Negative differential resistance and hysteresis in graphene-based organic light-emitting devices

Qin Zhang a, Shufen Chen *ab, Shuai Zhang a, Wenjuan Shang ac, Lihui Liu a, Minghao Wang a, Hongtao Yu a, Lingling Deng a, Guangqin Qi a, Laiyuan Wang a, Sanyang Han a, Bo Hu a, Qi Kang a, Yuejiao Liu a, Mingdong Yi a, Yanwen Ma a, Wenjing Yang a, Jing Feng c, Xiaogang Liu de, Hongbo Sun c and Wei Huang *ab
aKey Laboratory for Organic Electronics and Information Displays & Jiangsu Key Laboratory for Biosensors, Institute of Advanced Materials (IAM), Jiangsu National Synergetic Innovation Center for Advanced Materials (SICAM), Nanjing University of Posts & Telecommunications (NUPT), 9 Wenyuan Road, Nanjing 210023, China. E-mail:;
bShaanxi Institute of Flexible Electronics (SIFE), Northwestern Polytechnical University (NPU), 127 West Youyi Road, Xi'an 710072, Shaanxi, China
cState Key Laboratory on Integrated Optoelectronics, College of Electronic Science and Engineering, Jilin University, Changchun 130012, China
dDepartment of Chemistry, National University of Singapore, Singapore 117543, Singapore
eCentre for Functional Materials, NUS (Suzhou) Research Institute, Suzhou 215123, China

Received 11th November 2017 , Accepted 28th December 2017

First published on 1st January 2018

Here, we report the experimental observation of negative differential resistance (NDR) and hysteresis phenomena in graphene-based OLEDs and their effects on device performance. Our results reveal that the NDR and hysteresis mainly originate from the poly(methyl methacrylate) residue resting on graphene. We further demonstrate that current annealing is a facile and effective technique to remove the polymeric residue and eliminate the NDR, leading to the dramatically enhanced luminous efficiency from 30.9 to 41.6 cd A−1, and to 89.2 cd A−1 when equipped with a high index half-ball lens. The demonstrated pretreatment process of graphene establishes a new path for the construction of a wide variety of high performance optoelectronic devices.


The ultimate challenge in developing flexible displays and solid-state lighting based on organic light-emitting devices (OLEDs) is to overcome the drawback of brittleness inherent in conventional tin-doped indium oxide (ITO) electrodes. A fine control over the mechanical flexibility of OLEDs can be accessed by incorporating graphene-based materials into the devices because graphene offers a wonderful mechanical compliance, high transparency, and good conductivity.1–3 The urgent problems of current graphene-based OLEDs is that most devices suffer from relatively low electroluminescence (EL) performance, mainly arising from graphene's low work function and relatively high sheet resistance.1,2 Conventional approaches to resolving this problem include p-type doping of graphene with nitric acid,3,4 HAuCl4,5 trifluoromethanesulfonic acid,6 metal chloride (AuCl3,3,7 FeCl3,8 RhCl3,9–11 MoCl3,10 OsCl3,10 IrCl3,10,11 PdCl2,10 CuCl212), metal oxide (MoO3,13 WO3,14 V2O514,15), perfluorinated ionomer,3 or tetracyanoquinodimethane.16 For instance, Kim et al. employed a strong p-type chemical dopant bis(trifluoromethanesulfonyl)amide to dope graphene and obtained a polymer light-emitting diode with a maximum efficiency of 9.6 cd A−1 (10.5 lm W−1) and a brightness of 5 400 cd m−2 by significantly reducing five-layered graphene's sheet resistance from 240 to 90 Ω sq−1.17 Li et al. reported a significant enhancement in the work function of graphene by soaked it into a mixed solution of triethyloxonium hexachloroantimonate and dichloroethene to form a p-type doping.18 With this chemical treatment, they manufactured highly efficient green and white emissions with current efficiencies of 80 and 45 cd A−1 at 3000 cd m−2 (250 and 120 cd A−1 at 10[thin space (1/6-em)]000 cd m−2 with a high-index lens). In 2012, Han et al. modified graphene with HNO3 or AuCl3 to form a p-type doping and fabricated fluorescent and phosphorescent OLEDs with EL efficiencies of 37.2 and 102.7 lm W−1 when combined with the use of a gradient hole injection layer.3 Very recently, Lee's group has (part of the work cooperated with Ahn or Yoo's group) published series of green OLEDs with ultrahigh efficiencies of 168.4 cd A−1 (257.0 cd A−1 with half-ball lens) for single-junction and 205.9 cd A−1 (396.4 cd A−1 with half-ball lens) for tandem ones.3,4,19 By using gradient injection and microcavity technique, these devices’ luminous efficiencies reached the world's highest level among OLEDs with various anodes including ITO,20 carbon nanotubes,21 metal nanowires,22 metal grid23 and other conductive polymers.24,25 These excellent data indicate a bright application prospect of graphene electrode in high-performance OLEDs.

In this work, we report the observed phenomenon of negative differential resistance (NDR) and hysteretic behaviors in graphene-based OLEDs and their impact on emission stability and device performance. We found that cyclic scans of OLED's electrical characteristics in forward and reverse voltage directions or a fixed electric field treatment to device or even graphene sheet are beneficial to restraining electrical bistability and NDR, thus efficiently improving EL intensity and efficiency. As a proof of concept, we demonstrated that by suppressing the hysteresis and NDR of the devices over 156% and 35% increases in brightness and current efficiency can be obtained, respectively. These studies indicate that apart from sheet resistance and work function, NDR and hysteresis are two other limiting factors affecting the EL performance in graphene-based OLEDs. Hysteresis, especially NDR that originates from the poly(methyl methacrylate) (PMMA) residue can be eliminated with electrical scanning technique, which is particularly suitable for OLEDs with flexible substrates that cannot tolerate a high-temperature treatment.

Results and discussion

We used the trilayered graphene (TLG) as anode to fabricate OLEDs, with the transfer approach and cleaning process of TLG sheets described in detail in Experimental section. A 6 min UV O3 treatment was employed to decrease graphene's hydrophobicity before spin-coating poly(3,4-ethylenedioxythiophene):poly(4-styrenesulphonate) (PEDOT: PSS) hole injection layer. Nonetheless, the graphene-based OLEDs showed low fabrication yield and unstable EL characteristics, which can be attributed to the poor surface coverage of the PEDOT:PSS film and the PMMA residue absorbed on the graphene film, resulting in short circuit and leakage currents of the devices. The PEDOT:PSS coverage can be improved by spin-coating PEDOT:PSS twice through a two-step process. The surface roughness (RMS) of graphene electrode was reduced from 2.25 nm (naked graphene) to about 1.13 nm (surface morphology measured by atomic force microscopy (AFM), as shown in Fig. S1, ESI). In addition, the increased spin-coating steps decreased the graphene's sheet resistance from 470 ± 15 to 160 ± 10 Ω sq−1 (Fig. 1(a)), accompanied with a slight decrease in transparency from 91% (bare graphene) to 85% (after coating) at 550 nm (Fig. 1(b)). The decrease in the TLG's sheet resistance was mainly attributed to the p-doping effect of PEDOT:PSS.26 This was further confirmed by the Raman spectra in Fig. 1(c)–(e), of which the 2D band peak of the PEDOT:PSS-covered graphene was obviously blueshifted, as the statistic result indicating in Fig. 1(g). A thicker PEDOT:PSS layer on the graphene sheet resulted in a disappearance of both D, G and 2D peaks, as the case of spin-coating the PEDOT:PSS twice (Fig. 1(e)), which only showed the characteristic peaks of PEDOT:PSS (Fig. 1(f) and Fig. S2, ESI, 1366 and 1439 cm−1 come from Cβ[double bond, length as m-dash]Cβ and symmetric Cα[double bond, length as m-dash]Cβ(–O) stretches, while 1504 and 1566 cm−1 originate from C[double bond, length as m-dash]C and ring stretch modes). Based on the IG/I2D ratio calculated from more than 50 Raman spectra as shown in Fig. 1(h) and the transmissivity of 91% as shown in Fig. 1(b),18 the graphene used in this work was inhomogeneous and partially trilayer (3L).
image file: c7tc05148d-f1.tif
Fig. 1 Characterizations of graphene. (a) Sheet resistance and (b) transmission of graphene film (1), graphene/PEDOT:PSS (2; one-step coating) and graphene/PEDOT:PSS (3; two-step coating). (c–f) are Raman spectra of intrinsic graphene, PEDOT:PSS-coated graphene and PEDOT:PSS. (g) Is the distribution of 2D band peaks of graphene with and without PEDOT:PSS modification. (h) Statistics on the layer numbers by calculating IG/I2D ratio.

Orange-yellow OLEDs based on the phosphor emission of iridium(III)bis(4-phenylthieno[3,2-c]pyridinato-N,C2′)acetylacetonate (PO-01) were fabricated onto the TLG sheets with a schematic configuration shown in Fig. 2(a). Scanning results of current density–voltage (JV) characteristics showed that a NDR phenomenon occurred in JV curves with a very high current injection at a low driving voltage and then an abrupt drop at 5 V (black square curve in Fig. 2(b)). And no emission can be observed until the driving voltage reached at 3.9 V.

image file: c7tc05148d-f2.tif
Fig. 2 Schematic configuration and EL performances of OLEDs. (a) Schematic configuration of the graphene-based OLEDs, (b) JV, (c) luminance–voltage (LV) and (d) current efficiency–voltage (ηV) curves of OLEDs.

It is worth mentioning that the NDR behaviour usually emerges in memory devices with electrical bistability features,27–29 however, these two phenomena are rarely observed in OLEDs.30 The current–voltage (IV) curves were measured under both forward and reverse scanning (Fig. 3 and Fig. S3, ESI), and an obvious hysteretic behavior/bistability feature with a sudden increasing current at 3 V and an on/off ratio of four orders of magnitude for two conductivity states was found, which is similar to the flash memories. But read–write–erase cyclic measurements demonstrated that the graphene-based OLEDs do not possess stable memory ability, and the on/off ratio declines with increase in the read–write–erase measurement times, as shown in Fig. 3(b). But surprisingly, we found that IV curves pre-scanning before measuring emission performance can help to suppress hysteretic behavior/NDR, which can be regarded as a current annealing process of OLEDs. Herein, 5-repeated cyclic scans, following the sequence of 0 to 5 V, a reverse scan back to −5 V, and then another scan back to 0 V were used to sufficiently remove the hysteresis/NDR in our OLEDs. Along with eliminating the NDR, a decrease in Von from ∼3.9 V to ∼2.9 V and abrupt increases in luminance (L) and current efficiency (η) with 1–3 orders of magnitude at 4–5 V were respectively observed, as shown in Fig. 2(c) and (d). The maximum L and η are 59[thin space (1/6-em)]548 cd m−2 (14 V) and 41.6 cd A−1 (6 V), with the enhancement factors of 156% and 35% compared with the device without any scans of 23[thin space (1/6-em)]240 cd m−2 and 30.9 cd A−1. When a high refractive index half-lens (n = 1.922) was covered on the light output side of the glass substrate, an extremely high L and η of 76[thin space (1/6-em)]098 cd m−2 (13 V) and 89.2 cd A−1 (7 V) were achieved, amounting to one-fold enhancement over luminance and efficiency. This remarkable enhancement is due to an improved light outcoupling. The luminous efficiency in graphene-based OLED after cyclic scans maintained a high level, which is close to 52.2 and 89.0 (with lens) cd A−1 of ITO-based control device (Table 1).

image file: c7tc05148d-f3.tif
Fig. 3 IV curves in forward and reverse scan directions. (a and c) First- and (b and d) second-time scans on IV curves of OLEDs with graphene (a and b) and ITO (c and d) anodes.
Table 1 Summarized device performances
No. Device V on (V) L max @V (cd m−2@V) η max @V (cd A−1@V)
a V on: turn-on voltage. b Lmax: the maximum luminance. c η max: the maximum luminous efficiency. d OLED with high-index lens.
1 Without prescans 3.9 23[thin space (1/6-em)]240@14 30.9@8
2 With 5-repeated scans 2.9 59[thin space (1/6-em)]548@14 41.6@6
2d With lens and 5-repeated scans 2.8 76[thin space (1/6-em)]098@13 89.2@7
3 Clean graphene 2.9 55[thin space (1/6-em)]458@14 42.5@6
3d Clean graphene with lens 2.8 79[thin space (1/6-em)]159@13 94.0@7
4 ITO 3.0 81[thin space (1/6-em)]330@12 52.2@7
4d ITO with lens 3.0 147[thin space (1/6-em)]236@13 89.0@8

To investigate the origin of the NDR/hysteresis phenomenon in graphene-based OLEDs, we scanned the electrical properties of the ITO-based OLEDs (Fig. 3(c) and (d)) and found no NDR but weak hysteresis in these ITO devices, with difference in only one order of magnitude between the high and low conductivity state in the positive bias range. These results indicated that the NDR phenomenon is closely related to graphene, as further confirmed by investigating single-hole and single-electron devices with device structures and energy level diagrams shown in Fig. S4 (ESI). From the electrical features in Fig. S4(a) and (b) (ESI), we observed that the NDR characteristics in single-hole devices were similar to those in our graphene-based OLEDs, while the current in single-electron devices exhibited a normal increase against driving voltage.

Recently, the NDR phenomenon has been observed in graphene/insulator/graphene sandwiched devices through a resonant tunneling process.31–33 The low-temperature electrical characteristics of aforementioned single-hole device were conducted to verify the resonant tunneling process-induced NDR, which is universally observed under low temperature.34 From the low-temperature at 77 K data (Fig. S5, ESI), we did not observe any NDR in IV curves and thus excluded the possibility of resonant tunneling.

Although the NDR and hysteresis behaviours have never been reported in graphene-based OLEDs, hysteresis in conductance characteristics with gate is often observed in graphene-based field-effect transistors.35–38 Neighbouring adsorbates on graphene are generally considered to be the common cause of hysteresis. In our work, residual PMMA on the surface of graphene may generate the hysteresis and NDR due to the migration of oxygen ions in the polymer. To this end, the hysteresis and NDR depend on not only the sweep rate and time but also the temperature.39,40 Data analysis with fast- and slow-sweep rates (Fig. S6, ESI) suggests that slow sweep rate induced strong hysteresis and NDR, which can be restrained under a fast sweep rate. Moreover, both hysteresis and NDR almost eliminated at 77 K due to the migration difficulty of ionic under an extremely low temperature, irrespective of the scan rates (Fig. S5, ESI).

Here, an oxygen vacancy (VO) filament mechanism is proposed to illustrated the hysteresis/NDR phenomena in Fig. 4. When applying a positive driven bias onto graphene, an effective oxygen vacancy conductive path is created in PMMA due to migration of oxygen ions toward anode (equivalent to VO migration toward cathode direction), with which a large current (blue arrow in Fig. 4(b)) is injected into the graphene-based OLED. With a higher positive bias, the bottom layer of the PMMA island is essentially depleted of VO (Fig. 4(c)), thus resulting in a high overall resistance of the device and the occurrence of NDR (Fig. 2(b) and Fig. S3, ESI). The formation (Fig. 4(d)) and rupture (Fig. 4(e)) process of oxygen vacancy filament under a negative bias also generates high and low conduction states along with the NDR phenomenon (Fig. S3, ESI). After several cyclic JV scans to the as-fabricated OLEDs, the large electrical Joule heating generated inside the PMMA leads to the decomposition of PMMA.41 Additionally, a fixed electric field or pre-scanning graphene sheets with current has a similar effect on decomposing PMMA and eliminating hysteresis/NDR. This speculation was further confirmed by XPS characterization (Fig. 5), graphene sheets with and without 2 mA current pre-scanning treatment were prepared. XPS analysis performs the decreasing intensities of C–C (sp3 C, 285.6 eV), C–O (286.7 eV) and O–C[double bond, length as m-dash]O (288.35 eV) characteristic peaks after current scans, indicating the decomposition of PMMA.42,43 It's chemical structure and possible products were shown in the insets of Fig. 5(a) and (b). Furthermore, AFM characterization confirmed that the PMMA decomposition byproducts are likely to form aggregates under Joule heating (Fig. S7, ESI). Finally, it was observed that NDR phenomenon of 85% devices were effectively eliminated after current pre-scanning.

image file: c7tc05148d-f4.tif
Fig. 4 Schematic illustration of conduction mechanism of oxygen vacancy filament in graphene-based OLEDs. (a) No oxygen vacancy conductive path is formed in PMMA without any bias, (b) an effective oxygen vacancy conductive path is created by VO migration toward cathode and a large current (blue arrow) is injected into device when applying a positive bias onto graphene, (c) the bottom of PMMA island is depleted of VO with a high driven bias, thus the conductive path is blocked and the NDR occurs due to the rupture of oxygen vacancy filament, (d) when applying a negative bias onto graphene, an effective oxygen vacancy conductive path is created again with the migration of VOs toward anode, thus generating a large current in device, (e) at a high negative bias, the top of PMMA is depleted of VO and oxygen vacancy filament is broken.

image file: c7tc05148d-f5.tif
Fig. 5 XPS spectra of graphene. The C 1s spectra components (a) before and (b) after scans of TLG. The scanned current is 2 mA. Inset in (a) is the PMMA's schematic structure, while inset in (b) is the decomposed structure of PMMA after current treatment.

It is noticed that there was a non-NDR-type weak hysteresis existing in ITO-based OLEDs with decreasing ratio of high/low conductivity state under current scanning (Fig. 3(c) and (d)), which was related to the traps formation in bulk organic materials. Consequently, the effect of bulk traps in organic materials on the hysteresis behaviours of graphene-based OLEDs was also investigated. A theory model of trapped-charge-limited current was employed to calculate the trap concentrations in bulk organic materials from experimental IV curves. The fitted curves and the extracted parameters (Fig. S3 and Table S1, ESI) indicated that the trap concentration in the low-voltage NDR region was far higher than that in the high-voltage region (Table S1, ESI), which can decrease after IV curve scans. As result, both NDR and hysteresis were essentially eliminated after five consecutive scans. To summarize, the cyclic scan of IV curves can suppress NDR by removing the PMMA residue and eliminate hysteresis by filling the bulk traps in organic materials simultaneously.

Finally, we presented a direct comparison of the EL performances of graphene-based OLEDs devices employed TLG with some PMMA residue and clean surface (with cleaning process described in Experimental section), respectively. The devices with some PMMA residue on graphene were treated with above-mentioned current annealing technique. From results in Fig. 2(b)–(d) and Table 1, we observed that the device performance after current annealing treatment were similar to the pre-cleaning graphene device. In view of the challenge of PMMA residue totally removing from the graphene surface (see the XPS spectra in Fig. S8, ESI),44–48 the current annealing technique proposed in this work is much more effective and facile in graphene based OLEDs.


In conclusion, we have demonstrated that in addition to high sheet resistance and low work function, the NDR and the hysteresis are two equally important limiting factors that affecting the EL performance of graphene-based OLEDs through oxygen vacancy conductive filament and trapped-charge-limited current mechanisms. Importantly, it was found that the NDR and the hysteresis can be conveniently removed via current annealing technique, resulting in a significant increase in luminous efficiency with an enhancement factor of ∼35% and ∼189% using a high index half-ball lens. Compared with a high-temperature treatment approach, the current annealing technique in this work is more suitable for flexible devices with graphene electrodes because most of flexible substrates can’t withstand a high temperature sintering process. The cyclic scanning technique and the understanding gained from this study are expected to enlighten the construction of high performance optoelectronic devices in the fields of light-emitting diodes, solar cells, and potentially many others employing graphene electrodes.


Graphene transfer and characterization.

Graphene in this work was bought from ACS Material, which was synthesized with chemical vapor deposition method and transferred onto flexible substrates to form a trilayered structure. Before use, the TLG was dipped into water and then separated from the coarse flexible substrates, and finally it was transferred onto a target glass substrate. After that, it was washed with acetone at room temperature for 10 min to remove surface PMMA residue and this step was repeated three times. To obtain a clean graphene with less PMMA in the control graphene-based OLED, these graphene sheets were washed with hot acetone (60 °C) for 20 min and the cleaning process was repeated three times followed by a final DI water rinsing. The transmittance and sheet resistance of as-transferred TLG were measured with an ultraviolet-visible spectrophotometer (Shimadzu, UV-3600) and a four-point probe (RTS-9, China), while the surface morphology and roughness of TLG were measured with the AFM (Bruker, Dimension icon). The components of graphene were analyzed with X-ray photoelectron spectroscopy (PHI 5000 VersaProbe) and Raman spectrometer (Reinshaw, InVia). To observe the PMMA residue after a high temperature treatment, the graphene sheets were sintered under a high temperature of 500 °C for 4 h in a horizontal tube furnace with an Ar gas flow of 60 sccm. And then these sintered graphene was stored in N2 gas prior to a transfer to the vacuum chamber to take an XPS analysis.

Device fabrication

The as-transferred TLG was dried at 120 °C in an oven for 30 min before device fabrication. Then the graphene-covered substrate was treated with UV O3 for 6 min and transferred onto a spin-coater to spin-coat the PEDOT:PSS (Heraeus Clevios™ P VP AI 4083) with a two-step coating process (with a rotation rate of 2000 and 1000 rpm in the sequence, forming a total PEDOT:PSS thickness of 70 nm), followed by a dry process at 120 °C for 30 min. After that, the substrate was loaded into a glove box to spin-coat the first emitting layer 2,6-bis(3-(carbazol-9-yl)phenyl)pyridine(26DCzPPy):1,1-bis[4-[N,N-di(p-tolyl)amino]phenyl]cyclohexane (TAPC):PO-01 (∼26 nm) and then into a vacuum chamber to thermally deposit the second emitting layer 26DCzPPy:PO-01 (15 nm), the electron transport layer 1,3,5-tri(m-pyrid-3-yl-phenyl)benzene (TmPyPb, 50 nm), the electron injection layer LiF (0.5 nm) and the cathode Al (100 nm) in sequence. The first emitting layer was spin-coated with a rotation rate of 2000 rpm for 30 s and then dried at 120 °C for 30 min, where 26DCzPPy and TAPC were mixed with a weight ratio of 4[thin space (1/6-em)]:[thin space (1/6-em)]1 and a concentration of 10 mg ml−1 in chlorobenzene to act as a bipolarity co-host of the orange-yellow phosphor PO-01. The optimized doping concentration of PO-01 in 26DCzPPy:TAPC or 26DCzPPy was 4 wt%. The following inorganic/organic layers were thermally deposited in a high-vacuum system with a pressure of less than 5 × 10−4 Pa and deposition rates of 0.5–10 Å s−1 (5–10 Å s−1 for Al and 0.5–1 Å s−1 for other inorganic/organic layers). For comparison, another two groups of devices with common ITO and clean graphene anodes were also manufactured. The single-hole and single-electron devices used similar fabrication process with above OLEDs and their device structures were graphene/PEDOT:PSS (70 nm)/TAPC[thin space (1/6-em)]:[thin space (1/6-em)]26DCzPPy[thin space (1/6-em)]:[thin space (1/6-em)]PO-01 (26 nm, 4[thin space (1/6-em)]:[thin space (1/6-em)]1[thin space (1/6-em)]:[thin space (1/6-em)]0.04 for weight ratio)/26DCzPPy:4 wt% PO-01 (15 nm)/TAPC (30 or 50 nm)/Al (100 nm) (single-hole device) and graphene/LiF (0.5 nm)/TmPyPb (49 or 69 nm)/TAPC[thin space (1/6-em)]:[thin space (1/6-em)]26DCzPPy[thin space (1/6-em)]:[thin space (1/6-em)]PO-01 (26 nm, 4[thin space (1/6-em)]:[thin space (1/6-em)]1[thin space (1/6-em)]:[thin space (1/6-em)]0.04)/26DCzPPy:4 wt% PO-01 (15 nm)/TmPyPB (50 nm)/LiF (0.5 nm)/Al (100 nm) (single-electron device). Note that total organic layer thickness between two electrodes in single-hole devices was equal to that in single-electron ones and two groups of devices with total organic layer thicknesses of 141 and 161 nm were respectively fabricated.

Device characterization

The EL characteristics were measured with a PR655 spectrometer and a Keithley 2400 programmable voltage–current source. The electrical curve scans under a low driven bias were measured with a Keithley 4200 current source on Cascade probe station. The light output was improved by covering 5 mm diameter half-lens with a refractive index of 1.922 onto the glass substrates. All devices except single-hole device were directly measured at room temperature in ambient air without encapsulation. Single-hole devices were measured in liquid nitrogen.

Conflicts of interest

There are no conflicts to declare.


The authors acknowledge financial support from the National Foundation for Science and Technology Development (973 project, Grant No. 2015CB932203), the National Key Research and Development Program of China (Grant No. 2017YFB0404501), the National Natural Science Foundation of China (Grant No. 61274065, 61505086, 61705111, and 21404058), the Science Fund for Distinguished Young Scholars of Jiangsu Province of China (Grant No. BK20160039), the Natural Science Foundation of Jiangsu Province (Grant No. BM2012010 and BK20160088), the Priority Academic Program Development of Jiangsu Higher Education Institutions (Grant No. YX030002), the Jiangsu National Synergetic Innovation Center for Advanced Materials, the Synergetic Innovation Center for Organic Electronics and Information Displays, and the Open Foundation from Jilin University (Grant No. IOSKL2017KF04). We thank Dr Runfeng Chen for valuable discussion.

Notes and references

  1. J. B. Wu, M. Agrawal, H. A. Becerril, Z. N. Bao, Z. F. Liu, Y. S. Chen and P. Peumans, ACS Nano, 2010, 4, 43–48 CrossRef CAS PubMed.
  2. T. Sun, Z. L. Wang, Z. J. Shi, G. Z. Ran, W. J. Xu, Z. Y. Wang, Y. Z. Li, L. Dai and G. G. Qin, Appl. Phys. Lett., 2010, 96, 13301 CrossRef.
  3. T. H. Han, Y. Lee, M. R. Choi, S. H. Woo, S. H. Bae, B. H. Hong, J. H. Ahn and T. W. Lee, Nat. Photonics, 2012, 6, 105–110 CrossRef CAS.
  4. T. H. Han, M. H. Park, S. J. Kwon, S. H. Bae, H. K. Seo, H. Cho, J. H. Ahn and T. W. Lee, NPG Asia Mater., 2016, 8, e303 CrossRef CAS.
  5. Z. K. Liu, J. H. Li, Z. H. Sun, G. A. Tai, S. P. Lau and F. Yan, ACS Nano, 2012, 6, 810–818 CrossRef CAS PubMed.
  6. T. H. Han, S. J. Kwon, N. N. Li, H. K. Seo, W. T. Xu, K. S. Kim and T. W. Lee, Angew. Chem., Int. Ed., 2016, 55, 6197–6201 CrossRef CAS PubMed.
  7. H. J. Shin, W. M. Choi, D. Choi, G. H. Han, S. M. Yoon, H. K. Park, S. W. Kim, Y. W. Jin, S. Y. Lee, J. M. Kim, J. Y. Choi and Y. H. Lee, J. Am. Chem. Soc., 2010, 132, 15603–15609 CrossRef CAS PubMed.
  8. I. Khrapach, F. Withers, T. H. Bointon, D. K. Polyushkin, W. L. Barnes, S. Russo and M. F. Craciun, Adv. Mater., 2012, 24, 2844–2849 CrossRef CAS PubMed.
  9. D. H. Shin, J. H. Kim, J. H. Kim, C. W. Jang, S. W. Seo, H. S. Lee, S. Kim and S. H. Choi, J. Alloys Compd., 2017, 715, 291–296 CrossRef CAS.
  10. K. C. Kwon, K. S. Choi and S. Y. Kim, Adv. Funct. Mater., 2012, 22, 4724–4731 CrossRef CAS.
  11. M. H. Kang, W. I. Milne and M. T. Cole, ChemPhysChem, 2016, 17, 2545–2550 CrossRef CAS PubMed.
  12. K. C. Kwon, K. S. Choi, C. Kim and S. Y. Kim, Phys. Status Solidi A, 2014, 211, 1794–1800 CrossRef CAS.
  13. Z. Y. Chen, I. Santoso, R. Wang, L. F. Xie, H. Y. Mao, H. Huang, Y. Z. Wang, X. Y. Gao, Z. K. Chen, D. G. Ma, A. T. S. Wee and W. Chen, Appl. Phys. Lett., 2010, 96, 213104 CrossRef.
  14. A. Kuruvila, P. R. Kidambi, J. Kling, J. B. Wagner, J. Robertson, S. Hofmann and J. Meyer, J. Mater. Chem. C, 2014, 2, 6940–6945 RSC.
  15. H. Meng, J. X. Luo, W. Wang, Z. J. Shi, Q. L. Niu, L. Dai and G. G. Qin, Adv. Funct. Mater., 2013, 23, 3324–3328 CrossRef CAS.
  16. C. L. Hsu, C. T. Lin, J. H. Huang, C. W. Chu, K. H. Wei and L. J. Li, ACS Nano, 2012, 6, 5031–5039 CrossRef CAS PubMed.
  17. D. Kim, D. Lee, Y. Lee and D. Y. Jeon, Adv. Funct. Mater., 2013, 23, 5049–5055 CrossRef CAS.
  18. N. Li, S. Oida, G. S. Tulevski, S. J. Han, J. B. Hannon, D. K. Sadana and T. C. Chen, Nat. Commun., 2013, 4, 2294 Search PubMed.
  19. J. Lee, T. H. Han, M. H. Park, D. Y. Jung, J. Seo, H. K. Seo, H. Cho, E. Kim, J. Chung, S. Y. Choi, T. S. Kim, T. W. Lee and S. Yoo, Nat. Commun., 2016, 7, 11791 CrossRef CAS PubMed.
  20. J. X. Wang, J. S. Chen, X. F. Qjao, S. M. Alshehri, T. Aharnad and D. G. Ma, ACS Appl. Mater. Interfaces, 2016, 8, 10093–10097 CAS.
  21. L. P. Yu, C. Shearer and J. Shapter, Chem. Rev., 2016, 116, 13413–13453 CrossRef CAS PubMed.
  22. W. Gaynor, S. Hofmann, M. G. Christoforo, C. Sachse, S. Mehra, A. Salleo, M. D. McGehee, M. C. Gather, B. Lussem, L. Muller-Meskamp, P. Peumans and K. Leo, Adv. Mater., 2013, 25, 4006–4013 CrossRef CAS PubMed.
  23. H. Y. Xiang, Y. Q. Li, L. Zhou, H. J. Xie, C. Li, Q. D. Ou, L. S. Chen, C. S. Lee, S. T. Lee and J. X. Tang, ACS Nano, 2015, 9, 7553–7562 CrossRef CAS PubMed.
  24. K. Fehse, K. Walzer, K. Leo, W. Lovenich and A. Elschner, Adv. Mater., 2007, 19, 441–444 CrossRef CAS.
  25. Q. Dong, F. Tai, H. Lian, B. Zhao, Z. Zhong, Z. Chen, J. Tang and F. Zhu, Nanoscale, 2017, 9, 2875–2882 RSC.
  26. S. Bhaviripudi, X. T. Jia, M. S. Dresselhaus and J. Kong, Nano Lett., 2010, 10, 4128–4133 CrossRef CAS PubMed.
  27. L. P. Ma, J. Liu and Y. Yang, Appl. Phys. Lett., 2002, 80, 2997–2999 CrossRef CAS.
  28. L. P. Ma, S. Pyo, J. Ouyang, Q. F. Xu and Y. Yang, Appl. Phys. Lett., 2003, 82, 1419–1421 CrossRef CAS.
  29. L. D. Bozano, B. W. Kean, V. R. Deline, J. R. Salem and J. C. Scott, Appl. Phys. Lett., 2004, 84, 607–609 CrossRef CAS.
  30. R. J. Tseng, J. Ouyang, C. W. Chu, J. S. Huang and Y. Yang, Appl. Phys. Lett., 2006, 88, 123506 CrossRef.
  31. B. Fallahazad, K. Lee, S. Kang, J. M. Xue, S. Larentis, C. Corbet, K. Kim, H. C. P. Movva, T. Taniguchi, K. Watanabe, L. F. Register, S. K. Banerjee and E. Tutuc, Nano Lett., 2015, 15, 428–433 CrossRef CAS PubMed.
  32. A. Mishchenko, J. S. Tu, Y. Cao, R. V. Gorbachev, J. R. Wallbank, M. T. Greenaway, V. E. Morozov, S. V. Morozov, M. J. Zhu, S. L. Wong, F. Withers, C. R. Woods, Y. J. Kim, K. Watanabe, T. Taniguchi, E. E. Vdovin, O. Makarovsky, T. M. Fromhold, V. I. Fal'ko, A. K. Geim, L. Eaves and K. S. Novoselov, Nat. Nanotechnol., 2014, 9, 808–813 CrossRef CAS PubMed.
  33. U. Chandni, K. Watanabe, T. Taniguchi and J. P. Eisenstein, Nano Lett., 2016, 16, 7982–7987 CrossRef CAS PubMed.
  34. A. Nogaret, J. Appl. Polym. Sci., 2014, 131, 40169 CrossRef.
  35. S. Unarunotai, Y. Murata, C. E. Chialvo, H. S. Kim, S. MacLaren, N. Mason, I. Petrov and J. A. Rogers, Appl. Phys. Lett., 2009, 95, 202101 CrossRef.
  36. S. S. Sabri, P. L. Levesque, C. M. Aguirre, J. Guillemette, R. Martel and T. Szkopek, Appl. Phys. Lett., 2009, 95, 242104 CrossRef.
  37. H. M. Wang, Y. H. Wu, C. X. Cong, J. Z. Shang and T. Yu, ACS Nano, 2010, 4, 7221–7228 CrossRef CAS PubMed.
  38. V. Geringer, D. Subramaniam, A. K. Michel, B. Szafranek, D. Schall, A. Georgi, T. Mashoff, D. Neumaier, M. Liebmann and M. Morgenstern, Appl. Phys. Lett., 2010, 96, 082114 CrossRef.
  39. D. B. Strukov, G. S. Snider, D. R. Stewart and R. S. Williams, Nature, 2008, 453, 80–83 CrossRef CAS PubMed.
  40. Y. C. Yang, P. Sheridan and W. Lu, Appl. Phys. Lett., 2012, 100, 203112 CrossRef.
  41. J. Moser, A. Barreiro and A. Bachtold, Appl. Phys. Lett., 2007, 91, 163513 CrossRef.
  42. N. Peimyoo, T. Yu, J. Z. Shang, C. X. Cong and H. P. Yang, Carbon, 2012, 50, 201–208 CrossRef CAS.
  43. W. Ai, W. W. Zhou, Z. Z. Du, Y. P. Du, H. Zhang, X. T. Jia, L. H. Xie, M. D. Yi, T. Yu and W. Huang, J. Mater. Chem., 2012, 22, 23439–23446 RSC.
  44. W. J. Xie, L. T. Weng, K. M. Ng, C. K. Chan and C. M. Chan, Carbon, 2015, 94, 740–748 CrossRef CAS.
  45. Y. C. Lin, C. C. Lu, C. H. Yeh, C. H. Jin, K. Suenaga and P. W. Chiu, Nano Lett., 2012, 12, 414–419 CrossRef CAS PubMed.
  46. M. Regmi, M. F. Chisholm and G. Eres, Carbon, 2012, 50, 134–141 CrossRef CAS.
  47. K. Kumar, Y. S. Kim and E. H. Yang, Carbon, 2013, 65, 35–45 CrossRef CAS.
  48. J. H. Park, W. Jung, D. Cho, J. T. Seo, Y. Moon, S. H. Woo, C. Lee, C. Y. Park and J. R. Ahn, Appl. Phys. Lett., 2013, 103, 171609 CrossRef.


Electronic supplementary information (ESI) available: JV characteristics of single-hole and single-electron devices, IV curves of single-hole device with different sweep rates, AFM images, and XPS spectra. See DOI: 10.1039/c7tc05148d

This journal is © The Royal Society of Chemistry 2018