synthetic Metals 1, 101-118, 1980.

51. K. K. Kanazawa et al., “‘Organic metals’: polypyrrole, a stable synthetic ‘metallic’ polymer,” Journal of the Chemical Society, Chemical Communications, 854-855, 1979.

52. H. Shirakawa, S. Ikeda, “Preparation and morphology of as-prepared and highly stretch-aligned polyacetylene,” Synthetic Metals 1, 175-184, 1980.

53. O. Bubnova et al., “Semi-metallic polymers,” Nat Mater 13, 190-194, 2014. 54. N. Kim et al., “Highly Conductive All-Plastic Electrodes Fabricated Using a

Novel Chemically Controlled Transfer-Printing Method,” Advanced Materials 27, 2317-2323, 2015.

55. M. Berggren, D. Nilsson, N. D. Robinson, “Organic materials for printed electronics,” Nat Mater 6, 3-5, 2007.

56. G. Milczarek, O. Inganäs, “Renewable Cathode Materials from Biopolymer /Conjugated Polymer Interpenetrating Networks,” Science 335, 1468-1471, 2012.

57. B. Winther-Jensen, O. Winther-Jensen, M. Forsyth, D. R. MacFarlane, “High Rates of Oxygen Reduction over a Vapor Phase–Polymerized PEDOT Electrode,” Science 321, 671-674, 2008.

58. L. Groenendaal, F. Jonas, D. Freitag, H. Pielartzik, J. R. Reynolds, “Poly(3,4-ethylenedioxythiophene) and Its Derivatives: Past, Present, and Future,” Advanced Materials, 12, 481-494, 2000.

59. H. D. Wang, U. D. Ail, R. D. Gabrielsson, M. P. Berggren, X. Crispin, “Ionic Seebeck effect in conducting polymers,” Advanced Energy Materials, 5, 2015.

60. G. Wee et al., “Effect of the Ionic Conductivity on the Performance of Polyelectrolyte-Based Supercapacitors,” Advanced Functional Materials, 20, 4344-4350, 2010.

61. D. Hohnholz, H. Okuzaki, A. G. MacDiarmid, “Plastic Electronic Devices Through Line Patterning of Conducting Polymers,” Advanced Functional Materials, 15, 51-56, 2005.

61

62. A. A. Argun, J. N. Ashcraft, P. T. Hammond, “Highly conductive, methanol resistant polyelectrolyte multilayers,” Advanced Materials, 20, 1539-1543, 2008.

63. X. Gu, D. B. Knorr Jr, G. Wang, R. M. Overney, “Layered and interfacially blended polyelectrolyte multi-walled carbon nanotube composites for enhanced ionic conductivity,” Thin Solid Films, 520, 1872-1879, 2012.

64. C. A. Nguyen, S. Xiong, J. Ma, X. Lu, P. S. Lee, “High ionic conductivity P(VDF-TrFE)/PEO blended polymer electrolytes for solid electrochromic devices,” Physical Chemistry Chemical Physics, 13, 13319-13326, 2011.

65. H. Saruwatari, T. Kuboki, T. Kishi, S. Mikoshiba, N. Takami, “Imidazolium ionic liquids containing LiBOB electrolyte for lithium battery,” Journal of Power Sources, 195, 1495-1499, 2010.

66. S. S. Sekhon, “Conductivity behaviour of polymer gel electrolytes: Role of polymer,” Bulletin of Materials Science, 26, 321-328, 2003.

67. N. Kamaya et al., “A lithium superionic conductor,” Nature Materials, 10, 682-686, 2011.

68. H. Wada, O. Amiel, A. Sato, “Crystal structures and ionic conductivities of new silver hafnium sulfides.,” Journal of Alloys and Compounds, 219, 55-58, 1995.

69. V. Young, “Theory of electronic and ionic conductivity in particulate membrane ion selective electrode systems: All solid state silver sulfide electrode,” Solid State Ionics, 20, 277-282 1986.

70. Y. Inaguma et al., “High ionic conductivity in lithium lanthanum titanate,” Solid State Communications, 86, 689-693, 1993.

71. X. Ren, P. G. Pickup, “Coupling of ion and electron transport during impedance measurements on a conducting polymer with similar ionic and electronic conductivities. Journal of the Chemical Society,” Faraday Transactions, 89, 321-326, 1993.

72. G. L. Duffitt, P. G. Pickup, “Enhanced ionic conductivity of polypyrrole due to incorporation of excess electrolyte during potential cycling,” Journal of the Chemical Society, Faraday Transactions 88, 1417-1423, 1992.

73. H. C. Kang, K. E. Geckeler, “Enhanced electrical conductivity of polypyrrole prepared by chemical oxidative polymerization: Effect of the preparation technique and polymer additive,” Polymer, 41, 6931-6934, 2000.

74. E. B. online. (Encyclopaedia Britannica, 2015), vol. 2015. 75. R. A. Serway, Principles of Physics. (Saunders College Publishing, ed. 2nd,

1998). 76. G. Zotti et al., “Electrochemical and XPS Studies toward the Role of Monomeric

and Polymeric Sulfonate Counterions in the Synthesis, Composition, and Properties of Poly(3,4-ethylenedioxythiophene),” Macromolecules, 36, 3337-3344, 2003.

77. M. Henriksson, L. A. Berglund, P. Isaksson, T. Lindström, T. Nishino, “Cellulose Nanopaper Structures of High Toughness,” Biomacromolecules, 9, 1579-1585, 2008.

62

78. Lee, H.J., et al., “Enhanced thermoelectric performance of PEDOT: PSS/PANI–CSA polymer multilayer structures,” Energy & Environmental Science, 9(9): p. 2806-2811, 2016.

79. Cho, C., et al., “Outstanding Low Temperature Thermoelectric Power Factor from Completely Organic Thin Films Enabled by Multidimensional Conjugated Nanomaterials,” Advanced Energy Materials, 2016.

80. Zhang, K., et al., “Thermoelectric performance of p-type nanohybrids filled polymer composites,” Nano Energy, 13: p. 327-335, 2015

81. Fan, Z., et al., “Significantly Enhanced Thermoelectric Properties of PEDOT: PSS Films through Sequential Post‐Treatments with Common Acids and Bases,” Advanced Energy Materials, 2016.

82. Xia, Y., K. Sun, and J. Ouyang, “Solution‐processed metallic conducting polymer films as transparent electrode of optoelectronic devices,” Advanced Materials, 24(18): p. 2436-2440, 2012.

83. Bae, E.J., et al., “Enhancement of Thermoelectric Properties of PEDOT: PSS and Tellurium-PEDOT: PSS Hybrid Composites by Simple Chemical Treatment,” Scientific reports, 6: p. 18805-18805, 2015.

84. Fabiano, S., et al., “Poly (ethylene imine) Impurities Induce n‐doping Reaction in Organic (Semi) Conductors,” Advanced Materials, 26(34): p. 6000-6006, 2014

85. Bubnova, O., et al., “Optimization of the thermoelectric figure of merit in the conducting polymer poly (3, 4-ethylenedioxythiophene),” Nature materials, 10(6): p. 429-433, 2011.

86. Bubnova, O., et al., “Semi-metallic polymers,” Nature materials, 13(2): p. 190-194, 2014.

87. Kim, G.-H., et al., “Engineered doping of organic semiconductors for enhanced thermoelectric efficiency,” Nature Materials, 12(8): p. 719-723, 2013.

88. Søndergaard, R.R., et al., “Practical evaluation of organic polymer thermoelectrics by large‐area R2R processing on flexible substrates,” Energy Science & Engineering, 1(2): p. 81-88, 2013.

89. J. Choi, J.Y. Lee, S.S. Lee, C.R. Park, and H. Kim, “High‐Performance Thermoelectric Paper Based on Double Carrier‐Filtering Processes at Nanowire Heterojunctions,” Adv. Energy Mater., 6 1502181, 2016

90. C. Liu, B. Lu, J. Yan, J. Xu, R. Yue, Z. Zhu, S. Zhou, X. Hu, Z. Zhang, and P. Chen, “Highly conducting free-standing poly (3, 4-ethylenedioxythiophene)/poly (styrenesulfonate) films with improved thermoelectric performances,” Synth. Met. 160, 2481-2485, 2010.

91. Q. Jiang, C. Liu, H. Song, J. Xu, D. Mo, H. Shi, Z. Wang, F. Jiang, B. Lu, and Z. Zhu, “Free-standing PEDOT: PSS Film as Electrode for the Electrodeposition of Bismuth Telluride and Its Thermoelectric Performance,” Int. J. Electrochem. Sci., 9, 7540-7551, 2014.

92. Zaifang Li, Hengda Sunc, Ching-Lien Hsiao, Yulong Yao, Maryam Shahi, Alex Cruce, Xianjie Liu, Youyu Jiang, Wei Meng, Fei Qin, Thomas Ederth, Simone Fabiano, Weimin M. Chen, Jens Birch, Joseph W. Brill, Yinhua Zhou, Xavier

63

Crispin, and Fengling Zhang, “High power density free-standing PEDOT:PSS thermoelectric generator,” submitted to Energy & Environmental Science.

93. J. Liu, X. Wang, D. Li, N.E. Coates, R.A. Segalman, and D.G. Cahill, “Thermal conductivity and elastic constants of PEDOT: PSS with high electrical conductivity,” Macromolecules, 48, 585-591, 2015

94. Q. Wei, C. Uehara, M. Mukaida, K. Kirihara, and T. Ishida, “Measurement of in-plane thermal conductivity in polymer films,” AIP Adv., 6, 045315, 2016.

95. Z. Li, G. Ma, R. Ge, F. Qin, X. Dong, W. Meng, T. Liu, J. Tong, F. Jiang, and Y. Zhou, “Free‐Standing Conducting Polymer Films for High‐Performance Energy Devices,” Angewandte Chemie, 128, 991-994, 2016.

96. J.K. Carson, S.J. Lovatt, D.J. Tanner, and A.C. Cleland, “Thermal conductivity bounds for isotropic, porous materials,” Int. J. Heat Mass Transfer, 48, 2150-2158, 2005

97. R.H. Tarkhanyan and D.G. Niarchos, “Reduction of thermal conductivity in porous “gray” materials,” APL Mater., 2, 076107, 2014.

98. Q. Wei, M. Mukaida, K. Kirihara, and T. Ishida, “Experimental studies on the anisotropic thermoelectric properties of conducting polymer films,” ACS Macro Lett., 3, 948-952, 2014.

99. Measured impurity levels were ≤ 0.05%; consequently, if the Wiedemann–Franz law approximately holds and the electron mobility µ < 10 cm2/V·s, the electronic thermal conductivity < 0.01 mW/cm·K, orders of magnitude less than our measured values of κ.

100. Y. Diao, B. C.-K. Tee, G. Giri, J. Xu, D.H. Kim, H.A. Becerril, R.M. Stoltenberg, T.H. Lee, G. Xue, S.C. Mannsfeld, and Z. Bao, “Solution coating of large-area organic semiconductor thin films with aligned single-crystalline domains,” Nature Materials, 12, 665, 2013.

101. C. R. Swartz, S. R. Parkin, J. E. Bullock, J. E. Anthony, A. C. Mayer, G. G. Malliaras, “Synthesis and Characterization of Electron-Deficient Pentacenes,” Org. Lett., 7, 3163, 2005.

102. Yiqin Wang, M. Chung, T.N. O’Neal, and J.W. Brill, “Differential scanning calorimetry at charge-density-wave transitions,” Synth. Met., 46, 307, 1992.

103. O. Bubnova, Z.U. Khan, A. Malti, S. Braun, M. Fahlman, M. Berggren, and X. Crispin, “Optimization of the thermoelectric figure of merit in the conducting polymer poly(3,4-ethylenedioxythiophene),” Nature Materials, 10, 429, 2011.

104. M.B. Salamon, J.W. Bray, G. DePasquali, R.A. Craven, G. Stucky, and A. Schultz, “Thermal conductivity of tetrathiafulvalene-tetracyanoquinodimethane (TTF-TCNQ) near the metal-insulator transition,” Phys.Rev. B, 11, 619, 1975.

105. H. Grassi, A. Brau, and J.P. Farges, “Heat Pulse Diffusion in TEA(TCNQ)2,” Phys. Stat. Sol. (b), 112, 633, 1982.

106. K. Torizuki, H. Tajima, and T. Yamamoto, “Thermal conductivity of ( DMe − DCNQI )2 Li1−x Cux (0 ⩽ x ⩽ 0.14) : Phonon propagation and the spin-Peierls lattice distortion,” Phys. Rev. B, 71, 193101, 2005.

64

107. D. Wang, L. Tang, M. Long, and Z. Shuai, “Anisotropic Thermal Transport in Organic Molecular Crystals from Nonequilibrium Molecular Dynamics Simulations,” J. Phys. Chem. C, 115, 5940, 2011.

108. Y. Okada, M. Uno, Y. Nakazawa, K. Sasai, K. Matsukawa, M. yoshimura, Y. Kitaoka, Y. Mori, and J. Takeya, “Low-temperature thermal conductivity of bulk and film-like rubrene single crystals,” Phys. Rev. B, 83, 113305, 2011.

109. V.C. Sundar, J. Zaumseil, V. Podzorov, E. Menard, R.L. Willett, T. Someya, M.E. Gershenson, and J.A. Rogers, “Elastomeric Transistor Stamps: Reversible Probing of Charge Transport in Organic Crystals,” Science, 303, 1644, 2004.

110. A.S. Eggerman, S. Illig, A. Troisi, H. Sirringhaus, and P.A. Midgley, “Measurement of molecular motion in organic semiconductors by thermal diffuse electron scattering,” Nat. Mater., 12, 1045, 2013.

111. O. Kirstein, M. Prager, M. R. Johnson, and S. F. Parker, “Lattice dynamics and methyl rotational excitations of 2-butyne,” J. Chem. Phys., 117, 1313, 2002.

112. L.L. Espinoza, A. Toro, and P. Fuentealba, “Quadrupole moments, dipole quadrupole A and quadrupole C polarizabilities by means of perturbation theory,” Int. J. Quantum Chem. 16, 939, 1979.

113. R.C. Haddon, X. Chi, M.E. Itkis, J.E. Anthony, D.L. Eaton, T. Siegrist, C.C. Mattheus, and T.T.M. Palstra, “Band Electronic Structure of One- and Two-Dimensional Pentacene Molecular Crystals,” J. Phys. Chem. B, 106, 8288, 2002.

114. F. Rondelez, W. Urbach, and H. Hervet, “Origin of Thermal Conductivity Anisotropy in Liquid Crystalline Phases,” Phys. Rev. Lett., 41, 1058, 1978.

115. J. Epstein, W.-L Ong, C. Bettinger, and J.A. Malen, “Temperature Dependent Thermal Conductivity and Thermal Interface Resistance of Pentacene Thin Films with Varying Morphology,” ACS Appl. Mater. Interfaces, 8, 19168-19174, 2016.

116. R.J. Kline, S.D. Hudson, X. Zhang, D.J. Gundlach, A.J. Moad, O.D. Jurchescu, T.N. Jackson, S. Subramanian, J.E. Anthony, M.F. Toney, and L.J. Richter, “Controlling the Microstructure of Solution-Processable Small Molecules in Thin-Film Transistors through Substrate Chemistry,” Chem. Mater., 23, 1194-1203, 2011.

117. F. Reisdorffer, B. Garnier, N. Horny, C. Renaud, M. Chirtoc, and T.-P Nguyen, “Thermal conductivity of organic semi-conducting materials using 3omega and photothermal radiometry techniques,” EPJ Web of Conferences, 79, 02001, 2014.

118. X. Wang, K.D. Parrish, J.A. Malen, and P.K.L. Chan, “Modifying the thermal conductivity of small molecule organic semiconductor thin films with metal nanoparticles,” Scientific Reports, 5, 16095, 2015.

119. T. Borca-Tasciuc, A.R. Kumar, and G. Chen, “Data reduction in 3ω method for thin-film thermal conductivity determination,” Rev. Sci. Inst., 72, 2139-2147, 2001.

120. E.T. Swartz and R.O. Pohl, “Thermal resistance at interfaces,” Appl. Phys. Lett., 51, 2200-2202, 1987.

121. R.J. Stoner and H.J. Maris, “Kapitza conductance and heat flow between solids at temperatures from 50 to 300 K,” Phys. Rev. B, 48, 16373-16387, 1993.

65

122. H.-C. Chieh, D.-J. Yao, M.-J. Huang, and T.-Y. Chang, “Thermal conductivity measurement and interface thermal resistance estimation using SiO2 thin film,” Rev. Sci. Inst., 79, 054902, 2008.

123. Y. Jin, Y. Yadav, K. Sun, H. Sun, K.P. Pipe, and M. Shtein, “Thermal boundary resistance of copper phthalocyanine-metal interface ,” Appl. Phys. Lett., 98, 093305, 2011.

124. https://www.dowcorning.com/content/publishedlit/11-1712-01.pdf

66

VITA

Yulong Yao

EDUCATION

B.S. Physics………………………………………………..............................…. July 2011 Harbin Institute of Technology, Harbin, Heilongjiang, China

PROFESSIONAL POSITIONS HELD

Graduate Research Assistant………………………...…………….…... Jan 2014 – present University of Kentucky, Lexington, KY Teaching Assistant…………………………………………………. Aug 2011 – Dec 2013 University of Kentucky, Lexington, KY

PUBLICATIONS

Y. Yao, Maryam Shahi, Marcia M. Payne, J.E. Anthony, and J.W. Brill, “Thermal Resistances of Thin Films of Small Molecule Semiconductors”, J. Mater. Chem. C, 2016, 4, 8817-8821. Zaifang Li, Hengda Sun, Ching-Lien Hsiao, Yulong Yao, Maryam Shahi, Alex Cruce, Xianjie Liu, Youyu Jiang, Wei Meng, Fei Qin, Thomas Ederth, Simone Fabiano, Weimin M. Chen, Jens Birch, Joseph W. Brill, Yinhua Zhou, Xavier Crispin and Fengling Zhang, “High power density free-standing PEDOT:PSS thermoelectric generator”, submitted to Energy Environ. Sci. in March, 2017. Malti, J. Edberg, H. Granberg, Z.U. Khan, J. W. Andreasen, X. Liu, D. Zao, H. Zhang, Y. Yao, J.W. Brill, I. Engquist, M. Fahlman, L. Wagberg, X. Crispin, and M. Berggren, “An Organic Mixed Ion-Electron Conductor for Power Electronics”, Adv. Sci., 2015, 1500305. J.W. Brill, Maryam Shahi, Marcia M. Payne, Jesper Edberg, Y. Yao, Xavier Crispin, and J.E. Anthony," Frequency-Dependent Photothermal Measurement of Transverse Thermal Diffusivity of Organic Semiconductors", J. Appl. Phys., 2015, 118. 233501. H. Zhang, Y. Yao, Marcia M. Payne, J. E. Anthony, and J. W. Brill, "Thermal Diffusivities of Functionalized Pentacene Semiconductors", App. Phys. Lett., 2014, 105, 073302.

67

SOCIETY MEMBERSHIPS

American Physical Society Materials Research Society