2012-Small Molecule Semiconductors for High-efficiency Organic Photovoltaics
THERMAL CONDUCTIVITIES OF ORGANIC SEMICONDUCTORS
Transcript of THERMAL CONDUCTIVITIES OF ORGANIC SEMICONDUCTORS
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, [email protected] 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 [email protected].
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