Recent Progress in Oxide Semiconductor Based Thin-Film Transistors for Next-Generation Display

Byung Ha Kang, Kyungho Park, Jong Bin An, Hyun Jae Kim*
School of Electrical and Electronic Engineering, Yonsei University, 50 Yonsei-ro,
Seodaemun-gu, Seoul 03722, Republic of Korea
* Email: hjk3@yonsei.ac.kr

Introduction

At the same time that the 3rd industrial revolution gave rise to the current 4th industrial revolution of information and communication technology, display technology rapidly developed from cathode-ray tubes (CRTs) to flat panel displays (FPDs) that use liquid crystal displays  (LCDs)  and organic light  emitting diode (OLED) displays. Recently, the realization of immersive ultra-high resolution for augmented/virtual reality (AR/VR) and flexible, foldable, stretchable, and wearable capabilities are the predominant trends in display research and development.

To fully realize ultra-high resolution and flexible display technology, further development of backplane technology will be required. Because the thin-film transistor  (TFT),  which acts as a switch to control each pixel, is a crucial component of the backplane, the characteristics, performance, quantity, process conditions, and substrate have a direct effect on display resolution and flexibility. Generally, there are three types of TFTs applied to the display backplane, and they can be classified based on the active layer of material.  The main TFT types are amorphous silicon (a-Si) based TFTs, low temperature polycrystalline silicon  (LTPS)  based  TFTs,  and oxide semiconductor based TFTs. So far, a-Si-based TFT  and LTPS based TFT have mainly been used for display backplanes. The relatively low (0.5–1 cm2/Vs) field effect mobility of a-Si based TFT has led to the introduction of LTPS (a crystallized a-Si fabricated with an excimer laser. This improves mobility to 10–200 cm1/Vs, making high-resolution possible. While an improvement, these TFTs are still difficult to apply to flexible displays. Flexible displays require a plastic substrate with a low glass transition temperature due to the difficulty in controlling grain boundaries, low large-area uniformity, and high process temperature. Oxide semiconductor based TFTs such as indium gallium zinc oxide (IGZO) were developed in 2004, offering a potential solution to the limitations. The IGZO TFT has advantages such as high mobility (~10 cm2/Vs) despite its amorphous state, high uniformity over a large area, low leakage current (low power consumption),  and high transparency  (> 80%)  in the visible light region.1–3  Consequently,  IGZO  TFTs were used in commercialized LCD backplanes in 2012 and OLED display backplanes in 2013.4

In line with the development of OLED displays as a mainstream component of FPDs, oxide semiconductor based TFTs have also evolved to become an essential element of display backplane technology. Furthermore, they are expected to become the basis for micro-light emitting diode (μ-LED) displays, quantum nano emitting diode (QNED) displays, and for many next-generation displays, such as immersive, free-form, transparent, and  ultra- large displays. Various studies have been conducted to develop ultra-high resolution and flexible displays based on oxide semiconductor TFTs during the last several years. Therefore, we will review recent studies for achieving flexible oxide based TFTs, including both high mobility for ultra-high resolution and low- temperature process. Each subject will be divided into vacuum-processed and solution-processed TFTs.
 

Research on High-Performance Oxide Semiconductor Based TFT for Immersive Display

As previously mentioned, the display industry is gradually improving the resolution and frame speed of displays to better support immersive content. To satisfy the performance requirements of a display with a resolution of Super Hi-Vision (7,680 × 4,320) and a frame rate of 240 Hz, TFTs with a minimum mobility of 30 cm2/Vs or higher are required.5–6 However, the mobility of commercialized IGZO TFTs is around 10 cm2/Vs, which limits the implementation of high-performance
displays.7–8 Currently, various studies are underway to improve the mobility of oxide based TFTs and overcome this limitation.

In general, oxygen vacancies in the oxide semiconductor supply carriers that enable effective charge transport. To improve the electrical characteristics of TFT, the concentration of these vacancies must be optimized. In 2016, Wang et al. adjusted the deposition process parameters of tin-doped zinc oxide (TZO) to control the oxygen vacancy concentration, resulting in improved mobility. This method was conducted by adjusting the oxygen partial pressure of the plasma to  0%,  10%,  20%,  and  30% during the sputtering process. A TZO TFT fabricated under 0% oxygen partial pressure produced excess carriers due to a number of oxygen vacancies in the active layer. However, by gradually increasing the oxygen partial pressure, they showed improved transfer characteristics with an on/off ratio of more than 107 and a subthreshold swing (SS) of about 0.3 V/dec on average. Among them, a TZO TFT fabricated at 10%  oxygen partial pressure showed a low off current of 3 pA, an on/off ratio of 2×107, SS of 0.33 V/dec, and high mobility of 66.7 cm2/ Vs. As the oxygen partial pressure increased,  the interface-state density was reduced, and the electrical properties were improved due to suppression of excess carriers and a decrease in the surface roughness.9

In a study to control oxygen vacancy, a method to control cation doping and deposition process parameters was used. A typical method for controlling properties through cation doping uses a material with high standard electrode potential (SEP) compared to metal ions in an oxide semiconductor. Candidate groups with high SEP include Mg, Sr, Y, Zr, Ba, Hf, and
Nd.10–15 In 2015, Peng et al. reported improved electrical characteristics by doping neodymium, which has a high standard electrode potential, into indium oxide (In2O3) where excess carriers exist. Nd is a material with a strong binding force with oxygen (703 kJ/mol) and has the advantage of effectively controlling oxygen vacancies in the active layer. The formed indium neodymium oxide (InNdO) TFT showed a mobility of 20.4 cm2/Vs under the annealing condition of 450 °C.

The production of solution-processed oxide semiconductor thin films is divided into nanoparticle based methods and metal salt precursor based methods. Nanoparticle based  thin  films can be formed at low temperature and have excellent device performance, but are not primarily used as display backplanes due to their non-uniformity and low reliability. Therefore, in this review, we focus on the solution-processed oxide semiconductor TFT fabricated using the metal salt precursor based method. However, the metal salt precursor solution-processed oxide semiconductor based TFT must overcome low mobility compared to the vacuum-processed oxide semiconductor based TFTs currently used for commercial display manufacturing.

There are a wide variety of proposed technologies that can be used to overcome the low mobility of solution-processed oxide semiconductor TFTs. The first method is to introduce specific materials as supply carriers to the active layer of TFT to improve mobility.16 One example is the introduction of potassium superoxide (KO2) (Cat. No. 278904) to improve the mobility of zinc tin oxide  (ZTO)  TFT  based in a  solution process.  As shown in Figure 1A, KO2 is divided into positive and negative ions in the solution. Potassium cations increase the electron concentration by supplying electrons to the  ZTO  thin film,  and the anions cause a decrease in oxygen vacancies.  As a  result, the mobility of ZTO TFT treated with KO2 increased from 5.57 cm2/Vs to 8.74 cm2/Vs, as shown in Figure 1B.

 

Figure 1. A) Schematic diagram of the manufacturing process of TiO2:IZO TFT. Comparison of transfer characteristics between IZO TFTand TiO2:IZO TFT with annealing temperature of B) 280 °C and C) 230 °C.
 

Similarly, Figure 2 shows a study to improve the mobility of solution-processed IZO TFT by adding titanium dioxide (TiO2) (Cat. No. 1667585).17 As shown in Figure 2A, when TiO2 is added to the IZO solution and then irradiated with ultraviolet light, hydroxyl radicals are formed through a photocatalytic reaction with water accelerating the decomposition of organic compounds with  large  hydroxyl  radicals  and  forming  smaller organic molecules.  The  decomposed  small  organic  molecules  have a low mass and boiling point, reducing defects in the active oxide layer. The IZO TFT with reduced defects showed an increase  in mobility from 2.78 cm2/Vs to 7.82 cm2/Vs, as shown in Figure 2B.

 

Figure 2. A) Schematic diagram of the manufacturing process of TiO2:IZO TFT. Comparison of transfer characteristics between IZO TFT and TiO2:IZO TFT with annealing temperature of B) 280 °C and C) 230 °C.

 

Unlike the two methods introduced above, which add specific materials, the next method improves mobility through structural engineering. Figure 3 represents a solution-processed oxide semiconductor TFT with high mobility and excellent reliability based on the corrugated heterojunction oxide active layer structure.18 To control the electron concentration accumulated through charge modulation in the vertical region of the heterojunction, a corrugated active layer structure was formed by alternately using thin and thick ITZO/ IGZO. This optimized heterojunction ITZO/IGZO TFT showed a high mobility of 50 cm2/Vs.

 

Figure 3. A) Schematic diagram of corrugated heterojunction oxide semiconductor TFT. B) Energy band of ITZO/IGZO active layer by ITZO thickness. C) TCAD simulation result for the current density of the off-state TFT. Reprinted with permission from reference 18, copyright 2018 Wiley.

back to top

Research on flexible oxide semiconductor based TFTs for free-form display

In order to make a flexible, free-form display a reality, the backplane must be fabricated on a flexible substrate such as plastic rather than conventional glass. Representative flexible substrates include polyethylene terephthalate (PET) (Cat. No. GF14511215), polyethylene naphthalate (PEN) (Cat. No. GF23662043), polyethersulfone (PES) (Cat. No. GF98128109), polyimide (PI) (Cat. No. GF66798589), with glass transition temperatures of about 80 °C, 120 °C, 220 °C, and 360 °C, respectively. Although the process temperatures for a  device must be lower than the glass transition temperature of the substrate are typically processed at temperatures above 300 °C. Thus achieving a flexible, free-form display remains difficult due to the limitations and high cost of flexible plastic substrates, able to endure high process temperatures. As a result, it will be necessary to develop methods that obtain outstanding electrical characteristics and reliability using lower process temperature processed TFTs. This section introduces technologies that have the achieved process temperature below 250 °C through material, process, and structural engineering.

As more research to improve the flexibility of oxide TFTs based on the vacuum process have been reported, many of these methods of lowering the process temperature have included applying the flexible materials to the active layer and structure modification. For example, Kim et al. developed an open-air plasma treatment (OPT), which allows for dramatically decreased process temperature.19 OPT not only decomposes weak metal-oxygen bonds by collision of ions included in ultraviolet (UV) and plasma, but also creates strong metal-oxygen bonds through the reaction of radicals (NO•, O•, OH•) and thermal treatment by plasma. As a result, IGZO TFTs fabricated from OPT for 3 s at a low temperature below 300 °C showed electrical performances similar to conventional IGZO TFTs, allowing fabrication on a PI substrate, as shown in Figure 4B and Figure 4C.

Researchers have reported new approaches to fabricate flexible oxide TFTs in addition to the  low  temperature process.  Kim et al. demonstrated hybrid oxide TFTs with excellent mechanical characteristics by doping a  polytetrafluoroethylene  (PTFE)  with the Young’s modulus of 0.3 GPa on IGZO, as shown in Figure 4D and Figure 4E.20 The Young’s modulus of the IGZO:PTFE formed by plasma polymerization was 85  GPa,  less than the  IGZO Young’s modulus of 140 GPa. Additionally, since the fluorine (F) included in IGZO:PTFE has a strong electronegativity, it exhibits hydrophobicity by minimizing the number of coordination on the surface. As a result, the stability improved in ambient atmosphere compared to conventional TFTs.

Recently, Jang et al. IGZO TFTs fabricated on flexible plastic substrates applied on an etch-stopper (ES) - split active layer (SP) structure to improve the mechanical characteristics, as shown in Figure 4F.21 In this study, nitrogen trifluoride (NF3)/ hydrogen (H2) plasma dry etching was performed during the formation of ES-SP IGZO TFTs. At this time, the F atoms contained in the plasma remained in the ES and then flowed into both the ES-IGZO interface and bulk during the annealing process. Due to the previously mentioned characteristics of the F atom, it showed superior mechanical characteristics after bending up to 5,000 times at a radius of 1 mm by reducing the unsaturated bonds at the ES and IGZO interfaces as well as oxygen vacancies and hydrogen bonding in the IGZO, as shown in Figure 4G and Figure 4H.

 

Figure 4. A) Schematic of OPT system. B) Schematic and C) the transfer characteristic of the IGZO TFT on the PI substrate. D) Schematic of theIGZO:PTFE TFT fabricated by co-sputtering. E) Photograph, the cross-sectional HR-TEM image, and diffraction patterns of IGZO:PTFE TFTs on the PIsubstrate. F) Structure of the ES-SP IGZO TFT. G) Transfer characteristics of ES-SP IGZO TFTs with various radius. H) Photograph of flexible IGZO TFTarrays. Reprinted with permission from reference 21, copyright 2017 Wiley.


Unlike vacuum-processed oxide semiconductors, solution-processed oxide semiconductors are formed through decomposition, hydrolysis, dihydroxylation, and condensation reactions, requiring annealing temperature above 400 °C.22 Due to the higher annealing temperature compared to vacuum-processed oxide TFTs, research on solution-processed oxide TFTs has been focused on low annealing temperature oxide semiconductors.

Next, we will review the materials engineering of solution-processed oxide TFTs at low processing temperatures. Previously, several methods of reducing the formation temperature of oxide semiconductors have been proposed using different precursors. Among these, oxide TFTs fabricated by metal alkoxide precursors have been reported to have excellent electrical characteristics, including field effect mobility studies have shown that the processing temperatures can be reduced by doping precursors with various metal and non-metal materials.24–27 For example, when fabricating solution-processed IZO TFTs, precursors doped with the F atom show field-effect mobility of 4.1 cm2/Vs at a processing temperature of 200 °C. The F atom fills oxygen vacancies in oxide films, increasing the carrier concentration and strengthening bonding states of oxide films even at low processing temperatures.28

Unlike doping of inorganic materials in solution, the flexibility of the solution-processed oxide TFTs can be improved by doping with organic materials. As shown in Figure 5A and Figure 5B, In2O3 TFTs achieved a field effect mobility of above 10 cm2/Vs at a processing temperature of 225oC by doping a poly (4-vinyl phenol) (PVP) (Cat. No. 436224), a flexible insulating polymer.29 PVP has two advantages: i) high solubility in the In2O3solution, ii) enhanced flexibility through the hydroxyl group in the PVP, which can easily bind to the metal-oxygen lattice. Therefore, there is almost no change in the field effect mobility even after bending 100 times at a radius of 10 mm.

In addition to the already mentioned studies, other reports note that an additive capable of combustion can be added to the solution to cause an exothermic reaction on its own.30–32As shown in Figure 5C and Figure 5D, when an organic fuel(acetylacetone (Cat. No. 10916) or urea (Cat. No. U5128)and an oxidizer (in the form of a metal nitrate salt) (Cat. No. L060050) was mixed, the oxidation reaction reduced the energy barrier height to form an oxide thin film.30 Different types of additives display different degrees of combustion required to reduce the energy barrier height, so the effect of lowering the processing temperature was also different, as shown in Figure 5E. In Figure 5F, the electrical characteristics of IGZO TFTs fabricated at the same processing temperature of 300 °C tend to vary depending on the additives, with a field effect mobility of 5.7 cm2/Vs under optimized conditions.31 In this method, since the amount of emitted heat, gases, and impurities depends on the proportion of additives, optimization through appropriate trade-off was emphasized.

Figure 5. A) Schematic of solution-processed PVP doped In2O3 TFTs. B) Dependence of the field effect mobility on bending radius of both conventional In2O3 and PVP doped In2O3 TFTs. Reprinted with permission from reference 29, copyright 2015 Wiley. C) Schematic of reaction coordinate comparing the energy barrier height for combustion and conventional solution process. D) Example of combustion reaction usingan organic fuel and an oxidizer. Reprinted with permission from reference 30, copyright 2013 American Chemical Society. E) Comparing the energy barrier height according to additives. F) Transfer characteristics of IGZOTFTs depending on additives. Reprinted with permission from reference 31, copyright 2018 American Chemical Society.

 

So far, we have introduced methods for precursor engineering of the dissolved materials in solution. However, most of the solvents that are used in oxide semiconductor solutions are based on organic alcohol species and contain many impurities. Therefore, recent research investigated a carbon-free water solvent to enable lower process temperatures.33–34 Figure 6A and Figure 6B show the results of a study in which water solvent based ZnO solution was annealed at 180 °C to fabricate TFTs on flexible PEN substrates, achieving excellent mobility of 11 cm2/Vs.33 This is because the metal-amine of the aminehydroxyl readily decomposes using a water-based solvent and carrying out dehydration and condensation reactions at relatively low energy. Figure 6C and Figure 6D show theresults of a study in which water solvent based In2O3 TFTs were annealed at 100 °C and fabricated on PEN substrates, showing the mobility of 3.14 cm2/Vs.34 Using the thermo gravimetric analysis (TGA) measurement, researchers confirmed that thermal decomposition occurred at a lower temperature than the alcohol based solvent and that by-products were effectively removed through a sharp decrease in mass.

Figure 6. A) Image and B) transfer characteristics of solution-processed flexible ZnO TFT. Reprinted with permission from reference 33, copyright 2013 Wiley. C) Comparison of the TGA measurement on water and alcohol based In2O3 solution and D) image of the water solvent based flexible In2O3 TFT. Reprinted with permission from reference 34, copyright 2013 Nature Publishing Group.


Lastly, in terms of process, low temperature solution-processed oxide semiconductor based TFT studies can be classified according to the type of energy source  that  supplies  the activation energy required to form an oxide thin film. In using conventional heat treatment, high temperature is applied to the flexible substrate, which may cause damage to the substrate. Other forms of energy that can be used include optical, chemical, and physical energy sources. Photoactivation, plasma treatment, microwave treatment, and high-pressure annealing have most frequently been reported as representative energy sources for low temperature annealed solution-processed oxide semiconductor TFTs. Among them, photo-energy treatment will be discussed here. The photo-energy treatment process can support a wide range of process windows by controlling variables such as wavelength, intensity,  time,  and irradiation method (lamp or laser).

Photo-energy causes vibrating atoms and molecules to break or rearrange their bonds. A study on room-temperature fabricated IGZO TFTs using a  deep ultraviolet  (DUV)  photochemical activation process initiated active research on the annealing of solution-processed oxide thin films using various wavelengths of ultraviolet light.35 As the  DUV  was irradiated, alkoxy groups in the oxide thin film were removed, and the quality of the thin film was improved by a more robust metal-oxygen bond. As a result, it showed a mobility of 7 cm2/Vs.

Another study reported solution-processed oxide thin-film formation using instantaneously irradiating light energy from a Xenon flash lamp, as shown in Figure 7A.36 The ZnO thin film was annealed at 90 °C and irradiated with pulse light of a wavelength of 350–950 nm. Light irradiation in the form of a lamp is advantageous for large-area processes and improves yield since the post-treatment is completed in 15 s. The thin film forms rapidly, and the electrical performance was almost similar to that of the ZnO TFT annealed at 165C for 40 min, meaning that the metal-oxygen bonds were firmly formed in 15 s. In Figure 7B, a solution-processed IGZO TFT with a mobility of 7.7 cm2/Vs was fabricated by simultaneously applying near infrared (NIR) and DUV light energy at room temperature and sequentially applying a flashlight.37 In detail, NIR and DUV light caused decomposition, hydrolysis, and dehydration reactions in the oxide thin film, and the number of robust metal-oxygen bonds increased with irradiation by the flashlight. In addition to the lamp-type photo-energy treatment, laser type treatment, which can apply energy to only selective areas accurately and quickly, has also been developed. An In2O3 TFTannealed at 150 °C was fabricated using a krypton fluorine (KrF)excimer laser; the region that received laser beam irradiation selectively showed semiconducting properties. The unexposed region showed insulating properties.38 The TFTs fabricated by this technology showed higher mobility (13 cm2/Vs) thanthat of In2O3 TFT annealed at 250 °C. 

Figure 7C represents the schematic process of fabricating an IGZO TFT by inkjet deposition, post-treatment with 200 °C annealing, and selectiveNIR laser spike irradiation with a wavelength of 1,064 nm. The transfer characteristics are shown in Figure 7D by number of laser irradiations.39 Additionally, the laser-irradiated solutionprocessedIGZO TFT mobility was 1.5 cm2/Vs, demonstrating that the performance improves compared to the heat-only treatment process.

 

Figure 7. A) Schematic process of photo-energy treatment using Xenon flash light lamp. Reprinted with permission from reference 36, copyright 2015 Springer. B) Sequential process of NIR, DUV, and flash lamp photo-energy treatment. Reprinted with permission from reference 37,copyright 2019 American Chemical Society. C) Structure and D) transfer characteristics of IGZO TFT fabricated by inkjet process and NIR laser spike treatment. Reprinted with permission from reference 39, copyright 2017 Springer.

back to top

Conclusion

In summary, vacuum-processed oxide semiconductor TFTs will require further development to achieve sufficient mobility to become viable for use in next-generation immersive, flexible, and free-form displays. In particular, TFTs will need to achieve a mobility of at least 30 cm2/Vs or more as well as mechanical reliability when fabricated on flexible substrates for widespread adoption. Beyond the display backplane, vigorous research and development on vacuum-processed TFTs must be conducted for commercialization in various electronic devices such as integrated logic circuits, memory devices, bio-sensors,photo-sensors, and gas-sensors. On the other hand, further development of solution-processed oxide semiconductor TFTs will also be needed to match the electrical performance and stability vacuum-processed oxide TFTs. Also, low temperature fabrication processes that utilize additional energy sources will need to be developed. Although many challenges still exist, the cost and environmental advantages of solution-processed oxide TFTscompared to the vacuum-processed TFTs are expected to drive their widespread introduction and adoption of printing processes.

 

     

    Reference

  1. Klasens, H. A.; Koelmans, H. Solid-State Electron. 1964, 7, 701. 
  2. Nomura, K.; Ohta, H.; Takagi, A.; Kamiya, T.;  Hirano, M.; Hosono, H. Nature 2004, 432, 488.
  3. Kim, W.-G.; Tak, Y. J.; Kim, H. J. J. Inf. Disp. 2018, 19, 39.
  4. Park, J. W.; Kang, B. H.; Kim, H. J. Adv. Funct. Mater. 2020, 30, 1904632.
  5. Kamiya, T.; Nomura, K.; Hosono, H. Sci. Technol. Adv. Mater. 2010, 11, 044305.
  6. Kim, H. J.; Park, K.; Kim, H. J. J. Soc. Inf. Disp. 2020, 28, 591.
  7. Lee, H.; Chang, K. S.; Tak, Y. J.; Jung, T. S.; Park, J. W.; Kim, W.-G.; Chung, J.; Jeong, C. B.; Kim, H. J. J. Inf. Disp. 2017, 18, 131.
  8. Bebiche, S.; Wang, S.; Lei, L.; Wong, M.; Kwok, H.-S. SID Symposium Digest of Technical Papers 2018, 1587.
  9. Han, D.; Zhang, Y.; Cong, Y.; Yu, W.; Zhang, X.; Wang, Y. Sci Rep 2016, 6, 38984.
  10. Lim, H. S.; Rim, Y. S.; Kim, D. L.; Jeong, W. H.; Kim, H. J. Electrochem. Solid State Lett. 2012, 15, H78.
  11. Banger, K. K.; Peterson, R. L.; Mori, K.; Yamashita, Y.; Leedham, T.; Sirringhaus, H. Chem. Mat. 2014, 26, 1195.
  12. Jun, T.; Song, K.; Jung, Y.; Jeong, S.; Moon, J. J. Mater. Chem. 201121, 13524.
  13. Park, J.-S.; Kim, K.; Park, Y.-G.; Mo, Y.-G.; Kim, H. D.; Jeong, J. K. Adv. Mater. 2009, 21, 329.
  14. Hong, S.; Park, S. P.; Kim, Y.-G.; Kang, B. H.; Na, J. W.; Kim, H. J. Sci. Rep. 2017, 7, 16265.
  15. Lin, Z.; Lan, L.; Xiao, P.; Sun, S.; Li, Y.; Song, W.; Gao, P.; Ning, L. H.; Peng, J. Appl. Phys. Lett. 2015, 107, 112108.
  16. Jung, T. S.; Lee, H.; Kim, H. J.; Lee, J. H.; Kim, H. J. ACS Appl. Mater. Interfaces 2018, 10, 44554.
  17. Kang, J. K.; Park, S. P.; Na, J. W.; Lee, J. H.; Kim, D.; Kim, H. J. ACS Appl. Mater. Interfaces 2018, 10, 18837.
  18. Lee, M.; Jo, J. W.; Kim, Y. J.; Choi, S.; Kwon, S. M.; Jeon, S. P.; Facchetti, A.; Kim, Y. H.; Park, S. K. Adv. Mater. 2018, 30, 1804120. 
  19. Tak, Y. J.; Hilt, F.; Keene, S.; Kim, W.-G.; Dauskardt, R. H.; Salleo, A.; Kim, H. J. ACS Appl. Mater. Interfaces 2018, 10, 37223.
  20. Na, J. W.; Kim, H. J.; Hong, S.; Kim, H. J. ACS Appl. Mater. 2018, 10, 37207.
  21. Lee, S.; Shin, J.; Jang, J. Adv. Funct. Mater. 2017, 27, 1604921.
  22. Kim, G. H.; Shin, H. S.; Ahn, B. D.; Kim, K. H.; Park, W. J.; Kim, H. J. J. Electrochem. Soc. 2009, 156, H7.
  23. Banger, K.; Yamashita, Y.; Mori, K.; Peterson, R.; Leedham, T.; Rickard, J.; Sirringhaus, H. Nat. Mater. 2011, 10, 45.
  24. Park, S. Y.; Kim, B. J.; Kim, K.; Kang, M. S.; Lim, K. H.; Lee, T. I.; Myoung, J. M.; Baik, H. K.; Cho, J. H.; Kim, Y. S. Adv. Mater. 2012, 24, 834.
  25. Kim, K.; Park, S. Y.; Lim, K.-H.; Shin, C.; Myoung, J.-M.; Kim, Y. S. J. Mater. Chem. 2012, 22, 23120.
  26. Banger, K. K.; Peterson, R. L.; Mori, K.; Yamashita, Y.; Leedham, T.; Sirringhaus, H.; Chem. Mater. 2014, 26, 1195.
  27. Han, S.-Y.; Nguyen, M.-C.; Nguyen, A. H. T.; Choi, J.-W.; Kim, J.-Y.; Choi, R. Thin Solid Films 2017, 641, 19.
  28. Seo, J.-S.; Jeon, J.-H.; Hwang, Y. H.; Park, H.; Ryu, M.; Park, S.-H. K.; Bae, B.-S. Sci. Rep. 2013, 3, 2085.
  29. Yu, X.; Zeng, L.; Zhou, N.; Guo, P.; Shi, F.; Buchholz, D. B.; Ma, Q.; Yu, J.; Dravid, V. P.; Chang, R. P. Adv. Mater. 2015, 27, 2390.
  30. Hennek, J. W.; Smith, J.; Yan, A.; Kim, M.-G.; Zhao, W.; Dravid, V. P.; Facchetti, A.; Marks, T. J. J. Am. Chem. Soc. 2013, 135, 10729.
  31. Chen, Y.; Wang, B.; Huang, W.; Zhang, X.; Wang, G.; Leonardi, M. J.; Huang, Y.; Lu, Z.; Marks, T. J.; Facchetti, A. Chem. Mater. 2018, 30, 3323.
  32. Li, J.; Zhou, Y.-H.; Zhu, W.-Q.; Zhang, J.-H.; Zhang, Z.-L. Mater. Sci. Semicond. Process. 2019, 93, 201.
  33. Lin, Y. H.; Faber, H.; Zhao, K.; Wang, Q.; Amassian, A.; McLachlan, M.; Anthopoulos, T. D. Adv. Mater. 2013, 25, 4340.
  34. Hwang, Y. H.; Seo, J.-S.; Yun, J. M.; Park, H.; Yang, S.; Park, S.-H. K.; Bae, B.-S. NPG Asia Mater. 2013, 5, e45.
  35. Park, S.; Kim, K. H.; Jo, J. W.; Sung, S.; Kim, K. T.; Lee, W. J.; Kim, J.; Kim, H. J.; Yi, G. R.; Kim, Y. H. Adv. Funct. Mater. 2015, 25, 2807.
  36. Kim, D. W.; Park, J.; Hwang, J.; Kim, H. D.; Ryu, J. H.; Lee, K. B.; Baek, K. H.; Do, L.-M.; Choi, J. S. Electron. Mater. Lett. 2015, 11, 82. 
  37. Moon, C.-J.; Kim, H.-S. ACS Appl. Mater. Interfaces 2019, 11, 13380. 
  38. Dellis, S.; Isakov, I.; Kalfagiannis, N.; Tetzner, K.; Anthopoulos, T. D.; Koutsogeorgis, D. C. J. Mater. Chem. C 2017, 5, 3673.
  39. Huang, H.; Hu, H.; Zhu, J.; Guo, T. J. Electron. Mater. 2017, 46, 4497.