High-Mobility, Ultrathin Organic Semiconducting Films Realized by

Oct 5, 2017 - The functionality of common organic semiconductor materials is determined by their chemical structure ... Kwon, Kim, Choi, Jeong, Kim, L...
0 downloads 4 Views 2MB Size
Subscriber access provided by LAURENTIAN UNIV

Communication

High-mobility, Ultrathin Organic Semiconducting Films Realized by Surface-mediated Crystallization Ilja Vladimirov, Matthias Kellermeier, Thomas Geßner, Zarah Molla, Souren Grigorian, Ullrich Pietsch, Lilian Sophie Schaffroth, Michael Kühn, Falk May, and R. Thomas Weitz Nano Lett., Just Accepted Manuscript • DOI: 10.1021/acs.nanolett.7b03789 • Publication Date (Web): 05 Oct 2017 Downloaded from http://pubs.acs.org on October 5, 2017

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Nano Letters is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 22

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Nano Letters

High-mobility, Ultrathin Organic Semiconducting Films Realized by Surface-mediated Crystallization I. Vladimirov1,2, M. Kellermeier1, T. Geßner1, Z. Molla³, S. Grigorian³, U. Pietsch³, L.S. Schaffroth4, M. Kühn1,+, F. May1,2, R.T. Weitz1,2,4,5,# 1: BASF SE Carl-Bosch-Straße 38, 67063 Ludwigshafen am Rhein, Germany 2: InnovationLab GmbH Speyerer Str. 4, 69115 Heidelberg, Germany 3: Department of Physics, University of Siegen, Emmy-Noether-Campus, Walter-Flex-Str. 3, 57072 Siegen, Germany 4: Physics of Nanosystems, Physics Department, Ludwig Maximilians Universität München, Amalienstrasse 54, 80799 Munich, Germany 5: Nanosystems Initiative Munich (NIM) and Center for NanoScience (CeNS) Ludwig Maximilians Universität München, Schellingstraße 4, 80799 Munich, Germany

+ email: [email protected] # email: [email protected]

The functionality of common organic semiconductor materials is determined by their chemical structure and crystal modification. While the former can be fine-tuned via synthesis, a-priori control over the crystal structure has remained elusive. We show, that the surface tension is the main driver for the plate-like crystallization of a novel small organic molecule n-type semiconductor at the liquid-air interface. This interface provides an ideal environment for the growth of mm-sized semiconductor platelets that are only few nm thick and thus highly attractive for application in transistors. Based on the novel high-performance perylene-diimide, we show in as-grown, only 3 nm thin crystals electron mobilities of above 4 cm²/Vs and excellent bias stress stability. We suggest that the established systematics on solvent parameters can provide the basis of a general framework for more deterministic crystallization of other small molecules.

ACS Paragon Plus Environment

1

Nano Letters

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 22

Keywords: organic transistor, perylene-diimide, surface crystallization, bias stress stability

ACS Paragon Plus Environment

2

Page 3 of 22

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Nano Letters

Controlling the crystallization of organic materials is a key aspect in many fields of research and still far from being fully understood

1, 2

. In the field of organic electronics,

high-quality crystals are essential for two reasons. On the one hand, they serve as model systems for generating a basic understanding of the underlying charge transport properties 3. On the other hand, it is crucial to improve the performance and practical usability of organic semiconductors in applications such as organic transistors 4. This is why significant effort has been devoted to understanding and controlling the crystallization process of small molecules and polymers 5-7. Two popular methods for the realization of highly crystalline thin films via solution processing are either the solventshearing

8, 9

approach, where crystallization is guided via mechanical forces, or solvent-

antisolvent crystallization 5, where crystallization is induced via solubility differences. A common feature of the high crystal quality obtained using the two methods is, that the crystallization commences at the liquid-air interface. While the particular conditions for heterogeneous nucleation and growth have been studied in detail for inorganic minerals forming from aqueous solutions

10, 11

, much less is known about surface-mediated

crystallization of small organic molecules in non-aqueous media and has not been the focus of previous studies in organic thin films5-9. Here, we have systematically addressed this issue and show that the balance of interfacial tensions drives crystallization of organic semiconductors at the surface of an evaporating solvent droplet. Careful choice of solvents allowed us to tune the crystallinity of a novel perylene diimide n-channel semiconductor from nearly amorphous to highly crystalline films that can be as thin as only two molecular layers and yet extend over multiple hundreds of microns in lateral dimension. The highest quality crystals are characterized by high charge carrier mobility of above 4 cm²/Vs and extreme bias-stress stability.

ACS Paragon Plus Environment

3

Nano Letters

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 22

A schematic of the crystallization experiments is shown in Figure 1a where the semiconductor

PDI1MPCN2

(N,N´-di((S)-1-methylpentyl)-1,7(6)-dicyano-perylene-

3,4:9,10-bis(dicarboximide), Figure 1a inset, synthesis details are given in

12

) has been

dissolved in various solvents, dropcast onto a substrate and subsequently dried at a controlled temperature. The full experimental details are given in the methods section. We found via optical and atomic-force-microscopy (AFM) studies, that if the semiconductor is crystallized from solvents that possess relatively high viscosity (η ≥ 1.3 mPa*s) and surface tension (γLA ≥ 30 - 35 mN/m), mostly very thin crystals (2-5 nm) homogeneously covering large areas (several 100 µm²) of the substrate (Figure 1b and d) can be observed. Exemplary solvents are listed in Table 1, a full list is given in the supporting information (SI, Table S1). The largest flat crystals have been obtained with solvents from the phthalate family, for example dimethylphthalate (DMP) or diethylphthalate (DEP). Conversely, for solvents with lower viscosity and surface tension such as acetylacetone (AcAc) or amylacetate (AmAc), we only observed individual bulky 3D crystals (Figure 1c and e). These observations lead to the immediate question whether the viscosity or the surface tension of the liquid, or both, are the main driver for surface-mediated crystallization. Unfortunately, for most solvents that are suitable for semiconductor crystallization, viscosity and surface tension are either both low or both high (see Table S1 in the SI). In an attempt to disentangle the effects of viscosity and surface tension on crystallization, we have used toluene (a low viscosity solvent) and artificially increased its viscosity from 0.6 mPa*s to 8 mPa*s via the addition of polystyrene (PS). Over the entire range of viscosities, no sizable flat crystals but only small crystals with lateral dimensions in the 10 µm range were observed (see Table 1 and Figure S1 in the SI), while 2D ACS Paragon Plus Environment

4

Page 5 of 22

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Nano Letters

crystallization was clearly favored in pure solvents with η ≈ 8 mPa*s. Control experiments (see Figure S1 in the SI) have shown that the addition of PS to high viscosity solvents does not affect the crystallization behavior. Furthermore, we have determined that the surface tension γLA is not influenced by the addition of PS (γLA = 25,4 mN/m in pristine toluene, whereas γLA = 24,7 mN/m (24,9 mN/m) upon the addition of 0,66 wt% (1,3 wt%) of PS ). This is clear evidence that high viscosity alone is not enough to favor 2D crystallization. The role of the viscosity rather seems to be related to the hydrodynamic conditions during surface crystallization. It is well known, that in a drying droplet the temperature dependence of the liquid-air surface tension leads to a so-called Marangoni flow

13

. The velocity of this flow along the liquid-air boundary in the

limit of large contact angles increases with decreasing viscosity of the solvent

14

.

Another way of expressing the magnitude of Marangoni flow is the so-called Marangoni coefficient Ma

13, 14

, which is high in solvents that we obtained bulk crystals from (e.g.

Ma = 1.5 in toluene) and low in those where we observed surface crystallization (e.g. Ma = 0.08 in DMP). It is reasonable to assume that slower Marangoni flow is beneficial for controlled crystallization at the liquid-air interface. The crystallization behavior of solvents with a low viscosity will be again discussed below. After having established that the viscosity of the solvent does not play the dominating role in the competition between surface and bulk crystallization, we now turn to the role of the surface tension. From classical nucleation theory it is known, that heterogeneous nucleation i.e. nucleation at an interface, is governed by the balance of the three relevant interfacial energies: crystal (or nucleus)–liquid (γCL), crystal–air (γCA) and liquid– air (γLA also termed surface tension). Competition of the free energies favors nucleation

ACS Paragon Plus Environment

5

Nano Letters

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

at the liquid-air interface if γCL + γCA< γLA.

15

Page 6 of 22

The surface free energy of the nucleus (γCA)

can be approximated as that of the fully grown crystals (even though the structure of the nucleus is likely to be different from the bulk phase), which was determined to (28.6 ± 2.3) mN/m by contact angle measurements on thin crystalline films. The interfacial energy between the nucleus and the solvent (γCL) is accessible by homogeneous nucleation experiments, where induction times are measured for different levels of supersaturation (S) (Figure 2a).11 The interfacial energy is then obtained from a plot of the nucleation rate versus (ln(S))-2 (Figure 2b, also see SI Figure S2 for an experimental determination of S). Such analyses yields γCL values of below 0,2 mN/m for both high- and low-viscosity solvents. Similarly low values have been reported previously in e.g. salicylic acid16. The fact that the interfacial energy between the PDI1MPCN2 nucleus and the organic solvent does not show a pronounced dependence on the type of solvent (even though we used solvents that yield 3D crystals (AmAc) or thin films (DMP), respectively) leaves the surface tension of the solvent as sole parameter to decide whether surface or bulk crystallization is favored. This is in line with the data in Figure 1 and the SI, where solvents with γLA on the order of 30 -35 mN/m did produce thin films while those with γLA ≤ 30 mN/m did not. For solvents with high surface tension the condition γCL + γCA < γLA is met so that heterogeneous crystallization of the semiconductor at the liquid-air interface leads to a minimization of interfacial free energy. Similar arguments hold in principle also for crystal growth at the substrate-liquid interface which however was not observed in our experiments. While this could be due to lower interfacial energy, we believe the main reason is related to local supersaturation profiles: since crystallization is induced by evaporation of the solvent, the degree of supersaturation is highest near the liquid-air interface with the consequence that ACS Paragon Plus Environment

6

Page 7 of 22

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Nano Letters

nucleation and growth will predominantly occur at the droplet surface and not the liquidsubstrate interface. Finally, some solvents we tested, have a surface tension above the threshold of about 30 mN/m², but did predominantly not lead to the formation of thin crystalline films (e.g. chlorobenzene) (see SI, Table S1 for a complete list). This observation is on first sight at odds with our above reported observation that such solvents actually should allow for the molecules to crystalize at the liquid air interface. We currently believe, that the reason that such solvents show only bulky crystals lies in the role of the solvent viscosity. As can be seen in Table S1, the solvents that do not show surface crystallization have also a small viscosity. As mentioned above, the smaller the viscosity, also the more turbulent the liquid during the drying process and therefore possibly surface crystallization is suppressed – even though it should be possible from a purely surface tension point of view. We have therefore also increased the viscosity of a chlorobenzene solution via the addition of PS and observed small, potentially flat crystals (see Figure S1 in the SI) which give a first hint, that increasing the viscosity of a low viscosity but high surface tension solvent could yield also flat crystals. It is out of the scope of this study to answer the question if the observed crystals sizes can be made larger. Possibly we have not met the right processing conditions such as temperature, or the interplay between viscosity, surface tension and polymer additives is more complex than anticipated. The main disadvantage of using solvents with high surface tension is, that they usually also have high boiling points and low volatility, with the implication that typical drying times are usually very long, e.g. 1-2 days for a 1µL droplet of DMP at 70°. To improve

ACS Paragon Plus Environment

7

Nano Letters

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 8 of 22

the practical usability of our crystallization approach, we have investigated mixtures of a high- and a low boiling point solvent. Even though the resulting surface tension and viscosity of the mixtures are well below the above-determined threshold for surface crystallization in pure solvents, we obtained thin films with crystal domain sizes of multiple 100 µm. By adjusting the mixing ratio and the drying temperature, we were able to tune the size of our PDI1MPCN2 crystals from less than 1 µm to several 100 µm (see Figure S3 in the SI), while the growth rate, i.e. the progression of the crystal growth front across the droplet surface, can be varied between 0.5 and >2.5 µm/sec. With the higher growth rates achieved in solvent mixtures, it furthermore became possible to observe the crystallization process in situ with different experimental techniques and thus provide a direct proof for the claim that growth of our thin films indeed takes place at the liquid-air interface. The most direct way that has allowed us to observe the surface crystallization in-situ, is polarized optical microscopy in transmission mode. Figure 2d shows a series of in-situ polarized microscopy images of a DMP:Toluene mixture (1:3) at 70°C, where one can readily observe the progressing crystal front at the surface of the droplet (see also Figure S4 in the SI). An AFM image of a dried film is shown in Figure 2c. A movie (see supporting Movie S1) shows how the solvent slowly retracts from below the highly crystalline film. We additionally performed quartz-crystal microbalance measurements for different mixing ratios of highboiling point to low-boiling point solvent mixtures. These measurements are detailed in the SI (Figure S5) and give a further indication that growth of the thin film commences at the liquid-air interface. Finally, we have investigated the thin film both via grazingincidence x-ray diffraction (GIXD) as well as in-situ GIXD measurements in a drying film (data in SI Figure S6 and S7) where the most intense out-of-plane peak (001) was ACS Paragon Plus Environment

8

Page 9 of 22

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Nano Letters

recorded during crystallization and settling of the film onto the substrate. We observed, that the out-of-plane periodicity of the thin film does not change neither during the growth stage of the film on the liquid nor after when it has settled onto the substrate. This is another indication, that the crystallization at the liquid-air interface provides an ideal interface for the growth of highly crystalline films since it naturally confines the molecules towards a platelike 2D growth. In addition our interface growth technique allows for crystal growth independent of surface tension differences on the substrate (e.g. between metal contacts and dielectrics etc.). Next to the improvement of the practical usability of a shorter drying time, using binary solvent mixtures allowed us also to tune the crystallinity of our thin films between large extended thin crystals of multiple 100 µm in lateral dimension and crystal sizes of only few µm or 3D crystals. Using these thin films as channel material in an organic transistor allowed us to compare charge transport in the resulting thin films across a wide range of morphologies. The transfer and output characteristics of a field-effect transistor (schematic shown in inset to Figure 3e) employing a 3 nm, highly crystalline thin organic layer of PDI1MPCN2 as semiconductor grown from a 1:3 DMP:Toluene mix are shown in Figure 3a and b. We find an excellent electrical performance of µmax = 4.3 cm²/Vs (Figure 3c) and an Ion/off of 105. To the best of our knowledge, this is one of the highest n-channel mobilities for a solution-processed n-channel transistor and is especially remarkable given that the semiconducting film of that particular transistor is only two monolayers thin

17, 18

. Statistics of 84 transistors crystalized from 1:3 DMP:Toluene

solution which were measured on a substrate different from the one showing the highest mobility is given in the inset to Figure 3c. Their average mobility is 2.9 ± 0.5 cm²/Vs. More electrical statistics are given in the SI (Figure S8). While these measurements ACS Paragon Plus Environment

9

Nano Letters

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 22

were done using the films with the largest crystalline areas, we also used the versatility of our solution approach to tune the grain size across several orders of magnitude, showing a mobility increase for higher degree of crystallinity (i.e. a lower density of grain boundaries) (Figure 3d, see Figure S10 in the SI for transfer curves of thin films deposited from solvent mixtures). The grain sizes were determined via polarization microscopy (SI Figure S3). In general, the grain boundaries are known to act as traps for carriers thus limiting charge carrier mobility. This observed trend of increasing mobility with increasing average crystal grain size (i.e. smaller density of grain boundaries) is a logical consequence and consistent with previous investigations performed on similar perylene-diimide derivatives 19. Apart from high charge carrier mobility, also the electrical stability of the semiconductors under extended operation is critical for their practical usability. As shown in Figure 3e and f, the stability under electrical stress increases significantly as the crystalline quality increases

20

(i.e. the density of grain boundaries decreases) consistent with previous

reports. In our best films, it takes as long as 1013 s for the current to decay by 7% - as compared to 103 s in the films with the largest density of grain boundaries (see SI Figure S9 for data on the threshold voltage shift upon stress). We believe that our combined results on surface-mediated crystallization of nm-thin organic layers of high charge carrier mobility of above 4 cm²/Vs is a significant advancement to realize high performing thin organic films also from other organic materials. For example, while in the present work we have focused on the crystallization of PDI1MPCN2, surface crystallization seems to be a general concept as studies with other molecules (e.g. C8-BTBT) showed similar behavior 21, 22.

ACS Paragon Plus Environment

10

Page 11 of 22

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Nano Letters

Materials and Methods Preparation of thin films Crystalline thin films of the studied organic semiconductors were prepared via drop casting (typically 1-5 µL) of filtered solutions (0.2µm PTFE filter) at a concentration of 0.1 wt% onto a heated substrate (50°C to 90°C). The process took place at ambient conditions otherwise with no further steps until the full drying and crystallization of the solution. Prior to solution deposition the substrate surface energy was selectively structured using self-assembling monolayers (SAMs) with different hydrophobicities, namely tetradecyl phosphonic acid (TDPA, 22 mN/m) and 4-ethoxyphenylphosphonic acid (EPPA, 55 mN/m). We have used SiO2 wafers or glass slides as substrates. TDPA treatment was applied first and then selectively removed in a round, mm² sized area through a polydimethylsiloxane (PDMS) shadow mask via oxygen plasma treatment (5 min at 150 W). EPPA was deposited subsequently. The resulting local difference in surface energy leads to a confined wettability and enables to localize the droplet and resulting crystal area to the more hydrophilic part of the substrate23. The largest area we have covered with thin organic crystals was about 30 mm² (see Figure S11 in the SI). The SAM treatment did not show an impact on the overall morphological or electrical properties. Surface energies were determined via contact angle measurements using a Krüss DSA100 setup and three test liquids (water, diiodomethane and glycerol), which were applied to three different positions on the sample. The surface energy was subsequently calculated according to the established method of Owens and Wendt.24 The surface tension of selected solvents and solvent / polymer mixtures was determined

ACS Paragon Plus Environment

11

Nano Letters

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 22

with the pendant drop method using a Krüss DSA100 setup. Induction time measurements Samples for induction time measurements were prepared by equilibrating an excess of solid PDI1MPCN2 in the respective solvent at different temperatures and filtrating the hot supernatant. Rapid cooling to room temperature in a water bath then gave metastable solutions with different degrees of supersaturation (S = c / ceq), which were determined from the ratio of the refractive indices of the supersaturated (n) and corresponding saturated solution (neq; see Figure S2 in the SI). Subsequently, the turbidity of the freshly prepared supersaturated solutions was monitored over time at 23°C using a probe from Metrohm (Optrode). From the resulting data, the induction time is obtained by determining the onset of turbidity as shown in Figure 2a. According to classical nucleation theory, the interfacial energy between the forming crystal and the surrounding solvent (γCL) can be derived by plotting ( ) vs. (1/ ( )²), with 







 = (−  ) and  =      " !$% [ref 1]. Here, J0 is the nucleation rate (the #

inverse of the induction time times the volume of the beaker used for crystallization (10 ml in our case)), A and B are considered to be constants, ν is the molecular volume in the crystalline phase (607 g/mol, 6x10-28 m3) , T the temperature, kB the Boltzmann constant, and c a shape factor that was assumed to be 1 due to the lack of knowledge on the structure of the nucleus (note that for c > 1, as often reported in the literature

25

,

smaller values of γCL would be obtained). Corresponding fits of the experimental data measured at different levels of supersaturation in DMP and AmAc (Figure 2b) yielded the values shown in Table 2.

ACS Paragon Plus Environment

12

Page 13 of 22

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Nano Letters

Optical microscopy images The polarized transmission in-situ optical microscopy was performed on an Olympus BX51 microscope with the drops in a Linkam LTS420 variable temperature closed stage at a constant temperature. The movies were recorded with a digital camera and their contrast was enhanced to visualize the boundary between the liquid and the evolving crystal front.

Transistor preparation and electrical measurement For the preparation of thin film transistors we used degenerately doped Si wafers as substrate and gate electrode. The dielectric was a layer of 30 nm Al2O3 that had been grown via atomic layer deposition. Wafers were modified with a patterned SAM as described above where the transistors were fabricated in the places where the more hydrophilic parts where the crystals had settled. Gold source and drain contacts were thermally evaporated on top of the semiconducting film with a distance of 50 µm and a width of 200 µm through a shadow mask. All electrical measurements were performed in a Lakeshore variable temperature probe station in high vacuum. The dielectric constant of the bare Al2O3 films and the SAM-treated substrates were determined to be 7.4 and 6.6, respectively. Acknowledgement: M.Kuehn acknowledges funding from the NMP-20-2014: Widening materials models program (project MOSTOPHOS, grant 646259). Financial support from the German Ministry of Education and Research (BMBF) within the project KOSADIS (FKZ ACS Paragon Plus Environment

13

Nano Letters

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 14 of 22

13N10766) in the framework of the leading edge cluster “Forum Organic Electronics” is gratefully acknowledged

Supporting Information: Supporting text file: Details on solution processing as well as electrical measurements Supporting video: Surface crystallization of the PDI1MPCN2 molecules from a DMP:Toluene (1:3) solution at 70°C.

ACS Paragon Plus Environment

14

Page 15 of 22

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Nano Letters

Figure 1: Surface-mediated crystallization of the small-molecule organic semiconductor PDI1MPCN2. a-c) Schematic of the crystallization behavior observed in droplets of solvents with different surface tension γLA and viscosity η b) In solvents of high γLA and η, crystallization proceeds along the liquid-air interface and leads to thin, platelike structures. c) For solvents of low γLA and η, disconnected 3D crystals are formed. d) and e) polarized optical microscopy images of solvents leading to thin-film growth (d) and bulky, disconnected (e) crystals.

ACS Paragon Plus Environment

15

Nano Letters

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 16 of 22

Figure 2: a) Induction time measurements for supersaturated solutions of PDI1MPCN2 in DMP and AmAc. b) Plot of the nucleation rate as a function of supersaturation, allowing for the determination of the interfacial free energy γcs. c) AFM image of an as-formed thin film from DMP:toluene (1:3) including a linecut. d) Time-lapse sequence of images taken during evaporation-driven crystallization of PDI1MPCN2 in a DMP:Toluene (1:3) solution at 70°C (droplet deposition at 0 min). The corresponding movie of the drying sequence is shown in the SI.

ACS Paragon Plus Environment

16

Page 17 of 22

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Nano Letters

Figure 3: Electrical properties of transistors prepared with crystalline thin films of PDI1MPCN2. In cases where the solvent mixing ratio is not indicated it was 1:3. All films were dried at 70 °C at ambient atmosphere.a-c) Transfer, output and linear mobility measurement of a 3 nm thin, highly crystalline film of the semiconductor PDI1MPCN2 grown from a DMP:Toluene mixture. Currents are given normalized to the channel width of 200 µm (Inset) Statistics of the mobility of 84 transistors measured on the same substrate. d) Linear mobility of thin films with various crystallinity. e) Bias stress measurements of thin films with different crystallinity. (Inset) Schematic of thin-film transistor setup used for electrical characterization. f) Comparison of the time until the current has degraded 7% of its initial value as a function of thin film crystallinity and (Inset) versus the initial mobility of the thin films.

ACS Paragon Plus Environment

17

Nano Letters

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Solvent

Page 18 of 22

Viscosity [mPa*s]

Crystal habit

Dimethyl phthalate (DMP) (0 wt% PS)

14.4

2D

DMP (0.15 wt% PS)

20

2D

Dimethyl sulfoxide (DMSO) (0 wt% PS)

4

2D

Toluene (0 wt% PS)

0.6

3D

Toluene (0.66 wt% PS)

3.5

3D

Toluene (1.3 wt% PS)

8

3D

Amylacetate (AmAc) (0 wt% PS)

0.9

3D

Acetylacetone (AcAc) (0 wt% PS)

0.8

3D

(wt % of Polystyrene (PS))

Table 1 Viscosities and resulting morphologies of PDI1MPCN2 crystals for a selection of solvents used in this work, including mixtures with polystyrene to artificially increase the viscosity.

ACS Paragon Plus Environment

18

Page 19 of 22

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Nano Letters

A (m-3s-1)

B

γ CL(mN/m)

DMP

87±16

(1.97±0.65) E-8

0.2

AmAc

556±14

(2.4±0.1) E-9

0.03

Table 2: Fit values as described in text to data in Figure 2b.

ACS Paragon Plus Environment

19

Nano Letters

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 20 of 22

References 1. Davey, R. J.; Schroeder, S. L. M.; ter Horst, J. H. Angew. Chem., Int. Ed. Engl. 2013, 52, (8), 2166-2179. 2. Jiang, Y.; Gong, H. F.; Volkmer, D.; Gower, L.; Colfen, H. Adv. Mater. 2011, 23, (31), 3548-3552. 3. Sirringhaus, H. Adv. Mater. 2005, 17, (20), 2411-2425. 4. Klauk, H. Chem. Soc. Rev. 2010, 39, (7), 2643-2666. 5. Minemawari, H.; Yamada, T.; Matsui, H.; Tsutsumi, J. y.; Haas, S.; Chiba, R.; Kumai, R.; Hasegawa, T. Nature 2011, 475, (7356), 364-367. 6. Diao, Y.; Tee, B. C. K.; Giri, G.; Xu, J.; Kim, D. H.; Becerril, H. A.; Stoltenberg, R. M.; Lee, T. H.; Xue, G.; Mannsfeld, S. C. B.; Bao, Z. Nat. Mat. 2013, 12, (7), 665-671. 7. Shaw, L.; Hayoz, P.; Diao, Y.; Reinspach, J. A.; To, J. W. F.; Toney, M. F.; Weitz, R. T.; Bao, Z. N. ACS Appl. Mater. Interfaces 2016, 8, (14), 9285-9296. 8. Giri, G.; Verploegen, E.; Mannsfeld, S. C. B.; Atahan-Evrenk, S.; Kim, D. H.; Lee, S. Y.; Becerril, H. A.; Aspuru-Guzik, A.; Toney, M. F.; Bao, Z. A. Nature 2011, 480, (7378), 504-508. 9. Giri, G.; Li, R. P.; Smilgies, D. M.; Li, E. Q.; Diao, Y.; Lenn, K. M.; Chiu, M.; Lin, D. W.; Allen, R.; Reinspach, J.; Mannsfeld, S. C. B.; Thoroddsen, S. T.; Clancy, P.; Bao, Z. A.; Amassian, A. Nat. Commun. 2014, 5, 3573. 10. Hu, Q.; Nielsen, M. H.; Freeman, C. L.; Hamm, L. M.; Tao, J.; Lee, J. R. I.; Han, T. Y. J.; Becker, U.; Harding, J. H.; Dove, P. M.; De Yoreo, J. J. Faraday Discuss. 2012, 159, 509-523. 11. Hamm, L. M.; Giuffre, A. J.; Han, N.; Tao, J. H.; Wang, D. B.; De Yoreo, J. J.; Dove, P. M. Proc. Natl. Acad. Sci. U. S. A. 2014, 111, (4), 1304-1309. 12. Kastler, M.; Vaidyanathan, S.; Doetz, F.; Koehler, S.; Yan, H.; Facchetti, A.; Lu, S.; Zheng, Y. Perylene Semiconductors and Methods of Preparation and use thereof WO2009EP51313. 13. Hu, H.; Larson, R. G. J. Phys. Chem. B 2006, 110, (14), 7090-7094. 14. Hu, H.; Larson, R. G. Langmuir 2005, 21, (9), 3972-3980. 15. Kashchiev, D., Nucleation : basic theory with applications. Butterworth Heinemann: Oxford ; Boston, 2000. 16. Mealey, D.; Croker, D. M.; Rasmuson, A. C. CrystEngComm 2015, 17, (21), 3961-3973. 17. Ringk, A.; Li, X. R.; Gholamrezaie, F.; Smits, E. C. P.; Neuhold, A.; Moser, A.; Van der Marel, C.; Gelinck, G. H.; Resel, R.; de Leeuw, D. M.; Strohriegl, P. Adv. Funct. Mater. 2013, 23, (16), 2016-2023. 18. Sizov, A. S.; Agina, E. V.; Gholamrezaie, F.; Bruevich, V. V.; Borshchev, O. V.; Paraschuk, D. Y.; de Leeuw, D. M.; Ponomarenko, S. A. Appl. Phys. Lett. 2013, 103, (4), 043310.

ACS Paragon Plus Environment

20

Page 21 of 22

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Nano Letters

19. Weitz, R. T.; Amsharov, K.; Zschieschang, U.; Villas, E. B.; Goswami, D. K.; Burghard, M.; Dosch, H.; Jansen, M.; Kern, K.; Klauk, H. J. Am. Chem. Soc. 2008, 130, (14), 4637-4645. 20. Muller, S.; Baumann, R. P.; Gessner, T.; Weitz, R. T. Phys. Status Solidi-RRL 2016, 10, (4), 339-345. 21. Vladimirov, I.; Brill, J.; Freyberg, D.; Weitz, R. T.; Musiol, T.; Koehler, S. Organic Semiconductor Composition Comprising Liquid Medium WO2015IB56248. 22. Weitz, R. T.; Vladimirov, I.; Chiodo, T.; Seyfried, T. Process for Preparing Crystalline Organic Semiconductor Material WO2015IB56249. 23. Kim, Y. H.; Yoo, B.; Anthony, J. E.; Park, S. K. Adv. Mater. 2012, 24, (4), 497-502. 24. Owens, D. K.; Wendt, R. C. J. Appl. Polym. Sci. 1969, 13, (8), 1741-1747. 25. Garside, J.; Mersmann, A.; Nyvlt, J., Measurement of crystal growth and nucleation rates - 2nd edition. The Institution of Chemical Engineers: 2002.

ACS Paragon Plus Environment

21

drying

Nano Letters Page 22 of 22 thin-film growth

substrate large surface tension

3 monolayers

1 2 3 small ACS Paragon Plus Environment 1 monolayer 4 surface tension ultrathin, highly crystalline, high electron mobiltiy 5 3D crystal growth 6

2 µm