Direct Observation of the Dipole-Induced Energetic Disorder in

2 days ago - Beijing National Laboratory for Molecular Sciences, Key Laboratory of Organic Solids, Institute of Chemistry, Chinese Academy of Sciences...
0 downloads 4 Views 1MB Size
Subscriber access provided by Kaohsiung Medical University

Energy Conversion and Storage; Plasmonics and Optoelectronics

Direct Observation of the Dipole-Induced Energetic Disorder in Rubrene Single-Crystal Transistors by Scanning Kelvin Probe Microscopy Yuanyuan Hu, Lang Jiang, Qinjun Chen, Jing Guo, and Zhuojun Chen J. Phys. Chem. Lett., Just Accepted Manuscript • DOI: 10.1021/acs.jpclett.8b01274 • Publication Date (Web): 15 May 2018 Downloaded from http://pubs.acs.org on May 17, 2018

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 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 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.

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 21 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

The Journal of Physical Chemistry Letters

Direct Observation of the Dipole-Induced Energetic Disorder in Rubrene Single-Crystal Transistors by Scanning Kelvin Probe Microscopy Yuanyuan Hu1*, Lang Jiang2, Qinjun Chen1, Jing Guo1, Zhuojun Chen1 1

Key Laboratory for Micro-Nano Optoelectronic Devices of Ministry of Education, School of Physics and Electronics, Hunan University, Changsha 410082, China 2

Beijing National Laboratory for Molecular Sciences, Key Laboratory of Organic Solids, Institute of Chemistry, Chinese Academy of Sciences, Beijing 100190, China.

AUTHOR INFORMATION Corresponding Author *Email: [email protected]; Keywords: OFETs; dielectric dipoles; energetic disorder; charge transport; scanning Kelvin probe microscopy

ABSTRACT

ACS Paragon Plus Environment

1

The Journal of Physical Chemistry 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 21

It is commonly accepted that gate dielectric dipoles can induce energetic disorder in organic field-effect transistors. However, convincing experimental evidences that directly demonstrate this effect are still in lack. In this work, we present a combined experimental and theoretical study to reveal this effect. We have investigated the temperature-dependent mobility of two rubrene single-crystal transistors with different polymer dielectrics. Model fittings of the data indicate there is higher energetic disorder in the device on dielectric with larger permittivity. Scanning Kelvin probe microscopy was then employed to directly characterize the density of tail states, which is correlated with energetic disorder, in the devices. The results further confirm that dielectric dipoles can increase energetic disorder in organic semiconductors.

TOC GRAPHICS

As the most elementary device in organic electronics, organic field-effect transistors (OFETs) have been under intensive research in the past few decades.1-4 One well-known fact about OFETs is that the properties of dielectrics have a large influence on the electrical properties of OFETs. Early in 1990, Horowitz et al. reported the phenomena that there is a strong correlation between the insulator k (dielectric constant) values and the device mobility.5 Since then much efforts have been made to improve our understanding about the influence of the

ACS Paragon Plus Environment

2

Page 3 of 21 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

The Journal of Physical Chemistry Letters

permittivity of dielectrics on charge transport.6-8 The commonly accepted model for the explanation of the mobility decrease with increasing dielectric constant was proposed by Veres et al. in 2003.6 In this model, it is argued that static dipolar disorder in dielectrics would increase the energetic disorder in amorphous or polycrystalline organic semiconductors, and thus broaden the density of states (DOS) and enhance charge carrier localization, leading to lower mobility values.6,

9-10

Later, Richards et al. set up an analytical model to calculate and verify the

broadening of DOS caused by dipolar disorder.8 More recently, Minder et al. demonstrated the there is also dipole-induced disorder in PDIF-CN2 single-crystal (SC) FETs by model fittings.1112

Despite impressive progress in understanding of the influence of gate dielectrics on charge transport in OFETs, direct experimental evidence showing the induction of tail states or energetic disorder in organic semiconductors by dielectric dipoles is still in lack. This is mainly due to the challenges in accurately probing energetic disorder of organic semiconductors. Some interfering factors such as interface roughness/morphology at the semiconductor/dielectric interface that may contribute to energetic disorder in semiconductor layers also further complicates the experimental work of demonstrating this effect. Nevertheless, since this effect is such an essential effect in OFETs, it is crucial to reveal it directly from experimental measurements. With this goal in mind, in this work we compared the electrical performance of two rubrene SC-FETs fabricated on different polymer dielectrics, one with a lower dielectric constant (CYTOP, εins=2.1) and the other with a higher dielectric constant (PMMA, εins=3.5). SC-FETs were chosen for the study because the rubrene single crystals were physically transferred onto the polymer dielectrics, which excludes the possibility that dielectrics would influence the semiconductor morphology or structure at the semiconductor/dielectric interface.13,14 By fitting

ACS Paragon Plus Environment

3

The Journal of Physical Chemistry 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 21

the dependence of mobility on temperatures with a mobility-edge (ME) model, we were able to obtain information about energetic disorder in the two SC-FETs, which can provide important insights into the effect of dielectric dipoles on inducing energetic disorder in rubrene crystals. Moreover, we employed scanning Kelvin probe microscopy (SKPM) to probe tail states in rubrene SC-FETs with different polymer dielectrics. The analysis of our observation, in conjunction with the model fitting results, provide direct experimental evidence showing that dielectric-dipoles can increase energetic disorder in the semiconductor layers of OFETs. Figure 1(a) shows the bottom-gate, top-contact device structure used in this study, which is appropriate for carrying out SKPM measurements. The optical microscopy image of a typical rubrene SC-FET device is shown in Figure 1(b). The transfer characteristics of the two devices with CYTOP (ε=2.1) and PMMA (ε=3.5) as dielectrics measured in saturation regime (Vd = –60 V) at room temperature are presented in Figure 1(c). It should be noted that the current is

normalized as  = = ( −  ), in which Id is the originally measured drain current 

and  is the capacitance per unit area. It is seen that both the two devices exhibit quite ideal transfer characteristics, by which the device mobility can be reliably extracted.15-16 The saturation mobility of the devices was extracted using the aforementioned equation, which yields 5.33 cm2/Vs and 1.42 cm2/Vs for the devices on CYTOP and PMMA, respectively. The observation of a lower mobility for the device on PMMA, which has a higher dielectric constant, is fully in accordance with previous reports.11-12

ACS Paragon Plus Environment

4

Page 5 of 21 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

The Journal of Physical Chemistry Letters

Figure 1 (a) Molecule structure of rubrene and device structure of the SC-FETs used in this study. (b) Optical microscopy image of a rubrene SC-FET. (c) Transfer characteristics (measured at room temperature) of the rubrene SC-FETs fabricated on two dielectrics.

The lower mobility in the rubrene/PMMA device seems to indicate that there is higher energetic disorder in this device, which can be attributed to the stronger dielectric dipoles, just as expected. However, care should be taken for this conclusion because there is an alternative explanation for the dependence of mobility on dielectric properties of the gate insulator in SCFET devices. Hulea et al. proposed that the DOS in single crystals could be assumed to be extended states and thus the effect of DOS broadened by dipoles in dielectrics is weak. They reported the formation of Fröhlich polarons, which are caused by interaction of charges with induced dielectric dipole moments, to explain the mobility dependence on dielectrics in rubrene SC-FETs.17-18 One thing should be noted is that the dielectrics used in their work are mostly

ACS Paragon Plus Environment

5

The Journal of Physical Chemistry 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 6 of 21

high-k dielectrics, which results in strong coupling between charge carriers and the dielectric dipoles. For low-k dielectrics like the polymer dielectrics used in this study, Minder et al. argued that the formation of Fröhlich polarons is unlikely and the broadening of DOS, namely, the induction of more tail states or energetic disorder is still responsible for the dependence of mobility on dielectric permitivitties, even in organic single crystals.11 In this case, it is more appropriate to attribute the influence of gate dielectrics to the so-called dipolar disorder,6, 8 which originates from the random orientation of the dipolar components in polymeric dielectrics. In other words, the stronger dipolar disorder in PMMA induces more energetic disorder in the rubrene crystals, which is detrimental to charge transport and results in a lower mobility compared to the rubrene/CYTOP device. Furthermore, temperature-dependent electrical properties of the two devices were measured in order to get more intrinsic information about charge transport. The transfer characteristics at varied temperatures for the rubrene/CYTOP and rubrene/PMMA device are shown in Figure 2(a) and (b), respectively. When temperature goes below 200 K, the difference between the thermal expansion of the crystal and substrates might cause a cracking of the devices.17, 19 So, we only got limited data in the temperature region of 200-300 K for the device on CYTOP. The dependence of mobility on temperature for the two devices is illustrated in Figure 2(c). It is seen that band-like transport occurs between 300-240 K in the rubrene/CYTOP device. In comparison, the mobility of rubrene/PMMA device keeps decreasing with the lowering temperature, indicating a thermally activated transport. These results again imply that there are more tail states or localized states in the rubrene/PMMA device, which prevents band-like transport from occurring by localizing the charge carriers.

ACS Paragon Plus Environment

6

Page 7 of 21 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

The Journal of Physical Chemistry Letters

To quantitatively understand the difference in the energetic disorder between the two devices, we fitted the temperature-dependent mobility data using the mobility-edge (ME) model.11-12, 20-22 This model assumes that there is a defined energy in the DOS that separates mobile states (delocalized states) from tail states (localized states). Charge carriers contribute to charge transport only when they are in the mobile states and the intrinsic mobility for the mobile states is given by  () = α  with a material-specific parameter .22 Here we assume that the density of band states is constant to be  (for E < 0) while the tail states follow a Gaussian distribution:  ( ) =  !"# ( in which  and



% & %'&

) (for E > 0).

(1)

are the height and width of the distribution, respectively. The degree of

energetic disorder can be represented by the width of the Gaussian distribution

.

Since the total

amount of charge carriers in the device is the summation of mobile charges and trapped charges, which is equal to the amount of the gate voltage-induced interface charges, the Fermi level energy

(

can be obtained by solving the equation: 

) = )*+ + )- = .2  /( where /(

(,

(,

2

)1 + .  ( )/(

(,

)1

) is the Fermi-Dirac distribution. Once the Fermi level energy

(2)

(

is obtained, the

device mobility can be calculated as: () =  ()

34567 (8)

(3)

'4'96

Under these assumptions, we fitted the temperature-dependent mobility data with  =1015 cm-2 eV-1 and three fitting parameters α,  as well as

.

ACS Paragon Plus Environment

7

The Journal of Physical Chemistry 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 21

The fitting results are shown in Figure 2(c), in which it is seen that the experimental data for the two devices can be fitted reasonably well by the model. The parameters  and



extracted

from the fits are 4.7×1013 cm-2 eV-1 and 45 meV for CYTOP, and 9.5×1013 cm-2 eV-1 and 75 meV for PMMA, respectively, with a sharing parameter α = 8.51×105 cm2 K2 V-1 s-1 for the two fittings. The fitting results therefore validate the idea that dipoles present in polymer dielectrics can induce significant energetic disorder in organic semiconductors.

Figure 2 Temperature-dependent transfer characteristics (measured at Vd = − 60 V) for (a) rubrene/CYTOP and (b) rubrene/PMMA devices. (c) Temperature dependence of the mobility for devices on the two dielectrics. The mobility values are determined from the saturation transfer characteristics. The solid and dash lines are fits to data based on the ME model.

Having identified the effect of gate-dielectric dipoles on inducing energetic disorder in organic semiconductors by model fittings, we were eager to see if this effect can be directly observed by experiments. This is quite challenging since very delicate experiments are required to probe trivial changes in energetic disorder that is caused by dielectric dipoles. We realized that

ACS Paragon Plus Environment

8

Page 9 of 21 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

The Journal of Physical Chemistry Letters

SKPM could be an appropriate tool for this work since it has been shown to be powerful for directly measuring DOS in organic semiconductors with high accuracy and reproducibility,23-26 and for extracting energetic disorder in organic semiconductors.27 Ideally, the effect of dipoles on inducing disorder can be directly proved if the DOS in rubrene/CYTOP and rubrene/PMMA devices are extracted. However, the accurate measuring of DOS requires that the semiconductor is very thin so that the potential distribution can be approximately taken as uniform,25,

28

a

condition which is not satisfied in SC-FETs since the thickness of the crystals is over several tens of nms. Therefore, instead of extracting the DOS in the two devices accurately and discussing the dipole-induced disorder effect quantitatively, we seek to obtain more phenomelogical information regarding energetic disorder in the rubrene single crystals by SKPM. Similar to the way of extracting DOS by SKPM, we measured the dependence of surface potential Vsur on the applied gate voltages Vg in a grounding device, as shown in the inset in Figure 3(a). As Vg becomes more negative, there are more charge carriers induced into the channel, filling the DOS in the semiconductor and moving the Fermi energy level downwards relative to the highest occupied molecule orbitals (HOMO). Consequently, this Fermi level energy shift is reflected in the Vsur and thus can be captured by SKPM. Figure 3 (a) and (b) shows the gate-dependent potential profiles in rubrene/CYTOP and rubrene/PMMA device, respectively. Here, the scanning range is only 8 µm due to the limitation of our SKPM instrument. The tip was located around the source or drain electrodes allowing the tip to scan cross both the electrode and the channel. During the whole measurement, the electrodes were kept grounded so that the measured Vsur there was always a constant value at about –0.18 eV, which reflects the work function difference between the conductive tip and the gold electrode.

ACS Paragon Plus Environment

9

The Journal of Physical Chemistry 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 21

Figure 3 Measured variation of Vsur with Vg for (a) rubrene/CYTOP and (b) rubrene/PMMA device. Schematic diagram showing the SKPM measurements of surface potential profiles is illustrated in the inset of Figure 3(a). (c) Extracted dependence of Vsur on Vg for the two devices. The Vsur values were extracted by averaging the data points at the left part of each potential profile (1.5 µm long). The slope of the dash lines were used for extracting the density of tail states in the devices. Comparison of (d) energetic disorder and (e) trap density for the two devices. The data of energetic disorder shown in Figure 3(d) is from the fittings results of Figure 2(c).

The Vsur-Vg relationship for the two devices then can be extracted from these profiles, as seen in Figure 3(c). The Vsur was seen to respond well to the Vg variation at low Vg, indicating that charge carriers are filling the tail states in the bandgap of the semiconductor. Then the Vsur

ACS Paragon Plus Environment

10

Page 11 of 21 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

The Journal of Physical Chemistry Letters

tends to saturate at some high Vg (Vg < –20 V for CYTOP and Vg < –10 V for PMMA). This is because in such conditions the Fermi level is close to HOMO, where a large density of states exists.23 The variation of Vsur with Vg at low Vg actually reflects the DOS of tail states in terms of: ) ( ) =

:;(%) :%

=

 :