Structure-Function Relationships in Organic Charge-Transfer

STRUCTURE-FUNCTION RELATIONSHIPS IN ORGANIC CHARGE-TRANSFER

COMPLEXES

KATELYN PATRICIA GOETZ

A Dissertation Submitted to the Graduate Faculty of

WAKE FOREST UNIVERSITY GRADUATE SCHOOL OF ARTS AND SCIENCES

in Partial Fulﬁllment of the Requirements

for the Degree of

DOCTOR OF PHILOSOPHY

Physics

May 2016

Winston-Salem, North Carolina

Approved By:

Oana D. Jurchescu, Ph.D., Advisor

Laurie E. McNeil, Ph.D., Chair

Natalie A. W. Holzwarth, Ph.D.

George E. Matthews, Ph.D.

Richard T. Williams, Ph.D. Acknowledgments

Many people have been invaluable in the completion of my two Wake Forest University degrees. First and foremost, I would like to thank my advisor, Dr. Oana Jurchescu. She has been a great mentor in my academic career, and was always available to talk about science and life. I do not believe I would have had the opportunities I had during my graduate career if I had pursued it elsewhere. Thanks also to the entire Wake Forest Physics Department for inspiring and teaching me, and also for being living examples of what a good research career is and can be. I would especially like to thank Dr. Keith Bonin for (possibly unknowingly) providing the push I needed to major in physics by teaching an enthusiastic and interesting electronics class. Dr. Eric Carlson also deserves a thank you for organizing the society of physics students chapter in our department. I have very much enjoyed participating in science outreach through demo days at Sci Works and work with K-12 classrooms in the area. Eric Chapman has been invaluable throughout my nine years in the department for technical help and good conversations. The numerous people who have directly advanced my dissertation research through academic collaboration also have my gratitude. These include all members of the Ju- rchescu group from 2009-2016, but especially Jack Owen for being a good friend and fellow student. Additionally, my fellow graduate students Dr. Jeremy Ward, Dr. Yaochuan Mei, and Dr. Pete Diemer have set good examples as creative researchers and learners. Zach Lamport and Andrew Zeidell, later group members, have also been inspiring through good conversations, academic or otherwise. The papers I co- authored would not have been possible without the work of UNC researchers Dr. Laurie McNeil and Dr. Derek Vermeulen and the research group of Dr. Veaceslav Coropceanu at Georgia Tech. I would also like to thank the crystal growth group of Christian Kloc at Nanyang Technological University in Singapore for hosting and teaching me through the National Science Foundation (NSF) East Asia and Paciﬁc Summer Institute. Dr. Tatsuo Hasegawa and Dr. Jun’ya Tsutsumi taught me quite a bit about optical spectroscopy and alternative methods of device fabrication at the National Institute of Advanced Industrial Science and Technology (AIST) in Tsukuba, Japan. Dr. Curt Richter, Dr. David Gundlach, and Dr. Sujitra Pookpanratana at the National Institute of Standards and Technologies have earned my thanks for their collaborative work and conversations. Many thanks to all others who have worked on papers and projects with me. Finally, I would like to thank fellow graduate student Alex Taylor for being an excellent study partner and even better boyfriend. I hope to continue having fun

ii with you for years to come. I would also like to thank my entire family, extended and immediate, for being an unending positive presence in my life. This includes Mom, Kathryn Goetz, Dad, Devon Goetz, and siblings, Kirsten, Maggie, Jake, and Jeﬀ. Even our arguments are forces for bettering each other. To my siblings, I am excited to see what we can do as adults.

iii Table of Contents

Acknowledgments ii

List of Figures vii

List of Tables viii

List of Abbreviations ix

Abstract x

Chapter 1 Introduction 1 1.1 Introduction ...... 2 1.2 General Characteristics of Charge-Transfer Complexes ...... 6 1.2.1 Donors and Acceptors ...... 6 1.2.2 Crystal Structure ...... 9 1.3 Electronic Properties of CT Complexes ...... 12 1.4 The Interplay Between Degree of Charge Transfer and Electrical Prop- erties ...... 19 1.5 Outlook and Outline of This Thesis ...... 24 References ...... 27

Chapter 2 Crystal Growth of Organic Charge-Transfer Complexes 40 2.1 Introduction ...... 41 2.2 Single Crystal Growth of Charge-Transfer Complexes ...... 42 2.2.1 Solution Growth ...... 42 2.2.2 Vapor Growth ...... 43 2.3 Crystals of Novel Organic Charge-Transfer Complexes ...... 46 2.4 Thin-Film Growth of Organic Charge-Transfer Complexes ...... 49 2.5 Summary ...... 50 References ...... 53

Chapter 3 Electrical Characterization of Organic Charge-Transfer Com- plexes 57 3.1 Introduction ...... 58 3.2 Space-Charge-Limited Current ...... 59 3.2.1 Low-Voltage Regime ...... 60

iv 3.2.2 Space-charge-limited Current ...... 61 3.3 Ambipolar Organic Field-Eﬀect Transistors ...... 63 3.4 Comparison of Mobility Calculation by SCLC and OFET Measure- ments for an Ambipolar CT Complex ...... 69 3.5 Charge-Transport in Novel Organic Charge-Transfer Complexes . . . 72 3.5.1 Perylene–TCNQ in Three D:A Stoichiometries ...... 72 3.5.2 Anthracene and Pyrene with Acceptor PDIF-CN2 ...... 75 3.6 Summary ...... 77 References ...... 79

Chapter 4 The Eﬀect of Librational Motion on Charge Transport in Stilbene–F4TCNQ 81 4.1 Introduction ...... 82 4.2 Methods ...... 84 4.2.1 Crystal Growth ...... 84 4.2.2 Structure Determination ...... 84 4.2.3 Sample Fabrication and Electrical Characterization ...... 85 4.2.4 IR and Raman Spectroscopy ...... 86 4.2.5 Computational Methodology ...... 87 4.3 Results ...... 88 4.3.1 Electrical Properties ...... 88 4.3.2 Crystal Structure and Electronic Structure Calculations . . . . 90 4.3.3 Structure and Thermodynamics ...... 91 4.4 Discussion ...... 98 4.5 Conclusions ...... 101 References ...... 102

Chapter 5 The Eﬀect of Polymorphism on Charge Transport in the Charge-Transfer Complex DBTTF-TCNQ 107 5.1 Introduction ...... 108 5.2 Experiment and Results ...... 111 5.2.1 Structural Characterization of DBTTF–TCNQ ...... 111 5.2.2 Degree of Charge Transfer ...... 116 5.2.3 Charge Transport in DBTTF–TCNQ Polymorphs ...... 120 5.3 Discussion ...... 123 5.4 Conclusions ...... 127 References ...... 128

Chapter 6 Summary 133

Chapter 7 Curriculum Vitae 135

v List of Figures

1.1 DBTTF-TCNQ SC-OFETs with various contact metals ...... 3 1.2 Examples of donors and acceptors ...... 7 1.3 Charge-transfer complex energetics ...... 9 1.4 π-stacking motifs in 1:1 CT complexes ...... 10 1.5 OFET device geometries ...... 14 1.6 Organic metal electrodes for organic CT complexes in OFETs . . . . 15 1.7 Organic metal electrodes for organic CT complexes in OFETs . . . . 18 1.8 TCNQ with bonds sensitive to charge-transfer ...... 19 1.9 Mixed and segregated-stack BEDT-TTF–TCNQ ...... 21 1.10 Conductivity versus Degree of Charge Transfer ...... 22

2.1 Color change upon CT complex formation ...... 43 2.2 Vapor Growth Furnaces ...... 44 2.3 Crystal growth of Perylene–TCNQ in multiple stoichiometries . . . . 45 2.4 Green Anthracene-PDIF-CN2 Crystals ...... 47 2.5 Anthracene-PDIF-CN2 preliminary crystal structure ...... 48 2.6 Pyrene-PDIF-CN2 preliminary crystal structure ...... 48 3.1 Metal-Insulator-Metal Device Structures ...... 59 3.2 Current-Voltage characteristics for a metal-insulator-metal structure . 63 3.3 OFET Device Structure ...... 64 3.4 OFET Channel ...... 66 3.5 Ambipolar OFET Operation ...... 68 3.6 Ambipolar DBTTF-TCNQ OFET ...... 69 3.7 Ambipolar DBTTF-TCNQ SCLC Curve ...... 70 3.8 Crystals and structures of the Perylene–TCNQ CT complex system . 73 3.9 Current-Voltage characteristics the Perylene–TCNQ crystal system . 74 3.10 Optical reﬂectance of Anthracene and Pyrene–PDIF-CN2 ...... 76 3.11 Transfer integrals for Anthracene–PDIF-CN2 ...... 77 3.12 The Transport Properties of Pyrene–PDIF-CN2 ...... 78

4.1 Skeletal structures and crystals of stilbene, F4TCNQ, and STB–F4TCNQ 83 4.2 Thermal ellipsoid plots for STB–F4TCNQ ...... 86 4.3 Current-Voltage Characteristics for two-contact measurements of STB– F4TCNQ ...... 88 4.4 Temperature dependent electrical characteristics of STB–F4TCNQ . . 89

vi 4.5 Temperature dependence of the STB–F4TCNQ lattice constants and charge carrier eﬀective masses...... 90 4.6 Depiction of librationl motion in trans-stilbene ...... 92 4.7 CT thermal ellipsoids at 143 K, with and without disorder ...... 93 4.8 Fourier diﬀerence maps above and below the transition for STB–F4TCNQ 94 4.9 Libration of the C7=C7a stilbene bond in STB–F4TCNQ ...... 95 4.10 Degree of charge transfer (q) as a function of temperature...... 97 4.11 Transfer integral dependence on the phonon displacements ...... 99 4.12 Temperature dependent electrical characteristics of related compounds 101

5.1 Crystalline polymorphs of DBTTF–TCNQ ...... 110 5.2 Selected area electron diﬀraction images for DBTTF–TCNQ polymorphs112 5.3 XPS for DBTTF–TCNQ polymorphs ...... 113 5.4 Polarized absorption spectra for DBTTF–TCNQ polymorphs . . . . . 114 5.5 Polarized IR spectra for DBTTF–TCNQ polymorphs ...... 116 5.6 Molecular orientation of DBTTF and TCNQ with respect to crystal surface ...... 117 5.7 Raman spectra for α and β-DBTTF–TCNQ ...... 118 5.8 Electrical characteristics of α-DBTTF–TCNQ ...... 121 5.9 Electrical characteristics of β-DBTTF–TCNQ ...... 122 5.10 UPS of DBTTF–TCNQ Polymorphs ...... 123 5.11 DFT calculations for DBTTF–TCNQ Polymorphs ...... 125

vii List of Tables

1.1 Properties of CT complexes used in OFETs ...... 16 1.2 CT complexes listed by q and σ ...... 23

2.1 Solution growth methods of CT complexes ...... 51 2.2 Physical vapor transport growth methods of CT complexes and Parent Compounds ...... 52

viii List of Abbreviations

Abbreviation Meaning HOMO Highest Occupied Molecular Orbital LUMO Lowest Unoccupied Molecular Orbital XRD X-ray Diﬀraction SCLC Space-charge-limited current OFET Organic Field-eﬀect Transistor µ Charge-carrier mobility q Degree of charge transfer (unless otherwise listed) CT Charge Transfer D Donor A Acceptor

ix Abstract

Organic charge-transfer complexes are ordered combinations of charge-donating (D) and charge-accepting (A) compounds. The proximity of the D and A units results in a partial degree of ionicity and a band structure that is a hybrid of the parent compounds, allowing the complex to exhibit interesting physical characteristics. These include metallicity, ambipolar semiconductivity, photoconductivity, ferroelectricity, and more. The focus of this work is to identify several structure-function relationships in charge-transfer complexes pertaining to organic field-effect transistors (OFETs). The materials of focus in these studies are perylene- 7,7,8,8-tetracyanoquinodimethane (perylene–TCNQ), dibenzotetrathiafulvalene–TCNQ (DBTTF–TCNQ), and stilbene– F4TCNQ. We found that the perylene–TCNQ crystallizes in three D:A ratios — 1:1, 2:1, and 3:1. Single crystal X-ray diffraction revealed that each compound exhibits a mixed-stack structure with slight differences in D/A overlap. Though the differences are small, the charge-transfer was found to increase with increasing amount of donor. Fabrication and measurement of OFETs revealed that the 1:1 crystal was n-type, the 2:1 was ambipolar, and the 3:1 was p-type. DBTTF-TCNQ grows in two 1:1 polymorphs, one of which, the α-polymorph, is known and the other, the β-polymorph, was grown here for the first time. We found that the former exhibits a degree of charge transfer of 0.5e, while the latter is nearly neutral. OFET measurements revealed that with the same device structure, the α-polymorph is ambipolar with electron-dominant transport, while the β-polymorph exhibits hole-dominant ambipolar transport. The investigation of electrical properties of stilbene–F4TCNQ revealed that in this complex, the transport is thermally activated above 235 K, and temperature independent at low temperatures. The transition correlates with a freezing-in of orientational disorder caused by a mobile moiety on the donor. Together, these results provide insight into the interplay between crystal structure, ionicity, and charge transport.

x Chapter 1

Introduction

The discovery of the organic metal TTF–TCNQ in 1973 led to an explosion of research conducted on organic charge-transfer complexes. While these materials have been studied intensely for several decades, the research was mostly aimed at the discovery of materials with high room-temperature conductivity or high-temperature superconductivity. Recently, attention has turned to technologically-relevant properties of charge-transfer complexes, such as ambipolar transport, metallicity, photoconductivity, ferroelectricity, or magnetoresistance. This introduction chapter reviews the growth, structure, and properties of charge-transfer complexes and underlines recent progress in their application in organic devices. Their prospects in future applications are discussed, as well as the challenges yet to be overcome to understand the fundamental parameters governing their operation.

This portion of the text is adapted from K. P. Goetz, D. Vermeuelen, M. E. Payne, C. Kloc, L. E. McNeil, and O. D. Jurchescu, ”Charge-transfer complexes: new perspectives on an old class of compounds”, J. Mater. Chem. C 2, 3065-3076 (2014)1

1 1.1 Introduction

Organic electronics is often regarded as a young field due to the recent emergence of plastic electronics into the consumer market. While the past decade has seen a surge of research attention towards these materials, they have been of interest to chemists and physicists alike for over a century.2 In 1906 for example, Pochettino discovered the photoconductive properties of anthracene,3 and in 1948 Eley discovered the semiconducting properties of phthalocyanines.4 There now exists a wealth of publications on organic materials whose building blocks range in size from small molecules to polymers and whose electrical properties range from insulating to superconduct- ing.5–12 Of particular interest are the properties and processing of small-molecule organic semiconductors for application in organic field-effect transistors (OFETs), organic photovoltaic cells (OPVs), and organic light emitting diodes (OLEDs). Ex- amples include rubrene, pentacene, tetrathiafulvalene, and their derivatives.5, 13–22 While the electronic characteristics of these monomolecular compounds have reached impressive values with hole mobility values greater than 10 cm2V−1s−1, electron mobility values greater than 1 cm2V−1s−1, and solar cell efficiencies exceeding 10%,18–27 good performance alone is inadequate to address a large number of consumer applications. One barrier to their use is that these devices are often hybrids employing one or more layers of inorganic material, most often as a metal contact. This is problematic because inorganic metals can cause inefficient injection and high contact resistances in devices. Though they are not used as frequently as their inorganic counterparts, organic metal electrodes have several advantages. First, they can be processed at relatively low temperatures, which makes them amenable to use with flexible substrates. Second, several studies have shown that charge injection efficiency is improved at the organic metal/organic interface as compared to the inorganic metal/organic semiconductor interface. For example, the use of the organic metal TTF–TCNQ (tetrathiafulvalene – 7,7,8,8-tetracyanoquinodimethane) was shown to

2 cause a drastic improvement in electron mobility in an organic semiconductor compared to devices with either silver or gold contacts, even though the conductivity of the organic metal is about three orders of magnitude lower than that of either inorganic metal (Fig. 1.1).28 It was further shown that devices based on organic contacts not only exhibit superior electrical properties such as lower contact resistance and enhanced charge carrier mobilities, but the use of organic electrodes allows for fine tunability of the metal Fermi energy as a result of the chemical versatility of the organic molecules.29–34 These results suggest that manipulating organic/organic interfaces is a robust way to improve device performance, as exemplified by the current research efforts aimed towards the discovery of metallic organics.35 Examples include poly(3,4-ethylenedioxythiophene) poly(styrenesulfonate) (PEDOT:PSS), carbon nanotubes, and functionalized graphene.36–44 Besides modest device performance compared to inorganic semiconductors, another limitation suffered by most monomolecular compounds studied to date is that a majority of them are unipolar semiconductors, and in most cases they fall into the

Figure 1.1: Transfer characteristics of DBTTF-TCNQ SC-OFETs with the source and drain electrodes composed of Au, Ag, and TTF-TCNQ. (Taken from Takahashi, et al.28)

3 subset of p-type, or hole transporting, materials.6 This property is partially driven by extrinsic factors, such as available and applicable contact metals or trapping of one type of charge carrier due to structural imperfections.45, 46 Ambipolar transport, which implies simultaneous electron and hole transport, is relevant because it enables the engineering of low-power complementary metal-oxide semiconductor (CMOS)-like logic circuits. To ﬁll in these voids — i.e. to create the possibility for all-organic high- performance devices and to broaden the applicability of organic semiconductors — many research groups have recently focused their attention on a class of compounds that has been studied for several decades: organic charge-transfer (CT) compounds. Charge-transfer complexes are combinations of charge-donating (D) and charge- accepting (A) materials. While the parent compounds tend to be unipolar semiconductors, the CT complex can have entirely diﬀerent properties: it can be an ambipolar semiconductor, a metal, or even a superconductor.2, 47–54 Famous examples include

TTF–TCNQ, the ﬁrst organic metal discovered in 1973, and ((TMSTF))2–PF6, the organic superconductor made from donor tetramethyltetraselenafulvalene and acceptor hexaﬂuorophosphate discovered in 1980.55, 56 (Note that throughout this report, all CT complexes will be named as donor–acceptor.) Historically, these compounds created excitement because of predictions of high-temperature superconductivity or high conductivity. In addition, their crystal structures provide an excellent platform for theorists to verify ideas on one-dimensional systems, as these compounds can display quasi-one-dimensional structures with interesting phenomena, including charge density waves or Peierls transitions.2, 57 Several recent reports have caused a renewal of interest in CT complexes. These focus on interesting properties that arise at the interface between donor and acceptor single crystals that are not present in the individual parent compounds. For example, a study by Alves, et al., shows that metallic conduction is exhibited at the interface of a TCNQ single crystal and a TTF single crystal.58 A follow-up study by Mathis,

4 et al., suggested that this phenomenon occurs as a result of the adsorption of one compound into the other and the subsequent formation of the CT salt TTF–TCNQ at the interface rather than due to the physical contact between the D and A single crystals.59 Additional interface studies highlight the potential for CT materials to be

employed in thermoelectric devices, as in the case of pentacene and F4TCNQ (here thin-films were explored as opposed to single crystals), as well as their use as photoconductive materials, as in the case of rubrene and TCNQ,.60, 61 Photoconductivity is another property exhibited by several CT crystals that is of tremendous technological interest. Tsutsumi, et al., demonstrated that several CT complexes yield hole and electron photocurrents and the diffusion length of the photocurrent is directly dependent on the size of the CT gap energy.62, 63 Of further interest is the potential for ferroelectric behavior in CT compounds. In the past, this was only demonstrated in three different TTF-based CT compounds at cryogenic temperatures; recently, however, ferroelectricity was demonstrated in supramolecular CT complex networks under ambient conditions, as well as in CT complexes that behave as quantum magnets.64–67 Other exotic properties recorded in CT systems include magnetoconductance, field emission, photo-switching, and memory functions.68–72 They are thus an interesting testing medium for fundamental concepts that are important to organic electronics, such as charge transport and structure-function relationships. The novelty in the more recent research on CT complexes is the attempt to characterize them with applications in mind, such as employing them as contact materials in various types of organic devices or incorporating existing or novel CT complexes as the active metallic, semiconducting, or insulating layer in devices.

5 1.2 General Characteristics of Charge-Transfer Com- plexes

1.2.1 Donors and Acceptors

The choice of donor and acceptor and their organization within the CT crystal determine the electronic properties of the complex. The electronic coupling between the highest occupied molecular orbital (HOMO) of the donor and the lowest unoccupied molecular orbital (LUMO) of the acceptor molecules yields a partial degree of charge transfer (q) between the two, which results in a ground state characterized by partial ionicity, Dq+Aq−.73 (Note that in the literature this value is referred to as Z or ρ; here we use q to avoid confusion when referring to electrical resistivity later in this dissertation.) The value of q is dictated ﬁrst and foremost by the ionization potential of the donor (ID) and the electron aﬃnity of the acceptor (EA). Secondly, it is dictated by the proximity of the D and A units within the crystal structure of the

CT complex, manifested in the Madelung constant. This is given by EM , and reﬂects all of the electrostatic interactions of the long-range crystal structure. As discussed later in this dissertation, changes in crystal structure while holding the components constant cause changes in q, which are in turn responsible for the observed transport properties. A neutral or quasi-neutral q is derived from a D:A combination characterized by ID − EA EM , and is characterized by q < 0.5. On the contrary, when

D:A combinations for ID − EA EM are used, the resulting q is much greater and a quasi-ionic (D+A−) or ionic charge transfer is obtained. Several examples of donors and acceptors are listed in Fig. 1.2 with their chemical structures and abbreviations. Aside from the relative energetics of the D and A units, use of planar molecules is expected to enhance charge transfer.47 While each of these characteristics is a strong identiﬁer for a good choice of donor or acceptor, they are not entirely suﬃcient to complete the picture either on their own or together. A relation among the crystal structure, the degree of charge transfer, and the cor-

6 Figure 1.2: Skeletal structures for several materials typically used as donor or acceptor units. Donors are primarily aromatic hydrocarbons, while acceptors contain electron withdrawing groups such as CN, ﬂuorine, or oxygen.

7 responding electrical properties is not yet established. TCNQ and TCNE serve as excellent examples to highlight the limitations of the first quality. Initially synthe- sized in 1962, TCNQ is the most extensively studied of the acceptors employed in CT complexes.74 Its potential to form complexes was noted immediately; the synthesis paper highlighted its vivid blue color when reduced to a radical-anion, and researchers immediately followed up by employing it in CT complexes.74, 75 Table I in the paper by Melby, et al., notes the association constant for complexes of three different donors each with TCNE and TCNQ; in two cases, this value is higher for TCNE, and in one case, this value is higher for TCNQ, underlining the fact that there is more to consider than just the electron affinity when searching for a good donor-acceptor pair.75 As for planarity, several non-planar donors and/or acceptors have been shown to form CT complexes, including fullerenes and related compounds, so this is not always a limitation.76 What is important, however, is that the moieties of the donor and acceptors responsible for charge transport are not sterically hindered, i.e. bulky side groups do disrupt transport. One of the most interesting concepts related to donor and acceptor choice is the potential for band engineering. An illustration of this is shown in Fig. 1.3. Here the donor and acceptor HOMO-LUMO energy levels are depicted in blue and red respec-

tively. The ID and EA are also shown. The CT complex will have a band structure (shown in yellow) diﬀerent from, but related to, those of the parent compounds. This is easily observed when crystallizing a charge-transfer complex, as its formation is accompanied by a vivid color change. Several theoretical studies have indicated that the donor HOMO will have a larger contribution to the HOMO of the CT, whereas the CT LUMO correlates to the acceptor LUMO.77, 78 To exemplify this we included the charge-densities corresponding to the conduction and valence bands of tetracene– TCNQ in Fig. 1.3.77 The charge density of the conduction band is centered mostly at the acceptor molecule and has the shape of the LUMO charge-density of a single

8 Figure 1.3: The donor band structure (blue), acceptor band structure (red), and an approximation of the CT band structure (yellow). The charge densities corresponding to the HOMO and the LUMO bands for the tetracene–TCNQ CT complex are included (adapted from Shokaryev, et al.77).

TCNQ molecule. Similarly, the valence band charge density is centered on the donor molecule and has the shape of the HOMO of the tetracene molecule. This ﬁnding can potentially enable researchers to tune a material’s properties depending on the application where it is to be incorporated. As we will show later in this dissertation, other factors, such as crystal structure, play an important role in determining the electronic characteristics of the complex. To ﬁrst order approximation, though, knowledge of the parent compound energetics can allow CT materials to be designed to be metals, ambipolar semiconductors, unipolar semiconductors, or insulators.

1.2.2 Crystal Structure

For a CT complex given by D(n):A(m), n and m are integers and can have any value. For the 1:1 stoichiometry, two types of crystal structure predominate, as shown in Fig. 1.4. The ﬁrst, depicted in Fig. 1.4a, is the mixed-stack geometry. Here, the donor and acceptor units alternate in the π-stacking direction as ...A-D-A-

9 Figure 1.4: Crystalline packing of 1:1 charge-transfer compounds for (a) mixed stack and (b) segregated stack materials.

D...; examples of this crystal structure include Anthracene–PMDA, DBTTF–TCNQ, pyrene–TCNQ, perylene–TCNQ, and many more.2, 79–82 The second, depicted in Fig. 1.4b, is the segregated stack geometry. In this case, the donor and acceptor are each π-stacked separately as ...A-A-A-A... and ...D-D-D-D...; an example of this is the CT salt TTF–TCNQ.83 Of these two cases, the mixed-stack geometry occurs more frequently. Interestingly, several D-A pairs can be grown in both stacking geometries. This is the case for TMTSF–TCNQ, which forms a black segregated-stack metal or a red mixed-stack semiconductor.84, 85 Though there are generalities for each type of stacking geometry — for example, segregated stacking is a necessity for (but not an indicator of) room-temperature metallic conductivity — further study is needed to elucidate any speciﬁc relationship between the structure of the crystal and the electronic function. There are a few possible methods to evaluate this concept. First, the degree of tilt of the donor-acceptor pairs with respect to the stacking axis is likely

10 to be correlated with both the degree of charge transfer and the conductivity of the sample, as it is likely correlated with the coupling between the molecular orbitals of the molecules. Second, the overlap between the charge-donating and charge-accepting moieties of the parent compounds is also likely correlated with the degree of charge transfer between donor and acceptor pairs. These methods and other empirical ones have been suggested by Herbstein.47 More complex ratios, such as 2:1, 3:1, and 3:2, have been crystallized and studied, although Herbsteins count in 2005 suggested that the 1:2 and 2:1 ratios make up roughly 5% of all known CT compounds, while other ratios are even less common.47 In these cases, extra donor or acceptor molecules have been shown to exist either within the stack (and, as a result, participate in the charge transfer process), or to lie within the interstitial space between stacks. This is the case with the Perylene– TCNQ system, which grows in 1:1, 2:1, and 3:1 stoichiometries.79, 86, 87 Here, the 1:1 ratio is a classic example of mixed stacking. In the case of the 2:1 ratio, an additional perylene molecule lies in the interstitial space between stacks. This molecule does not participate in the charge transfer. The 3:1 complex is simply the 2:1 with an additional perylene molecule in the stack axis, such that it alternates D-D-A-D-D-A. The growth and electrical characterization of this CT complex system will be detailed further in Chapters 2 and 3. Because the crystal structure determines the ultimate electronic properties of a material, routes towards control of this feature are an important area of research. As will be discussed in chapters 2 and 5 of this dissertation, polymorphism and stoichiometry can be controlled by growth temperature and solvent choice. Interestingly, Black, et al., have identiﬁed hydrogen bonding as a tool towards design control of organic charge-transfer complexes. Here, knowing the nature of the hydrogen bonds between the donor and acceptor units ﬁxes stoichiometry and stack geometry, and allows for tuning of the properties simply by tuning the size of the donor or acceptor

11 molecular core.88, 89 Finally, control over crystal structure may be achieved through the use of alkyl-chains that promote solubility in either the donor or acceptor unit. The solubilizing moieties exist at the periphery of the molecule, and do not participate in charge transfer or transport. They can, however, tune the separation of the donor and acceptor molecules and create two-dimensional layered stacks that promote charge-transport.90

1.3 Electronic Properties of CT Complexes

CT complexes vary in band gap from insulators to metals, and can even be superconductors. Their temperature-dependent electrical properties are fundamentally interesting, as several CT complexes exhibit such phenomena as metal-to-insulator transitions, charge-density waves, and Peierls transitions; however, these primarily occur at very low temperatures, well below the operating regimes of most applicable devices. Here, we will limit the discussion to their room-temperature electrical properties. Their electrical conductivities are measured from several device geometries (often by simply applying contacts perpendicular to the axis of measurement), and range from less than 10−12 Ω−1cm−1 for materials such as pyrene–TCNQ to approximately 2000 Ω−1cm−1 for HMTSF–TCNQ at room temperature.75, 91 It should be noted that the one-dimensional nature of the CT complex crystal structures results in highly anisotropic electrical parameters.78 In TTF–TCNQ, this manifests itself as a decrease from up to 650 Ω−1cm−1 in the direction parallel to the stacks to 1 Ω−1cm−1 in the direction perpendicular to the stacks.55 While high conductivity is important for metals, a ﬁgure of merit for evaluating the performance of semiconductors is the mobility of holes and electrons as it is directly proportional to the transconductance, or switching capability, of the material. Experimentally, the mobility in CT complexes was evaluated using a wide range of methods, however, we will limit the discussions

12 in this report to the three most common methods — space-charge-limited current (SCLC), time-of-flight (TOF), and organic field-effect transistor (OFET) measurements. SCLC measurements make use of two contacts, and are useful for measuring the bulk properties of unipolar compounds (in the case of ambipolar compounds, the mobility will be a combination of hole and electron mobilites, and include a factor for recombination of carriers).2 Coronene–TCNQ was studied by this method, and was determined to have a mobility of 0.3 cm2V−1s−1.92 TOF is another two-contact measurement, but can be used to extract a charge-carrier-specific mobility, and can therefore be used on ambipolar materials. Anthracene–PMDA was studied by this method, and was found to have an electron mobility of 0.15 cm2V−1s−1 parallel to the stacking direction.93 With the increased interest towards technologically-relevant properties of CT complexes, OFET measurements have emerged as a powerful tool to characterize such materials. They also offer a different, comparative look at electronic properties of CT salts, as they are much more dependent on surface properties than are the bulk- dependent SCLC and TOF measurements. A study by Hasegawa, et al., was one of the first to examine the use of a CT complex as the active layer in an OFET.94 Here,

BEDT-TTF–F2TCNQ (where BEDT-TTF is bis(ethylenedithio)tetrathiafulvalene) was contacted with gold top contacts and a parylene-C top-gate dielectric and exhibited hole and electron mobilities on the order of 10−3 at 40 K. More recent examples of studies focused on OFETs based on CTs include the evaluation of microcrystals nucleated from solution directly on the substrate; the materials of interest here are based on the donor sulfur-bridged annulene and the acceptors TCNQ, DTTCNQ, and fullerene. All were reported to exhibit ambipolar transport with a bottom-gate

(SiO2 or SiO2/OTS gate dielectric), top-contact (Au) structure, with the DTTCNQ- based crystals achieving the highest performance at 0.54 cm2V−1s−1 for holes and 0.19 cm2V−1s−1 for electrons.76, 95, 96 We note that the OFET device structure, of

13 which there are four (Fig. 1.5), can suppress the charge-carrier mobility of both charge carriers. Both charge carriers are aﬀected by the use of top-contacts, as the evaporation of metals directly onto the organic semiconductor can cause damage at the interface. This, however, is the only option when using organic metal electrodes, as their rough evaporated surface is too rough for use as bottom-contacts. Dielectric choice particularly aﬀects electron transport — use of SiO2 reduces electron mobility, but can be passivated by the use of a self-assembled monolayer (SAM) such as octyltrichlorosilane (OTS). This and the energetic relationship between the contacts and active layer makes comparison between materials tricky. All four geometries have been used to study CT complexes.

Figure 1.5: The four OFET device geometries: Bottom gate, bottom contact; Bot- tom gate, top contact; Top gate, top contact; and Top gate, bottom contact.

DBTTF–TCNQ in particular has received quite a bit of recent research interest. In 2005 it was reported to have an electron mobility of 1 cm2V−1s−1 when contacted with TTF–TCNQ source and drain electrodes in an OFET conﬁguration.28 When contacted with gold electrodes, DBTTF–TCNQ was reported to be ambipolar, with an electron mobility of 0.13 cm2V−1s−1 and a hole mobility of 0.04 cm2V−1s−1.97 One study in particular both highlights the ability to engineer electrical properties in CT complexes, and leads into an interesting discussion on metallic behavior. Here, DBTTF–TCNQ was contacted with six types of organic metal — TTF–TCNQ, TTF–

14 Figure 1.6: Transfer characteristics of DBTTF–TCNQ single-crystal field effect transistors with source and drain electrodes composed of (a) TTF–TCNQ, (b) TTF–F1TCNQ, (c) TTF–F2TCNQ, (d) TSF–F1TCNQ, (e) TSF–F2TCNQ, and (f) 29 DBTTF–F4TCNQ. This figure is taken from Takahashi, et al.

29 F1TCNQ, TTF–F2TCNQ, TSF–F1TCNQ, TSF–F2TCNQ, and DBTTF–F4TCNQ. As shown in Fig. 1.6, this enabled the electrical properties of the same organic semiconductor to be tuned from entirely n-type to ambipolar to p-type by shifting the metal Fermi energy with respect to the semiconductor band energy. This parameter was fine-tuned over a broad range, from the conduction-band edge of the semiconductor to its valence-band edge by varying the type and stoichiometry of the D and A species forming the CT. The resulting device mobility was higher for the cases when an ohmic contact was established (unipolar device operation, Fig. 1.6a, f) and was diminished when ambipolar transport was reached because of the inherent charge injection inefficiency correlated with the location of the metal Fermi level in the middle of the semiconductor band gap. This study simultaneously highlights the contact- dependent properties of semiconducting materials in OFET devices and the potential for tunability of Fermi levels in organic metals. This finding is of critical importance in organic device research, as establishing an ohmic contact to organic semiconductors has proven a very challenging task.6, 35, 46, 98, 99 In summary, the OFET mobilities of several CT complexes are summarized in Table 1.1. This list is likely not complete, but includes many examples. Comparison between the injection efficiencies at the organic metal/organic semi-

15 2 −1 −1 2 −1 −1 CT Complex Structure, Dielectric, Contacts µe(cm V s ) µh(cm V s ) 95 DPTTA–TCNQ BGTC, SiO2/OTS, Au 0.03 0.04 76 DPTTA–C60 BGTC, SiO2, Au 0.01 0.3 76 DPTTA–C70 BGTC, SiO2, Au 0.05 0.07 96 DPTTA–DTTCNQ BGTC, SiO2,Au 0.19 0.54 DBTTF–TCNQ28 TGTC, Parylene-C, TTF-TCNQ 1.0 — DBTTF-TCNQ100 TGBC, Parylene-N, Au 0.5 0.05 90 −4 −4 diC8-BTBT–TCNQ BGBC, Cytop, Au 4 x 10 4 x 10 16 90 diC8-BTBT–F2TCNQ BGBC, Cytop, Au 0.4 — 90 diC8-BTBT–F4TCNQ BGBC, Cytop, Au 0.1 — Anthracene–TCNQ101 TGTC, Parylene-C, TTF-TCNQ 4.5 x 10−5 — 4M-DSB–CN-TFPA102 BGTC, PMMA, Au 6.7 x 10−2 6.7 x 10−3

Table 1.1: Properties of charge-transfer complexes used as the active layer in organic field-effect transistors. The material is listed with its device structure, electron mobility, and hole mobility. Abbreviations are: BGTC is bottom gate, top contact, BGBC is bottom gate, bottom contact, TGTC is top gate, top contact, TGBC is top gate, bottom contact, as shown in Fig. 1.5. DPTTA is meso-diphenyl tetrathia[22]annulene[2,1,2,1], diC8-BTBT is 2,7-dialkyl[1]benzothieno[3,2-b][1]benzothiophene. conductor interface versus the inorganic metal/organic semiconductor interface were not only done in combination with CT semiconductors, but also for the case of monomolecular crystals. When used to contact pentacene thin films, the contact resistance of both top and bottom contacts was reported to be comparable to that of top contact gold — about 10 kΩ cm — whereas the bottom-contact gold on pentacene was reported to be 10 kΩ cm in the optimal case but on the order of 1000 kΩ cm in the reproducible case.30, 103 A further method for improving injection, the so-called “self-contact” method, takes advantage of the metallic conducting layer formed at the interface of donor and acceptor crystals,58 as well as the possibility that this property arises from diffusion of one layer into the other.59 Here, devices are designed by evaporation of an acceptor molecule through a shadow mask onto the already-formed donor molecule thin-film. A post-evaporation annealing step allows for formation of the CT salt at the areas where the acceptor contacts the donor, creating charge-transfer metal electrodes with the donor molecule used as the active semiconducting layer. This method has proved successful for donors TMTTF and HMTTF with the acceptor TCNQ, where devices exhibited low contact resistances and high hole mobilities of 0.5 cm2V−1s−1 and 1 cm2V−1s−1 respectively.104, 105 The self-contacting of HMTTF with TCNQ is shown in Fig. 1.7 for both top (Fig. 1.7a) and bottom (1.7b) contact geometries, as well as for an all-organic device (Fig. 1.7c).104 The versatility of this method is a topic of great research interest, and it has recently proven possible to self-contact devices by solution deposition methods.106 Similar to the small-molecule CT complexes outlined in this review, the poly- meric semiconductors can be blended to achieve charge transfer between a D and A species and tune physical properties by controlling this transfer. When compared to conducting polymers such as PEDOT:PSS, CT complexes can exhibit the same order of magnitude of conductivity (1000 Ω−1cm−1), though the polymer exhibits a

17 Figure 1.7: Self-contacting of HMTTF with TCNQ allows for organic metal HMTTF-TCNQ to contact p-type active layer HMTTF. This is possible for (a) top- contact OFETs, (b) bottom-contact OFETs, and (c) all-organic devices. This ﬁgure is taken from Tamura, et al.104

lower sheet resistance in the extreme case (on the order of 100 Ω sq−1, whereas TTF– TCNQ was reported to have a sheet resistance of 4000 Ω sq−1).31, 103 The advantage of polymers is their solution processability, which allows for reduced complexity in processing. For this reason, they have been extensively used as contact materials in molecular electronic devices, hole extraction layer (HEL) in organic photovoltaics, or source-drain electrodes in FETs.6, 38, 40, 107–109

18 1.4 The Interplay Between Degree of Charge Trans- fer and Electrical Properties

The degree of charge transfer from the donor to the acceptor is a defining feature of CT complexes. Numerous physical properties are dependent on this parameter — the lattice energy, conductivity, nature of a possible Peierls transition, and the type of ESR spectrum are a few that are well known.47 Unfortunately, at the moment this value is only determinable within a wide margin of error (roughly 0.1e) for compounds containing only a few types of donor or acceptor; this is most often TCNQ, but methods for determination of degree of charge transfer in TMPD and TTF containing compounds have been discussed as well.47, 84, 110, 111 Two methods have been mostly employed to determine the degree of charge transfer. The first is diffraction — this can be X-ray, diffuse X-ray or neutron scattering. Here, the bond lengths of molecules are known to change as a function of the degree of charge transfer. In the case of TCNQ, four specific bonds are measurably sensitive to the addition of charge. These are shown in Fig. 1.8. Four empirical relations that we are aware of have been reported for TCNQ, and one each for donors TTF and TMPD.47, 84, 110 The relation most commonly used (though there is not much reason to use one over the other) is shown in equation 1.1.

Figure 1.8: The TCNQ molecule is shown with bonds a, b, c, and d labeled. These are measurably sensitive to the addition of charge on TCNQ and used in the calculation of degree of charge transfer.

1 b − c d − c q = 1 − CT CT + 1 − CT CT (1.1) 2 bN − cN dN − cN

19 Here, bond lengths b, c, and d in the CT crystal (e.g. bCT ) are compared to the neutral case (N). In addition to evaluating bond length changes by diﬀraction, spectroscopic studies (e.g. Raman or IR spectroscopy) can be used. Here the dependence of speciﬁc stretching frequencies on moiety charge is assumed to be linear.111–113 In the case of

Raman spectroscopy, the frequency of the TCNQ ν4 mode, which is exocyclic C=C stretching (i.e. the aromatic ring containing bonds a and b in Fig. 1.8) occurs at 1454 cm−1 in neutral TCNQ and 1395 cm−1 in ionic TCNQ; thus, measuring the frequency of this mode in the CT complex of interest allows for calculation of q by equation 1.2.112

q = −0.0169 ∗ ν4 + 24.6 (1.2)

Similarly, shifts in the IR active C≡N stretching mode, ω0, can be used to calculate q via equation 1.3.111 ω 2227 q = − 0 + (1.3) 44 44

Inconsistencies noted between different proposed methods show that there is a clear need for both increased accuracy of determination of this value, as well as a need for methods to evaluate a wider array of compounds. A review by Torrance has suggested that the Coulombic interactions within the crystal account for the broad range of electrical properties reported in CT complexes. In this band picture, large Coulombic energies compared to the bandwidth of the solid localize the charges and a Mott insulator state is obtained.49, 114 When this energy is on the order of the bandwidth, the complex can become metallic. Later the same author identified the dependence of conductivity in organic semiconductors and metals on the reduction potential of the cation. Additional theoretical calculations, however, have shown that the relationship between charge transfer and the donor ionization potential and acceptor electron affinity is non-monotonic.115 The example of BEDT-TTF–TCNQ highlights the problems with this type of categorization further.

20 Figure 1.9: Mixed and segregated-stack BEDT-TTF–TCNQ. The former is a semiconductor with a low degree of charge transfer and conductivity, while the latter is a metal with a mid-range degree of charge-transfer and metallic conductivity.

It is known to exhibit two 1:1 polymorphs — one that is a mixed-stack semiconductor and another that is a segregated-stack metal — but while the energetics of the donor and acceptor remain constant, the properties are remarkably different.84 The crystal structure, degree of charge transfer, and conductivity are shown for each polymorph of this complex in Fig. 1.9. For this reason, we plot the conductivity (σ) of D:A CT complexes as a function of the degree of charge transfer, q. This plot is shown in Fig. 1.10, with Table 1.2 listing the conductivities and degrees of charge transfer in line with the compound. In the case where the degree of charge transfer was not reported, we estimated it by the method of Flandrois, et al., using the crystal structure available in the literature. We observe that the value of q gives rise to large differences in the measured conductivities. Two distinct groups can be identified in this graph: at low and high values of q, the materials are electrically insulators or semiconductors (low conductivity, σ). At intermediate q, in a small window highlighted by a red circle and corresponding to 0.5 < q < 0.74, the CT complexes exhibit metallic properties. The exception to this is phenazine–TCNQ (W); here, the mixed-stack crystal structure (Fig. 1.4a) prohibits metallic conductivity, even if an optimal q is present. All

21 other crystals within this category show segregated stacking (Fig. 1.4b). Note that these metals convert to insulators at low temperatures via a metal insulator phase transition.

Figure 1.10: The conductivity of several CT complexes versus the degree of charge transfer. Values are tabulated in Table X. The red circle indicates metallic complexes (all exhibit segregated stacking), which lie between 0.5e < q < 0.75e.

CT Complex q σ (Ω−1cm−1)

A DTE–TCNQ 0.003(110, 116) 2 x 10−8(116)

B STB–TCNQ 0.063(110, 116 2 x 10−10(116)

C DTT–TCNQ 0.03(110, 116 1 x 10−6(116)

D TMS–TCNQ 0.075(110, 116 2 x 10−9(116)

E Anthracene–TCNQ 0.12(110, 117 1 x 10−11(75)

F Naphthalene–TCNQ 0.075(110, 118 3 x 10−12(119)

G BEDT-TTF–TCNQ (insulator) 0.2(120 1 x 10−6(120–123 )

H Coronene–TCNQ 0.3(92) 1 x 10−8(92)

22 CT Complex q σ (Ω−1cm−1)

I pyrene–TCNQ 0.3(80, 110) 1 x 10−12(75)

J BMDTP–TCNQ (insulator) 0.31(124) 3.7 x 10−6(124)

K Perylene–TCNQ 0.46(77) 1 x 10−6(75)

L DBTTF–TCNQ 0.47(125) 1 x 10−6(126)

M TMTSF–DMTCNQ 0.5(127) 1 x 102(128)

N BEDT-TTF–TCNQ 0.5(129) 1 x 101(121)

O TTF–TCNQ 0.59(130) 7 x 102(55)

P TMTSF–TCNQ (metal) 0.6(84) 1.2 x 103(131)

Q NMP–TCNQ 0.62(111) 3 x 102(132)

R TSF–TCNQ 0.63(133) 8 x 102(134)

S BMDTP–TCNQ (metal) 0.71(124) 1.1 x 102(124)

T TMTTF–TCNQ 0.71(84) 4 x 102(135)

U HMTTF–TCNQ 0.72(136) 4 x 102(137)

V HMTSF–TCNQ 0.74(132) 2 x 103(91)

W Phenazine–TCNQ 0.75(110, 138) 3 x 10−9(119)

(139) −5(139) X TMTSF–CF3TCNQ 1 1.2 x 10

(139) −6(139) Y TTF–CF3TCNQ 1 1.5 x 10

(139) −7(139) Z HMTTF–CF3TCNQ 1 1 x 10

(139) −8(139) AA TMTTF–CF3TCNQ 1 1 x 10

(139) −8(139) BB OMTTF–CF3TCNQ 1 1 x 10

(139) −8(139) CC TMPD–CF3TCNQ 1 1 x 10

(139) −8(139) DD Me2P–CF3TCNQ 1 1 x 10

Table 1.2: The compounds corresponding to Fig. 1.10, listed with their degree of charge transfer (q), conductivity (σ), and respective references. In the case that the crystal structure was reported but not a degree of charge transfer, it was calculated based on the method by Flandrois, et al.110

23 This plot highlights several important features of CT complexes. First, it is clear that segregated stacking is necessary for metallic conductivity. Second, an optimal ionicity correlates with metallic conductivity. This lies somewhere between completely neutral and completely ionized because both extremes localize available charge carriers. Simultaneously, this plot implies areas of need when classifying and understanding charge-transfer complexes. The plot is relatively scattered, primarily because of the empirical nature of the methods available to determine q and the sample-dependent nature of the conductivity. Other reasons for error in this plot may originate from the fact that we reported the maximum value for conductivity found in the literature, but depending on the crystal orientation (anisotropy in electrical properties), the type of crystal growth technique and unknown factors such as impurity concentration, a given CT complex can exhibit conductivities ranging over a few orders of magnitude. Unavoidable experimentation artifacts, such as contact resistance, also diminish the conductivity from the intrinsic value of the crystal. Therefore, in order to improve this plot and thus our understanding of the relationship between charge-transfer and properties, methods for accurately determining both q and σ need to be improved.

1.5 Outlook and Outline of This Thesis

In summary, CT complexes have been studied for several decades as a result of the rich physical properties that they display and the intriguing fundamental questions that they raise. With the emergence of organic electronics into consumer applications in more recent years, a large number of materials from this class have begun to be applied in devices such as ﬁeld-eﬀect transistors, and have demonstrated properties that cannot be provided by monomolecular compounds. They have tremendous

24 potential to serve as ambipolar semiconductors, and while their mobilities are cur- rently somewhat low (likely as a combination of processing limitations as well contact injection barriers), they provide a versatile approach to precisely control hole and electron injection and transport by band engineering. Metallic CT complexes have demonstrated their ability to serve as superior electrode materials, with tunable Fermi energies. More in-depth study of this concept and its application in conjunction with other organic semiconductors may provide a solution for one of the most difficult bottlenecks in organic device research: establishing efficient contacts. In addition, CT complexes are beginning to be of interest in emerging concepts, such as high temperature ferroelectric materials or biomaterials.52, 67 With regard to the latter, they may prove useful as sensors, and there is the additional concept that charge transfer is not only limited to the electron, but can also include proton transfer. This opens the door to investigation of a wide array of biologically relevant molecules, such as amino acids, and even DNA.52 Further progress is necessary in several areas. First, only a limited amount of information is available relating the structure of CT complexes to their properties. Further analysis of donor-acceptor overlap, stacking structure, spacing, and degree of charge transfer between the donor and acceptor molecules could yield much insight into the structure-function relationship of these materials and ultimately a rational design of CT complexes with electrical properties ranging from insulating to metallic. At the same time, development of less-complex processing methods for such materials is desired in order for them to be able to sustain low-cost fabrication of organic devices. With a few exceptions, most CT crystals are fabricated using methods which are only applicable for low-yield device production. Transition to methods that allow high-throughput deposition is urgently needed and much of the present research efforts are aimed towards that aim.21, 140, 141 The aim of this dissertation reflects some of the discussion in section 1.4 on the relationship between the degree of charge transfer and electrical properties. We have

25 grown single crystals of three diﬀerent crystal systems: Perylene–TCNQ, Stilbene–

F4TCNQ, and DBTTF–TCNQ. Prior to our discussion of these complexes, the general principles of crystal growth of CT complexes will be discussed, as well as considerations in measuring the electrical characteristics of ambipolar materials. We find that the Perylene–TCNQ system grows in three different D:A polymorphs — 1:1, 2:1, and 3:1, all displaying different transport properties and different degrees of charge transfer. The Stilbene–F4TCNQ complex grows in a 1:1 crystal and exhibits a temperature-dependent librational mode that affects the electrical characteristics and the degree of charge transfer. The DBTTF–TCNQ system is found to grow in two polymorphs, the first of which, the α-polymorph, is known and has been discussed extensively in this chapter. The second polymorph is previously unreported will be called the β-polymorph. Though these complexes are both mixed-stack ambipolar semiconductors in the same device structure, they exhibit slightly different transport properties and very different values of q. Together, these systems provide insight into the structure-function relationships in organic charge-transfer complexes.

26 References

[1] K. P. Goetz, D. Vermeulen, M. E. Payne, C. Kloc, L. E. McNeil, and O. D. Jurchescu. Charge-transfer complexes: new perspectives on an old class of compounds. Journal of Materials Chemistry C, 2:3065–3076, 2014.

[2] M. Pope and C. E. Swenberg. Electronic Processes in Organic Crystals and Polymers. Oxford University Press, New York, second edition, 1999.

[3] A. Pochettino. E. Acad. Lincei. Rend., 15:355, 1906.

[4] D. D. Eley. Phthalocyanines as semiconductors. Nature, 162:819, 1948.

[5] M. E. Gershenson, V. Podzorov, and A. F. Morpurgo. Colloquium: Electronic transport in single-crystal organic transistors. Reviews of Modern Physics, 78(3):973–989, 2006.

[6] H. Klauk. Organic thin-ﬁlm transistors. Chemical Society Reviews, 39:2643– 2666, 2010.

[7] H. Klauk, M. Halik, U. Zschieschang, G. Schmid, W. Radlik, and W. Weber. High-mobility polymer gate dielectric pentacene thin ﬁlm transistors. Journal of Applied Physics, 92(9):5259, 2002.

[8] E. J. Meijer, D. M. de Leeuw, S. Setayesh, E. van Veenendaal, B. H. Huisman, P. W. M. Blom, J. C. Hummelen, U. Scherf, J. Kadam, and T. M. Klapwijk. Solution-processed ambipolar organic ﬁeld-eﬀect transistors and inverters. Na- ture Materials, 2(10):678–682, 2003.

[9] Y. Kim, S. Cook, S. M. Tuladhar, S.A. Choulis, J. Nelson, J. R. Durrant, D. D. C. Bradley, M. Giles, I. McCulloch, C.-S. Ha, and M. Ree. A strong regioregularity effect in self-organizing conjugated polymer films and high- efficiency polythiophene:fullerene solar cells. Nature Materials, 5(3):197–203, 2006.

[10] Z. Bao. Materials and fabrication needs for low-cost organic transistor circuits. Advanced Materials, 12(3):227–230, 2000.

[11] A. F. Hebard, M. J. Rosseinsky, R. C. Haddon, D. W. Murphy, S. H. Glarum, T. T. M. Palstra, A. P. Ramirez, and A. R. Kortan. Superconductivity at 18 K in potassium-doped C60. Nature, 350:600–601, 1991.

27 [12] M. J. Rosseinsky, A. P. Ramirez, S. H. Glarum, D. W. Murphy, R. C. Haddon, A. F. Hebard, T. T. M. Palstra, A. R. Kortan, S. M. Zahurak, and A. V. Makhija. Superconductivity at 28 K in Rb₂C₆₀. Physical Review Letters, 66(21):2830–2832, 1991.

[13] V. Podzorov, S. E. Sysoev, E. Loginova, V. M. Pudalov, and M. E. Gershenson. Single-crystal organic ﬁeld eﬀect transistors with the hole mobility 8 cm[sup 2]/Vs. Applied Physics Letters, 83(17):3504, 2003.

[14] V. Podzorov, V. M. Pudalov, and M. E. Gershenson. Field-eﬀect transistors on rubrene single crystals with parylene gate insulator. Applied Physics Letters, 82(11):1739–1741, 2003.

[15] J. E. Anthony. Functionalized acenes and heteroacenes for organic electronics. Chemical Reviews, 106(12):5028–5048, 2006.

[16] H. Najafov, B. Lee, Q. Zhou, L. C. Feldman, and V. Podzorov. Observation of long-range exciton diﬀusion in highly ordered organic semiconductors. Nature Materials, 9(11):938–943, 2010.

[17] 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(5664):1644–6, 2004.

[18] O. D. Jurchescu, M. Popinciuc, B. J. van Wees, and T. T. M. Palstra. Interface- controlled, high-mobility organic transistors. Advanced Materials, 19(5):688– 692, 2007.

[19] M. Mas-Torrent, M. Durkut, P. Hadley, X. Ribas, and C. Rovira. High Mobility of Dithiophene-Tetrathiafulvalene Single-Crystal Organic Field Eﬀect Transis- tors. Journal of the American Chemical Society, 126(4):984–985, 2004.

[20] K. P. Goetz, Z. Li, J. W. Ward, C. Bougher, J. Rivnay, J. Smith, B. R. Con- rad, S. R. Parkin, T. D. Anthopoulos, A. Salleo, J. E. Anthony, and O. D. Jurchescu. Eﬀect of acene length on electronic properties in 5-, 6-, and 7-ringed heteroacenes. Advanced Materials, 23(32):3698–3703, 2011.

[21] Y. Mei, M. A. Loth, M. Payne, W. Zhang, J. Smith, C. S. Day, S.R. Parkin, M. Heeney, I. McCulloch, T. D. Anthopoulos, J.E. Anthony, and O. D. Ju- rchescu. High mobility ﬁeld-eﬀect transistors with versatile processing from a small-molecule organic semiconductor. Advanced Materials, 25(31):4352–4357, 2013.

[22] J. E. Anthony, A. Facchetti, M. Heeney, S. R. Marder, and X. Zhan. N-Type organic semiconductors in organic electronics. Advanced Materials, 22(34):3876– 3892, 2010.

28 [23] M. Mas-Torrent, P. Hadley, S.T. Bromley, X. Ribas, J. Tarrés, M. Mas, E. Molins, J. Veciana, and C. Rovira. Correlation between crystal structure and mobility in organic field-effect transistors based on single crystals of tetrathiafulvalene derivatives. Journal of the American Chemical Society, 126(27):8546–53, 2004. [24] E. Menard, V. Podzorov, S.-H. Hur, A. Gaur, M. E. Gershenson, and J. A. Rogers. High-performance n- and p-type single-crystal organic transistors with free-space gate dielectrics. Advanced Materials, 16(23-24):2097–2101, 2004. [25] J. Soeda, Y. Hirose, M. Yamagishi, A. Nakao, T. Uemura, K. Nakayama, M. Uno, Y. Nakazawa, K. Takimiya, and J. Takeya. Solution-crystallized organic field-effect transistors with charge-acceptor layers: high-mobility and low- threshold-voltage operation in air. Advanced Materials, 23(29):3309–14, 2011. [26] H. Minemawari, T. Yamada, H. Matsui, J. Tsutsumi, S. Haas, R. Chiba, R. Kumai, and T. Hasegawa. Inkjet printing of single-crystal films. Nature, 475(7356):364–367, 2011. [27] NREL. Best research-cell efficiencies. [28] Y. Takahashi, J. Hasegawa, Y. Abe, Y. Tokura, K. Nishimura, and G. Saito. Tuning of electron injections for n-type organic transistor based on charge- transfer compounds. Applied Physics Letters, 86(6):063504, 2005. [29] Y. Takahashi, T. Hasegawa, Y. Abe, Y. Tokura, and G. Saito. Organic metal electrodes for controlled p- and n-type carrier injections in organic field-effect transistors. Applied Physics Letters, 88(7):073504, 2006. [30] K. Shibata, H. Wada, K. Ishikawa, H. Takezoe, and T. Mori. (Tetrathiafulva- lene)(tetracyanoquinodimethane) as a low-contact-resistance electrode for organic transistors. Applied Physics Letters, 90(19):193509, 2007. [31] K. Shibata, K. Ishikawa, H. Takezoe, H. Wada, and T. Mori. Contact resistance of dibenzotetrathiafulvalene-based organic transistors with metal and organic electrodes. Applied Physics Letters, 92(2):023305, 2008. [32] X. Xian, K. Yan, W. Zhou, L. Jiao, Z. Wu, and Z. Liu. Unipolar p-type single- walled carbon nanotube field-effect transistors using TTF-TCNQ as the contact material. Nanotechnology, 20(50):505204, 2009. [33] M. Kraus, S. Richler, A. Opitz, W. Brütting,S. Haas, T. Hasegawa, A. Hin- derhofer, and F. Schreiber. High-mobility copper-phthalocyanine field-effect transistors with tetratetracontane passivation layer and organic metal contacts. Journal of Applied Physics, 107(9), 2010. [34] B. Mukherjee, M. Mukherjee, K. Sim, and S. Pyo. Solution processed, aligned arrays of TCNQ micro crystals for low-voltage organic phototransistor. Journal of Materials Chemistry, 21(6):1931, 2011.

29 [35] R. Pfattner, C. Rovira, and M. Mas-Torrent. Organic metal engineering for enhanced field-effect transistor performance. Phys. Chem. Chem. Phys., 17(40):26545–26552, 2015. [36] A. Pierre, M. Sadeghi, M. M. Payne, A. Facchetti, J. E. Anthony, and A. C. Arias. All-printed flexible organic transistors enabled by surface tension-guided blade coating. Advanced Materials, 26:5722–5727, 2014. [37] H. Rost, J. Ficker, J. S. Alonso, L. Leenders, and I. McCulloch. Air-stable all-polymer field-effect transistors with organic electrodes. Synthetic Metals, 145:83–85, 2004. [38] H. Sirringhaus, T. Kawase, R. H. Friend, T. Shimoda, M. Inbasekaran, W. Wu, and E. P. Woo. High-resolution inkjet printing of all-polymer transistor circuits. Science, 290:2123–2126, 2000. [39] P. G. Taylor, J.-K. Lee, A. A. Zakhidov, M. Chatzichristidi, H. H. Fong, J. A. DeFranco, G. G. Malliaras, and C. K. Ober. Orthogonal patterning of PE- DOT:PSS for organic electronics using hydrofluoroether solvents. Advanced Materials, 21(22):2314–2317, 2009. [40] Y. F. Lim, S. Lee, D. J. Herman, M. T. Lloyd, J. E. Anthony, and G. G. Malliaras. Spray-deposited poly(3,4-ethylenedioxythiophene): Poly(styrenesulfonate) top electrode for organic solar cells. Applied Physics Letters, 93(19), 2008. [41] I. Valitova, M. Amato, F. Mahvash, G. Cantele, A. Maffucci, C. Santato, R. Martel, and F. Cicoira. Carbon nanotube electrodes in organic transistors. Nanoscale, 5(11):4638–46, 2013. [42] G. Eda, Y. Y. Lin, C. Mattevi, H. Yamaguchi, H. A. Chen, I. S. Chen, C. W. Chen, and M. Chhowalla. Blue photoluminescence from chemically derived graphene oxide. Advanced Materials, 22(4):505–509, 2010. [43] G. Eda and M. Chhowalla. Chemically derived graphene oxide: Towards large- area thin-film electronics and optoelectronics. Advanced Materials, 22(22):2392– 2415, 2010. [44] F. Cicoira, N. Coppede, S. Iannotta, and R. Martel. Ambipolar copper phthalocyanine transistors with carbon nanotube array electrodes. Applied Physics Letters, 98(18):183303, 2011. [45] L. Chua, J. Zaumseil, J. Chang, E. C.-W. Ou, P. K.-H. Ho, H. Sirringhaus, and R. H. Friend. General observation of n-type field-effect behaviour in organic semiconductors. Nature, 434(7030):194–199, 2005. [46] D. J. Gundlach, L. Zhou, J. A. Nichols, T. N. Jackson, P. V. Necliudov, and M. S. Shur. An experimental study of contact effects in organic thin film transistors. Journal of Applied Physics, 100(2):024509, 2006.

30 [47] F. H. Herbstein. Crystalline molecular complexes and compounds: structures and principles. Oxford University Press, Oxford, 2005. [48] J. B. Torrance. An Overview of organic charge-transfer solids: insulators, metals, and the neutral-ionic transition. Molecular Crystals and Liquid Crystals, 126(1):55–67, 1985. [49] J. B. Torrance. The difference between metallic and insulating salts of tetra- cyanoquinodimethone (TCNQ): how to design an organic metal. Accounts of Chemical Research, 12(3), 1979. [50] S. Horiuchi, T. Hasegawa, and Y. Tokura. Molecular donor-acceptor compounds as prospective organic electronics materials. Journal of the Physical Society of Japan, 75(5):051016, 2006. [51] G. Saito and Y. Yoshida. Frontiers of Organic Conductors and Superconductors. Top Curr Chem, 312:67–126, 2011. [52] G. Saito and Y. Yoshida. Development of Conductive Organic Molecular As- semblies: Organic Metals, Superconductors, and Exotic Functional Materials. Bulletin of the Chemical Society of Japan, 80(1):1–137, 2007. [53] D. Jérome. Organic conductors: from charge density wave TTFTCNQ to su- perconducting (TMTSF)2PF6. Chemical Reviews, 104(11):5565–5592, 2004. [54] T. Mori and T. Kawamoto. Organic conductors - from fundamentals to nonlinear conductivity. Annual Reports Section ”C” (Physical Chemistry), 103:134, 2007. [55] J. Ferraris, D. O. Cowan, V. Jr Walatka, and J. H. Perstein. Electron transfer in a new highly conducting donor-acceptor complex. Journal of the American Chemical Society, 95(3):948–949, 1973. [56] D. Jerome, A. Mazaud, M. Ribault, and K. Bechgaard. Superconductivity in a synthetic organic conductor (TMTSF)2PF6. J. Physique Lett., 41:95–98, 1980. [57] S. Brown and G. Gruner. Charge and spin density waves. Scientific American, 270(4):50–56, 1994. [58] H. Alves, A.S Molinari, H. Xie, and A. F. Morpurgo. Metallic conduction at organic charge-transfer interfaces. Nature Materials, 7(7):574–580, 2008. [59] T. Mathis, K. Mattenberger, P. Moll, and B. Batlogg. Tetrathiofulvalene and tetracyanoquinodimethane crystals: Conducting surface versus interface. Ap- plied Physics Letters, 101(2):023302, 2012. [60] K. Harada, M. Sumino, C. Adachi, S. Tanaka, and K. Miyazaki. Improved thermoelectric performance of organic thin-film elements utilizing a bilayer structure of pentacene and 2,3,5,6-tetrafluoro-7,7,8,8-tetracyanoquinodimethane (F4TCNQ). Applied Physics Letters, 96(25):253304, 2010.

31 [61] H. Alves, R. M Pinto, and Ermelinda S Ma¸cˆoas.Photoconductive response in organic charge transfer interfaces with high quantum eﬃciency. Nature Com- munications, 4:1842, 2013.

[62] J. Tsutsumi, T. Yamada, H. Matsui, S. Haas, and T. Hasegawa. Com- petition between charge-transfer exciton dissociation and direct photocarrier generation in molecular donor-acceptor compounds. Physical Review Letters, 105(22):226601, 2010.

[63] J. Tsutsumi, H. Matsui, T. Yamada, R. Kumai, and T. Hasegawa. Genera- tion and diﬀusion of photocarriers in molecular donoracceptor systems: dependence on charge-transfer gap energy. The Journal of Physical Chemistry C, 116(45):23957–23964, 2012.

[64] K. Kobayashi, S. Horiuchi, R. Kumai, F. Kagawa, Y. Murakami, and Y. Tokura. Electronic ferroelectricity in a molecular crystal with large polarization directing antiparallel to ionic displacement. Physical Review Letters, 108(23):237601, 2012.

[65] F. Kagawa, S. Horiuchi, M. Tokunaga, J. Fujioka, and Y. Tokura. Ferroelectric- ity in a one-dimensional organic quantum magnet. Nature Physics, 6(3):169– 172, 2010.

[66] F. Kagawa, S. Horiuchi, H. Matsui, R. Kumai, Y. Onose, T. Hasegawa, and Y. Tokura. Electric-Field Control of Solitons in a Ferroelectric Organic Charge- Transfer Salt. Physical Review Letters, 104(22):227602, 2010.

[67] A. S. Tayi, A. K. Shveyd, A. C.-H. Sue, J. M. Szarko, B. S. Rolczynski, D. Cao, T. J. Kennedy, A. A. Sarjeant, C. L. Stern, W. F. Paxton, W. Wu, S. K. Dey, A. C. Fahrenbach, J. R. Guest, H. Mohseni, Lin X Chen, K. L. Wang, J. F. Stoddart, and S. I. Stupp. Room-temperature ferroelectricity in supramolecular networks of charge-transfer complexes. Nature, 488(7412):485–489, 2012.

[68] T.-H. Lee, J.-H. Li, W.-S. Huang, B. Hu, J. C. A. Huang, T.-F. Guo, and T.-C. Wen. Magnetoconductance responses in organic charge-transfer-complex molecules. Applied Physics Letters, 99(7):073307, 2011.

[69] Y. Liu, L. Jiang, H. Dong, Z. Tang, and W. Hu. Large-area single-crystalline nanocone arrays of an organic charge-transfer complex: controlling growth, characterization, and applications. Small, 7(10):1412–1415, 2011.

[70] Y. Liu, M. He, Q. Meng, Z. Tang, L. Li, and W. Hu. Mass-production of single- crystalline device arrays of an organic charge-transfer complex for its memory nature. Small, 8(4):557–560, 2012.

[71] R. S. Potember, T. O. Poehler, and D. O. Cowan. Electrical switching and memory phenomena in CuTCNQ thin ﬁlms. Applied Physics Letters, 405(1979):405– 407, 1979.

32 [72] H. Liu, Q. Zhao, Y. Li, Y. Liu, F. Lu, J. Zhuang, S. Wang, L. Jiang, D. Zhu, D. Yu, and L. Chi. Field emission properties of large-area nanowires of organic charge-transfer complexes. Journal of the American Chemical Society, 127(4):1120–1121, 2005.

[73] Z. G. Soos. Phenazine cation radical salts: charge-transfer complexes with tcnq. Annals of the New York Academy of Sciences, 313:442–458, 1978.

[74] D. S. Acker and W. R. Hertler. Substituted quinodimethans. I. Preparation and chemistry of 7, 7, 8, 8-tetracyanoquinodimethan. Journal of the American Chemical Society, 84(665):3370–3374, 1962.

[75] L.R. Melby, R.J. Harder, W.R. Hertler, W. Mahler, R.E. Benson, and W.E. Mochel. Substituted Quinodimethans. II. Anion-radical Derivatives and Com- plexes of 7,7,8,8-Tetracyanoquinodimethan. Journal of the American Chemical Society, 84:3374–3387, 1962.

[76] J. Zhang, J. Tan, Z. Ma, W. Xu, G. Zhao, H. Geng, C. Di, W. Hu, Z. Shuai, K. Singh, and D. Zhu. Fullerene/sulfur-bridged annulene cocrystals: two- dimensional segregated heterojunctions with ambipolar transport properties and photoresponsivity. Journal of the American Chemical Society, 135(2):558– 561, 2013.

[77] I. Shokaryev, J. Buurma, O.D. Jurchescu, M. A. Uijttewaal, G. A. de Wijs, T. T. M. Palstra, and R. A. de Groot. Electronic band structure of tetracene- TCNQ and perylene-TCNQ compounds. The Journal of Physical Chemistry A, 112(11):2497–502, 2008.

[78] L. Zhu, Y. Yi, Y. Li, E. G. Kim, V. Coropceanu, and J.-L. Br´edas. Pre- diction of remarkable ambipolar charge-transport characteristics in organic mixed-stack charge-transfer crystals. Journal of the American Chemical So- ciety, 134(4):2340–2347, 2012.

[79] I. J. Tickle and C. K. Prout. Molecular complexes. Part XVII. Crystal and molecular structure of perylene7, 7, 8, 8-tetracyanoquinodimethane molecular complex. Journal of the Chemical Society, Perkin Transactions 2, 6:720–723, 1973.

[80] C. K. Prout, I. J. Tickle, and J. D. Wright. Molecular complexes. Part XVI. Crystal structure of the 1:1 molecular complex of pyrene and 7,7,8,8- tetracyanoquinodimethane. Journal of the Chemical Society, Perkin Transac- tions 2, (5):528–530, 1973.

[81] T. J. Kistenmacher, T. J. Emge, F. M. Wiygul, W. A. Bryden, J. S. Chappell, J. P. Stokes, L.-Y. Chiang, and D. O. Cowan. DBTTF-TCNQ: A fractionally- charged organic salt with a mixed-stack crystalline motif. Solid State Commu- nications, 39:415–417, 1981.

33 [82] J. C. A. Boeyens and F. H. Herbstein. Molecular compounds and complexes . II. Exploratory crystallographic study of some donor-acceptor molecular compounds. The Journal of Physical Chemistry, 69(7):2153–2159, 1965.

[83] T. J. Kistenmacher, T. E. Phillips, and D. O. Cowan. The Crystal structure of the 1:1 radical cation-radical anion salt of 2,2’-Bis-1,3-dithiole (TTF) and 7,7,8,8-tetracyanoquinodimethane (TCNQ). Acta Crystallographia B, 30:763– 768, 1974.

[84] T. J. Kistenmacher, Thomas J Emge, A. N. Bloch, and Dwaine O Cowan. Structure of the red, semiconducting form of 4,4’,5,5’-tetramethyl-∆2,2’-bi-1,3- diselenole-7,7,8,8-tetracyano-p-quinodimethane, TMTSFTCNQ. Acta Crystal- lographica Section B, 38(4):1193–1199, 1982.

[85] K. Bechgaard, T. J. Kistenmacher, A. N. Bloch, and D. O. Cowan. The Crystal and molecular structure of an organic conductor from 4,4’,5,5 ’ - tetramethyl- delta2,2’-bis-1,3-diselenole and 7,7,8,8-tetracyano-p-quinodimethane (TMTSF- TCNQ). Acta Crystallographica B, 33:417–422, 1977.

[86] D. Vermeulen, L. Y. Zhu, K. P. Goetz, P. Hu, H. Jiang, C. S. Day, O. D. Jurchescu, V. Coropceanu, C. Kloc, and L. E. McNeil. Charge Transport Prop- erties of PeryleneTCNQ Crystals: The Eﬀect of Stoichiometry. The Journal of Physical Chemistry C, 118(42):24688–24696, 2014.

[87] A. W. Hanson. 7,7,8,8-Tetracyanoquinodimethane(perylene)2-Perylene. Acta Crystallographia B, 34:2339–2341, 1978.

[88] H. T. Black and D. F. Perepichka. Crystal engineering of dual channel p/n organic semiconductors by complementary hydrogenbonding. Angewandte Chemie, 53(8):2138–2142, 2014.

[89] H. T. Black, H. Lin, F. B´elanger-Gari´epy, and D. F. Perepichka. Supramolecular control of organic p/n-heterojunctions by complementary hydrogen bonding. Faraday discussions, pages 1–16, 2014.

[90] J. Tsutsumi, S. Matsuoka, S. Inoue, H. Minemawari, T. Yamada, and T. Hasegawa. N-type ﬁeld-eﬀect transistors based on layered crystalline donoracceptor semiconductors with dialkylated benzothienobenzothiophenes as electron donors. Journal of Materials Chemistry C, 3(9):1976–1981, 2015.

[91] A. N. Bloch, D. O. Cowan, K. Bechgaard, R. E. Pyle, R. H. Banks, and T. O. Poehler. Low-temperature metallic behavior and resistance minimum in a new quasi one-dimensional organic conductor. Physical Review Letters, 34(25):1561– 1564, 1975.

[92] X. Chi, C. Besnard, V. K. Thorsmølle, V. Y. Butko, A. J. Taylor, T. Siegrist, and A. P. Ramirez. Structure and Transport Properties of the Charge-Transfer Salt CoroneneTCNQ. Chemistry of Materials, 16(26):5751–5755, 2004.

34 [93] N. Karl and J. Ziegler. Generation and transport of charge carriers in the charge-transfer complex anthracene-pyromellitic-dianhydride. Chemical Physics Letters, 32(3):438–442, 1975. [94] T. Hasegawa, K. Mattenberger, J. Takeya, and B. Batlogg. Ambipolar field-effect carrier injections in organic Mott insulators. Physical Review B, 69(24):245115, 2004. [95] J. Zhang, H. Geng, T. S. Virk, Y. Zhao, J. Tan, C. Di, W. Xu, K. Singh, W. Hu, Z. Shuai, Y. Liu, and D. Zhu. Sulfur-bridged annulene-TCNQ co-crystal: a self- assembled ”molecular level heterojunction” with air stable ambipolar charge transport behavior. Advanced Materials, 24(19):2603–7, 2012. [96] Y. Qin, J. Zhang, X. Zheng, H. Geng, G. Zhao, W. Xu, W. Hu, Z. Shuai, and D. Zhu. Charge-transfer complex crystal based on extended-π-conjugated acceptor and sulfur-bridged annulene: Charge-transfer interaction and remarkable high ambipolar transport characteristics. Advanced Materials, 26(24):4093– 4099, 2014. [97] H. Wu, F. Wang, Y. Xiao, and G. Pan. Preparation and ambipolar transistor characteristics of co-crystal microrods of dibenzotetrathiafulvalene and tetracyanoquinodimethane. Journal of Materials Chemistry C, 1(12):2286–2289, 2013. [98] D.J. Gundlach, L. Jia, and T. N. Jackson. Pentacene TFT with improved linear region characteristics using chemically modified source and drain electrodes. IEEE Electron Device Letters, 22(12):571–573, 2001. [99] L. C. Teague, B. H. Hamadani, O. D. Jurchescu, S. Subramanian, J. E. Anthony, T. N. Jackson, C. A. Richter, D.J. Gundlach, and J. G. Kushmerick. Surface Potential Imaging of Solution Processable Acene-Based Thin Film Transistors. Advanced Materials, 20(23):4513–4516, 2008. [100] K. P. Goetz, J. Tsutsumi, S. Pookpanratana, J. Chen, C. A. Richter, C. A. Hacker, T. Hasegawa, and O. D. Jurchescu. Polymorphism in the 1:1 Charge- Transfer Complex DBTTF-TCNQ and Its Effects on Optical and Electronic Properties. Submitted, 2016. [101] S. Yokokura, Y. Takahashi, H. Nonaka, H. Hasegawa, J. Harada, T. Inabe, R. Kumai, H. Okamoto, M. M. Matsushita, and K. Awaga. Switching of Trans- fer Characteristics of an Organic Field-Effect Transistor by Phase Transitions: Sensitive Response to Molecular Dynamics and Charge Fluctuation. Chemistry of Materials, 27:4441–4449, 2015. [102] S. K. Park, S. Varghese, J.H. Kim, S.-J. Yoon, O. K. Kwon, B. K. An, J. Gier- schner, and S. Y. Park. Tailor-made highly luminescent and ambipolar transporting organic mixed stacked charge-transfer crystals: An isometric donor- acceptor approach. Journal of the American Chemical Society, 135(12):4757– 4764, 2013.

35 [103] Y. H. Kim, C. Sachse, M. L. Machala, C. May, L. M¨uller-Meskamp, and K. Leo. Highly Conductive PEDOT:PSS Electrode with Optimized Solvent and Ther- mal Post-Treatment for ITO-Free Organic Solar Cells. Advanced Functional Materials, 21(6):1076–1081, mar 2011.

[104] S. Tamura, T. Kadoya, and T. Mori. All-organic self-contact transistors. Applied Physics Letters, 105:023301, 2014.

[105] S. Tamura, T. Kadoya, T. Kawamoto, and T. Mori. Self-contact thin-ﬁlm organic transistors based on tetramethyltetrathiafulvalene. Applied Physics Let- ters, 102(6):063305, 2013.

[106] T. Kadoya, S. Tamura, and T. Mori. Energy-level engineering in self-contact organic transistors prepared by inkjet printing. Journal of Physical Chemistry C, 118(40):23139–23146, 2014.

[107] H. B Akkerman, P. W. M. Blom, D. M. de Leeuw, and B. de Boer. To- wards molecular electronics with large-area molecular junctions. Nature, 441(7089):69–72, 2006.

[108] Z. Nie and E. Kumacheva. Patterning surfaces with functional polymers. Nature Materials, 7(4):277–290, 2008.

[109] L. Nyholm, G. Nystrom, A. Mihranyan, and M. Stromme. Toward ﬂexible polymer and paper-based energy storage devices. Advanced Materials, 23(33):3751– 3769, 2011.

[110] S. Flandrois and D. Chasseau. Longueurs de liaison et transfert de charge dans les sels du t´etracyanoquinodim´ethane (TCNQ). Acta Crystallographica Section B, 33:2744–2750, 1977.

[111] T. C. Umland, S. Allie, T. Kuhlmann, and P. Coppens. Relation between geometry and charge transfer in low-dimensional organic salts. The Journal of Physical Chemistry, 92(22):6456–6460, 1988.

[112] S. Matsuzaki, R. Kuwata, and K. Toyoda. Raman spectra of conducting TCNQ salts; estimation fo the degree of charge transfer from vibrational frequencies. Solid State Communications, 33:403–405, 1980.

[113] J. S. Chappell, A.N. Bloch, W. A. Bryden, M. Maxﬁeld, T. O. Poehler, and D. O. Cowan. Degree of charge transfer in organic conductors by infrared absorption spectroscopy. Journal of the American Chemical Society, 103(9):2442– 2443, 1981.

[114] J. B. Torrance, B. A. Scott, and F. B. Kaufman. Optical properties of charge transfer salts of tetracyanoquinodimethane (TCNQ). Solid State Communica- tions, 17:1369–1373, 1975.

36 [115] L. Zhu, E. Kim, Y. Yi, and J.-L. Br´edas. Charge transfer in molecular complexes with 2,3,5,6-tetraﬂuoro-7,7,8,8-tetracyanoquinodimethane (F4-TNCQ): A density functional theory study. Chemistry of Materials, 23(23):5149–5159, 2011.

[116] D. Zobel and G. Ruban. The structures of some charge-transfer complexes containing TCNQ as acceptor and their electrical anisotropy. Acta Crystallo- graphica Section B, 39:638–645, 1983.

[117] R. M. Williams and S. C. Wallwork. Molecular complexes exhibiting polarization bonding. XI. The crystal and molecular structure of the 7,7,8,8- tetracyanoquinodimethaneanthracene complex. Acta Crystallographica Section B, 24(2):168–174, 1968.

[118] B. Shaanan, U. Shmueli, and D. Rabinovich. Structure and packing arrangement of molecular compounds. VII. 7,7,8,8-tetracyanoquinodimethane- naphthalene (1:1). Acta Crystallographica B, 32:2574–2580, 1976.

[119] V. R. Gakel, B. P. Bespalov, and A. A. Pankratov. Investigation of electric conductivity of weak molecular complexes of tetracyanoquinodimethane. The- oretical and Experimental Chemistry, 13(6):635–637, 1978.

[120] T. Mori and H. Inokuchi. Crystal structure of the mixed-stacked salt of bis (ethylenedithio)-tetrathiafulvalene (BEDT-TTF) and tetracyanoquinodimethane (TCNQ). Bulletin of the Chemical Society of Japan, 60(1):402–404, 1987.

[121] T. Mori and H. Inokuchi. Structural and electrical properties of (BEDT- TTF)(TCNQ). Solid State Communications, 59(6):355–359, 1986.

[122] G. Saito, H. Hayashi, T. Enoki, and H. Inokuchi. The Study of Charge Transfer Complexes of Bedt-TTF Derivatives. Molecular Crystals and Liquid Crystals, 120(1):341–344, 1985.

[123] M. Mizuno, A. F. Garito, and M. P. Cava. Organic metals’: alkylthio sub- stitution eﬀects in tetrathiafulvalenetetracyanoquinodimethane charge-transfer complexes. Journal of the Chemical Society, Chemical Communications, (1):18– 19, 1978.

[124] K. Nakasuji, M. Sasaki, T. Kotani, I. Murata, T. Enoki, K. Imaeda, H. Inokuchi, A. Kawamoto, and J. Tanaka. Methylthio- and ethanediyldithio-substituted 1,6-dithiapyrenes and their charge-transfer complexes: New organic molecular metals. Journal of the American Chemical Society, 109:6970–6975, 1987.

[125] T. J. Kistenmacher, T. J. Emge, F. M. Wiygul, W. A. Bryden, J. S. Chappell, J. P. Stokes, L-Y. Chiang, and D. O. Cowan. DBTTF-TCNQ: A fractionally- charged organic salt with a mixed-stack crystalline motif. Solid State Commu- nications, 39:415–417, 1981.

37 [126] G. S. Bajwa and K. Darrell Berlin. Synthesis of delta2,2’-bis(1,3- benzodithiolidine) derivatives and complex salts therefrom with 7,7,8,8- tetracyanoquinodimethane. The Journal of Organic Chemistry, 41(1):145–148, 1976.

[127] A. Andrieux, C. Duroure, D. Jerome, and K. Bechgaard. The metallic state of the organic conductor TMTSF-DMTCNQ at low temperature under pressure. Le Journal De Physique - Lettres, 40(1):L381–L383, 1979.

[128] A. Andrieux, P. M. Chaikin, C. Duroure, D. Jerome, C. Weyl, K. Bechgaard, and J. R. Andersen. Transport properties of the metallic state of TMTSF- DMTCNQ. Le Journal De Physique, 40:1199–1206, 1979.

[129] Y. Iwasa, K. Mizuhashi, T. Koda, Y. Tokura, and G. Saito. Metal-insulator transition and antiferromagnetic order in bis(ethylenedithio)tetrathiafulvalene tetracyanoquinodimethane (BEDT-TTF)(TCNQ). Physical Review B, 49(5):3580–3583, 1994.

[130] R. Comes, S. M. Shapiro, G. Shirane, A. F. Garito, and A. J. Heeger. Neutron-scattering study of the 38- and 54-K phase transitions in deuterated tetrathiafulvalenetetracyanoquinodimethane (TTF-TCNQ). Physical Review Letters, 35(22):1518–1521, 1975.

[131] K. Bechgaard, D. O. Cowan, and A. N. Bloch. Stabilization of the organic metallic state: The properties of two substituted tetraselenafulvalenes and their TCNQ salts. Molecular Crystals and Liquid Crystals, 32:227–230, 1976.

[132] A. J. Epstein, E. M. Conwell, and J. S. Miller. Charge Transport in Molecular Conductors: Role of Mobility. Annals of the New York Academy of Sciences, 313:183–209, 1978.

[133] C. Weyl, E.M. Engler, K. Bechgaard, G. Jehanno, and S. Etemad. Diﬀuse X-Ray Scattering in the Metallaic State of TSeF-TCNQ and HMTSeF-TCNQ. Solid State Communications, 19:925–930, 1976.

[134] E. M. Engler, B. A. Scott, T. Penney, and V. V. Patel. Organic Alloys: Synthesis and Properties of Solid Solutions of TSef-TCNQ and TTF-TCNQ. Journal of the American Chemical Society, 99(18):5909–5916, 1977.

[135] D. O. Cowan, A. Bloch, T. Poehler, T. Kistenmacher, J. Ferraris, K. Bechgaard, R. Gemmer, C. Hu, P. Shu, W. Krug, R. Pyle, V. Walatka, T. Carruthers, T. Phillips, and R. Banks. The Organic Metallic State. Molecular Crystals and Liquid Crystals, 32(1):223–225, 1976.

[136] J. B. Torrance, J. J. Mayerle, K. Bechgaard, B. D. Silverman, and Y. Tomkiewicz. Comparison of two isostructural organic compounds, one metallic and the other insulating. Physical Review B, 22(10):4960–4965, 1980.

38 [137] R. L. Greene, J. J. Mayerle, R. Schumaker, G. Castro, P. M. Chaikin, S. Etemad, and S. J. LaPlaca. The structure, conductivity, and thermopower of HMTTF- TCNQ. Solid State Communications, 20(10):943–946, 1976.

[138] H. Endres, H. J. Keller, W. Moroni, and D. N¨othe. Highly conducting phenazine-doped 5,10-dihydro-5,10-dimethylphenaziniumylTCNQ. Physical properties and crystal and molecular structure. Acta Crystallographica Section B, 36(6):1435–1440, jun 1980.

[139] G. Saito, H. Ikegami, Y. Yoshida, O. O. Drozdova, K. Nishimura, S. Horiuchi, H. Yamochi, A. Otsuka, T. Hiramatsu, M. Maesato, T. Nakamura, T. Aku- tagawa, and T. Yumoto. Ionicity Phase Diagram of Triﬂuoromethyl-TCNQ (CF3TCNQ) Charge-Transfer Solids. Bulletin of the Chemical Society of Japan, 83(12):1462–1480, 2010.

[140] A. C. Arias, J. D. MacKenzie, I. McCulloch, J. Rivnay, and A. Salleo. Materials and applications for large area electronics: Solution-based approaches. Chemical Reviews, 110(1):3–24, 2010.

[141] N. A. Azarova, J. W. Owen, C. A. McLellan, M. A. Grimminger, E. K. Chap- man, J. E. Anthony, and O.D. Jurchescu. Fabrication of organic thin-ﬁlm transistors by spray-deposition for low-cost, large-area electronics. Organic Elec- tronics: physics, materials, applications, 11(12):1960–1965, 2010.

39 Chapter 2

Crystal Growth of Organic Charge-Transfer Complexes

The electronic properties and device functionality of a charge-transfer complex — or any solid — are a result of its structure, which is in turn a function of growth conditions. The details of the growth may strongly impact the crystal’s structure and quality, as well as its appearance, size, roughness, and more. Single crystals are most appropriate for fundamental studies, such as those in this thesis, where their nearly defect-free nature allows them to reveal the intrinsic properties of the material. In this work we discuss the growth of perylene–TCNQ in various stoichiometries as well as DBTTF–TCNQ polymorphs by physical vapor transport. For stilbene– F4TCNQ, crystals were grown by solution methods. Solution growth was also used in an exploratory search for CT complexes with atypical acceptors (for example PDIF- CN2). Solution-deposited crystals and either vapor- or solution-deposited thin-ﬁlms are innately more useful in device fabrication than single crystals; therefore, eﬀorts in the research community to adapt CT complexes to high-throughput processing are discussed and reviewed here.

40 2.1 Introduction

The method of crystal growth for charge-transfer complexes is important to take into consideration as it can have an enormous impact on stoichiometry, crystal structure, size, appearance, and roughness, and therefore on the compound’s electronic performance and applicability. Single crystals are best for fundamental studies as they are low in defects and perform closer to the intrinsic capabilities of the material.1–3 For this reason, the studies included in this thesis were performed on single crystals grown by either solution or vapor methods. Through physical vapor transport (PVT), we were able to grow Perylene–TCNQ in three donor-to-acceptor (D:A) stoichiometries — 1:1, 2:1, and 3:1.4, 5 In addition, we were able to grow DBTTF–TCNQ in two polymorphs, one of which is the same structure as that in the literature6, 7 and the other of which is new. The stilbene–F4TCNQ complex was grown by solution methods, as the varying vapor pressures of the donor and acceptor did not lend themselves to growth from the gas phase. While such single crystals exhibit the highest possible performance, their use in devices is limited by their time-consuming growth and technically difficult fabrication procedures. To take full advantage of the processing power of organic materials, solution-deposited thin-films are desired — they can be processed at room temperature and ambient pressure, and are amenable to large-area deposition techniques. This chapter will therefore also include a brief discussion of thin-film growth of organic CT complexes.

41 2.2 Single Crystal Growth of Charge-Transfer Com- plexes

2.2.1 Solution Growth

The general method for solution growth is a seemingly simple three step process: the donor and acceptor are dissolved individually in either the same or in diﬀerent solvents, the solutions are supersaturated and mixed at elevated temperatures, and slow cooling allows CT crystals form. In practice, however, each step can be complicated, starting with the choice of solvent. Here, relative solubilities of the donor and the acceptor in a given solvent (or two) will determine the stoichiometric outcome of the crystal. This is apparent for the Perylene–TCNQ system, where the 1:1 complex grows when the donor and acceptor are dissolved in solvents such as toluene, but the 3:1 complex grows when benzene is used. As Hu, et al., discuss, this is possibly due to the fact that the 1:1 complex is less soluble than the 3:1 in toluene, whereas the reverse is true with benzene.5 Dissolving equimolar amounts of Perylene and TCNQ in a 5:1 volumetric ratio of benzene to acetonitrile results in the 1:1 ratio, however. This indicates that subtle changes in solution composition drastically change the resulting crystals.8 The beauty (and intriguing physics) of these complexes is evident in the mixing phase of the crystal growth. As mentioned in chapter one, the formation of the charge-transfer complex results in a smaller, hybrid band structure that is primarily comprised of the donor HOMO and the acceptor LUMO. Thus, mixing a transparent donor and a transparent acceptor immediately results in a color change. In the case of pyrene or anthracene when complexed with pyromellitic dianhydride (PMDA), a transparent donor and acceptor mix to form vivid orange complexes. Anthracene, PMDA, and the CT complex crystals are shown in Fig. 2.1. Depending on the solubility of the complex, the crystals may quickly precipitate out of solution at the point of mixing. The resulting crystals are typically rough, and

42 Figure 2.1: The color change upon CT formation. Anthracene and PMDA crystals are transparent, while the CT crystals are orange. unsuitable for device fabrication. Thus, it is important to heat the solvent to ensure saturation. Slow-cooling then allows for slow formation of the complex and, ideally, highly ordered single crystals. Based on observation, the crystal morphology of 1:1 complexes is more needle-like and the m:n complexes are more sheet-like or block-like, as in the case of 1:1 versus 3:1 Perylene–TCNQ.5 Solvent-growth was possible for all crystals mentioned in this thesis except the 2:1 Perylene–TCNQ complex and the β-DBTTF–TCNQ, which have only been grown from the vapor phase.

2.2.2 Vapor Growth

Crystal growth from the vapor phase has several advantages. First, the lack of solvents involved in the growth process means that the crystals can achieve higher purity. Second, the resulting crystals are often thinner, flatter, and more plate-like than when they are grown from solvent. This is true even in the case of the 1:1 complexes, which tend to be needle-like when grown from solution. If the crystals remain needle-like when grown by vapor, they still possess smoother surfaces resulting from prolonged growth in a clean environment. Such crystals are ideal for organic field-effect transistors because the smooth, thin, flexible crystals can be electrostatically adhered to a dielectric, reducing the charge scattering effects that would have occurred in the presence of a rougher surface.9 To grow single crystals from the gas phase, either vacuum sublimation or physical

43 vapor transport (PVT) can be used.10, 11 To accomplish the former, a clean quartz growth tube with one end sealed and the other end attached to a vacuum pump is placed in a temperature gradient. Fig. 2.2 depicts the furnaces used at Wake Forest University. As is shown, small tubes can be placed inside the larger one to avoid the need to break the growth tube. Heating the clean, empty tubes to a temperature well above the crystal growing temperature (usually 350 ◦C) with an inert gas ﬂowing or under vacuum prior to placing the starting material ensures that the tube is suﬃ- ciently clean.12 The CT crystals can be grown either by co-sublimation of the parent compounds, or by the growth of the binary CT crystal from solution followed by its sublimation. For the co-sublimation approach, powders of the donor and acceptor molecules are placed at temperatures such that their sublimation rates will be compa-

Figure 2.2: The vapor growth furnaces used at Wake Forest University. Graphs show the variation in temperature from the left to right sides of the furnace. Furnace 1 has shallower temperature gradient than Furnace 2.

44 rable, and convection currents cause the sublimed materials to recrystallize in a cooler region of the temperature gradient. For PVT, the only difference is that an inert gas at ambient pressure carries the sublimed material to a cooler recrystallization zone near the end of the tube.13–16 Because the process takes place at ambient pressure, a higher temperature is necessary to sublimate the materials. The typical outcome of this method is that there will be a section in the growth tube where only crystals of the donor compound will form, sections of different CT stoichiometries (when applicable), and a section of only the acceptor crystals. For example, perylene-TCNQ grows in the order of perylene, followed by the 1:1 complex, the 2:1, and the 3:1, and then TCNQ (Fig. 2.3). Some mixing of the zones will occur. In the case that the vapor pressures of the donor and acceptor are roughly the same, the CT already formed from solution can be sublimed to re-grow the crystal from a vapor phase. This works for DBTTF–TCNQ, Pyrene–PMDA, and Anthracene–PMDA, for example. In the case that the vapor pressures of the donor and acceptor are extremely different, vapor growth may not be possible. This is the case with STB–F4TCNQ, where the

◦ ◦ stilbene begins to sublimate around 100 C but F4TCNQ grows at 220 C. Experience has shown that the bulk crystalline morphology is dependent on the

Figure 2.3: The result of growing Perylene–TCNQ by PVT. Yellow crystals are either perylene or TCNQ, while darker crystals are the CT complex in either 1:1, 2:1, or 3:1 stoichiometry. The 1:1 complex is needle-like and dark blue-green, while the 2:1 and 3:1 complexes are both green platelets.

45 dimensions of the growth tubes and the nature of the temperature gradient. In the case of rubrene, for example, some groups have had success growing thin, flexible crystals within a half-hour of reaching the growth temperature.17 Such fast growth has not been possible on the furnaces at Wake Forest University. While this fact often goes unpublished, anecdotally, this phenomenon is likely due to differences in the steepness of the temperature gradient (with a very steep gradient allows for fast growth), as tube sizes are comparable in this particular instance. For the furnaces used in this thesis (Fig. 2.2), both are set to use 30 inch ( 76 cm) long tubes with a 22 mm inner diameter. As it can be seen, Furnace 2 has a steeper temperature gradient than Furnace 1. The impact of temperature gradient on crystal growth is especially notable when growing DBTTF–TCNQ. Here, using the same starting material and growing temperature with an argon flow of 150 mL/min, the α-polymorph grows to higher yield in Furnace 2 than Furnace 1. The reverse is true for the β-polymorph. This is because the α-polymorph crystallizes at a lower temperature (room temperature- 45 ◦C) than the β-polymorph (40-65 ◦C), and can reach its crystallization zone in a much shorter distance in Furnace 2 than in Furnace 1.

2.3 Crystals of Novel Organic Charge-Transfer Com- plexes

A majority of the studies on charge-transfer complexes made use of the acceptor

TCNQ and its ﬂuorinated derivatives, such as F4TCNQ. It has been shown, however, that materials such as C60 and pyromellitic-dianhydride (PMDA) can also be used as electron acceptors. In the case of the former, at least one device study has been published.18 In the case of the latter, most literature has focused on the crystal growth of these complexes,19–28 with several spectroscopic studies being published on anthracene–PMDA.29 Though no device studies are known at this point in time,

46 theoretical calculations indicate that PMDA-containing complexes may exhibit high electron and hole mobilities.30 Another potential candidate for an electron acceptor is the n-type material PDIF-CN2 (N,N-bis(n-alkyl)-(1,7 and 1,6)-dicyanoperylene- 3,4;9,10-bis(dicarboximide)) (Fig. 2.5a), which has been shown to have an electron mobility of up to 6 cm2V−1s−1in single crystals.31 This compound is of interest because its large, planar aromatic backbone can promote charge-transfer, and the electron-withdrawing ﬂuorine side-chains are oﬀ-set enough to not prohibit complex formation.

The backbone of PDIF-CN2 is based on perylene. Therefore, perylene and several smaller aromatic compounds that typically exhibit the tendency to donate an electron were chosen to for CT complexes with this acceptor. Though we noted a color change upon mixing a perylene solution with the acceptor, indicating the formation of a complex, the crystals that precipitated were too small to collect structure information or make devices. Vapor growth was attempted but was unsuccessful. Greater success was attained with donors anthracene and pyrene when crystallized from chlorobenzene or dichlorobenzene. Crystals of anthracene, PDIF-CN2, and the CT complex are shown in Fig. 2.4. As it can be seen, anthracene is transparent,

PDIF-CN2 is red-orange, and the resulting CT crystal is green. Pyrene–PDIF-CN2 is also green. Preliminary crystal structures of these complexes are shown in Fig.

Figure 2.4: Anthracene crystals are transparent, PDIF-CN2 crystals are red-orange 31 (picture taken from Molinari, et al. ), and the complex is green. Pyrene-PDIF-CN2 is also green.

47 2.5 and Fig. 2.6, though we caution that these are only reﬁned to a value of R = 20%. As is evident in reﬂectance measurements which will be discussed in chapter 3, this is due to the inherent disorder present in these complexes that has not yet been overcome by crystal growth methods.

Figure 2.5: Anthracene-PDIF-CN2 skeletal and crystal structures. (a) The skeletal structure of the acceptor, PDIF-CN2. (b) The skeletal structure of the donor, anthracene. (c) The preliminary crystal structure of the mixed-stack CT complex. (d) The donor-acceptor overlap of the mixed-stack CT complex.

Figure 2.6: Pyrene-PDIF-CN2 skeletal and crystal structures. (a) The skeletal structure of the donor, pryene. The skeletal structure of the acceptor is shown in Fig. 2.5a. (b) The preliminary crystal structure of the mixed-stack CT complex. (c) The donor-acceptor overlap of the mixed-stack CT complex, showing disorder.

48 2.4 Thin-Film Growth of Organic Charge-Transfer Complexes

Although single crystals are preferable for fundamental studies of organic materials because they offer increased access to the intrinsic properties of the material, many applications of CT materials necessitate their growth in a thin-film form. These are typically polycrystalline in nature and exhibit diminished performance when compared to single crystals, but are compatible with high-throughput growth methods and in some cases low-temperature processing. Thin-films are most often encountered for organic metals which are patterned to contact organic semiconductors (either monomolecular or binary). In this case, the metallic film can be grown directly onto a single crystal or semiconducting thin film to be used as a top contact. Bottom-contact structures are not favorable as the surface of the organic metal contact is rough. The method most widely employed to fabricate electrodes of organic metallic films is thermal evaporation under vacuum, primarily because of the reduced solubility characteristics of CTs, which precludes solution processability. When the sublimation temperatures of the D and A materials are comparable, the powders are mixed in one crucible and simultaneous direct thermal evaporation is performed. This method was used for complexes like TTF–TCNQ, TTF–F1TCNQ, TTF–F2TCNQ, TSF–F1TCNQ, TSF–F2TCNQ, and HMTTF–TCNQ.13, 14, 32–35 For D/A combinations that exhibit significant differences in the sublimation temperatures, as is the case of DBTTF–F4TCNQ complex, two-source thermal evaporation is employed.14 Several strategies were introduced recently to allow deposition of organic metallic electrodes using solution-based approaches and reduce the complexity of processing. Most notable is double-shot inkjet printing, which relies on deposition of device electrodes by printing the D and A inks on identical spots using an inkjet printer equipped with double heads, followed by rapid liquid intermixing which results in the instantaneous formation of the CT complexes. This method was successfully adopted for fabrication of TTF–TCNQ,

49 TTF–F1TCNQ, TTF–F2TCNQ, TSF–F1TCNQ, and TSF–F2TCNQ contacts with conductivities approaching those obtained in vacuum-deposited films of the same compounds.36, 37 Efforts to grow thin-films of the active layer are a current area of research.

2.5 Summary

In summary, the method of crystal growth and speciﬁcs of the crystal growth ap- paratus can have a direct impact on the resulting crystal structure of the complex. In particular, varying solvents can lead to varying stoichiometries, as is the case for Perylene–TCNQ. Within the PVT process, the steepness of the temperature gradient can lead to varying prevalence of one polymorph over another, as is the case for DBTTF–TCNQ. Attempts to explore the use of non-TCNQ materials as electron acceptors yielded some success with n-type material PDIF-CN2; this crystallized with donors antrhacene and pyrene. Further work will be needed to grow disorder-free crystals of this complex system. All growth methods used in this thesis are summarized in Tables 2.1 and 2.2.

50 Material Donor Solvent, Color Acceptor Solvent, Color Max. Temp. CT Morphology, Color Perylene–TCNQ, 1:1 toluene, yellow toluene, yellow 80 ◦C Blue-green, needle-like Perylene–TCNQ, 3:1 benzene, yellow benzene, yellow 80 ◦C blue-green, block-like ◦ Stilbene–F4TCNQ, 1:1 acetonitrile, clear acetonitrile, yellow 80 C green, needle-like α-DBTTF–TCNQ, 1:1 xylenes, yellow acetonitrile, yellow 80 ◦C dark red, needle-like ◦ Anthracene–PDIF-CN2, 1:1 chlorobenzene, clear chlorobenzene, red-orange 100 C green, needle-like 51 ◦ Pyrene–PDIF-CN2, 1:1 dichlorobenzene, clear chlorobenzene, red-orange 120 C green, needle-like Pyrene–PMDA, 1:1 2,butanone, clear 2,butanone, clear 70 ◦C orange, elongated blocks Anthracene–PMDA, 1:1 2,butanone, clear 2,butanone, clear 70 ◦C orange, elongated blocks Stilbene–PMDA, 1:1 2,butanone, clear 2,butanone, clear 70 ◦C orange, thin plates

Table 2.1: Solution growth methods for the organic charge-transfer complexes used in this thesis. Note that the PDIF-CN2 and PMDA complexes exhibit some disorder when grown from solution. Material Placement Temperature Carrier Gas, Crystal Color, Flow Rate Morphology Perylene Max Temp, 200 ◦C Argon, Yellow, Furnace 1 or 2 100 mL/min Platelets TCNQ Max Temp, 160-180 ◦C Argon, Yellow, Depending on purity, Furnace 1 or 2 100 mL/min Small blocks or Needles ◦ F4TCNQ Max Temp, 200-220 C Argon, Yellow, Furnace 1 or 2 100 mL/min Needles Stilbene Max Temp, 100 ◦C Argon, Transparent, Furnace 1 or 2 100 mL/min Kite-shaped Platelets PMDA Max Temp, 100 ◦C Argon, Transparent, Furnace 1 or 2 100 mL/min Small Blocks DBTTF Max Temp, 225 ◦C Argon, Yellow, Furnace 1 or 2 50 mL/min Platelets 52 Perylene–TCNQ D at max temp, A 10-12 cm past, 200 ◦C Argon, Dark Green-Blue, Furnace 1 100 mL/min Needles & Plates of 1:1, 2:1, 3:1 α-DBTTF–TCNQ CT complex at max temp, 160 ◦C Argon, Dark Red, Furnace 2 150-200 mL/min Elongated Platelets β-DBTTF–TCNQ CT complex at max temp 160 ◦C Argon, Light Red, Furnace 1 150-200 mL/min Elliptical Platelets Anthracene–PMDA Mixture of D and A at max temp 120 ◦C Argon, Orange, Furnace 1 100 mL/min Needles Pyrene–PMDA Mixture of D and A at max temp 130 ◦C Argon, Orange, Furnace 1 100 mL/min Small Blocks

Table 2.2: PVT growth conditions for organic charge-transfer complexes, donors, and acceptors with which the author has experience. In the case that the temperature gradient makes a diﬀerence, the furnace used (1 or 2 as show in Fig. 2.2) is noted. Otherwise, the diﬀerence is not appreciable. Note that Stilbene and PMDA sublimate at low temperatures. Crystals of these material are not stable at atmospheric pressure and temperature and will quickly disintegrate. References

[1] M. E. Gershenson, V. Podzorov, and A. F. Morpurgo. Colloquium: Elec- tronic transport in single-crystal organic transistors. Reviews of Modern Physics, 78(3):973–989, 2006. [2] W. L. Kalb, S. Haas, C. Krellner, T. Mathis, and B. Batlogg. Trap density of states in small-molecule organic semiconductors: A quantitative comparison of thin-film transistors with single crystals. Physical Review B, 81(15):155315, 2010. [3] D. Braga and G. Horowitz. High-Performance organic field-effect transistors. Advanced Materials, 21:1473–1486, 2009. [4] D. Vermeulen, L. Y. Zhu, K. P. Goetz, P. Hu, H. Jiang, C. S. Day, O. D. Jurch- escu, V. Coropceanu, C. Kloc, and L. E. McNeil. Charge Transport Properties of PeryleneTCNQ Crystals: The Effect of Stoichiometry. The Journal of Physical Chemistry C, 118(42):24688–24696, 2014. [5] P. Hu, L. Ma, K. Tan, H. Jiang, F. Wei, C. Yu, K. P. Goetz, O. D. Jurchescu, L. E. McNeil, G. G. Gurzadyan, and C. Kloc. Solvent-Dependent Stoichiometry in Perylene7,7,8,8-Tetracyanoquinodimethane Charge Transfer Compound Single Crystals. Crystal Growth & Design, 14(12):6376–6382, 2014. [6] T. J. Emge, F. M. Wiygul, J. S. Chappell, A. N. Bloch, J. P. Ferraris, D. O. Cowan, and T. J. Kistenmacher. Crystal structures for the electron donor dibenzotetrathiafulvalene DBTTF, and its mixed-stack charge-transfer salts with the electron acceptors 7,7,8,8-tetracyano-p-quinodimethane, TCNQ, and 2,5- difluoro-7,7,8,8-tetracyano-p-quinodimethane, 2,5-TCNQF2. Molecular Crystals and Liquid Crystals, 87:137–161, 1982. [7] H. Kobayashi and J. Nakayama. The Crystal structure of the charge-transfer compex of dibenzotetrathiafulvalene-tetracyanoquinodimethane, DBTTF-TCNQ. Bulletin of the Chemical Society of Japan, 54(8):2408–2411, 1981. [8] K. D. Truong and A.D. Bandrauk. A new TCNQ complex: (Perylene)3 TCNQ. Chemical Physics Letters, 44(2):232–235, 1976. [9] H. Sirringhaus. Device Physics of Solution-Processed Organic Field-Effect Tran- sistors. Advanced Materials, 17(20):2411–2425, 2005.

53 [10] R. A. Laudise, C. Kloc, P. G. Simpkins, and T. Siegrist. Physical vapor growth of organic semiconductors. Journal of Crystal Growth, 187:449–454, 1998.

[11] C. Kloc, P. G. Simpkins, T. Siegrist, and R. A. Laudise. Physical vapor growth of centimeter-sized crystals of alpha-hexathiophene. Journal of Crystal Growth, 182:416–427, 1997.

[12] O. D. Jurchescu, J. Baas, and T. T. M. Palstra. Eﬀect of impurities on the mobility of single crystal pentacene. Applied Physics Letters, 84(16):3061, 2004.

[13] Y. Takahashi, J. Hasegawa, Y. Abe, Y. Tokura, K. Nishimura, and G. Saito. Tun- ing of electron injections for n-type organic transistor based on charge-transfer compounds. Applied Physics Letters, 86(6):063504, 2005.

[14] Y. Takahashi, T. Hasegawa, Y. Abe, Y. Tokura, and G. Saito. Organic metal electrodes for controlled p- and n-type carrier injections in organic ﬁeld-eﬀect transistors. Applied Physics Letters, 88(7):073504, 2006.

[15] J. Tsutsumi, H. Matsui, T. Yamada, R. Kumai, and T. Hasegawa. Genera- tion and diﬀusion of photocarriers in molecular donoracceptor systems: dependence on charge-transfer gap energy. The Journal of Physical Chemistry C, 116(45):23957–23964, 2012.

[16] A. J. C. Buurma, O. D. Jurchescu, I. Shokaryev, J. Baas, A. Meetsma, G. A. de Wijs, R. A. de Groot, and T. T. M. Palstra. Crystal growth, structure, and electronic band structure of TetraceneTCNQ. The Journal of Physical Chemistry C, 111(8):3486–3489, 2007.

[17] M. A. Reyes-Martinez, A. Ramasubramaniam, A. L. Briseno, and A.J. Crosby. The intrinsic mechanical properties of rubrene single crystals. Advanced Materi- als, 24(41):5548–52, 2012.

[18] J. Zhang, J. Tan, Z. Ma, W. Xu, G. Zhao, H. Geng, C. Di, W. Hu, Z. Shuai, K. Singh, and D. Zhu. Fullerene/sulfur-bridged annulene cocrystals: two- dimensional segregated heterojunctions with ambipolar transport properties and photoresponsivity. Journal of the American Chemical Society, 135(2):558–561, 2013.

[19] F. H. Herbstein. Crystalline molecular complexes and compounds: structures and principles. Oxford University Press, Oxford, 2005.

[20] T. Kodama and S. Kumakura. The Crystal structure of the 1: 1 complex of pyromellitic dianhydride with trans-stilbene. Bulletin of the Chemical Society of Japan, 47(5):1081–1084, 1974.

[21] B. E. Robertson and J. J. Stezowski. The crystal structure of the π-molecular complex of anthracene with pyromellitic dianhydride at 120C. Acta Crystallo- graphica Section B, 34(10):3005–3011, 1978.

54 [22] F. H. Herbstein and J. A. Snyman. The Crystal Structures at 110 and 300 K of the Equimolar Molecular Compound of Pyrene and Pyromellitic Dianhydride. Philosophical Transactions of the Royal Society of London. Series A, Mathemat- ical and Physical Sciences, 264(1157):635–662, 1969. [23] F. H. Herbstein, R. E. Marsh, and S. Samson. X-ray diffraction study of the crystal structure of the pi-molecular compound pyrene ... pyromellitic dianhydride at 19 K. Acta Crystallographica Section B, 50:174–181, 1994. [24] F. H. Herbstein and S. Samson. X-ray diffraction study of the disorder-to-order transition at ˜160 K in the pi-molecular compound pyrene...pyromellitic dianhydride. Acta Crystallographica Section B, 50(2):182–191, 1994. [25] C. C. Allen, J. C. A. Boeyens, and D. C. Levendis. Analysis of the disorder in room-temperature pyrene-pyromellitic dianhydride. South African Journal of Chemistry, 42(1):38–43, 1989. [26] I. V. Bulgarovskaya, V. K. Belsky, and V. M. Vozzhennikov. Structure of the 1:1 pi-molecular complex of 9,10-Dibromoanthracene with 1,2:4,5-Pyromellitic Dianhydride. Acta Crystallographica C, 43:768–770, 1987. [27] I. V. Bulgarovskaya, V. E. Zavodnik, and V. M. Vozzhennikov. Structure of the 1:1 pi -molecular complex of chrysene with 1,2:4,5-pyromellitie dianhydride. Acta Crystallographica C, 43:766–768, 1987. [28] J. C. A. Boeyens and F. H. Herbstein. Molecular compounds and complexes. III. The Crystal structures of the equimolar pi-molecular compounds of anthracene and perylene with pyromellitic dianhydride. The Journal of Physical Chemistry, 168(8):2160–2176, 1964. [29] N. Karl and J. Ziegler. Generation and transport of charge carriers in the charge- transfer complex anthracene-pyromellitic-dianhydride. Chemical Physics Letters, 32(3):438–442, 1975. [30] L. Zhu, Y. Yi, A. Fonari, N. S. Corbin, V. Coropceanu, and J.-L. Brédas. Electronic properties of mixed-stack organic charge-transfer crystals. Journal of Physical Chemistry C, 118:14150–14156, 2014. [31] A. S. Molinari, H. Alves, Z. Chen, A. Facchetti, and A. F. Morpurgo. High electron mobility in vacuum and ambient for PDIF-CN2 single-crystal transistors. Journal of the American Chemical Society, 131(7):2462–2463, 2009. [32] K. Shibata, K. Ishikawa, H. Takezoe, H. Wada, and T. Mori. Contact resistance of dibenzotetrathiafulvalene-based organic transistors with metal and organic electrodes. Applied Physics Letters, 92(2):023305, 2008. [33] K. Shibata, H. Wada, K. Ishikawa, H. Takezoe, and T. Mori. (Tetrathiaful- valene)(tetracyanoquinodimethane) as a low-contact-resistance electrode for organic transistors. Applied Physics Letters, 90(19):193509, 2007.

55 [34] T. Hasegawa and J. Takeya. Organic ﬁeld-eﬀect transistors using single crystals. Science and Technology of Advanced Materials, 10(2):024314, 2009.

[35] M. Kanno, Y. Bando, T. Shirahata, J. Inoue, H. Wada, and T. Mori. Stabilization of organic ﬁeld-eﬀect transistors in hexamethylenetetrathiafulvalene derivatives substituted by bulky alkyl groups. Journal of Materials Chemistry, 19(36):6548– 6555, 2009.

[36] M. Hiraoka, T. Hasegawa, Y. Abe, T. Yamada, Y. Tokura, H. Yamochi, G. Saito, T. Akutagawa, and T. Nakamura. Ink-jet printing of organic metal electrodes using charge-transfer compounds. Applied Physics Letters, 89(17):2004–2007, 2006.

[37] M. Hiraoka, T. Hasegawa, T. Yamada, Y. Takahashi, S. Horiuchi, and Y. Tokura. On-substrate synthesis of molecular conductor ﬁlms and circuits. Advanced Ma- terials, 19(20):3248–3251, 2007.

56 Chapter 3

Electrical Characterization of Organic Charge-Transfer Complexes

Determination of the charge-carrier mobility (or mobilities, in the case of ambipolar materials) is an important topic in the field of organic electronics. Here we discuss the calculation of mobility by space-charge-limited current (SCLC) measurements and organic field-effect transistor measurements (OFET). The former is a two contact measurement best suited for unipolar materials, while the latter is suitable for both unipolar and ambipolar materials. We compare the results of each type of measurement for an ambipolar material, α-DBTTF–TCNQ. We also use this analysis to discuss the optoelectronic properties of several CT crystal systems — the Perylene– TCNQ system, of which the 2:1 D:A ratio is novel, and two materials based on the acceptor PDIF-CN2.

57 3.1 Introduction

The charge-carrier mobility (µ) of an organic semiconductor is often regarded as a figure-of-merit for the material. This is due to its direct correspondence with switching speed when the material is used as the active layer in an organic field-effect transistor (OFET). Several methods are commonly used to determine the charge-carrier mobility in organic materials, with time-of-flight, space-charge-limited-current (SCLC), and OFET measurements being the most common. In this thesis, materials are characterized by the latter two methods — SCLC and OFET measurements. In addition to being relevant to technological applications, OFETs are preferable for mobility extraction because they provide a clear distinction between hole and electron transport. This type of characterization was used in this thesis for DBTTF–TCNQ polymorphs

(Chapter 5), Perylene–TCNQ crystals (Chapter 3), and Pyrene–PDIF-CN2 (Chapter 3). OFET fabrication is only possible, however, for single crystals with at least one smooth surface. The metal-insulator-metal structures necessary for SCLC analysis, on the other hand, can be fabricated with crystals of any size or shape. In this work, for example, the solution-grown STB–F4TCNQ crystal surfaces were too rough for the fabrication of OFETs; therefore, SCLC measurements were used. The disadvan- tage, though, is that the Mott-Gurney model derived to explain the dependence of current on voltage depends on several factors, including trap densities and type of charge-carrier (electrons, holes, or both). In the case that the latter concept is not clear for a particular material, errors can arise, as discussed in section 3.4. This chapter has two main objectives. First, the calculation of mobility by SCLC and OFET methods will be described. Second, we will include measurements on several materials systems containing novel CT crystals. These measurements were either not published or were part of a large collaboration of which I was not the lead author.

58 Figure 3.1: Metal-insulator-metal device structures. The sandwich structure is shown in (a), with L representing the distance between the contacts. The planar structure is shown in (b), with L representing the distance between the contacts and t representing the thickness of the channel.

3.2 Space-Charge-Limited Current

The SCLC method is a two-contact current-voltage measurement.1 The optimal device structure for this technique is the sandwich structure; it allows for current injection on the same crystal axis along which the electric field is induced by the application of voltage (Fig. 3.1a). The electric field is large and homogenous throughout the sample. With most organic crystals, however, the axis of interest is the π-stacking direction. This is typically the fast-growth direction of the crystal, making a sandwich structure along π-stack axis difficult to achieve. In this case, a pseudo-planar structure can be fabricated (Fig. 3.1b). Here, the potential difference applied via the planar contacts causes the electric field to have some non-direct elements, ne- cessitating the consideration of both the channel geometry and the anisotropy of the conductivity. In the case that the anisotropy is large and the ratio between the length (L) of the channel and thickness (t) of the crystal (L/t) is low, the Mott-Gurney law is applicable to the current-voltage relationship.2 As L/t increases, however, the so- called Geurst model is valid.3, 4 For pentacene single crystals, a cross-over from one regime to the other was shown, with the Mott-Gurney law being valid for L/t < 15 and the Geurst law for L/t > 250.5 For CT complexes, the electrical anisotropy is high, meaning the crossover will take place at higher values of L/t. For example,

the conductivity of DBTTF–TCNQ varies as 2 : 30 : 1 for σa : σb : σc, with the

59 b-axis being the stack axis of the crystal.6 Thus, only the Mott-Gurney law will be considered here.

3.2.1 Low-Voltage Regime

If the work function of the metal electrodes matches either the HOMO or LUMO of the semiconductor, they provide an Ohmic contact, and there is no injection barrier (i.e. Schottky barrier). In this case, assuming unipolar injection, (hole- or electron- only conduction), the low-voltage current regime follows the drift-diﬀusion equation, as shown for holes in eq. 3.1.

dV dp J = −qµ p − qD (3.1) p dx p dx

dV Here, −qµpp dx is the drift component of the equation, and the current varies as a function of electric ﬁeld. Variables in this equation are q, which is elementary charge,

dV µp, which is the hole mobility, p, which is the hole concentration, and E = dx , which is the electric ﬁeld (assumed to vary uniformly from one electrode to the other). The latter part of the equation represents the diﬀusion component, where current varies

dp as a function of carrier concentration ( dx ). The mobility is given by the Einstein

µpkT relation, Dp = q , where k is Boltzmann’s constant and T is temperature. Though there is some indication that diﬀusion provides a non-trivial contribution to current,7 in most cases it is signiﬁcantly smaller than the drift component and can therefore be ignored, leaving the current to consist of only the drift component (eq. 3.2).

J = qµppE (3.2)

Thus, knowledge of the charge-carrier mobility allows for calculation of the carrier concentration or vice-versa.

60 3.2.2 Space-charge-limited Current

At higher voltages, the density of injected carriers is high and the semiconductor is locally charged. This limits the current, hence this current-voltage regime is called “space-charge-limited”. As opposed to the Ohmic regime, where current is linear with respect to voltage (J ∝ V ), the current has quadratic dependence on voltage (J ∝ V 2). The formal relationship between them — i.e. the Mott-Gurney law — is derived below. First, consider that all injected charges are free, such that the charge density is given by ρ(x) = qp(x). As with the Ohmic regime, diﬀusion currents are disregarded, and

J = qµpp(x)E(x). (3.3)

The Poisson equation then becomes

dE(x) ρ(x) qp(x) = = , (3.4) dx εrε0 εrε0

where εr is the relative permittivity of the semiconductor and ε0 is the vacuum permittivity. Solving this for charge-carrier density (p(x)) and substituting it into 3.3 yields 1 dE2(x) J = µ ε ε . (3.5) 2 p r 0 dx

Integration over x from 0 to x with the assumption that E(0) = 0 gives a solution of

s 2Jx E(x) = . (3.6) µpεrε0

Finally, integration of the electric ﬁeld (V (x) = − R E(x0)dx0) yields the voltage of the channel. Imposing the boundary conditions implied by the measurement such that V (0) = V and V (L) = 0 (i.e. the grounded electrode) and solving for J gives

61 the well-known Mott-Gurney Law for SCLC conductivity:8, 9

9 V 2 J = µ ε ε Θ . (3.7) 8 p r 0 L3

The factor of Θ is the ratio of free to total charge carriers, and is added to account for trapping. It is typically estimated to be one, lowering the value for mobility. Several assumptions have been made to arrive at this current-voltage relation:

1. The transport is unipolar. Thus, only one type of charge-carrier contributes to the current — either holes or electrons, but not both.

2. All charges are injected, and the material has no intrinsic free charges.

3. Charge-carrier mobility (µp) and dielectric constant (εrε0) are constant throughout the sample.

4. The electric ﬁeld at the charge-injecting electrode is zero.

5. Diﬀusion current is neglected.

6. The metal injecting and collecting electrodes provide an Ohmic contact to the semiconductor.

In summary, the current at low voltages behaves according to Ohm’s law, while the current at higher voltages behaves according to the Mott-Gurney law. This is shown in Fig. 3.2. The crossover from one regime into the other can be understood if both types of current are considered present at all voltages. Then, the current can be divided as |JΩ JSCLC |JΩ = JSCLC |JΩ JSCLC |. The crossover-voltage (VΩ) can be found as shown in equation 3.9.

V 9 V 2 qpµ = ε ε µ (3.8) p d 8 r 0 p L3

62 Figure 3.2: Current-voltage characteristics of a unipolar semiconductor with Ohmic contacts. The black regime is Ohmic, while the red regime varies as the Mott-Gurney Law.

8 qpL2 VΩ = .. (3.9) 9 εrε0

If the device does not break down ﬁrst due to high voltage, increasing V past the SCLC regime will result in the ﬁlling of all traps, and a sudden increase of several orders of magnitude in current will occur. This current regime also has a quadratic voltage dependence.

3.3 Ambipolar Organic Field-Eﬀect Transistors

Charge-carrier mobilty can also be calculated based on the current-voltage characteristics of OFETS following the gradual-channel approximation as derived for inorganic metal-oxide-semiconductor ﬁeld-eﬀect transistors (MOSFETs).10 The use of this approximation for ambipolar transistors has been thoroughly discussed by several researchers;11–13 therefore, we only give an overview here. A bottom-gate, bottom- contact OFET is show in Fig. 3.3, with the channel length labeled as L and the channel width labeled as W . This device is operated by applying a potential dif-

ference between the gate and source (VGS) as well as between the drain and source

63 Figure 3.3: Organic Field-Eﬀect Transistor Schematic

(VDS) such that the source is grounded. The current in the channel is described by Ohm’s law, described here for electrons (for now, we assume unipolar electron-only operation):

J = qµenE. (3.10)

Here, µe is electron mobility, n is the electron density at the accumulation layer formed during operation, q is elementary charge, and E is electric ﬁeld. The value for n is

determined by the capacitance of the dielectric (Ci) and the eﬀective voltage, which is the gate voltage (VG) less the threshold voltage for the particular charge carrier (VT h,e)

and the voltage at a particular position in the channel(V (x)) (VG − VT h,e − V (x)).

Thus, n(x) = Ci(VG − VT h,e − V (x))/qt, where t is the accumulation layer thickness. Given the definition of current density J = I/W t (referencing Fig. 3.3 for W) and the definition of the electric field E = dV/dx, equation 3.10 becomes

I(x)dx = W µeCi(VG − VT h,e − V (x))dV. (3.11)

Integration of this equation over the length of the channel under the assumption that the potential varies linearly from source — x = 0 and V (0) = 0 — to drain — x = L

and V (L) = VDS — yields

W V 2 I = µ C (V − V )V − D . (3.12) D,lin L e,lin i G T h,e D 2

64 If the drain voltage is considerably smaller than the gate voltage, the quadratic term can be disregarded and

W I = µ C (V − V )V . (3.13) D,lin L e,lin i G T h,e D

This is the linear regime of the device operation, where ID ∝ VG. In this regime, the accumulation layer is roughly constant from the source to the drain; this is depicted in Fig. 3.4a, and is true for VD < VG − VT h,e. At VG < VT h,e the unipolar device is off, and no current flows. As VG increases and eventually VD = (VG − VT h,e), the channel pinches off at the drain electrode due to the opposing charges between the two (3.4b).

Beyond this point, as VG continues to increase, the current remains constant as the pinch-oﬀ point moves further towards the middle of the device (3.4c). This is the saturation regime and the current here is given by

W I = µ C (V − V )2. (3.14) D,sat 2L e,sat i G T h,e

When the bandgap of an organic semiconductor is small enough that the same type of source/drain electrode metal allows for injection of both holes and electrons, or if diﬀerent metals are used for the two contacts — one hole-injecting and the other electron-injecting — the device is ambipolar. To understand the charge-transport in this case, ﬁrst consider that there are now two threshold voltages — one for electrons

(VT h,e) that is usually positive and one for holes (VT h,h) that is usually negative.

Now, for VD < (VG − VT h,e), the device exhibits unipolar, linear electron transport, just as in the discussion above. This process is depicted in Fig. 3.5, for the device

with the red channel. As VD approaches VG − VT h,e and eventually comes to equal it (VD = VG − VT h,e), the electron current saturates, again as it did in the above discussion (Fig. 3.5, cyan channel). The threshold voltage for holes is close enough

65 Figure 3.4: A qualitative depiction of the accumulation layer. The channel at low drain voltages (the linear regime) is shown in (a), the pinch-oﬀ point is shown in (b), and the saturation regime is shown in (c). This ﬁgure was inspired by Zaumseil, et al.12

to that for electrons, though that further increases in VD cause holes to be injected from the opposite electrode, and the current is now a sum of the saturated hole and saturated electron currents (i.e. both are pinched-oﬀ at a point between the electrodes, as in Fig. 3.5, purple). Quantitatively, the current in the linear electron transport regime is given by equation 3.12 and the current in the saturated electron transport regime is given by equation 3.14. The linear hole current, then, is

W V 2 I = µ C (V − V )V − D , (3.15) D,lin L h,lin i G T h,h D 2

66 and the saturation current is

W I = µ C (V − V − V )2. (3.16) D,sat 2L h,sat i G T h,h D

Note that this is reverse hole-only operation and is true for positive VG — hence the

dependence on drain voltage. If the device were biased such that VG were negative, the current would be independent of the drain voltage, as it is for electrons in equation 3.14. The ambipolar current, then, is a sum of the saturation current for each charge carrier:

W W I = µ C (V − V − V )2 + µ C (V − V )2. (3.17) D,sat 2L h,sat i G T h,h D 2L e,sat i G T h,e

These equations were developed with several assumptions, the most obvious being the assumption of a linear electric ﬁeld across the channel. This is most certainly not the case.14 In addition, the hole and electron mobilities may not be constant with increasing drain voltage, which would cause further deviation from the model. These and other considerations are taken into account by Smits, et al., and Kang, et al.11, 13

67 Figure 3.5: Ambipolar OFET operation. The graph in (a) shows the transport features, with the red current being linear electron transport, the cyan current being saturated electron transport, and the purple current being ambipolar transport. The graph in (b) shows the transfer features, with the orange current being linear hole current and the green current being saturated hole current. The other colors are the same as in (a). This plot was developed for a device with equal hole and electron 13 mobilities, VT h,e = 5V , and VT h,h = −5V . The graphs were inspired by Kang, et al. The accumulation layers for the condition shown in the transport plot are depicted in the devices to the right.

68 3.4 Comparison of Mobility Calculation by SCLC and OFET Measurements for an Ambipolar CT Complex

A concluding point to take from sections 3.2 and 3.3 is that OFET analysis allows for easy determination of hole versus electron transport in the case of an ambipolar device, whereas SCLC measurements are suitable for unipolar materials. In the case that an SCLC measurement is performed on an ambipolar material, however, it may not be evident that the model fails, as the current may still follow the power law depicted in Fig. 3.2. The material α-DBTTF–TCNQ lends itself to this discussion. OFETs based on single crystals of this material were fabricated in the top-gate, bottom- contact structure, with a 650 nm Parylene-N gate dielectric, and Ti/Au source and drain electrodes (5nm,e-beam/45nm,thermal evaporation). The gate electrode was gold. The resulting current dependences on drain voltage and gate voltage are shown in Fig. 3.6a and b respectively. Analysis of these by the discussion given in section 3.3

2 −1 −1 2 −1 −1 yielded charge-carrier mobilities of µe = 0.15 cm V s and µh = 0.008 cm V s . Using this same device but leaving the gate electrode at a ﬂoating potential allowed us to measure the two-contact current-voltage dependence as discussed in section 3.2.

Figure 3.6: Transport (a) and transfer (b) plots for a DBTTF-TCNQ single crystal. The drain voltage in (b) is VD = 60V .

69 These results are shown in Fig. 3.7. As it can be seen, current density closely follows the power law described previously, despite the fact that two charge-carriers were present in the OFET measurement. Use of the Mott-Gurney law then yields an absurdly high mobility of 85 cm2V−1s−1(estimating the thickness of the crystal to be 1 µm, with L = 100 µm and W = 90 µm). In fact, in other samples, a mobility upwards of 200 cm2V−1s−1 was calculated. Several methods of reasoning may apply towards understanding this measurement. First, as Kao and Hwang discuss,15 the current for two injected carriers where one is signiﬁcantly larger than the other may go as equation 3.18, where the eﬀective mobility is simply the larger of the two.

9 V 2 J = µ ε ε (3.18) 8 eff r 0 L3

In the case of this crystal, the electron mobility is one order of magnitude larger than the hole mobility. Furthermore, the transport plot in Fig. 3.6 indicates that the electron injection occurs with no barrier (otherwise the low-voltage regime would

Figure 3.7: SCLC curve for a DBTTF-TCNQ single crystal. The low-voltage regime appears to be Ohmic with a slope of 1, while the high-voltage regime exhibits J ∝ V 2. The mobility calculated based on the Mott-Gurney law is 85 cm2V−1s−1.

70 be non-linear). Given a bandgap of about 0.8 eV,16 if there is no barrier present for electrons, there is likely a non-trivial barrier present for hole injection. The electron mobility, then, would be 85 cm2V−1s−1, while the hole mobility could only be measured by the use of an electron blocking layer or contacts with a work function close to the HOMO of DBTTF–TCNQ. If this is the case, the question of why the electron mobility is three orders of magnitude larger in the bulk (Mott-Gurney) case than in the OFET case must be explained. Indeed, signiﬁcantly more scattering events take place at the OFET dielectric/semiconductor interface than in the bulk crystal, which would imply that a lower mobilty is expected. It is not possible, however, that this phenomenon can explain the three order-of-magnitude observed diﬀerence between the FET and SCLC mobilities. Therefore, we looked into other possible explanations. In the case of both electrons and holes being present, the current in the channel is not only dependent on the applied bias, but also on recombination events that occur in the bulk of the crystal. This is discussed by Pope and Swenburg.2 Here, they consider that the drift current is now given by equation 3.19:

J = qE(µhpf + µenf ), (3.19)

where pf and nf are the densities of free holes and electrons respectively. Derivation of a current-voltage dependence again starts with the use of Poisson’s equation, as in section 3.2. The ﬁnal result, however, is one of about 300 equations; thus we will not present the full derivation here. A possible relationship is given by equation 3.20:

1/2 2 (µeµh)(µh + µe) V J = 1.31εrε0 3 , (3.20) µ0 L

where µ0 is the recombination mobility, given in terms of the velocity (v) and scat-

71 tering cross-sections (s) of the charge carriers such that

ε ε vs µ = r 0 . (3.21) 0 2q

An estimation of µ0 based on equation 3.20, the hole and electron mobilities from OFET measurements, and ε = 3 (a typical value for organic semiconductors), yields a

−9 2 −1 −1 value of µ0 = 2x10 cm V s . Further measurements are planned towards relating this value with the physical processes taking place in this material. Although, as we discussed previously, the gold contacts appear to be Ohmic for electrons, the band-bendings in the space-charge accumulation regime would be large enough for holes to overcome the energetic barrier to injection, explaining their presence in the measurement. This result is promising, but needs further analysis. In summary, obeying the power law alone is not enough reason to employ the Mott-Gurney law to determine a value for mobility. Further analysis of the sample in question is needed to understand the results.

3.5 Charge-Transport in Novel Organic Charge- Transfer Complexes

3.5.1 Perylene–TCNQ in Three D:A Stoichiometries

Following our discussion of the device physics of mobility calculation, in this section we present results on novel CT crystals systems that were not included in a stand- alone chapter. As discussed in Chapter 2, the CT complex Perylene–TCNQ can be grown in three diﬀerent D:A ratios. The ratios are 1:1, 2:1, and 3:1. The 1:1 and 3:1 ratios were reported previously; the 2:1 ratio had not been reported prior to 2014.17–20 Crystals of each of these stoichiometries and their corresponding structures as determined by X-ray diﬀraction (XRD) are shown in Fig. 3.8. As it can be seen,

72 Figure 3.8: Crystals and structures of the Perylene–TCNQ CT complex system. The 1:1 complex grows as green-blue needles, and exhibits classic mixed-stack D/A organization. The 2:1 complex grows as green, elongated platelets and exhibits a mixed-stack structure with an extra donor molecule in the stack. The 3:1 complex grows as green platelets or blocks, and exhibits a structure similar to the 2:1 complex but with an additional molecule in the interstitial space between D/A stacks. the crystals exhibit mixed-stack D/A organization; the diﬀerences are that the 2:1 complex has an additional donor molecule in the interstitial stack space, and the 3:1 complex has an additional donor molecule in the stack axis.

Electrical measurements were performed by preparing OFETs from the crystals (Fig. 3.8). Several device architectures were fabricated for each crystal stoichiometry; the combination that yielded the best performance will be discussed. The 2:1 and 3:1 crystals were ﬂat, thin platelets, and were therefore amenable to a pre-fabricated bottom-gate bottom-contact OFET geometry. Here, highly n-doped Si wafers were employed as the gate electrode, with 200-nm thermally-oxidized SiO2 as the gate dielectric. Gold source and drain contacts (60 nm, with a 10 nm Ti adhesion layer) were deﬁned by photolithography and deposited by e-beam evaporation. These test

73 beds were cleaned for 10 min in a hot acetone bath and 10 min in a hot isopropanol bath. This was followed by 10 min in a UV Ozone cleaner and a thorough rinse with DI water. After cleaning, crystals were laminated by hand on the surface of the test beds, adhering electrostatically to the dielectric surface. The 1:1 crystals grew as needles and were too thick to laminate, so a bottom-gate top-contact OFET geometry was employed instead of the bottom-gate bottom-contact geometry used for the platelets. A similar structure was also tested for the case of the 2:1 complex. Here, the crystals were placed on heavily-doped Si wafers (the gate electrode) with

200 nm of thermally-oxidized SiO2 for the gate dielectric (cleaned in the same way as the previous substrates), and Ag epoxy was painted as the source and drain top contacts on the crystals. The samples were electrically characterized in air at room temperature using an Agilent 4155C Semiconductor Parameter Analyzer. All crystals were evaluated for both electron and hole transport. The charge-carrier mobility of the transistors was calculated as described in the section 3.3. As can be observed from the transfer curves depicted in Fig. 3.9, we found that the electrical properties of the perylene–TCNQ charge-transfer compounds were strongly dependent on stoichiometry. The 1:1 crystals (in this particular example, L = 160 µm and W = 60 µm) exhibited an electron

−3 2 −1 −1 mobility of µe = 1.2 x 10 cm V s when measured in air, with no detectable

Figure 3.9: Current-voltage characteristics for the Perylene–TCNQ crystal system. The 1:1 crystals were n-type, the 2:1 crystals were ambipolar, and the 3:1 crystals where p-type.

74 activity for hole transport. The 2:1 crystals (device dimensions L = 100 m, W = 275 m) displayed ambipolar charge transport characteristics, with similar electron and

−5 2 −1 −1 −5 2 −1 −1 hole mobilities: µe = 2.9 x 10 cm V s and µh = 7.4 x 10 cm V s . Note that P2T1 exhibited balanced transport but lower mobilities when Au contacts were

−7 2 −1 −1 −7 2 −1 −1 used (µe = 2.1 x 10 cm V s and µh = 7.5 x 10 cm V s ), probably because of contact resistance and lamination imperfections. Unfortunately, the limited number of available P2T1 crystals precluded extensive device optimization. The 3:1 crystal (device dimensions L = 80 m, W = 65 m) exhibited a hole mobility of µh = 5.5 x 10−5 cm2V−1s−1and no electron conduction. With the exception of the 2:1 crystals the mobilities reported here are representative values for each stoichiometry and were measured in several crystals. Interestingly, the degree of charge-transfer (q) changes as a function of stoichiometry. The 1:1 complex appeared almost neutral with q = 0.01 as measured by the method of Flandrois, et al.21 The 2:1 complex demonstrated q = 0.12, and the 3:1 complex was almost double the 2:1 at q = 0.23. Values calculated from Raman spectroscopy were comparable.19

3.5.2 Anthracene and Pyrene with Acceptor PDIF-CN2

The crystal growth of CT complexes Anthracene–PDIF-CN2 and Pyrene–PDIF-CN2 was discussed in Chapter 2. Here, we discuss the optoelectronic properties of each complex. Due to the novelty of these complexes, we were interested in obtaining information about their electronic structures. This could be done through the use of optical reﬂectance. Here, a combined grating monochromator and Cassegrain- type microscope was used to shine light on the crystal surface ranging in photon energy from about 1 eV to about 3 eV — i.e. infrared through violet light. The measured reﬂected light increases in intensity as the energy approaches that of the material bandgap, allowing us to estimate the HOMO-LUMO gap of the complex.

75 Figure 3.10: Optical reﬂectance of Anthracene (bottom) and Pyrene–PDIF-CN2 (top). The black lines correspond to incident light polarized such that the spectra is maximized, while the red lines correspond to the minimum spectra.

The results for both complexes are shown in Fig. 3.10 for light polarized so that the reﬂectance is maximized (black, parallel to the long crystal axis) and minimized (red, parallel to the short crystal axis). As it can be seen, the charge-transfer band for each complex is approximately 1.5 eV. For comparison, the charge-transfer band of DBTTF–TCNQ, is roughly half this value at 0.8 eV.16 This somewhat large value indicates that ambipolar transistors will not be easily fabricated from these materials. Interestingly, the peak does not extinguish as the polarization is rotated through the crystal axis. As discussed in Chapter 1, the bandgap of a CT complex is the energetic gap between the donor HOMO and the acceptor LUMO to ﬁrst-order approximation. Thus, the corresponding transition should be observed only in the stack-axis direction, and not perpendicular to it, as we see here. This result indicates that there is likely disorder present in the crystals. Calculation of the transfer integrals along the stack axis and from stack to stack agree with this assumption. As shown in Fig. 3.11, for

76 Figure 3.11: Transfer integrals parallel to (t1) and perpendicular to (t2) the D/A stack axis in anthracene–PDIF-CN2.

Anthracene–PDIF-CN2, the transfer integral is about 25 times higher in the direction of the D/A stacks than perpendicular to it. Results were similar for Pyrene–PDIF-

CN2. We attempted to fabricate OFETs based on these materials, and succeeded for

Pyrene–PDIF-CN2. These results are shown in Fig. 3.12. The current in these devices is low; this can be due to large contact resistances, as suggested by the transport graph, or related to the large amount of disorder implied by the reﬂectance spectra. The estimated mobility is 10−4 cm2V−1s−1. Thus, while the novelty of the complexes makes them interesting, their resulting optoelectronic properties suggest that they are not strong candidates for OFETs.

3.6 Summary

In summary, we described the calculation of charge-carrier mobility for two-contact metal-insulator-metal samples by SCLC measurements and for unipolar and ambipolar semiconductors through the use of OFETs. The former will be used in the following chapter to assess the temperature-dependent electrical characteristics of the CT complex STB–F4TCNQ. The latter will be used to compare polymorphs of the CT

77 Figure 3.12: The transfer (left) and transport (right) properties of Pyrene–PDIF- CN2, showing low electron currents. complex DBTTF–TCNQ in Chapter 5, as the crystals were amenable to OFET fabrication. We were also able to fabricate OFETs of the Perylene–TCNQ CT complex system, and found that the transport properties varied from n-type, to ambipolar, to p-type with the addition of donor. Finally, we discussed the optoelectronic properties of Anthracene and Pyrene–PDIFCN2 and found that the crystals were dominated by disorder. This does not, however, rule out perylene-diimide-based molecules as electron acceptors in other CT crystals.

78 References

[1] O.D. Jurchescu. Conductivity measurements of organic materials using ﬁeld- eﬀect transistors (FETs) and space-charge-limited current (SCLC) technique. In Handbook of Organic Materials for Optical and (Opto)electronic Devices, pages 377–397. Elsevier, 2013.

[2] M. Pope and C. E. Swenberg. Electronic Processes in Organic Crystals and Polymers. Oxford University Press, New York, second edition, 1999.

[3] J. A. Geurst. Theory of space-charge-limited currents in thin semiconductor layers. Physica Status Solidi, 15(1):107–118, 1966.

[4] R. Zuleeg and P. Knoll. Space-charge-limited currents in heteroepitaxial ﬁlms of silicon grown on sapphire. Applied Physics Letters, 11(6):183–185, 1967.

[5] O. D. Jurchescu and T. T. M. Palstra. Crossover from one- to two-dimensional space-charge-limited conduction in pentacene single crystals. Applied Physics Letters, 88(12):122101, 2006.

[6] Y. Takahashi, J. Hasegawa, Y. Abe, Y. Tokura, K. Nishimura, and G. Saito. Tun- ing of electron injections for n-type organic transistor based on charge-transfer compounds. Applied Physics Letters, 86(6):063504, 2005.

[7] J. Dacu˜naand A. Salleo. Modeling space-charge-limited currents in organic semiconductors: Extracting trap density and mobility. Physical Review B, 84(19):195209, 2011.

[8] N. F. Mott and R. W. Gurney. Electronic Processes in Ionic Crystals. Oxford University Press, 2nd edition, 1948.

[9] K. C. Kao and W. Hwang. Electronic Transport in Solids. Pergamon Press, 1st edition, 1981.

[10] S.M. Sze and Kwok K. Ng. Physics of Semiconductor Devices. 3rd edition, 2007.

[11] E. Smits, T. Anthopoulos, S. Setayesh, E. van Veenendaal, R. Coehoorn, P. Blom, B. de Boer, and D. de Leeuw. Ambipolar charge transport in organic ﬁeld-eﬀect transistors. Physical Review B, 73(20):205316, may 2006.

79 [12] Jana Zaumseil and Henning Sirringhaus. Electron and ambipolar transport in organic ﬁeld-eﬀect transistors. Chemical Reviews, 107(4):1296–1323, 2007.

[13] M. S. Kang and C. D. Frisbie. A pedagogical perspective on ambipolar FETs. ChemPhysChem, 14(8):1547–1552, 2013.

[14] E. C. P. Smits, S. G. J. Mathijssen, M. Cölle,A. J. G. Mank, P. A. Bobbert, P. W. M. Blom, B. De Boer, and D. M. De Leeuw. Unified description of potential profiles and electrical transport in unipolar and ambipolar organic field-effect transistors. Physical Review B, 76(12):125202, 2007.

[15] KC Kao and W Hwang. Electrical Transport in Solids. Pergamon Press, 1981.

[16] Y. Takahashi, T. Hasegawa, Y. Abe, Y. Tokura, and G. Saito. Organic metal electrodes for controlled p- and n-type carrier injections in organic ﬁeld-eﬀect transistors. Applied Physics Letters, 88(7):073504, 2006.

[17] I. J. Tickle and C. K. Prout. Molecular complexes. Part XVII. Crystal and molecular structure of perylene7, 7, 8, 8-tetracyanoquinodimethane molecular complex. Journal of the Chemical Society, Perkin Transactions 2, 6:720–723, 1973.

[18] A. W. Hanson. 7,7,8,8-Tetracyanoquinodimethane(perylene)2-Perylene. Acta Crystallographia B, 34:2339–2341, 1978.

[19] D. Vermeulen, L. Y. Zhu, K. P. Goetz, P. Hu, H. Jiang, C. S. Day, O. D. Jurch- escu, V. Coropceanu, C. Kloc, and L. E. McNeil. Charge Transport Properties of PeryleneTCNQ Crystals: The Eﬀect of Stoichiometry. The Journal of Physical Chemistry C, 118(42):24688–24696, 2014.

[20] P. Hu, L. Ma, K. Tan, H. Jiang, F. Wei, C. Yu, K. P. Goetz, O. D. Jurchescu, L. E. McNeil, G. G. Gurzadyan, and C. Kloc. Solvent-Dependent Stoichiometry in Perylene7,7,8,8-Tetracyanoquinodimethane Charge Transfer Compound Single Crystals. Crystal Growth & Design, 14(12):6376–6382, 2014.

[21] S. Flandrois and D. Chasseau. Longueurs de liaison et transfert de charge dans les sels du t´etracyanoquinodim´ethane(TCNQ). Acta Crystallographica Section B, 33:2744–2750, 1977.

80 Chapter 4

The Eﬀect of Librational Motion on Charge Transport in Stilbene–F4TCNQ

The crystalline structure of organic materials dictates their physical properties, but while signiﬁcant research is geared towards understanding structure-property relationships in such materials, the details remain unclear. Many organic crystals exhibit transitions in their electrical properties as a function of temperature. One example is the 1:1 charge-transfer complex trans-stilbene–2,3,5,6-tetraﬂuoro-7,7,8,8- tetracyanoquinodimethane. Here we show that the mobility and resistivity of this material undergo a transition from being thermally activated at temperatures above 235 K to being temperature independent at low temperatures. On the basis of our experimental and theoretical results, we attribute this behavior to the presence of a glass-like transition and the accompanied freezing-in of orientational disorder of the stilbene molecule.

This portion of the text is adapted from K. P. Goetz, A. Fonari, D. Vermeulen, P. Hu, H. Jiang, P. J. Diemer, J.W. Ward, M. E. Payne, C. S. Day, C. Kloc, V. Coropceanu, L. E. McNeil, and O. D. Jurchescu. “Freezing in orientational disorder induces crossover from thermally-activated to temperature-independent transport in organic semiconductors, Nature Communications 5, 5642 (2014). The author was responsible for or participated in the crystal growth, electrical characterization, structural characterization, combining all parts of the project, and writing the manuscript.1

81 4.1 Introduction

Charge-transfer complexes are often subject to phase transitions as a function of temperature and pressure, resulting in changes in structre and/or electrical properties at a critical temperature or pressure. Examples include metal-insulator transitions, neutral-ionic transitions, Peierls transitions, and transitions to a ferroelectric state.2–11 This chapter discusses the effect of thermally-induced static and dynamic structural modifications on the electronic and electrical properties of the CT complex composed of the donor trans-stilbene (STB) (Fig. 4.1a, Fig. 4.1b) and the acceptor 2,3,5,6- tetrafluoro-7,7,8,8-tetracyanoquinodimethane (F4TCNQ) (Fig. 4.1c, Fig. 4.1d) by using a combined experimental and theoretical approach. This binary crystal grows in a 1:1 ratio, crystallizes in a mixed-stack array along the a-axis, and will be referred to as STB–F4TCNQ (Fig. 4.1e, Fig. 4.1f).

We have discovered that STB–F4TCNQ exhibits a crossover from temperature- independent to thermally-activated charge transport near 235 K. Similar transitions have been observed recently in a series of semiconducting polymers12, 13 and in the small-molecule organic semiconductor 6,13-bis-triisopropyl-silylethynyl pentacene

14, 15 (TIPS-pentacene). In STB–F4TCNQ, by combining a previous thermodynamic study on this system16 with our X-ray diffraction (XRD) and optical measurements, we are able to attribute the observed behavior to a glass-like transition. (Here, the term “glass-like transition” is adopted for a crystalline solid with some degrees of freedom frozen-in, as described in detail in other reports.17–23) This transition does not distort the crystal lattice, but is accompanied by the freezing in of orientational disorder related to the torsional oscillation (libration) of the double bond connecting the C7 and C7a carbons in the STB molecule (Fig. 4.1a). Below the transition temperature the librational motion is not sufficiently strong to mediate transitions between the states of different orientations. Librational motion is ubiquitous

82 Figure 4.1: (a) The donor stilbene skeletal structure with the atoms labeled to corresponde with the crystallographic structure. (b) An optical image of a stilbene single crystal. (c) The skeletal structure of the acceptor F4TCNQ with bond lengths relevant to the calculation of degree of charge transfer (q) labeled. (d) An optical image of F4TCNQ crystals. (e) The CT complex STB–F4TCNQ crystal structure that exhibits mixed stacking in the a direction. (f) An optical image of the STB-F4TCNQ crystal. in organic semiconductors, and was predicted to localize charge carriers in organic crystals.24 While it was recently measured in TIPS-pentacene by thermal diffuse electron scattering, this motion has been difficult to connect with experimental observations of a materials electronic properties.25 We find that this librational motion and temperature-dependent phase transition is also present in neat stilbene crystals and another stilbene-based CT complex. Our results provide direct insight into the effect of libration on electrical properties of organic molecular crystals, and show that its change in amplitude can result in a crossover from thermally-activated to a temperature-independent charge transport.

83 4.2 Methods

4.2.1 Crystal Growth

The physical vapor transport crystal growth method was employed to grow single crystals when possible, as discussed in chapter 2.26, 27 This method was successful for

the growth of F4TCNQ (Fig. 4.1d), TCNQ, and stilbene single crystals (Fig. 4.1b).

Here, the commercial F4TCNQ (J&K Scientific or Sigma Aldrich) was heated to 220 ◦C under Argon flow. Platelet-like, yellow single crystals of a few millimeters in size formed in a quartz tube at the place where the temperature abruptly fell from 220 ◦C to room temperature. TCNQ similarly grew at 200 ◦C, and stilbene grew at 120 ◦C. The vapor pressures of the donor and acceptor were too different to grow the STB–

F4TCNQ crystals from the gas phase, so they were grown from solution instead. A few mg of trans-Stilbene (Sigma Aldrich) and F4TCNQ were weighed separately so that their molar masses were approximately 1:1. Each was dissolved in hot acetonitrile. The solutions were mixed and loosely covered to allow for slow solvent evaporation. Small, green, needle-like crystals of the 1:1 CT complex (Fig. 4.1f) formed after one- two days. The crystals were washed with acetonitrile and ﬁltered prior to use. We conﬁrmed the structures of F4TCNQ and STB–F4TCNQ using single-crystal XRD measurements.28

4.2.2 Structure Determination

Multiple datasets for single crystals of F4TCNQ and STB–F4TCNQ were collected over a temperature range from 123 K to 297 K. Crystals of F4TCNQ are orthorhombic, group Pbca with Z=4; crystals of STB–F4TCNQ are monoclinic, space group P 21/n

(an alternate setting of P 21/c) with Z= 2 (formula units). A full-hemisphere of diﬀracted intensities (2424 total 20-, 30- or 40-second frames with an ω scan width of 0.30◦) was measured for each dataset using graphite-monochromated MoKα¯ radiation

84 (λ = 0.71073 A)˚ on a Bruker APEX CCD Single Crystal Diffraction System. X-rays were provided by a fine-focus sealed x-ray tube operated at 50 kV and 30 mA. Lattice constants were determined with the Bruker APEX2 software package and integrated reflection intensities having 2θ (MoKα¯) ≤ 60.00◦ were produced for each dataset. The structures were solved using “direct methods” techniques and all stages of

2 weighted full-matrix least-squares refinement were conducted using Fo data with the SHELXTL software package.29 The resulting structural parameters for all datasets were refined to convergence using counter-weighted full-matrix least squares techniques and structural models which incorporated anisotropic thermal parameters for all non-hydrogen atoms. In STB–F4TCNQ, ring hydrogen atoms were included in the structural model as fixed atoms (using idealized sp2- hybridized geometry and C-H bond lengths of 0.93 – 0.95 A)˚ “riding” on their respective carbon atoms. The hydrogen atom on the ethylene group was located from a difference Fourier map and was refined as an independent isotropic atom. The isotropic thermal parameters for the ring hydrogen atoms were fixed at a value 1.2 times the equivalent isotropic thermal parameter of the carbon atom to which they are covalently bonded. For the F4TCNQ and STB–F4TCNQ structures, totals of 91 and 158 parameters, respectively, were re-

ﬁned using no restraints. Diﬀerence electron density contour maps (Fo –Fc) were generated using the Bruker program XP. Probability ellipsoid plots are included in Fig. 4.2.

4.2.3 Sample Fabrication and Electrical Characterization

Carbon paste (IPA base, Ted Pella) was painted on both ends of the long axis of the crystals to form an in-plane measurement structure. The crystals were placed on clean Si/SiO2 to provide good electrical insulation. Current-voltage characteristics were measured in vacuum (10−6 Torr) and in the dark using an Agilent 4155C Semiconductor Parameter Analyzer. Current was measured while decreasing the tem-

85 Figure 4.2: Thermal ellipsoid plots of STB–F4TCNQ. perature at 0.5 K/min; measurements were taken every 10 K outside the region of the phase transition and with higher resolution around the phase transition. The crystals were stable through multiple rounds of measurement. Neat STB and STB–TCNQ crystals were prone to disintegration at atmospheric pressure and room temperature — the stilbene molecule is volatile and leaves the lattice even in the case of the CT crystal. Therefore, these crystals were cooled as soon as possible and measurements were taken while heating. The STB–F4TCNQ crystal is not prone to this.

4.2.4 IR and Raman Spectroscopy

Infrared absorption measurements were made using a Nicolet Nexus 670 FTIR spectrometer with a spectral range of 400-4000 cm−1. The small size of the crystals

86 made single-crystal measurements impractical, so STB–F4TCNQ crystals were lightly crushed together and placed on a copper TEM grid with 54 micron grid-hole size. Temperature-dependent measurements were made using an MMR Technologies Joule- Thomson refrigerator over the temperature range 80-300 K. Temperature-dependent Raman measurements were carried out with an excitation wavelength of 532 nm using an MMR Technologies Joule-Thomson refrigerator over the temperature range 80-300 K. The needle-shaped crystals were placed on the cooling stage and held in place with a speck of vacuum grease. Very low laser powers were used (0.5 mW over a 10 µm spot) because the crystals are sensitive to laser-induced damage.

4.2.5 Computational Methodology

Geometry optimizations of the crystal structures of STB–F4TCNQ at different temperatures were performed at the B3LYP/6-31G level of theory. During the optimizations, the unit cell parameters were kept fixed at the experimental values. A uniform 6x8x4 Monkhorst-Pack k-point grid was employed. Effective masses were obtained by diagonalizing the inverse effective mass tensor. Energy derivatives with respect to the k-point that form the tensor were evaluated using the finite-difference method with 0.01 Bohr−1 step. The -point phonon modes were obtained using the finite- difference method with a 0.003 Aatomic˚ displacement step. These calculations were carried out using the CRYSTAL09 package.30 Transfer integrals (electronic couplings) between the HOMO of stilbene and the LUMO of F4TCNQ were calculated at the B3LYP/6-31G(d,p) level using a fragment orbital approach in combination with a basis set orthogonalization procedure31 as implemented in the development version of the NWChem package.32

87 4.3 Results

4.3.1 Electrical Properties

We evaluate the electrical properties of the STB–F4TCNQ crystals by space-charge- limited current (SCLC) measurements, as described in Chapter 3. Although more than ﬁve crystals have been studied, we report here the results from one representative sample. The resistivity (ρ) is calculated from the ohmic region of the I-V curve (Fig. 4.3). The mobility (µ) is determined from the Mott-Gurney law, shown in equation

2 4.1, in the region where the current density, JSCLC , is proportional to V :

9 V 2 J = µεε θ (4.1) SCLC 8 0 L3

Here, ε is the relative permittivity of the semiconductor (approximated to be 3),

ε0 is the permittivity of free space, and θ is the ratio of free charge carriers to total charge carriers (estimated to be 1, so the resulting mobility is a low estimation). As seen in Fig. 4.3, the proportionality of J to V validates our use of both Ohms law and

Figure 4.3: Current-voltage characteristics of STB–F4TCNQ contacted with two electrodes at 300 K. The slope in the low-voltage region is 0.85 and the slope in the high-voltage region is 2.01.

88 the Mott-Gurney Law. The value of L/t (where L is the channel length and t is the thickness of the crystal) lies in the range 10-30, which is suﬃciently low to validate our use of the Mott-Gurney law instead of the Guerst law.33 Figure 4.4 shows the resistivity (blue, right axis) and mobility (black, left axis) of

STB–F4TCNQ as a function of temperature. The resistivity and the mobility each independently exhibit an inﬂection point around T = 235 K. In the high-temperature region (T > 235 K), labeled I, the resistivity decreases with increasing temperature and the mobility is thermally activated with an activation energy EA = 0.17 eV. At low temperatures (T < 235 K), marked by region II, the charge transport is temperature independent. Interestingly, Saito and coworkers discovered a signature for a thermodynamic transition near the same temperature by measuring the heat

16 capacity of STB–F4TCNQ as a function of temperature. It is therefore plausible to assume that both the change in electrical properties and the phase transition occur as a result of the same mechanism.

Figure 4.4: The temperature dependence of the electrical characteristics of STB– F4TCNQ. The regions here are divided into I, where the transport is activated, and II, where the charge transport is temperature independent. Right axis, in blue: resistivity; left axis, in black: mobility.

89 4.3.2 Crystal Structure and Electronic Structure Calcula-

tions

To determine the nature of the transition, we ﬁrst evaluate the crystal structure of the CT complex as a function of temperature using X-ray diﬀraction. The temperature dependence of the lattice parameters of STB–F4TCNQ is plotted in Fig. 4.5a. At all investigated temperatures, the system crystallizes in the monoclinic space group

P21/n with two STB molecules and two F4TCNQ molecules per unit cell, each located at an inversion center. The unit cell varies by the expected linear thermal expansion; 1.1% along all three unit cell axes and 3.1% for the total unit cell volume, with no evidence of the presence of a structural phase transition. Subsequently, based on the temperature-dependent structures, we have performed band structure calculations and evaluated the effective masses for both holes and electrons. The results are shown in Fig. 4.5b. Our calculations indicate that charge carriers have small effective masses only along the stacking directions, and that both hole and electron effective masses increase

Figure 4.5: The temperature dependence features of the CT complex lattice. (a)The temperature dependence of the lattice parameters of STB–F4TCNQ. The lattice expansion is approximately linear with temperature. (b) The temperature dependence of the smallest component of the eﬀective masses of holes and electrons as derived from band-structure calculations using the lattice parameters.

90 systematically with the increase in temperature. This dependence is consistent with the linear thermal expansion of the crystal. As the cell parameters increase, the

distance between the STB and F4TCNQ molecules also increases. Consequently, the

calculations suggest that the transfer integral (tHL) from the HOMO (highest occupied molecular orbital) of STB to the LUMO (lowest unoccupied molecular orbital) of the

neighboring F4TCNQ molecule decreases, resulting in an increase in the eﬀective mass of both holes and electrons. Thus, the temperature dependence of the transfer integrals and eﬀective masses derived from the static structure are not able to provide an explanation for the features observed at T = 235 K in the electrical measurements.

4.3.3 Structure and Thermodynamics

Stilbene and related compounds have been the subject of intense research scrutiny for a number of years. In particular, static and dynamical disorder are known to play a significant role in the structure and properties of stilbene-containing compounds. Single-crystal XRD measurements of neat STB and the CT complex STB–TCNQ show that the STB molecule can adopt two orientations that are related to one another by a two-fold rotation about the longest molecular axis.34–36 Librational motion can responsible for inter-conversion between these two conformers and can be viewed as a rotation of the central C=C stilbene moiety at fixed orientation of the benzene rings relative to the crystal lattice and is referred to as a pedal-like (crankshaft) motion (Fig. 4.6). In the neat STB crystal, this motion results in an orientational disorder.37, 38 Variable-temperature structural studies indicate that this disorder is dynamic — i.e. conformational inter-conversion is temperature dependent (and as a result the conformer populations are also), with the occupancy factor of the minor, energetically-unfavorable conformer, decreasing as the temperature is lowered.34, 37–39 Below 170 K the pedal motion becomes inactive and thus the conformational disorder does not undergo any significant changes. The quantity of the minor conformer

91 Figure 4.6: The librational motion in stilbene that may cause interconversion between conformers. present at low temperatures can be increased by increasing the cooling rate, but always remains very low (< 10% for ﬂash-cooled crystals and < 5% for slow cooled crystals). Freezing of the conformational changes generates thermodynamic non- equilibrium states that persist below the transition temperature and represents a signature of a glass-like transition. The anomaly observed in calorimetry measurements at 170 K in STB was therefore attributed to this glass-like transition due to the freezing of the crankshaft motion.23 It is important to note that the potential energy along the torsional coordinate connecting the two conformers of the stilbene molecules can be described by a double-well potential. The ensemble of systems described by such energy potentials are well-known models for glasses.40, 41 In addition to these observations in neat STB, thermodynamic transitions are also found in calorimetry measurements of CT complexes of STB: near 250 K in STB–

16 TCNQ and near 240 K in STB–F4TCNQ. The transitions are similar in shape to that seen in neat stilbene and are therefore interpreted to be caused by the freezing of

92 crankshaft motion; however, these materials have not been reported to exhibit temperature dependent changes in conformer populations as stilbene does. In the CT complexes, while the C=C bond undergoes a similar librational motion, a quantitative analysis of the conformational inter-conversion was impossible under regular labora- tory conditions. Our temperature-dependent STB–F4TCNQ structure determination indicates that the disorder is low and the population of the minor conformer is below 5% at all temperatures, as show in Fig. 4.7, a plot of the thermal ellipsoids in the CT crystal modeled with and without disorder. Our results are in agreement with Saito et al.who concluded based on the heat capacity measurements that while there is disorder present in STB–F4TCNQ, the population of the misoriented STB molecules is less than 10%, even at room temperature.16 The lack of signiﬁcant changes in the crystal structure as the system passes through the transition temperature coupled with the stepped-like feature in the heat capacity measurements are clear signatures of a glass-type transition17, 18 As a result, we believe that the mechanism for the observed transition in the transport properties of STB-F4TCNQ crystal is related to the transition into a glassy crystal, where the constituent STB molecules exhibit freezing of the orientational disorder and the thermal equilibrium is no longer attained. In addition, the existence of a glass-like transition is supported by a signiﬁcant thermal history dependence which is also observed in our measurements.

Figure 4.7: Thermal ellipsoids for STB–F4TCNQ modeled with and without disorder.

93 Interestingly, the Fourier maps do show a qualitative diﬀerence in electron density surrounding the C=C bond below vs above the transition. At 143 K, below the transition temperature (Fig. 4.8a), the density appears constant throughout the molecule, with a slight excess of charge density around the center ethylene bond, suggesting the presence of the minor conformer. Above the transition, at 253 K (Fig. 4.8b), the density shows a signature of the increasing torsional motion of this bond, in agreement with our observations described below. We gained further understanding of the dynamics of the central C=C bond in STB by analyzing its change in length as a function of temperature. It is known that this bond can show apparent shrinking as a result of disorder or large-amplitude torsional motion.39 In the present case the C7=C7a (Fig. 4.1a) bond length gradually decreases with increasing temperature once the amplitude of the libration becomes large. Speciﬁcally, this bond length lies between the standard ethylene bond length of 1.34 Aand˚ the fully-conjugated benzene bond length of 1.39 Aat˚ low temperatures,

Figure 4.8: Fourier difference maps above and below the transition temperature T=235K. (a) The Fourier difference map calculated for a temperature in region II (143K). (b) The Fourier difference map calculated in region I (253K). These show a redistribution of charge density around C7 and C7a. Positive contour lines were calculated at intervals of 0.05 A˚

94 and it shortens to 1.322(6) Aat˚ 273 K (Fig. 4.9a). Sato, et al., reports a further shortened bond length of 1.315(4) Aat˚ 293 K, also included in 4.9a.28 Fig. 4.9b shows a schematic of the librational motion, with the C=C bond able to rotate above and below the plane created by the benzene rings. The calculations of the Γ-point phonons indicate that there are two vibrational modes with frequencies of 279 cm−1 (Fig. 4.9c) and 285 cm−1 (Fig. 4.9d) related to the libration of the STB bridge. As seen in Fig. 4.9c, and d, the benzene rings are not involved in these modes – i.e. these librations occur along the crankshaft trajectory. Assuming that the bond modiﬁcation is due to librational motion, the root-mean-square libration amplitude, ϕ, can be obtained from the bond shortening by means of equation 4.2:39

r ϕ = cos−1 obs (4.2) rSTB

Here (see the inset in Fig. 4.9a) robs is the apparently shortened stilbene C7=C7a length as observed by diﬀraction, and rSTB is the expected C=C distance. Our calculations show that below the transition temperature the libration is small, reaching the

Figure 4.9: Libration of the C7=C7a stilbene bond in STB–F4TCNQ. (a) The C7=C7a bond length is given as a function of temperature with the libration schematic as an inset. The error bars were calculated based on the standard deviation of the bond lengths. (b) The oscillation of the C7 and C7a atoms above and below the plane formed by the benzene rings. (c) The eigendisplacements of the 279 cm−1 phonon mode. (d) The eigendisplacements of the 285 cm−1 phonon mode. Only one DA complex from the unit cell is shown.

95 detection limit, whereas above the transition it has an amplitude of approximately 10◦, with the change taking place gradually, as the temperature is increased. This result shows a quantitative analysis of the temperature-dependent libration in organic crystals and supports the conclusion that the amplitude of this motion can directly aﬀect charge transport in this class of materials.

In addition to its effect on electrical properties of the STB–F4TCNQ crystals, libration significantly affects the partial degree of charge transfer (q) between the donor and acceptor molecules. In general, in organic CT complexes the lattice expansion that occurs with increasing temperature yields larger intermolecular distances between the donor and acceptor species, resulting in weaker interactions between them and thus a lower q. The consequent decrease in tHL as described above is expected

to lead to a decrease in the amount of charge transfer from STB to F4TCNQ in the

ground state. Indeed, the lattice parameters and tHL follow the conventional trend; however, XRD, Raman and IR spectroscopy indicate an increase in q with increasing temperature, as described below. Similar to its unﬂuorinated analogue, TCNQ,

the bonds in F4TCNQ are sensitive to the value of q, allowing us to estimate this value based on bond lengths and vibrational frequencies. Thus for our estimation of q from XRD measurements, we adapted the method reported for TCNQ-containing CT materials.42 The bond lengths a and c (Fig. 4.1a) expand with increasing degree of charge transfer, while the bond lengths b and d shrink. The value for q can be

calculated when these bond lengths are compared to neutral F4TCNQ, as in equation 4.3:

1 b − c d − c q = 1 − CT CT + 1 − CT CT (4.3) 2 bN − cN dN − cN

Here, bond lengths with the subscript CT correspond to those of the acceptor

within the CT complex STB–F4TCNQ, while the bond lengths with subscripts of

N correspond to neutral F4TCNQ. Note that bond lengths for the neutral and CT

96 F4TCNQ are compared at the same temperature so that thermal expansion does not aﬀect the value for q. In Fig. 4.10 we plot the results as a function of temperature, in black. Even though the structural changes follow the conventional behavior in which the lattice expands with temperature, q nevertheless increases with temperature and mirrors the changes in the libration of the C7=C7a bond. This suggests that the libration increases the coupling between the donor and acceptor units and is responsible for the observed increase in q, in spite of the expansion of the unit cell. As Fig. 4.10 shows, q ≈ 0 for T < 235 K, where the libration is reduced in amplitude. As the temperature increases and the system passes through the transition, q increases to about 0.1e. This conclusion is also supported by our calculations of the degree of charge transfer using optical measurements (Fig. 4.10, in grey). It is known both experimentally and theoretically that, as for TCNQ, the frequencies of vibrations of the cyano groups

Figure 4.10: Degree of charge transfer (q) as a function of temperature. Values are derived from bond lengths (black, squares) and IR mode B1u (grey, triangles). All measurements show an increase in degree of charge transfer as a function of temperature. The error bars were calculated based on the standard deviation of the bond lengths.

97 43–45 of F4TCNQ vary linearly with the degree of charge transfer q. There are four

cyano-group vibrations; two are Raman active (Ag and B3g) and two are infrared

−1 active (B1u and B2u), with only the Ag mode (2225 cm ) and the B1u mode (2227 cm−1) being experiementally observed. The atoms in motion in these two vibrational modes are in close proximity to the surrounding molecules and may therefore be sensitive to the crystalline environment as well as to the degree of charge transfer.44

The shift in frequency of the B1u mode between the F4TCNQ neutral molecule and

−1 −1 44 anion is 34 cm whereas the shift in frequency of the Ag mode is only 6 cm .

Therefore, we use only the B1u mode for the determination of the degree of charge transfer, shown in Fig. 4.10. The temperature dependence of q as determined from the optical measurements shows the same behavior as was determined from the XRD, although the values are somewhat smaller. The charge transfer is negligible at low temperatures and increases with temperature to a value not larger than 0.1 at room temperature, in agreement with the XRD results. The Raman measurements of the

Ag mode agree with this ﬁnding.

4.4 Discussion

The observed increase in q with the increase of libration amplitude can be explained in terms of non-local (Peierls) electron-phonon coupling, which is the dependence of the transfer integrals on the displacements of the phonon modes.46 The transfer integral, tHL computed as a function of the phonon displacements of the librational modes discussed above is shown in Fig. 4.11. For small phonon displacements,

47 tHL shows the expected linear behavior. However, for larger displacements, such as those expected above the transition temperature, tHL exhibits a signiﬁcant non- linear increase. Therefore, the increase in the transfer integrals due to these librations surpasses the decrease related to the thermal expansion, leading to an overall increase

98 Figure 4.11: Transfer integral dependence on the phonon displacements. The results for the 285 and 279 cm−1 modes are indicated by black squares and red circles, respectively; the phonon displacements, Q, are given in dimensionless units.

of the tHL and, consequently, of the q with the increase in temperature. Our model thus aligns very well with the experimental observations. With a better understanding of the dynamics of the librational motion as a function of temperature, we are now able to describe the mechanism that can explain the evolution of the electrical properties of the STB–F4TCNQ crystals. One possible explanation for the crossover from tunneling (and therefore temperature independence) at low temperatures to activated charge transport at higher temperatures invokes interaction with local electron-phonon coupling (modulation of the site energies by vibrations), known as the Holstein small polaron model.46, 48 The Holstein polaron

binding energy (Epol) of 0.14 eV estimated for STB–F4TCNQ in our previous study,

however, results in an activation energy of (Ea) of 0.07 eV (Ea = Epol/2) which is signiﬁcantly smaller than the observed activation energy of 0.17 eV.49 We note that

the above estimation of the Epol and Ea is based on the coupling with intra-molecular vibrations only, i.e. the contribution due to the electron-phonon interactions with

99 intermolecular vibrations (lattice phonons) is neglected. Although the latter contribution to Holstein polaron binding energy in single-component systems is believed to be small, it may be different in the case of charge-transfer complexes. Therefore while we cannot rule it out, we do not have at this point clear evidence that would support the implication of the Holstein-type electron phonon coupling in the explanation of the observed crossover from temperature-independent to activated transport. A second possible explanation for the crossover could be related to the non-local electron- phonon coupling. This interaction can lead to an increase of the transfer integrals and subsequently to an enhancement of the transport properties (phonon-assisted).50 At the same time, the Peierls-type coupling is a source of dynamical disorder and thus of charge localization.24 If the former effect, which increases the mobility, were to dominate at high temperature, and the latter effect, which decreases the mobility, were to dominate at low temperatures, a crossover in the temperature dependence of the transport could result. Previous studies49 indicate that the Peierls-type polaron binding energy in STB–F4TCNQ is smaller by an order of magnitude than the electronic coupling, and it is also much smaller than the experimental activation energy observed for the mobility. Therefore, we believe that the Peierls electron-phonon interaction is not the cause of the observed temperature dependence of the transport. We suggest that the observed transport behavior is due to the double-well potential related to the librational motion of the center STB ethylene bond, which is also responsible for the glass-like transition in STB–F4TCNQ. At temperatures below the transition, the temperature-independent structural disorder is frozen in and the transport can take place only via tunneling of this manifold, therefore being temperature-independent. At higher temperatures, when inter-conversion motion is fast, a transition to temperature-activated hopping transport occurs. This proposed model is also supported by the measurements performed on analog compounds which exhibit orientational disorder at low temperatures: for both STB and STB–TCNQ we

100 Figure 4.12: Current value at 100V (normalized to the room temperature value) as a function of temperature for compounds related to STB–F4TCNQ. (a) The temperature profile for neat trans-stilbene. (b) The temperature profile for CT complex STB–TCNQ. (c) The temperature profile for neat TCNQ. observe a clear transition from a temperature-independent to activated transport at the same temperature where the glass-transition occurs (Fig. 4.12a,b). The transition is absent in TCNQ crystals(Fig. 4.12c).

4.5 Conclusions

In conclusion, we show that the electrical properties of STB–F4TCNQ undergo a crossover from thermally-activated at high temperatures to temperature-independent below T = 235 K. We attribute these changes to the presence of a glass-like transition, where a significant amount or orientational disorder of the STB molecule is frozen in at low temperature upon reducing in amplitude of the libration of a molecular moiety within this molecule. When the libration is large, the structural disorder resulting from it yields a thermally-activated charge transport. At low temperatures, the transport is temperature-independent and proceeds via tunneling between the states created by the available conformers that generate the static orientational disorder. This transition has further implications within the class of organic charge transfer complexes, as it affects the degree of charge transfer. These findings provide important insight into the effects that molecular dynamics — a ubiquitous feature in organic molecular materials — have on charge transport.

101 References

[1] K. P. Goetz, A. Fonari, D. Vermeulen, P. Hu, H. Jiang, P. J. Diemer, J. W. Ward, M. E. Payne, C. S. Day, C. Kloc, V. Coropceanu, L. E. McNeil, and O. D. Jurchescu. Freezing-in orientational disorder induces crossover from thermally- activated to temperature-independent transport in organic semiconductors. Na- ture Communications, 5:5642, 2014. [2] F. Kagawa, S. Horiuchi, M. Tokunaga, J. Fujioka, and Y. Tokura. Ferroelectricity in a one-dimensional organic quantum magnet. Nature Physics, 6(3):169–172, 2010. [3] F. Kagawa, S. Horiuchi, H. Matsui, R. Kumai, Y. Onose, T. Hasegawa, and Y. Tokura. Electric-Field Control of Solitons in a Ferroelectric Organic Charge- Transfer Salt. Physical Review Letters, 104(22):227602, 2010. [4] J. B. Torrance. An Overview of organic charge-transfer solids: insulators, metals, and the neutral-ionic transition. Molecular Crystals and Liquid Crystals, 126(1):55–67, 1985. [5] J. B. Torrance. Spin waves, scattering at 4kf, and spin-Peierls fluctuations in an organic metal: tetrathiafulvalene-tetracyanoquinodimethane (TTF-TCNQ). Physical Review B, 17(8):3099–3103, 1978. [6] J. B. Torrance, Y. Tomkiewicz, R. Bozio, C. Pecile, C. R. Wolfe, and K. Bechgaard. Magnetic properties of an organic Mott insulator. Sepa- rate donor and acceptor phase transitions in hexamethylenetetrathiafulvalene tetrafluorotetracyano-quinodimethane (HMTTF-TCNQF4). Physical Review B, 26(4):2267–2270, 1982. [7] J. B. Torrance, J. E. Vazquez, J. J. Mayerle, and V. Y. Lee. Discovery of a neutral-to-ionic phase transition in organic materials. Physical Review Letters, 46(4):253–257, 1981. [8] S. Horiuchi, T. Hasegawa, and Y. Tokura. Molecular donor-acceptor compounds as prospective organic electronics materials. Journal of the Physical Society of Japan, 75(5):051016, 2006. [9] S. Horiuchi, T. Hasegawa, and Y. Tokura. Neutral-ionic transition, ferroelectricity, and field-effect transistors based on molecular donor-acceptor compounds. Molecular Crystals and Liquid Crystals, 455:295–304, 2006.

102 [10] D. Jérome.Organic conductors: from charge density wave TTFTCNQ to super- conducting (TMTSF)2PF6. Chemical Reviews, 104(11):5565–5592, 2004. [11] G. Mihaly, Y. Kim, and G. Gruner. Crossover in low-temperature collective spin-density-wave transport. Physical Review Letters, 228(19):2713–2716, 1991. [12] K. Asadi, A.J. Kronemeijer, T. Cramer, L. J. A. Koster, P. W. M. Blom, and D. M. de Leeuw. Polaron hopping mediated by nuclear tunnelling in semiconducting polymers at high carrier density. Nature Communications, 4:1710, apr 2013. [13] J. D. Yuen, R. Menon, N. E. Coates, E. B. Namdas, S. Cho, S. T. Hannahs, D. Moses, and A. J. Heeger. Nonlinear transport in semiconducting polymers at high carrier densities. Nature Materials, 8:572–575, 2009. [14] T. Sakanoue and H. Sirringhaus. Band-like temperature dependence of mobility in a solution-processed organic semiconductor. Nature Materials, 9:736–40, 2010. [15] J. H. Worne, J. E. Anthony, and D. Natelson. Transport in organic semiconductors in large electric fields: From thermal activation to field emission. Applied Physics Letters, 96:053308, 2010. [16] K. Saito, M. Okada, H. Akutsu, A. Sato, and M. Sorai. Freezing of Crankshaft Motion of trans-Stilbene Molecule in Charge-Transfer Complexes, STB-TCNQ and STB-TCNQF4. The Journal of Physical Chemistry B, 108:1314–1320, 2004. [17] H. Suga. Thermodynamic aspects of glassy crystals: Approaching the equilibrium glass. Annals of the New York Academy of Sciences, 484:248–263, 1986. [18] H. Suga and S. Seki. Thermodynamic investigation on glassy states of pure simple compounds. Journal of Non-Crystalline Solids, 16(2):171–194, 1974. [19] A. F. Wright, A. N. Fitch, and B. E. F. Fender. How disordered can a crystallized glass be? Annals of the New York Academy of Sciences, 484:54–65, 1986. [20] J. P. Sethna. Glassy crystals: Low-frequency and low-temperature properties. Annals of the New York Academy of Sciences, 484:130–149, 1986. [21] F. Gugenberger, R. Heid, C. Meingast, P. Adelmann, M. Braun, H. Wühl, M. Haluska, and H. Kuzmany. Glass transition in single-crystal C60 studied by high-resolution dilatometry. Physical Review Letters, 69(26):3774–3777, 1992. [22] F. J. Bermejo, M. Jiménez-Ruiz,A. Criado, G. J. Cuello, C. Cabrillo, F. R. Trouw, R. Fernández-Perea, H. Löwen, and H. E. Fischer. Rotational freezing in plastic crystals: a model system for investigating the dynamics of the glass transition. Journal of Physics: Condensed Matter, 12:391–397, 2000. [23] K. Saito, Y. Yamamura, K. Kikuchi, and I. Ikemoto. Glass transition due to freezing of intramolecular motion: Crystalline trans-azobenzene and trans-stilbene. Journal of Physics and Chemistry of Solids, 56(6):849–857, 1995.

103 [24] A. Troisi and G. Orlandi. Charge-transport regime of crystalline organic semiconductors: Diffusion limited by thermal off-diagonal electronic disorder. Physical Review Letters, 96(8):086601, 2006. [25] A. S. Eggeman, S. Illig, A. Troisi, H. Sirringhaus, and P. A. Midgley. Measure- ment of molecular motion in organic semiconductors by thermal diffuse electron scattering. Nature Materials, 12(11):1045–1049, 2013. [26] R. A. Laudise, C. Kloc, P. G. Simpkins, and T. Siegrist. Physical vapor growth of organic semiconductors. Journal of Crystal Growth, 187:449–454, 1998. [27] A. J. C. Buurma, O. D. Jurchescu, I. Shokaryev, J. Baas, A. Meetsma, G. A. de Wijs, R. A. de Groot, and T. T. M. Palstra. Crystal growth, structure, and electronic band structure of TetraceneTCNQ. The Journal of Physical Chemistry C, 111(8):3486–3489, 2007. [28] A. Sato, M. Okada, K. Saito, and M. Sorai. The charge-transfer complex trans- STB-TCNQF4. Acta Crystallographica Section C, C57:564–565, 2001. [29] G. M. Sheldrick. A short history of SHELX. Acta Crystallographica Section A, 64(1):112–122, 2007. [30] R. Dovesi, R. Orlando, B. Civalleri, C. Roetti, V. R. Saunders, and C. M. Zicovich-Wilson. CRYSTAL: a computational tool for the ab initio study of the electronic properties of crystals. Zeitschrift fur Kristallographie, 220:571–573, 2005. [31] E. F. Valeev, V. Coropceanu, D. A. da Silva Filho, S. Salman, and J.-L. Brédas. Effect of electronic polarization on chrage-transport parameters in molecular organic semiconductors. Journal of the American Chemical Society, 128(8):9882– 9886, 2006. [32] M. Valiev, E. J. Bylaska, N. Govind, K. Kowalski, T. P. Straatsma, H. J. J. Van Dam, D. Wang, J. Nieplocha, E. Apra, T. L. Windus, and W. A. De Jong. NWChem: A comprehensive and scalable open-source solution for large scale molecular simulations. Computer Physics Communications, 181(9):1477–1489, 2010. [33] O. D. Jurchescu and T. T. M. Palstra. Crossover from one- to two-dimensional space-charge-limited conduction in pentacene single crystals. Applied Physics Letters, 88(12):122101, 2006. [34] J. A. Bouwstra, A. Schouten, and J. Kroon. Structural studies of the system trans-azobenzene/trans-stilbene. II. A reinvestigation of the disorder in the crystal structure of trans-stilbene, C14H12. Acta Crystallographica Section C, 40(3):428–431, 1984. [35] J. Bernstein and K. Mirsky. Order and disorder in molecular crystals: trans- stilbene. Acta Crystallographica Section A, 34:161–165, 1978.

104 [36] D. Zobel and G. Ruban. The structures of some charge-transfer complexes containing TCNQ as acceptor and their electrical anisotropy. Acta Crystallographica Section B, 39:638–645, 1983.

[37] J. Harada and K. Ogawa. Invisible but common motion in organic crystals: a pedal motion in stilbenes and azobenzenes. Journal of the American Chemical Society, 123(44):10884–10888, 2001.

[38] J. Harada and K. Ogawa. Pedal motion in crystals. Chemical Society Reviews, 38:2244–2252, 2009.

[39] K. Ogawa, T. Sano, S. Yoshimura, Y. Takeuchi, and K. Toriumi. Molecular structure and intramolecular motion of (E)-stilbenes in crystals. An interpreta- tion of the unusually short ethylene bond. Journal of the American Chemical Society, 28:1041–1051, 1992.

[40] P. W. Anderson, B. I. Halperin, and C. M. Varma. Anomalous low-temperature thermal properties of glasses and spin glasses. Philosophical Magazine, 25(1):1–9, 1972.

[41] W. A. Phillips. Tunneling states and the low-temperature thermal expansion of glasses. Journal of Low Temperature Physics, 11(5-6):757–763, 1973.

[42] S. Flandrois and D. Chasseau. Longueurs de liaison et transfert de charge dans les sels du t´etracyanoquinodim´ethane(TCNQ). Acta Crystallographica Section B, 33:2744–2750, 1977.

[43] L. Zhu, E. Kim, Y. Yi, and J.-L. Br´edas. Charge transfer in molecular complexes with 2,3,5,6-tetraﬂuoro-7,7,8,8-tetracyanoquinodimethane (F4-TNCQ): A density functional theory study. Chemistry of Materials, 23(23):5149–5159, 2011.

[44] M. Meneghetti and C. Pecile. Chargetransfer organic crystals: Molecular vibrations and spectroscopic eﬀects of electronmolecular vibration coupling of the strong electron acceptor TCNQF4. The Journal of Chemical Physics, 84(8):4149– 4162, 1986.

[45] J. S. Chappell, A.N. Bloch, W. A. Bryden, M. Maxﬁeld, T. O. Poehler, and D. O. Cowan. Degree of charge transfer in organic conductors by infrared absorption spectroscopy. Journal of the American Chemical Society, 103(9):2442–2443, 1981.

[46] V. Coropceanu, J. Cornil, D. A. da Silva Filho, Y. Olivier, R. Silbey, and J. L. Bredas. Charge transport in organic semiconductors. Chemical Reviews, 107:926– 952, 2007.

[47] V. Coropceanu, R. S. Sanchez-Carrera, P. Paramonov, G. M. Day, and J.-L. Bredas. Interaction of charge carriers with lattice vibrations in organic molecular semiconductors: Naphthalene as a case study. The Journal of Physical Chemistry C, 113(11):4679–4686, mar 2009.

105 [48] T. Holstein. Studies of polaron motion part I. The molecular-crystal model. Annals of Physics, 8(3):325–342, 1959.

[49] L. Zhu, Y. Yi, Y. Li, E. G. Kim, V. Coropceanu, and J.-L. Br´edas. Prediction of remarkable ambipolar charge-transport characteristics in organic mixed-stack charge-transfer crystals. Journal of the American Chemical Society, 134(4):2340– 2347, 2012.

[50] P. Gosar and S.-I. Choi. Linear-Response Theory of the Electron Mobility in Molecular Crystals. Physical Review, 150(2):529–538, 1966.

106 Chapter 5

The Eﬀect of Polymorphism on Charge Transport in the Charge-Transfer Complex DBTTF-TCNQ

The organic charge-transfer (CT) complex dibenzotetrathiafulvalene – 7,7,8,8- tetra-cyanoquinodimethane (DBTTF–TCNQ) is found to crystallize in two polymorphs when grown by physical vapor transport: the known α-polymorph and a new structure, the β-polymorph. Structural and elemental analysis via selected area electron diffraction (SAED), X-ray photoelectron spectroscopy (XPS), and polarized IR spectroscopy reveal that the complexes have the same stoichiometry with a 1:1 donor:acceptor ratio, but exhibit different unit cells. Though the structural differences are small, they result in significant differences in the optoelectronic properties of the crystals, as observed in our experiments and electronic-structure calculations. Raman spectroscopy shows that the α-polymorph has a degree of charge transfer of about 0.5e while the β-polymorph is nearly neutral. Organic field-effect transistors fabricated on these crystals reveal that in the same device structure both polymorphs show ambipolar charge transport, but the α-polymorph exhibits electron-dominant transport while the β-polymorph is hole-dominant. Together, these measurements imply that the transport features result from differing donor-acceptor overlap and consequential varying in frontier molecular orbital mixing, as suggested theoretically for charge-transfer complexes.

This portion of the text is adapted from K. P. Goetz, Junya Tsutsumi, Sujitra Pookpanratana, Jihua Chen, Nathan S. Corbin, Rakesh K. Behera, Veaceslav Coropceanu, Curt A. Richter, Christina A. Hacker, Tatsuo Hasegawa, and Oana D. Jurchescu. The author of this dissertation is responsible for or participated in the crystal growth, structural characterization by Raman, IR, and UV-Vis absorption spectroscopies, and electrical characterization. She fabricated all samples for SAED, UPS, and XPS. In preparation (2016).1

107 5.1 Introduction

The crystal structure of organic crystalline solids is mediated by the interplay between van der Waals interactions and dipoles or multipoles (either permanent or induced).2 Because of the weak energies of these interactions, variations in growth conditions can result in polymorphism, where the crystal constituents are the same but the crystal structure is diﬀerent. This is the case for several monomolecular compounds, such as 5,11-bis(triethylsilylethynyl) anthradithiophene (TES-ADT), where the structure is dictated by the type of solvent used for crystallization.3 Rubrene, one of the highest performance materials, grows in several polymorphs,4, 5 two of which can be tuned based on the pressure of the sublimation chamber when grown from a gas phase.6–8 Other examples include pentacene, TMS-DBC (7,14bis((trimethylsilyl) ethynyl)dibenzo [b,def]-chrysene), DBTTF (dibenzotetrathiafulvalene), and related DTTTF (dithiophene-tetrathiafulvalene).9–13 When electrical measurements are performed on these materials, large diﬀerences in charge-transport characteristics are often observed. In the case of TMS-DBC, for example, the red needle-like polymorph that exhibits a one-dimensional π-stacked crystal structure has a hole mobility on the order of 10−3 cm2V−1s−1, whereas the yellow polymorph that exhibits two-dimensional π-stacking has a hole mobility on the order of 1 cm2V−1s−1. Several theoretical studies have shed some light on why the electronic properties vary with crystal structure. For example, the energy splitting between the HOMO and LUMO levels for pentacene was shown to modulate based on the overlap or separation between to units in the π-stacking direction. This is important because the energy splitting is directly related to the transfer integral, which is in turn directly related to the hole and electron mobility.14 Although charge transport in most CT complexes is thought to take place by a superexchange mechanism as opposed to directly as in monomolecular compounds,15 polymorphism is also known to result in varying electrical characteristics in these two-

108 component systems. Interestingly, it also results in changes in the charge transfered from donor to acceptor (q).16, 17 As noted in the introductory chapter to this thesis, this value is governed by the diﬀerence between the ionization potential of the donor

(ID) and the electron aﬃnity of the acceptor (EA) in relation to the Madelung energy

(EM ) of the crystal. A neutral or quasi-neutral q is derived from a D:A combination characterized by ID − EA EM , and is characterized by q < 0.5, whereas ID −

EA EM results in quasi-ionic or ionic charge transfer. In the case of polymorphs, where the D and A units remain constant, changing the crystal structure is entirely responsible for any resulting changes in q. Such is the case with BEDT-TTF–TCNQ (where BEDT-TTF is the donor bis(ethylenedithio)-tetrathiafulvalene and TCNQ is the acceptor 7,7,8,8-tetracyanoquinodimethane). This complex has two metallic phases, both exhibiting a room-temperature electrical conductivity of 10 Ω−1cm−1: one crystallizes in a triclinic (β) segregated-stack structure and has a degree of charge transfer of 0.5e (where e is the elementary charge),18–20 while the other crystallizes in a triclinic (β) segregated-stack structure and is characterized by a 0.74e degree of charge transfer.21, 22 In contrast, its monoclinic, mixed-stack structure is a semiconductor with seven orders of magnitude lower conductivity and a degree of charge transfer of q = 0.1e.23 While the differences are clear in the case of segregated-stacked metals versus mixed-stack semiconductors, studies on different phases of mixed-stack crystals where the D and A units alternate are much less common. Here, we study the effect of polymorphism on the electrical properties in mixed-stack CT complexes by examining two polymorphs of the CT complex dibenzotetrathiafulvalene 7,7,8,8-tetracyanoquinodimethane (DBTTF–TCNQ) (Fig. 5.1a, 5.1b). We will refer to the known, triclinic polymorph as the α-polymorph.24–26 Crystals of this structure adopt a rectangular shape (Fig. 5.1c), whereas the crystals of the new, previously-unreported polymorph (which we will call the β-polymorph), are elliptical in shape (Fig. 5.1d). The thin

109 Figure 5.1: Crystalline polymorphs of DBTTF–TCNQ. (a) The skeletal structure of the donor dibenzotetrathiafulvalene. (b) The skeletal structure of the acceptor TCNQ. (c) An α-DBTTF–TCNQ crystal. (d) A β-DBTTF–TCNQ crystal.

nature of the β-crystals made it impossible to completely determine its structure by X-ray diffraction (XRD). Therefore, a combination of selected area electron diffraction (SAED), IR and UV-vis absorption spectroscopies was used to confirm the structure of the α-polymorph and estimate the structure of the new polymorph. DFT calculations allowed us to further refine this crystal structure and compare features of its electronic structure to those of the α-polymorph.15 To evaluate the optoelectronic properties of the two polymorphs we incorporated their single crystals in organic field-effect transistors (OFETs). We observed that the differences between the two systems are very subtle, both exhibiting ambipolar- semiconducting electrical properties with gold source/drain contacts. The α-polymorph shows superior electron transport while, in contrast, the β-polymorph exhibits hole dominant transport. Raman spectroscopy suggests that the degree of charge transfer between the D and A molecules in the α-polymorph approaches q = 0.5e, whereas the β-polymorph is almost neutral, with q = 0.1e. Our in-depth investigations of the structural, optical, and electronic properties of DBTTF–TCNQ crystals provide con-

110 clusive evidence of the impact of crystalline packing on the properties of mixed-stack CT complexes and validate the theoretical calculations which predict that the charge transport can lose its balanced nature if other orbitals besides the donor HOMO and the acceptor LUMO participate in charge transport.17 We propose that the more favorable charge transport pathway for holes in β-DBTTF-TCNQ compared to the α-polymorph is a consequence of varying frontier orbital mixing between the two polymorphs.

5.2 Experiment and Results

5.2.1 Structural Characterization of DBTTF–TCNQ

Crystals of α and β-DBTTF–TCNQ were grown as discussed in chapter 2 of this thesis. The crystal structure of the α-polymorph was confirmed by X-ray diffraction (XRD). The obtained unit cell parameters coincide with the previously reported ones25 with a = 7.7576 A,˚ b = 8.3622 A,˚ c = 10.40 A,˚ α = 72.14◦, β= 109.77◦, and γ= 110.19◦. The β-polymorph crystals were extremely thin, and even synchrotron XRD measurements were unsuccessful in determining their structure. Therefore, a combination of SAED, IR spectroscopy, XPS, optical absorbance, and DFT calculations was used to determine the unit cell and orientation of the bulk crystal, evaluate the D:A ratio, and to estimate the crystal structure. SAED experiments were performed in a Zeiss Libra 120 transmission electron microscope at 120kV using a LaB6 filament and low electron dose conditions. Al (111) of 0.234 nm d spacing was used as the electron diffraction calibration standard. The SAED results for the two DBTTF–TCNQ polymorphs are shown in Fig. 5.2, where the unit cell parameters are also listed. We note that the parameters for the α-polymorph differ slightly from those previously reported for DBTTF–TCNQ,25, 26 possibly due to differences in crystal growth conditions, but they are very similar to the triclinic unit cell reported by Kobayashi, et al.,

111 Figure 5.2: SAED patterns for the α-DBTTF–TCNQ (left) and β-DBTTF-TCNQ single crystals. Measured lattice parameters are listed below the corresponding patterns, with α-DBTTF–TCNQ showing a 124 ◦ alpha angle and β-DBTTF–TCNQ exhibiting a 90◦ alpha angle. where the space-group is with b = 0.828 nm, c = 0.773 nm, and α = 110.03◦. SAED measurements on microrods of DBTTF-TCNQ showed a unit cell structure similar to the α-polymorph discussed here.27 In contrast, the β-polymorph has a unit cell of monoclinic symmetry at a minimum, as it exhibits one unit cell angle of 90◦ with lattice parameters found to be b = 0.997 nm and c = 0.652 nm. X-ray photoelectron spectroscopy (XPS) identifies and quantifies the chemical composition of the crystals, and can also provide information about the degree of charge transfer between the donor and acceptor in the complex, as we will show in the next section. XPS measurements were performed in a commercial instrument (base pressure 1 x 10−9 Torr or better) equipped with a hemispherical electron energy analyzer and monochromatized Al Kα photons (1486.6 eV). Spectral fitting was performed to identify the different states in which the S atoms exist. For the β-DBTTF–TCNQ, the S 2p fits were performed by using one pair of Voigt lineshapes for the 2p3/2 and 2p1/2, fixing the spin orbit area ratio (2p3/2:2p1/2 as 2:1), and coupling the Gaussian widths. Due to the complicated S 2p spectral shape of

112 the α-DBTTF–TCNQ, three pairs of Voigt lineshapes were used while coupling the Gaussian and Lorentzian widths, fixing the area ratio (2:1), and fixing the spin-orbit splitting as 1.14 eV (determined from the fit of β-DBTTF–TCNQ). It was necessary to use three different chemical environments as it was clear that using one or two chemical environments poorly described the spectrum. The N 1s spectrum of both polymorphs were deconvoluted by coupling the Gaussian widths, while the Lorentzian widths were allowed to vary to account for the different electronic states. Crystals of both polymorphs were laminated onto gold substrates and contacted to the substrate using silver epoxy. In this study, we are able to use the S 2p spectra to represent features of the donor moiety because DBTTF is sulfur-containing and TCNQ is not (Fig. 5.3a). Similarly, the N1s spectrum is representative of the acceptor (Fig. 5.3b). Relative sulfur and nitrogen content was estimated by XPS by using the S 2s and N 1s photoemission

Figure 5.3: XPS for α-DBTTF–TCNQ (top) and β-DBTTF–TCNQ (bottom). The S 2p spectra are shown in (a), and the N 1s spectra are shown in (b). Data (open- circles) and spectral ﬁts (as solid lines) of neutral and CT moietties as well as the shake-up satellite are shown. The horizontal aroow line in (b) indicate the estimation of the HOMO-LUMO transition, in the presence of a core-hole, of the TCNQ moiety in each polymorph.

113 (PES) lines and correcting by the photoionization cross sections and spectrometer transmission function. The sulfur fraction, [S]/([S]+[N]), of each polymorph is 0.56 ± 0.1 for the α-crystals and 0.59 ± 0.1 for the β-crystals. Within experimental error, this result confirms that each polymorph has a 1:1 donor-to-acceptor ratio. While the α-polymorph is known to have a 1:1 ratio from X-ray diffraction, this finding is not trivial, as we have shown that different D:A ratios can be obtained for other systems by tuning the growth conditions.28 In order to align the crystals such that we access the D/A charge-transfer stack in our electrical measurements, we measured the optical absorbance of each crystal type between 0.5 eV and 3.0 eV (near IR to near UV) by using a combined grating monochromator and Cassegrain-type microscope. The results are shown in Fig. 5.4. In these spectra, a charge-transfer band is present around 0.7 eV for each crystal type. This band exhibits high intensity when the incident light is polarized in the direction of the stack axis (i.e. the direction of charge-transfer), and negligible intensity when light is polarized perpendicularly to the stack axis. We concluded that the stack axis

Figure 5.4: The optical absorbance of the α–DBTTF-TCNQ (black) and β- DBTTF–TCNQ (blue). The charge-transfer peak of the former peaks at 0.78 eV while the latter peaks at 0.73 eV.

114 is the long crystal axis of the α-polymorph, in agreement with a previous report,29 and the short axis of the elliptical β-polymorph. Polarized IR spectroscopy was applied to each crystal using a Thermofisher Sci- entific continuum IR microscope equipped with an iS50R FT-IR spectrometer. The anisotropies of the spectra were measured by polarizing the incident light with a wire-grid polarizer. For the spectral assignment, key vibrations were simulated by using GAUSSIAN 09 with the 6-31G(d) basis set at the B3LYP level. The results are shown in Fig. 5.5. The anisotropy of the spectra shown in this figure points to another key difference between the two polymorphs. For the α-polymorph, the observation of the vibrations labeled c, d, e, and f, which are each parallel to the long axes of the molecular DBTTF and TCNQ, indicates that these axes are oriented parallel to the surface of the crystal. The out-of-plane C-H stretching of the DBTTF molecule (vibration a), which is perpendicular to the molecular plane, shows strong anisotropy and is most pronounced when the light is polarized along the long axis of the bulk crystal surface. Vibrations e and f, and stretching on the TCNQ molecule respectively, occur in plane and have a transition moment parallel to the molecular long axis. They are strongest when the light is polarized parallel to the short axis of the bulk crystal. Together, a, e, and f show that the molecular CT stack axis is parallel to the long axis of the crystal, in agreement with the absorption data. The β-polymorph, on the other hand, shows strong anisotropy in the vibration labeled b, and weak intensity in all directions for vibrations c, d, e, f, and g. This indicates that the short molecular axis of the DBTTF and TCNQ is parallel to the bulk crystal surface, and the stack axis is parallel to the short axis of the crystal, confirming our optical absorbance measurements. DFT calculations were performed on each crystal polymorph to further optimize the structure. The α-polymorph was optimized at the PBE30/6-31G level with lattice parameters fixed at experimental values using CRYSTAL14 . For the β-polymorph,

115 Figure 5.5: IR spectra for each type of crystal, polarized along both the long and short axes. Vibrations a-g are identiﬁed within the spectra, and pictured at the right. CN stretching (f) frequency shifts with respect to TCNQ charge.

lattice parameters obtained from SAED were fully optimized at the PBE30/6-31G level with a semi-empirical dispersion correction. This correction is based on Grimmes D2 model,30 but uses standard van der Waals radii as reported by Bondi31 with the exception of hydrogen.32 A summary of D and A molecular orientation within the single crystals of each polymorphs, as well as with respect to the crystal edge parallel to the substrate, as obtained by combining the above mentioned measurements is included in Fig. 5.6.

5.2.2 Degree of Charge Transfer

Charge transfer complexes are characterized by partial ionicity. Organic metals fall in the 0.5e–0.75e range, although having such degree of charge transfer does not guarantee metallicity.33 Techniques to quantify the degree of charge transfer exploit

116 Figure 5.6: Molecular orientation of DBTTF and TCNQ with respect to the crystal surface, and long and bulk crystal axes. The xy plane is drawn as a guide, and is oriented to the crystal face. The α-polymorph crystal structure is used here, while the β-polymorph is approximated based on spectroscopy and DFT. the change in the bond lengths and vibration frequencies of the acceptor/donor as a result of the partial addition/loss of charge. In this study we determined the degree of charge-transfer via Raman spectroscopy by evaluating the dependence of the TCNQ

34 ν4 (exocyclic C=C stretching) vibration frequency on the surrounding charge. The measurements were performed using a Renishaw inVia Raman microscope with a 532 nm excitation laser polarized along different crystal axes. The results are shown in Fig. 5.7. The α-polymorph exhibited a shift to 1426 cm−1 in our measurements, corresponding to a degree of charge transfer of 0.5e. This is in agreement with the value reported in the literature based on IR spectroscopy. The β-polymorph appears more neutral, showing a very modest shift to 1448 cm−1, which corresponds to q = 0.13e. XPS provides a complementary measurement to detect elements in different electronic states, but it only probes the first 5-10 nm of the crystal surface. As shown in Fig. 5.3a, the spectral shapes of the DBTTF in the α and β structures have

117 Figure 5.7: The Raman spectrum for each polymorph from 1000 to 1500 cm−1

pronounced differences. The S atoms in the α-polymorph display a shoulder from 165-166 eV and show higher intensity at 164.5 eV when compared to the β-DBTTF– TCNQ polymorph. To understand the spectra of the polymorphs, the spectra were deconvoluted and we applied the methods used by Medjanik, et al35 for spectral assignment. The states of the sulfur atoms here are adequately described by two pairs of lineshapes, which suggests that the atoms detected at the surface of the α-polymorph are in two different chemical (or electronic) environments one charge-transfer state and one neutral. In contrast, the S 2p spectrum of the β-DBTTF–TCNQ shows that the S atoms are in one chemical (or electronic) state. The single state present in the β-DBTTF–TCNQ spectrum is assigned to neutral DBTTF and this indicates that the degree of CT at the surface of the crystal is zero. The fraction of charged DBTTF molecules in the α-polymorph based on the fitted results of the S 2p spectrum is 0.31 ± 0.02. While we cannot confidently report q based on this measurement, it is clear that the α-polymorph contains more charged S atoms than the β-polymorph. The

118 N 1s spectra for the polymorphs also convey the diﬀerences between the two types of crystal, and are shown in Fig. 5.3b. The α-polymorph is described by three lineshapes: one centered at 398.6 eV, one centered at 399.6 eV, and one centered at 402.0

36 37 eV. Similar to the studies focusing on TTF–TCNQ and F4TCNQ, we assign the low-binding-energy feature at 398.6 eV as the TCNQ moiety within the CT complex that does undergo CT and the high-binding-energy feature at 399.6 eV as the TCNQ moiety that does not undergo CT. The N 1s spectrum of the β-DBTTF–TCNQ dis- plays a prominent peak at 399.2 eV, and a less intense, broad peak at 401.8 eV. This spectrum is very similar to earlier reports of TCNQ,38 and we assign the main peak as the direct N 1s photoemission of electrons in the TCNQ moiety of the CT crystal while the broad feature is assigned as a shake-up satellite feature. The fraction of charged TCNQ in the same polymorph is estimated from the fit results in Fig. 5.3b as (0.23 ± 0.02). The fraction of DBTTF molecules that undergo CT is greater than TCNQ within the α-DBTTF–TCNQ polymorph. This provides structural insight into the localization of the charge within the α-DBTTF–TCNQ polymorph, where a greater amount of charge is near the thiophene group relative to the cyano group. The XPS results also provide additional evidence of the different electronic structures between the two polymorphs, where CT is detected at the surface in the α-DBTTF–TCNQ structure and not in the β-DBTTF–TCNQ. The complementary measurements described in this section confirm that the α- polymorph exhibits a high degree of charge transfer of 0.5e, while the β-polymorph is nearly neutral. Such a difference results from the shift in the molecular overlap of the donor and acceptor molecules, in agreement with the theoretical work suggesting that q, similar to charge carrier mobility, depends on the molecular overlap.16

119 5.2.3 Charge Transport in DBTTF–TCNQ Polymorphs

In this section we present the effect of the differences in the crystal structure and partial ionicity on the charge transport in this material system. We measured the electrical characteristics of the two DBTTF–TCNQ polymorphs by fabricating single crystal OFETs of the top-gate, bottom-contact design. The device structure is included as an inset in Fig. 5.8a. At least 30 crystals of each polymorph were investigated and all measurements were performed in the D/A stack direction. To fabricate the field-effect transistors, source/drain electrodes (5 nm Ti/45 nm Au) were pat-

terned by photolithography and deposited by e-beam evaporation on a 200 nm SiO2 substrate that served the role of an insulating substrate. Following cleaning with hot acetone, hot isopropanol, and a UV/ozone treatment, the crystals were laminated by hand onto the substrate as they were thin enough to stick by electrostatic adhesion. The crystals were aligned with the contacts so that the charge-transfer stack axis was directed across the OFET channel. To achieve the top-gate conﬁguration, 300-750

nm N-parylene (εr = 2.65, thickness determined from capacitance measurements) was deposited onto the crystals following a procedure described elsewhere,39, 40 and silver (60 nm) was thermally evaporated to act as a gate contact. Electrical properties were measured in air and in the dark by using an Agilent 4155C semiconductor parameter analyzer. Current-voltage (I-V) characteristics are shown for a typical device in Fig. 5.8 for the α-polymorph and Fig. 5.9 for the β-polymorph. Here, panels (a) and (b) show the

transfer characteristics (the evolution of the drain current (ID) with the gate-source

voltage (VDS)), when positive (a) or negative (b) source-drain bias (VDS) is applied and the device operates in saturation mode (VDS = ±60 V). The left axis shows the square root of ID, and the hole and electron mobilities, µh and µe, were calculated from the slope of these plots by using standard procedures.41–43 The right axis, in blue,

shows ID on a logarithmic scale. The eﬀects of a sweeping VDS on the ID for diﬀerent

120 applied VGS, namely the output characteristics, are shown in panels c and d. The non-linear behavior of the low VDS region of the output characteristics originates from the injection barriers between the metal contacts and the semiconductor materials, which are inherent in ambipolar FET devices. Ambipolar transport is observed in both polymorphs, but the diﬀerences between hole and electron enhancement are clear. In the α-polymorph the electron transport

2 −1 −1 is dominant, with the average electron mobility µeα = 0.4 ± 0.2 cm V s being one order of magnitude greater than the average hole mobility in the same crystals, µhα = 0.04 ± 0.02 cm2V−1s−1. These results are in agreement with those reported pre-

Figure 5.8: Electrical characteristics for α-DBTTF–TCNQ, with the device structure shown as an inset of plot (a). The channel geometry is L/W = 0.85. The transfer characteristics for negative drain voltage are shown in (a) and for positive drain voltage in (b), with the black axis showing the square root of the drain current versus the voltage and the blue axis shoing a log-scale plot of the drain current versus voltage. The output characteristics for the α-polymorph are shown for negative and positive gate voltage in (c) and (d) respectively, indicating electron-favored transport.

121 Figure 5.9: Electrical characteristics for the β-polymorph. The channel geometry is L/W = 0.75. The transfer characteristics for negative drain voltage are shown in (a) and for positive drain voltage in (b). Panels (c) and (d) show the output characteristics for the β-polymorph.

viously for solution-grown microrods of the α-polymorph DBTTF–TCNQ on similar contacts, though the electron mobility in our devices is slightly higher.27 This is to be

44 expected, as the use of SiO2 in earlier reports inhibits electron transport. We note, however, that another study showed only electron transport in the α-polymorph, with a similar device structure, but at much lower applied voltages (linear regime of device operation).45 We believe hole transport is observed in our studies because the devices are biased at higher drain voltages. The β-polymorph also shows ambipolar behavior

2 −1 −1 but favors hole transport with mobilities of µhβ = 0.1 ± 0.07 cm V s and µeβ = 0.03 ± 0.02 cm2V−1s−1, as shown in Fig. 5.9. These trends are consistent for all the crystals we investigated in this study.

122 5.3 Discussion

It is well known that the differences in crystal structure that are inherent in polymorphs result in modifications in the band structures and physical properties.10, 46 This has been measured, for example, in thin films of diindenoperylene by a combination of UPS, IPES (inverted photoemission spectroscopy), low-energy electron diffraction (LEED), and XPS.47 In order to understand the differences induced in the band structure of the DBTTF–TCNQ polymorphs, we performed both UV-Vis- NIR absorption spectroscopy (Fig. 5.4) and UPS measurements (Fig. 5.10). UPS

Figure 5.10: UPS spectra of the HOMO region for the α-polymorph (top) and the β-polymorph (bottom) with respect to the Fermi energy (EF). The data of the polymorphs (open circles) are shown after subtracting the background contribution from the Au substrate, and a ﬁt of one Gaussian lineshape (red line) is used to estimate the center of the HOMO. The dashed line is a guide to the eye for the small, but measurable shift of the HOMO between the two polymorphs.

123 measurements were performed in the same XPS vacuum system stated previously by using an He discharge lamp (21.2 eV) as the excitation source and applying a -5 V bias to the sample to clear the work function of the spectrometer. The energy scale was calibrated by measuring the Fermi step in a bare Au sample. In each spectrum (Fig. 5.10), the background contribution from the Au was removed in order to clearly resolve the HOMO feature (open circles). Finally, one Gaussian lineshape was used to estimate the center of the HOMO (solid red line). The optical absorption measurements show that the α-polymorph CT band peaks at 0.78 eV, while the β-polymorph peaks at 0.73 eV. Each of these absorption bands results from the CT excitations from the HOMO level of the DBTTF to the LUMO level of TCNQ and thus provide an estimate for the band-gap of each DBTTF–TCNQ polymorph, suggesting a nar- rower value for the β-polymorph. These results demonstrate that (1) polymorphism in molecular crystals results in changes in HOMO/LUMO levels which may impact charge injection, and (2) the band-gap formation of the CT salt is more complex than simply mixing the levels of the donor and acceptor species. Based on the UPS results, we find a HOMO shift with respect to the Fermi energy level of 1.03 ± 0.07 eV for the α-polymorph and 1.10 ± 0.05 eV for the β-polymorph, which implies that the different polymorphs have different hole-injection barriers when contacted to Au source/drain electrodes. While we cannot unambiguously quantify the difference in HOMO between the two structures because the standard deviations in these mea-

surements are of similar magnitude to the energy shift value (∆EHOMO = 0.07eV), the UPS measurements suggest that the shift is present. Such a small shift, however, is not sufficient to explain the differences observed in charge transport, as it can be overcome by band bending. The existence of distinct unit cells implies that the molecular overlap between the donor (DBTTF) and the acceptor (TCNQ) is different between the two polymorphs (see Fig. 5.6). Calculations and experimental work previously performed on mono-

124 Figure 5.11: Band structure and density of states for the α- (left) and β-polymorphs (right) of DBTTF–TCNQ. The points of high symmetry in the ﬁrst Brillouin zone are labeled as follows: Γ = (0,0,0), X = (0.5,0,0), Y = (0,0.5,0), Z = (0,0,0.5), A = (0.5,0.5,0), B = (0.5,0,0.5), C = (0,0.5,0.5), D = (0.5,0.5,0.5).

molecular crystals show that very small changes in the intermolecular overlap of the adjacent organic semiconductor molecules can yield dramatic changes in intermolecular electronic coupling, which directly impact charge transport properties.14, 48, 49 In order to evaluate this eﬀect in CT complexes, in parallel to our experimental measurements aimed at understanding the electronic structure of the polymorphs, we performed DFT calculations on the known α-structure and the estimated β-structure. Transfer integrals for the α- and β- polymorphs were calculated by extracting the monomer, dimer, or trimer geometries from the optimized crystal and running at the B3LYP/6-31G level in Gaussian09. These results are presented in Fig. 5.11. The transfer integrals were found to be approximately equal for the hole and electron transport in α-polymorph, with the value for holes (th = 87 meV ) slightly higher

15 than that of electrons (te = 73 meV), in agreement with previous report. The β- polymorph transfer integrals were also found to be approximately equal, with th =

8 meV and te = 19 eV. These results suggest a more eﬃcient charge transport in the α-polymorph, in agreement with our experimental results. In addition, they also point to an intrinsic balanced transport in both polymorphs, i.e. similar mobilities

125 for electrons and holes in each type of crystal. In devices, however, the presence of extrinsic factors such as the injection barriers formed at the metal/semiconductor interface or polarization at the semiconductor/dielectric interface may, and indeed have, led to different results. Takahashi, et al., showed that in OFETs fabricated on the α-polymorph,45, 50 electron-only transport, ambipolar transport, and hole-only transport can be achieved by selecting source-drain electrodes of different work functions. Since we used the same type of metal (gold) to contact both polymorphs in the present study, contact choice alone did not cause the differences we observed. A possible reason for the unbalanced electrical properties of the two crystalline structures may be related to their different responses to the presence of the trap states and defects, as we have demonstrated for another system.10 A more significant contribution to these differences was suggested by a recent theoretical study, which showed17 that the effect of molecular overlap on electronic coupling in CTs is not necessarily mirrored for holes and electrons — i.e, their transfer integrals and therefore their mobilities, do not increase and decrease simultaneously. Consequently, the bulk material may exhibit primarily hole, primarily electron, or ambipolar transport, depending on the strength of interactions of molecular orbitals participating in charge transport. For example, if the electronic coupling is dominated by the interaction between the donor HOMO and the acceptor LUMO, balanced electron/hole transport is expected. Otherwise, electron or hole dominant transport, or even unipolar transport may be achieved. In addition, the shifts in molecular overlap are also expected to induce modulation of q, as proposed by Sini, et al.,16 and confirmed here experimentally. The higher degree of charge transfer observed in the α-DBTTF–TCNQ suggests a higher crystal binding energy in the α-polymorph than the β-polymorph, as these two features are directly related.51 Therefore, we attribute the differences in transport — electron-favored transport in the α-polymorph and hole-favored transport in the β-polymorph — to different D/A electronic coupling and the presence of different

126 degrees of charge transfer between the D and A species, both caused by diﬀerences in the D/A overlap.

5.4 Conclusions

In conclusion, we discovered a new polymorph of the charge transfer complex DBTTF– TCNQ, which we refer to as the β-polymorph. The existence of two distinct crystalline structures with the same D/A ratio provides us with a unique system for the study of the impact of the donor-acceptor interactions on the optoelectronic properties of organic charge transfer complexes. We find that the α-polymorph shows a large degree of charge transfer between the constituent molecules of about 0.5e, while in the β-polymorph the charge transfer is weak, quantified by a degree of charge transfer of about 0.1e. Interestingly, while both crystals exhibit ambipolar transport, electron- transport dominates in the α-polymorph, whereas hole-dominant charge transport is observed in the β-polymorph. The changes in the electronic structure as a result of different crystalline structures are minimal, as determined by UPS. We conclude that the differences detected in the electrical properties are a result of the interplay between the effect of the varying degree of charge transfer and electronic coupling between the donor and acceptor. These differences likely result in a variance in the molecular orbitals contributing to transport due to differing band structures between the two polymorphs.

127 References

[1] K. P. Goetz, J. Tsutsumi, S. Pookpanratana, J. Chen, C. A. Richter, C. A. Hacker, T. Hasegawa, and O. D. Jurchescu. Polymorphism in the 1:1 Charge- Transfer Complex DBTTF-TCNQ and Its Effects on Optical and Electronic Properties. Submitted, 2016. [2] C. A. Hunter and J. K. M. Sanders. The Nature of pi-pi interactions. Journal of the American Chemical Society, 112(2):5525–5534, 1990. [3] J. Chen, M. Shao, K. Xiao, A. J. Rondinone, Y. Loo, P. R. C. Kent, B. G. Sumpter, D. Li, J. K. Keum, P. J. Diemer, J. E. Anthony, O. D Jurchescu, and J. Huang. Solvent-type-dependent polymorphism and charge transport in a long fused-ring organic semiconductor. Nanoscale, 6(1):449–56, 2014. [4] W. H. Taylor. X-Ray measurements on diflavylene, rubrene, and related compounds. Zeitschrift fur Kristallographie, 93(151-155), 1936. [5] Z. A. Akopyan, R. L. Avoyan, and Yu T. Struchkov. Crystallographic data on certain sterically strained naphthacene derivatives. Journal of Structural Chem- istry, 3(5):576–579, 1962. [6] D. E. Henn, W. G. Williams, and D. J. Gibbons. Crystallographic data for an orthorhombic form of rubrene. Journal of Applied Crystallography, 4(3):256–256, 1971. [7] I. Bulgarovskaya, V. Vozzhenikov, and S. V. Aleksandrov. No Title. Latv PSR Zinat Akad Vestis Fiz Teh Zinat Ser, 4:53–59, 1983. [8] O. D. Jurchescu, A. Meetsma, and T. T. M. Palstra. Low-temperature structure of rubrene single crystals grown by vapor transport. Electrochemical and Solid- State Letters, 9(5):330–334, 2006. [9] C. C. Mattheus, A. B. Dros, J. Baas, G. T. Oostergetel, A. Meetsma, J. L. De Boer, and T. T. M. Palstra. Identification of polymorphs of pentacene. Synthetic Metals, 138(3):475–481, 2003. [10] L. A. Stevens, K. P. Goetz, A. Fonari, Y. Shu, R. M. Williamson, J.-L. Brédas, V. Coropceanu, O. D. Jurchescu, and G. E. Collis. Temperature-mediated polymorphism in molecular crystals: The impact on crystal packing and charge transport. Chemistry of Materials, 27(1):112–118, 2015.

128 [11] A. Brillante, I. Bilotti, R. G. Della Valle, E. Venuti, and A. Girlando. Probing polymorphs of organic semiconductors by lattice phonon Raman microscopy. CrystEngComm, 10(8):937, 2008.

[12] A. Brillante, I. Bilotti, R. G. Della Valle, E. Venuti, S. Milita, C. Dionigi, F. Bor- gatti, A. N. Lazar, F. Biscarini, M. Mas-Torrent, N. S. Oxtoby, N. Crivillers, J. Veciana, C. Rovira, M. Leufgen, G. Schmidt, and L. W. Molenkamp. The four polymorphic modiﬁcations of the semiconductor dibenzo-tetrathiafulvalene. CrystEngComm, 10(12):1899, 2008.

[13] R. Pfattner, M. Mas-Torrent, I. Bilotti, A. Brillante, S. Milita, F. Liscio, F. Biscarini, T. Marszalek, J. Ulanski, A. Nosal, M. Gazicki-Lipman, M. Leuf- gen, G. Schmidt, W. M. Laurens, V. Laukhin, J. Veciana, and C. Rovira. High-performance single crystal organic ﬁeld-eﬀect transistors based on two dithiophene-tetrathiafulvalene (DT-TTF) polymorphs. Advanced Materials, 22(37):4198–4203, 2010.

[14] J.-L. Br´edas,J. P. Calbert, D. A. da Silva Filho, and J. Cornil. Organic semiconductors: a theoretical characterization of the basic parameters governing charge transport. Proceedings of the National Academy of Sciences, 99(9):5804–9, 2002.

[15] L. Zhu, Y. Yi, Y. Li, E. G. Kim, V. Coropceanu, and J.-L. Br´edas. Prediction of remarkable ambipolar charge-transport characteristics in organic mixed-stack charge-transfer crystals. Journal of the American Chemical Society, 134(4):2340– 2347, 2012.

[16] G. Sini, J. S. Sears, and J.-L. Bredas. Evaluating the performance of DFT functionals in assessing the interaction energy and ground-state charge transfer of donor/acceptor complexes: Tetrathiafulvalenetetracyanoquinodimethane (TTFTCNQ) as a model case. Journal of Chemical Theory and Computation, 7(3):602–609, mar 2011.

[17] H. Geng, X. Zheng, Z. Shuai, L. Zhu, and Y. Yi. Understanding the Charge Transport and Polarities in Organic Donor-Acceptor Mixed-Stack Crystals: Molecular Insights from the Super-Exchange Couplings. Advanced Materials, 27(8):1443–1449, 2015.

[18] T. Mori and H. Inokuchi. Structural and electrical properties of (BEDT- TTF)(TCNQ). Solid State Communications, 59(6):355–359, 1986.

[19] S. Flandrois and D. Chasseau. Longueurs de liaison et transfert de charge dans les sels du t´etracyanoquinodim´ethane(TCNQ). Acta Crystallographica Section B, 33:2744–2750, 1977.

[20] T. C. Umland, S. Allie, T. Kuhlmann, and P. Coppens. Relation between geometry and charge transfer in low-dimensional organic salts. The Journal of Physical Chemistry, 92(22):6456–6460, 1988.

129 [21] H. M. Yamamoto, M. Hagiwara, and R. Kato. New phase of (BEDT- TTF)(TCNQ). Synthetic Metals, 133-134:449–451, 2003.

[22] P Guionneau, C. J. Kepert, G. Bravic, D. Chasseau, M. R. Truter, M. Kurmoo, and P. Day. Determining the charge distribution in BEDT-TTF salts. Synthetic Metals, 86(1 -3 pt 3):1973–1974, 1997.

[23] T. Mori and H. Inokuchi. Crystal structure of the mixed-stacked salt of bis (ethylenedithio)-tetrathiafulvalene (BEDT-TTF) and tetracyanoquinodimethane (TCNQ). Bulletin of the Chemical Society of Japan, 60(1):402–404, 1987.

[24] T. J. Kistenmacher, T. J. Emge, F. M. Wiygul, W. A. Bryden, J. S. Chappell, J. P. Stokes, L-Y. Chiang, and D. O. Cowan. DBTTF-TCNQ: A fractionally- charged organic salt with a mixed-stack crystalline motif. Solid State Commu- nications, 39:415–417, 1981.

[25] H. Kobayashi and J. Nakayama. The Crystal structure of the charge-transfer compex of dibenzotetrathiafulvalene-tetracyanoquinodimethane, DBTTF-TCNQ. Bulletin of the Chemical Society of Japan, 54(8):2408–2411, 1981.

[26] T. J. Emge, F. M. Wiygul, J. S. Chappell, A. N. Bloch, J. P. Ferraris, D. O. Cowan, and T. J. Kistenmacher. Crystal structures for the electron donor dibenzotetrathiafulvalene DBTTF, and its mixed-stack charge-transfer salts with the electron acceptors 7,7,8,8-tetracyano-p-quinodimethane, TCNQ, and 2,5- diﬂuoro-7,7,8,8-tetracyano-p-quinodimethane, 2,5-TCNQF2. Molecular Crystals and Liquid Crystals, 87:137–161, 1982.

[27] H. Wu, F. Wang, Y. Xiao, and G. Pan. Preparation and ambipolar transistor characteristics of co-crystal microrods of dibenzotetrathiafulvalene and tetracyanoquinodimethane. Journal of Materials Chemistry C, 1(12):2286–2289, 2013.

[28] D. Vermeulen, L. Y. Zhu, K. P. Goetz, P. Hu, H. Jiang, C. S. Day, O. D. Jurch- escu, V. Coropceanu, C. Kloc, and L. E. McNeil. Charge Transport Properties of PeryleneTCNQ Crystals: The Eﬀect of Stoichiometry. The Journal of Physical Chemistry C, 118(42):24688–24696, 2014.

[29] J. Tsutsumi, T. Yamada, H. Matsui, S. Haas, and T. Hasegawa. Competition between charge-transfer exciton dissociation and direct photocarrier generation in molecular donor-acceptor compounds. Physical Review Letters, 105(22):226601, 2010.

[30] S. Grimme. Semiempirical GGA-type density functional constructed with a long- range dispersion correction. Journal of Computational Chemistry, 27(15):1787– 1799, 2006.

130 [31] A. Bondi. van der Waals Volumes and Radii. The Journal of Physical Chemistry, 68(3):441–451, 1964.

[32] B. Civalleri, C. M. Zicovich-Wilson, L. Valenzano, and P. Ugliengo. B3LYP aug- mented with an empirical dispersion term (B3LYP-D*) as applied to molecular crystals. CrystEngComm, 10(4):405–410, 2008.

[33] K. P. Goetz, D. Vermeulen, M. E. Payne, C. Kloc, L. E. McNeil, and O. D. Jurch- escu. Charge-transfer complexes: new perspectives on an old class of compounds. Journal of Materials Chemistry C, 2:3065–3076, 2014.

[34] S. Matsuzaki, R. Kuwata, and K. Toyoda. Raman spectra of conducting TCNQ salts; estimation fo the degree of charge transfer from vibrational frequencies. Solid State Communications, 33:403–405, 1980.

[35] K. Medjanik, a. Gloskovskii, D. Kutnyakhov, C. Felser, D. Chercka, M. Baum- garten, K. M¨ullen,and G. Sch¨onhense. Charge transfer in the novel donoracceptor complexes tetra- and hexamethoxypyrene with tetracyanoquinodimethane studied by HAXPES. Journal of Electron Spectroscopy and Related Phenomena, 185(3-4):77–84, apr 2012.

[36] C. Rojas, J. Caro, M. Grioni, and J. Fraxedas. Surface characterization of metallic molecular organic thin ﬁlms : tetrathiafulvalene tetracyanoquinodimethane. Surface Science, 482-485:546–551, 2001.

[37] N. Koch, S. Duhm, J.P. Rabe, A. Vollmer, and R. L. Johnson. Optimized Hole Injection with Strong Electron Acceptors at Organic-Metal Interfaces. Physical Review Letters, 95(23):237601, 2005.

[38] J. M. Lindquist and J. C. Hemminger. High-Resolution Core Level Photoelectron Spectra. The Journal of Physical Chemistry, 92(6):1394–1396, 1988.

[39] V. Podzorov, V. M. Pudalov, and M. E. Gershenson. Field-eﬀect transistors on rubrene single crystals with parylene gate insulator. Applied Physics Letters, 82(11):1739–1741, 2003.

[40] M. A. Reyes-Martinez, A. J. Crosby, and A. L. Briseno. Rubrene crystal ﬁeld- eﬀect mobility modulation via conducting channel wrinkling. Nature Communi- cations, 6(May):6948, 2015.

[41] H. Klauk. Organic thin-ﬁlm transistors. Chemical Society Reviews, 39:2643–2666, 2010.

[42] M. S. Kang and C. D. Frisbie. A pedagogical perspective on ambipolar FETs. ChemPhysChem, 14(8):1547–1552, 2013.

[43] E. Smits, T. Anthopoulos, S. Setayesh, E. van Veenendaal, R. Coehoorn, P. Blom, B. de Boer, and D. de Leeuw. Ambipolar charge transport in organic ﬁeld-eﬀect transistors. Physical Review B, 73(20):205316, may 2006.

131 [44] L. Chua, J. Zaumseil, J. Chang, E. C.-W. Ou, P. K.-H. Ho, H. Sirringhaus, and R. H. Friend. General observation of n-type ﬁeld-eﬀect behaviour in organic semiconductors. Nature, 434(7030):194–199, 2005.

[45] Y. Takahashi, J. Hasegawa, Y. Abe, Y. Tokura, K. Nishimura, and G. Saito. Tun- ing of electron injections for n-type organic transistor based on charge-transfer compounds. Applied Physics Letters, 86(6):063504, 2005.

[46] A. Troisi and G. Orlandi. Band structure of the four pentacene polymorphs and eﬀect on the hole mobility at low temperature. The Journal of Physical Chemistry B, 109:1849–1856, 2005.

[47] S. Krause, a. Sch¨oll,and E. Umbach. Interplay of geometric and electronic structure in thin ﬁlms of diindenoperylene on Ag(111). Organic Electronics, 14(2):584–590, 2013.

[48] O. Kwon, V. Coropceanu, N. E. Gruhn, J. C. Durivage, J. G. Laquindanum, H. E. Katz, J. Cornil, and J.-L. Br´edas.Characterization of the molecular parameters determining charge transport in anthradithiophene. The Journal of Chemical Physics, 120(17):8186–8194, 2004.

[49] V. Coropceanu, J. Cornil, D. A. da Silva Filho, Y. Olivier, R. Silbey, and J. L. Bredas. Charge transport in organic semiconductors. Chemical Reviews, 107:926– 952, 2007.

[50] Y. Takahashi, T. Hasegawa, Y. Abe, Y. Tokura, and G. Saito. Organic metal electrodes for controlled p- and n-type carrier injections in organic ﬁeld-eﬀect transistors. Applied Physics Letters, 88(7):073504, 2006.

[51] J. B. Torrance and B. D. Silverman. Charge transfer and ionic bonding in organic solids with segregated stacks. Physical Review B, 15(2):788–801, 1977.

132 Chapter 6

Summary

In summary, the three crystal systems presented in this dissertation illustrate some of the structure-fucntion relationships of mixed-stack organic charge-transfer complexes. The perylene–TCNQ system grows in three D:A stoichiometries, with the 1:1 exhibiting n-type transport, the 2:1 exhibit ambipolar transport, and the 3:1 exhibiting p-type transport in similar device structures. The degree of charge transfer increases from approximately 0 to 0.1 to 0.2 for the 1:1, 2:1, and 3:1 respectively.

The stilbene–F4TCNQ complex exhibits temperature-dependent charge transport and charge transfer that is associated with the freezing-in of a librational moiety within the donor molecule. Above the glass transition, charge transport is thermally activated and the degree of charge transfer is 0.1, but below, the charge transport is temperature independent and the degree of charge transfer is neutral. Finally, the DBTTF–TCNQ polymorphs demonstrate how small changes in molecular overlap without changing D:A ratio can impact charge transport and charge transfer. The α-polymorph exhibits electron-favored ambipolar transport and a degree of charge transfer of 0.5, while the β-polymorphs exhibits hole-favored ambipolar transport and a degree of charge transfer of 0.1. Together, these systems simultaneously verify some of the theoretical models in place for organic materials while highlighting challenges yet to be overcome. The stilbene–F4TCNQ system shows how librational modes can aﬀect transport. These

133 are present in all materials, but especially come into play with the so-called ”greasy groups”, or the functional groups bonded to an aromatic backbone to solubilize the molecule. The DBTTF–TCNQ system shows how crystals containing the same D and A constituents can demonstrate different electrical properties based on subtle shifts in molecular arrangement. Still lacking, however, is an agreement between theory and experiment on the exact nature of charge transport in these materials. Each of these materials was predicted to show ambipolar transport along the stack axis, with low effective masses for both holes and electrons. The experimental measurements, however, did not correlate well with predictions. Part of this stems from the lack of perfect crystals in nature, while theoretical predictions are based on defect-free structures. In addition, the need for metal contacts to measure a given material adds an extrinsic factor to the transport properties, increasing the difficulty in comparing experiment and theory. Future experiments should seek to bridge the gap between model and measurement, possibly by developing methods to incorporate disorder in the case of the former and by developing contact-less measurements in the case of the former. Such work would be a step towards making the dream of material design with predictable properties a reality.

134 Chapter 7

Curriculum Vitae

Education Ph.D., Physics May, 2016 Wake Forest University Winston-Salem, NC

B.S. in Physics with Honors, Minor in Mathematics May, 2011 Wake Forest University Winston-Salem, NC

Honors and Awards 2014 National Science Foundation Graduate Research Opportunities Worldwide (GROW) — travel award for research at the National Institute for Advanced Industrial Science and Technology (AIST) in Tsukuba, Japan, with Dr. Tatsuo Hasegawa

2014 Peer Mentor Award, Wake Forest University Department of Physics

2013 National Science Foundation Graduate Research Fellowship Program (GRFP)

2012 National Science Foundation East Asia and Paciﬁc Summer Institute (EAPSI) - travel award for research at Nanyang Technological University (NTU), Singapore, with Dr. Christian Kloc

2011 Undergraduate Degree Completed with Honors in Physics

2010 Wake Forest Fellowship for Summer Undergraduate Research

2010 Wake Forest Fellowship for Participation in the 2010 Plastic Electronics Conference in Dresden, Germany

135 Publications

1. K. P. Goetz, S. Pookpanratana, J. Tsutsumi, J. Chen, C. A. Richter, C. A. Hacker, T. Hasegawa, and O. D. Jurchescu. “Polymorphism in the 1:1 charge-transfer complex DBTTF–TCNQ and its impact on optical and electrical properties. Submitted, Adv. Elec. Mater. (2016).

2. J. L. Marshall, K. Uchida, C. K. Frederickson, C. Schutt, A. M. Zeidell, K. P. Goetz, L. N. Zakharov, C. Risko, R. Herges, O. D. Jurchescu, and M. M. Haley. “Indacenodi(benzothiophene)s: A Comprehensive study of the synthesis, optoelectronic properties, solid-state characterization and materials applications of a class of molecules possessing pronounced antiaromatic character, Accepted, Chemical Science (2016).

3. A. Fonari, N. Corbin, D. Vermeulen, K. P. Goetz, O. D. Jurchescu, L. E. Mc- Neil, J. L. Brdas, and V. Coropceanu. “Vibrational properties of organic donor- acceptor molecular crystals: Anthracene-pyromellitic dianhydride (PMDA) as a case study, J. Chem. Phys. 143, 224503 (2015).

4. P. J. Diemer, Z. A. Lamport, Y. Mei, J. W. Ward, K. P. Goetz, W. Li, M. M. Payne, M. Guthold, J. E. Anthony, and O. D. Jurchescu. “Quantitative analysis of the density of trap states at the semiconductor-dielectric interface in organic ﬁeld-eﬀect transistors, Appl. Phys. Lett. 107, 103303 (2015).

5. L. A. Stevens, K. P. Goetz, A. Fonari, Y. Shu, R. M. Williamson, J. L. Brdas, V. Coropceanu, O. D. Jurchescu, and G. E. Collis. “Temperature-mediated polymorphism in molecular crystals: The Impact on crystal packing and charge transport, Chem. Mater. 27, 112-118 (2015).

6. K. P. Goetz, A. Fonari, D. Vermeulen, P. Hu, H. Jiang, P. J. Diemer, J.W. Ward, M. E. Payne, C. S. Day, C. Kloc, V. Coropceanu, L. E. McNeil, and O. D. Jurchescu. “Freezing of librational modes induces crossover from thermally- activated to temperature-independent transport in organic semiconductors, Nat. Commun. 5, 5642 (2014).

7. P. Hu, L. Ma, K. J. Tan, H. Jiang, F. Wei, C. Yu, K. P. Goetz, O. D. Jurchescu, L. E. McNeil, G. G. Gurzdyan, and C. Kloc. “Solvent-dependent stoichiometry in perylene-7,7,8,8-tetracyanoquinodimethane charge transfer compound single crystals, Cryst. Growth Des. 14, 6376-6382 (2014).

8. D. Vermeulen, L. Y. Zhu, K. P. Goetz, P. Hu, H. Jiang, C. S. Day, O. D. Jurchescu, V. Coropceanu, C. Kloc, and L. E. McNeil. “Charge transport properties of perylene-TCNQ crystals: The Eﬀect of stoichiometry, J. Phys. Chem. C 118, 24688-24696 (2014).

9. J. W. Ward, K. P. Goetz, A. Obaid, M. M. Payne, P. J. Diemer, C. S. Day, J. E. Anthony, and O. D. Jurchescu. “Low-temperature phase transitions in a

136 soluble oligoacene and their effect on device performance and stability, Appl. Phys. Lett. 105, 083305 (2014). 10. M. E. Payne, K. P. Goetz, C. S. Day, and O. D. Jurchescu. “The 1:1 charge- transfer complex dibenzotetrathiafulvalene-pyromellitic dianhydride (DBTTF- PMDA), Acta Crystallogr. E 70, o844-o845 (2014). 11. K. P. Goetz, D. Vermeulen, M. E. Payne, C. Kloc, L. E. McNeil, and O.D. Jurchescu, “Charge-transfer complexes: new perspectives on an old class of compounds, J. Mater. Chem. C 2, 3065-3076 (2014). • Part of the emerging investigators edition of the Journal of Materials Chem- istry C • One of the 30 most downloaded papers of 2014 for this journal 12. D. Lehnherr, A. R. Waterloo, K. P. Goetz, M. M. Payne, F. Hampel, J. E. Anthony, O. D. Jurchescu, and R. R. Tykwinski, “Isomerically Pure syn- Anthradithiophenes: Synthesis, Properties, and FET Performance, Org. Lett. 14 (14), 3660 (2012). 13. K. P. Goetz, Z. Li, J. W. Ward, C. Bougher, J. Rivnay, J. Smith, B. R. Conrad, S. R. Parkin, T. D. Anthopoulos, A. Salleo, J. E. Anthony and O. D. Jurchescu, “Effect of Acene Length on Electronic Properties in 5-, 6-, and 7-Ringed Heteroacenes, Adv. Mater. 23, 3698-3703 (2011). 14. M. Coll, K. P. Goetz, B. R. Conrad, C. A. Hacker, D. J. Gundlach, C. A. Richter, and O. D. Jurchescu, Flip chip lamination to electrically contact organic single crystals on flexible substrates, Appl. Phys. Lett. 98, 163302 (2011).

Presentations

1. K. P. Goetz, J. Tsutsumi, S. Pookpanratana, J. Chen, C. A. Richter, C. A. Hacker, T. Hasegawa, and O. D. Jurchescu. “Polymorphism in the organic charge-transfer complex dibenzotetrathiafulvalene – 7,7,8,8-tetracyanoquino-dimethane (DBTTF–TCNQ) and its eﬀect on optical and electrical properties. Contributed Talk, SPIE Optics + Photonics, San Diego, California, August, 2015. 2. K. P. Goetz, D. Vermeulen, A. Fonari, P. Hu, H. Jiang, P. J. Diemer, J.W. Ward, M. E. Payne, C. S. Day, C. Kloc, V. Coropceanu, L. E. McNeil, and O. D. Jurchescu. “Temperature Activated Transport Tuned by Libration in the Charge-Transfer Salt STB–F4TCNQ, Contributed Talk, American Physical Society (APS) March Meeting, Denver, Colorado, March, 2014. 3. K. P. Goetz, D. Vermeulen, M. E. Payne, C. S. Day, V. Coropceanu, L. E. McNeil, C. Kloc, and O. D. Jurchescu. “The Impact of Structure and Donor- to-Acceptor Ratio on Electrical Properties in Charge Transfer Complexes, Con- tributed Talk, Electronic Materials Conference (EMC), South Bend, Indiana, June, 2013.

137 4. K. P. Goetz, M. E. Payne, D. Vermeulen, P. J. Diemer, J. W. Ward, C. S. Day, V. Coropceanu, L. E. McNeil, C. Kloc, and O. D. Jurchescu, “Electronic Properties of Organic Charge-Transfer Compounds, Poster Presentation, Mate- rials Research Society (MRS) Spring Meeting, San Francisco, California, April, 2013.

5. K. P. Goetz, Z. Li, J. W. Ward, C. Bougher, J. Rivnay, J. Smith, B. R. Conrad, S. R. Parkin, T. D. Anthopoulos, A. Salleo, J. E. Anthony, and O. D. Jurchescu. “Electronic Response to Structural Changes in 5-, 6-, and 7-Ringed Soluble Heteroacenes, Contributed Talk, Materials Research Society (MRS) Fall Meeting in Boston, Massachusetts, November, 2011.

6. K. P. Goetz, J. W. Owen, J. W. Ward, C. Bougher, Z. Li, Eric K. Chapman, B. R. Conrad, J. E. Anthony, and O. D. Jurchescu. “Tuning Crystal Packing to Produce High Mobility Organic Field Eﬀect Transistors. Contributed Talk, Plastic Electronics Conference in Dresden, Germany, October, 2010.

138