Molecular-switch-embedded organic Schottky barrier transistors for a high switching ratio

Abstract

Research on organic semiconductors has increasingly focused on developing multifunctional devices, leveraging their inherent advantages such as lightweight properties, low-cost fabrication, and tunable optoelectronic characteristics. Within this context, the exploration of molecular switches, particularly diarylethenes (DAEs), for precise current modulation in transistors has garnered significant interest. While the optimal molecular switch design has been extensively studied, advancements in transistor architecture have remained limited. This work introduces a novel approach utilizing organic Schottky barrier transistors (OSBTs), a type of organic vertical transistor featuring a distinct operation mechanism in terms of conductive channel formation and charge injection, enabling superior hole-trapping efficiency compared with conventional organic field-effect transistors (OFETs). By incorporating a dielectric/metal/dielectric transparent electrode to mitigate light-irradiation limitations, we successfully integrated DAEs into OSBTs, achieving a record-high photoprogrammable switching ratio exceeding 6.4 × 10⁴ at a 30 wt% DAE concentration. The physics underlying the superior performance of OSBTs compared to that of OFETs is explained, with a focus on the distinct gate-field effect. Furthermore, stable switching performance was maintained over 100 repeated cycles, demonstrating exceptional fatigue resistance. This innovative architecture paves the way for the development of high-performance photoprogrammable transistors.

---

New concepts

This work overcomes the limitations of previous molecular-switch-embedded organic transistors, highlighting how structural innovations in device architecture—specifically the vertical layout—can dramatically improve the efficiency of photoprogrammable organic transistors. Unlike conventional lateral organic field-effect transistors (OFETs), a vertical organic Schottky barrier transistor (OSBT) allows DAE molecules dispersed throughout the entire active layer to actively participate in charge trapping and to influence the Schottky barrier. This results in a significant increase in the switching ratio, surpassing 64 000 and allows the device to be completely turned off. The integration of a transparent dielectric/metal/dielectric electrode mitigates light obstruction, ensuring effective photoisomerization.

Introduction

Organic materials have gained significant attention in recent research due to their lightweight properties, cost-effectiveness, tunable electrical and optical characteristics via chemical structure modifications, and ability to be fabricated on flexible substrates using low-temperature processes.^1–6 Recent trends have shifted from merely enhancing charge mobility and optimizing device structures to the development of multifunctional devices with broader capabilities.^7–10 In this context, molecular switches controlled by light stimuli have emerged as one of the simplest and most attractive approaches.^11–15 Among them, diarylethene (DAE) stands out as a prominent example of a photochromic molecule, exhibiting binary isomerization between two distinct states: the open-ring isomer (DAE_o) and the closed-ring isomer (DAE_c). These isomers differ in conjugation length and exhibit excellent fatigue resistance and thermal stability in both states.^16–22 When exposed to specific wavelengths of light, DAEs can reversibly modify their electronic properties, including shifts in the highest occupied molecular orbital (HOMO) and lowest unoccupied molecular orbital (LUMO) energy levels. Consequently, when incorporated into a semiconductor with compatible energy levels, DAEs can function as trap states that can be regulated by non-destructive light. This functionality makes them particularly suitable for applications in organic field-effect transistors (OFETs), enabling photoprogrammable current modulation.

In photoprogrammable OFETs, two key performance parameters are fatigue resistance, which represents the number of operational cycles before failure, and the switching ratio, defined as I_DS (DAE_o)/I_DS (DAE_c), which quantifies the degree of current modulation in the on-state of the transistor according to the isomerization of DAE. Since the initial development of these devices, extensive research has focused on enhancing these parameters, primarily through the optimization of DAE chemical structures and the development of polymers suitable for DAE switching. Given that the active layer, comprising a mixture of the semiconductor and molecular switch, requires direct irradiation, research has predominantly concentrated on bottom-gate, bottom-contact (BGBC) structures that lack an upper electrode. Consequently, alternative device architectures have remained unexplored.

Against this backdrop, we propose the first investigation into the structural modification of photoprogrammable organic transistors using a vertical structure. Organic vertical transistors can readily achieve channel lengths on the order of tens of nanometers, as the active layer thickness defines the channel length, potentially overcoming the charge mobility limitations of organic semiconductors.^23–26 Among various organic vertical transistor structures, organic Schottky barrier transistors (OSBTs) are particularly suitable for integrating DAEs due to two key characteristics: (1) in OSBTs, charge transport occurs throughout the entire bulk of the active layer.^27–29 As a result, all molecular switches embedded in the semiconductor film participate in trapping, maximizing current modulation induced by the molecular switches. In contrast, in lateral OFETs, the conductive channel induced by the gate field is confined to a few nanometers at the dielectric/semiconductor interface,^30,31 meaning that only a very small fraction of the molecular switches contribute to current modulation. (2) Unlike in OFETs, charge accumulation at the dielectric/semiconductor interface in OSBTs plays a crucial role in lowering the Schottky barrier at the source/semiconductor interface, which is also significantly influenced by DAEs, thereby further boosting the photoprogrammable switching ratio.

In this study, we demonstrate an unprecedentedly high photoprogrammable switching ratio by integrating DAEs into OSBTs utilizing Ag nanowires (AgNWs) as the source electrode. We employ poly[2,5-(2-octyldodecyl)-3,6-diketopyrrolopyrrole-alt-5,5-(2,5-di(thien-2-yl)thieno[3,2-b]thiophene)] (DPP-DTT) as the semiconductor, given its proven compatibility with both OSBTs and photoprogrammable OFETs.^29,32 At the interface with AgNWs, DPP-DTT forms a Schottky barrier with a barrier height of approximately 0.75 eV, which is essential for OSBT operation, and the suitability of its energy levels has been verified in our previous study.²⁹ Unlike conventional molecular switch-based OFETs adopting BGBC geometry—where the active layer is exposed and light irradiation is not hindered—the OSBT architecture inherently suffers from light obstruction by the opaque top electrode. To resolve this issue, we introduced a dielectric/metal/dielectric (DMD) transparent electrode (MoO₃/Ag/MoO₃) as the drain, which allows photons to penetrate the electrode and reach the embedded molecular switches to induce their photoisomerization, while maintaining low series resistance. Additionally, we utilize 1,2-bis(4-metyl-2-propyl-5-(4-hexylphenyl)-3-thienyl)cyclopentene (DAE-316), which demonstrated the highest switching ratio when combined with DPP-DTT in previous research.³³ As the first study incorporating DAEs into vertical transistors, we conduct a systematic investigation of key parameters influencing the switching ratio, including channel length, molecular switch concentration, the role of the DMD transparent electrode, and irradiation time requirements for isomerization.

Results and discussion

Mechanism of hole trapping in the polymer/DAE film

Fig. 1a presents the chemical structure of DAE isomers and DPP-DTT, illustrating the hole-trapping and release mechanism that enables photoprogrammable switching within the polymer/DAE system, thereby regulating the transistor's current (I_D) level. To optimize performance, we selected DPP-DTT and DAE-316, as this combination demonstrated the highest switching ratio in our previous research on OFETs.³³ DPP-DTT has a suitable HOMO level of −5.2 eV, positioned between the HOMO levels of DAE_o and DAE_c. In its open-ring form, DAE exhibits a deeper HOMO level of −5.6 eV, preventing efficient hole migration from the DPP-DTT matrix to DAE due to the energy barrier. Conversely, in its closed-ring form, DAE allows hole trapping due to its shallower HOMO level of −4.8 eV. This hole-trapping behavior in the closed form reduces hole transport through the DPP-DTT matrix, ultimately leading to a decrease in the I_D level.³³

Fig. 1 (a) Schematic illustration for the hole-trapping mechanism of DAE-316, including the chemical structures of DPP-DTT and DAE-316 and their HOMO levels. (b) UV-vis absorption spectra of the DPP-DTT/DAE-316 film. (c) Device geometry of the photoprogrammable OSBT. (d) Transmittance characteristics of a DMD electrode.

To investigate the photochromic properties of DAE, UV-vis absorption spectra were obtained for both pure DAE and polymer/DAE films. As shown in Fig. S1, the open-form DAE initially exhibited a broad absorption peak centered around 286 nm. Upon UV light irradiation at 312 nm, DAE transitioned from the open form to the closed form, causing the initial peak to disappear while new absorption peaks emerged at 367 nm and 543 nm. This spectral shift became more pronounced with increased UV irradiation time, reaching the photostationary state (PSS) of the closed form after 30 s of irradiation. Similarly, Fig. 1b displays the UV-vis absorption spectra of the DPP-DTT/DAE film containing 30 wt% DAE, which was identified as the optimal concentration in our experiment. The DPP-DTT/DAE film exhibited the same absorption changes upon UV irradiation, confirming the successful photoisomerization of DAE within the DPP-DTT/DAE system. More detailed photophysical properties of DAE-316 are available in our previous work.³³

Fig. 1c illustrates a schematic representation of the OSBT device architecture, while an optical microscope (OM) image of the device configuration is provided in Fig. S2. OSBTs function as vertical transistors, where an active layer is positioned between the two electrodes, forming a diode-like structure on a bottom-gate dielectric layer. Charge injection in OSBTs is modulated by adjusting the Schottky barrier between the source electrode and the semiconductor. The gate electric field, transmitted through the porous source electrode, enables charge injection, after which the charge travels through the polymer bulk to the drain electrode. The thickness of the active layer determines the channel length.

For effective photoisomerization of DAE, both UV and visible light must penetrate the active layer. However, since the active layer in conventional OSBTs is sandwiched between electrodes, light irradiation is obstructed. To address this challenge, a DMD structure was introduced as the drain electrode. Fig. 1d shows the transmittance characteristics of the DMD electrode. The transmittance at 520 nm, which facilitates the transition of DAE from the closed form to the open form, was relatively high at 74%. Meanwhile, the transmittance at 312 nm, which induces the transition from the open form to the closed form, was lower at 33%. Despite this limitation, it does not pose a significant concern, as the time required to reach the PSS under UV light irradiation is relatively short. Thus, extending the irradiation duration can effectively compensate for the lower transmittance, ensuring optimal device performance.

Crystalline ordering of DPP-DTT as a function of DAE concentration

The crystalline ordering of the polymer is crucial not only for the electrical properties of the transistor but also for the photochromism of DAE, as effective molecular switching occurs more readily in the amorphous regions of the polymer matrix.^18,33 Additionally, as the concentration of DAE increases, the number of trap states also increases; however, this results in a deterioration in the crystalline ordering of the polymer matrix, creating a trade-off that requires optimization.

To analyze the crystalline ordering of the DPP-DTT/DAE mixture thin films as a function of DAE concentration, we conducted 2D grazing-incidence X-ray diffraction (2D-GIXD) measurements. Notably, all samples were not subjected to annealing treatment to ensure effective switching, as sufficient free volume in the amorphous regions is essential for facilitating molecular switching. The (h00) peaks observed along the out-of-plane direction in the 2D-GIXD patterns shown in Fig. 2a–e indicated that the films predominantly exhibited edge-on orientations, regardless of the DAE concentration. As the DAE concentration increased from 0 wt% to 40 wt%, the q_z value of the (100) peak decreased, as shown in the out-of-plane line-cut profiles and the inset in Fig. 2f. This decrease corresponds to an increase in the d-spacing value, which increased from 22.13 Å to 23.24 Å. Fig. S3 presents the line-cut profiles of the in-plane direction in the 2D-GIXD patterns. With an increase in DAE concentration up to 30 wt%, the π–π distance remained relatively constant; however, at 40 wt%, it increased from 3.81 Å to 3.84 Å. The values of the π–π distance and d-spacing at different DAE concentrations are summarized in Table S1. As the DAE concentration increased, the d-spacing became larger, while the π–π distance remained relatively constant up to 30 wt%. This suggested that, at this concentration, the addition of DAE did not significantly affect the π–π interactions, allowing for improvements in switching performance with effective charge transport. However, at 40 wt%, the increased π–π distance weakened the π–π interactions, leading to a decrease in electrical properties.

Fig. 2 2D-GIXD pattern images of the DPP-DTT film: (a) without DAE, (b) 10 wt%, (c) 20 wt%, (d) 30 wt%, and (e) 40 wt% DAE-316. (f) Corresponding line-cut profiles along the q_zaxis (out-of-plane direction). Inset image shows the enlarged (100) peaks.

We also conducted atomic force microscopy (AFM) measurements to assess the changes in the surface morphology and thickness of the thin films as the concentration of DAE varied, as shown in Fig. S4. All samples were fabricated on AgNW films deposited on octadecyltricholorosilane (ODTS)-treated SiO₂ substrates, the same as the device fabrication process, with roughness measurements conducted in a 3 μm × 3 μm area while avoiding the area covered with AgNWs. The measured root-mean-square roughness values increased from 0.68 nm without DAE to 1.04 nm at 10 wt%, indicating an increase due to DAE incorporation. This value then reached 1.09 nm at 20 wt%, showing little change at this concentration. However, relatively large increases were observed at 30 wt% and 40 wt%, with roughness values of 1.27 nm and 1.53 nm, respectively. This increase in roughness, especially at 40 wt%, indicates that an excessive amount of DAE additive is not beneficial for vertical charge transport for OSBT. The thickness of the thin films was 52.5 nm without DAE, which increased to 69.3 nm at 10 wt% and 91.0 nm at 20 wt%. At 30 wt% and 40 wt%, the thickness values were 91.4 nm and 94.7 nm, respectively, showing little difference starting from 20 wt%. Since the thickness of the thin films determines the channel length in OSBTs, this increase also affects the electrical properties of OSBTs, along with changes in crystalline ordering corresponding to the amount of DAE.

Device performance of molecular-switch-embedded OSBTs

The performance of molecular-switch-embedded OSBTs was evaluated to assess the impact of DAE concentration on their optoelectronic characteristics. As shown in the transfer curve in Fig. S5, a DPP-DTT solution concentration of 9 mg mL⁻¹ was selected, as it exhibited both a high on-current and a high on/off ratio. Meanwhile, a concentration of 0.25 mg mL⁻¹ was used for the AgNW solution.²⁹

Fig. S6 presents the time required for DAE to facilitate the transition of the transistor to the photoprogrammable off state under UV light irradiation, as well as the time needed to restore it to the photoprogrammable on state under visible light exposure. The transistor took 90 s to fully turn off upon UV irradiation, while it required 7 min to completely recover to the on state under visible light. These durations are longer than the intrinsic switching times of DAE in the DPP-DTT film, where the transition from the open form to the closed form under UV light irradiation occurs in 30 s, and the reverse transition under visible light takes 5 min.³³ The device-level switching time reflects not only the photochemical isomerization of DAE but also the optical delivery of photons to the active layer and the electrical equilibration of the vertical stack. Although the DMD shows 74% transmittance at 520 nm, the cycloreversion (closed to open) exhibits a smaller effective rate constant due to the lower absorption cross-section/quantum yield in this band and additional optical losses (reflection/interference), which, together with trap-limited charge (de)accumulation and Schottky barrier reconfiguration, prolongs the recovery. By contrast, the cyclization (open to closed) step benefits from a larger photochemical rate and thus reaches the off state faster despite the lower nominal transmittance.

Fig. 3a–d presents the photoprogrammable switching performance of OSBTs utilizing DPP-DTT/DAE as a function of DAE concentration. Regardless of the DAE content, all OSBTs exhibited high-performance transfer curves with an on/off ratio exceeding 10⁵. However, as the DAE concentration increased from 10 wt% to 40 wt%, the on-current significantly decreased from 2.56 × 10⁻⁶ A to 6.27 × 10⁻⁷ A. At 10 wt% DAE, a high current level was observed due to the short channel length resulting from the thin film thickness. As the DAE content increased to 20 wt% and 30 wt%, the film thickness increased, lengthening the channel and reducing the current. Notably, despite similar film thicknesses, the current further decreased, and unstable transfer characteristics appeared at 40 wt%. This decline in on-current can be attributed to reduced crystalline ordering and an increased channel length caused by an excessive amount of DAE.

Fig. 3 All transfer characteristics were measured at V_DS = −30 V. Transfer characteristics of the OSBTs with varying concentrations: (a) 10 wt%, (b) 20 wt%, (c) 30 wt%, and (d) 40 wt% DAE-316, measured both before and after UV exposure.

Upon UV irradiation, the on-current further decreased, and unlike in the case of DAE-embedded OFETs, the threshold voltage (V_th) also shifted. This behavior is inherent to the operating principle of OSBTs, where charge injection is regulated through Schottky barrier height modulation. As the DAE concentration increased, this effect became more pronounced, with the transistor completely turning off at concentrations of 30 wt% and higher. Despite this, the higher initial on-current observed at 30 wt% prior to UV irradiation enabled the achievement of a switching ratio exceeding 64 000—the highest reported to date among photoprogrammable OFETs. We summarize the transistor performance in this study in Table S2 and compare it with that of other reported molecular-switch-embedded transistors in Table S3 to highlight the superior performance of OSBTs for this application. Fig. S7 presents the reproducibility of the devices, showing measurements of 10 devices for each DAE concentration. The data, presented as box plots, illustrates the distribution and consistency of the on-current and switching ratio across multiple samples. The output characteristics of OSBTs as a function of DAE-316 concentration are provided in Fig. S8.

As previously discussed, OSBTs exhibit significantly higher trapping efficiency by DAE_c compared to conventional lateral OFETs when similar amounts of DAE are incorporated. This enhanced trapping efficiency not only enables the achievement of a record-high photoprogrammable switching ratio but also results in a gradual negative shift in V_th (in the case of p-type transistors), ultimately leading to the realization of a fully off state. To illustrate this mechanism, Fig. 4 presents a schematic comparison of the quantitative effects of DAE_c on the photoprogrammable switching performances of OSBTs and OFETs. In the case of OFETs, as well documented in the literature, the accumulation layer is confined to a two-dimensional region within approximately 5 nm of the dielectric interface, meaning that not all DAE_c molecules dispersed throughout the thin film contribute to charge trapping (Fig. 4a). Additionally, the injection barrier between the source electrode and the semiconductor remains unaffected by DAE_c, making it difficult to achieve a significant V_th shift and a fully off state.

Fig. 4 Schematic illustration of the charge transport region in (a) OFETs and (b) OSBTs. (c) Schematic representation of the band diagram and operating mechanism in OSBTs with and without DAE.

In contrast, OSBTs utilize the entire thin film as the channel, as depicted in Fig. 4b. Consequently, all DAE_c molecules dispersed throughout the film contribute to charge trapping, significantly enhancing the switching effect. Furthermore, as illustrated in the band diagram and operating mechanism of OSBT in Fig. 4c, the application of a gate voltage induces charge accumulation at the AgNW (source)/DPP-DTT (semiconductor) interface. This accumulation gives rise to interfacial band bending, which lowers the effective Schottky barrier height and thereby enables efficient hole injection. However, when DAE molecules are incorporated, charge trapping by the DAE_c suppresses charge accumulation in response to gate voltage, thereby reducing the degree of band bending and hindering the lowering of the Schottky barrier. These characteristics result in a pronounced shift in the V_th and enable the realization of a fully off state. Such a mechanism highlights the advantage of OSBTs, as they exhibit ideal photoprogrammable switching behavior that cannot be achieved in conventional lateral OFET structures.

Fig. 5a displays the transfer curves of the OSBT utilizing DPP-DTT with DAE 30 wt% under UV light and visible light irradiation, with measurements taken up to 100 steps. Due to the high fatigue resistance of DAE, stable switching was observed throughout the 100 steps, demonstrating that this characteristic is applicable in OSBTs as well. Fig. 5b summarizes the results of the 100 switching steps. I_on (DAE_o) decreased from 8.72 × 10⁻⁷ A to 4.79 × 10⁻⁷ A, and the V_th shifted from 2.1 V to −1 V. After UV irradiation, it turned off completely, which resulted in Ion (DAE_c) remaining at a similarly low level.

Fig. 5 All transfer characteristics were measured at V_DS = −30 V, with a fixed UV irradiation time of 90 s and visible light irradiation time of 7 min. (a) Photoprogrammable switching behavior of OSBT based on DPP-DTT with 30 wt% DAE-316 after irradiation with UV (312 nm) and visible light (520 nm) for up to 100 steps. (b) Summary of the current levels corresponding to these 100 steps, extracted from the transfer characteristics at V_GS = −30 V and V_DS = −30 V.

Conclusions

We successfully integrated DAE molecular switches into a vertical OSBT architecture, achieving enhanced photoprogrammable switching performance. Compared to lateral OFET structures, the vertical OSBT architecture offers important advantages for more effective photoprogrammable switching: (1) DAE dispersed in the entire active layer contributes to trapping and (2) DAE also affects the Schottky barrier lowering behavior. To overcome the challenge of light obstruction by the top electrode, we introduced a DMD electrode, enabling a record-high switching ratio exceeding 64 000 at 30 wt% DAE. Additionally, the OSBT with a DAE exhibited excellent fatigue resistance, maintaining stable performance over 100 repeated switching cycles. Whereas previous studies on molecular-switch-embedded photoprogrammable OFETs were constrained by the physical, chemical, and crystallographic interactions between the molecular switch and the semiconductor, limiting improvements in the switching ratio, this study is the first to demonstrate that structural innovation in the transistor can overcome the performance limitations of photoprogrammable OFETs.

Author contributions

Conceptualization: H. R. Sim, S. Z. Hassan, and D. S. Chung; data curation: H. R. Sim; formal analysis: H. R. Sim, S. Z. Hassan, S. Lee, J. Kwon, G.-H. Nam, S. Baek, C. So, and Y. G. Lee; funding acquisition: D. S. Chung; investigation: H. R. Sim and S. Z. Hassan; methodology: H. R. Sim and S. Z. Hassan; project administration: D. S. Chung; resources: S. Z. Hassan; supervision: D. S. Chung; validation: H. R. Sim, S. Z. Hassan, S. Lee, and J. Kwon; visualization: H. R. Sim; writing – original draft: H. R. Sim; writing – review & editing: H. R. Sim and D. S. Chung.

Conflicts of interest

There are no conflicts to declare.

Data availability

The data supporting this article have been included as part of the supplementary information (SI). The SI includes detailed experimental procedures, additional UV-vis spectra, 2D-GIXD data, AFM images, device transfer/output curves, and tables summarizing electrical performance. See DOI: https://doi.org/10.1039/d5mh01504a.

Acknowledgements

This study was supported by the National R&D Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Science and ICT (RS-2022-NR068160 and RS-2024-00406548).

References

1. H. Bronstein, C. B. Nielsen, B. C. Schroeder and I. McCulloch, Nat. Rev. Chem., 2020, 4, 66.
2. Y. Wang, L. Sun, C. Wang, F. Yang, X. Ren, X. Zhang, H. Dong and W. Hu, Chem. Soc. Rev., 2019, 48, 1492.
3. S. Yuvaraja, A. Nawaz, Q. Liu, D. Dubal, S. G. Surya, K. N. Salama and P. Sonar, Chem. Soc. Rev., 2020, 49, 3423.
4. C. Liao, M. Zhang, M. Y. Yao, T. Hua, L. Li and F. Yan, Adv. Mater., 2014, 27, 7493.
5. H. Ling, S. Liu, Z. Zheng and F. Yan, Small Methods, 2018, 2, 1800070.
6. K. Liu, B. Ouyang, X. Guo, Y. Guo and Y. Liu, npj Flex. Electron., 2022, 6, 1.
7. Y. Zhang, Y. Wang, C. Gao, Z. Ni, X. Zhang, W. Hu and H. Dong, Chem. Soc. Rev., 2023, 52, 1331.
8. J. Wu, F. Xu, S. Li, P. Ma, X. Zhang, Q. Liu, R. Fu and D. Wu, Adv. Mater., 2019, 31, 1802922.
9. X. Yao, H. Ma, X. Wang, H. Wang, Q. Wang, X. Zou, Z. Song, W. Jia, Y. Li, Y. Mao, M. Singh, W. Ye, J. Liang, Y. Zhang, Z. Liu, Y. He, J. Li, Z. Zhou, Z. Zhao, Y. Zhang, G. Niu, C. Yin, S. Zhang, H. Shi, W. Huang and Z. An, Nat. Commun., 2022, 13, 4890.
10. G. P. Kini, S. J. Jeon and D. K. Moon, Adv. Funct. Mater., 2021, 31, 2007931.
11. S. H. Yu, S. Z. Hassan, C. So, M. Kang and D. S. Chung, Adv. Mater., 2023, 35, 2203401.
12. L. Kortekaas and W. R. Browne, Chem. Soc. Rev., 2019, 48, 3406.
13. R. Klajn, Chem. Soc. Rev., 2014, 43, 148.
14. M. Gobbi, S. Bonacchi, J. X. Lian, A. Vercouter, S. Bertolazzi, B. Zyska, M. Timpel, R. Tatti, Y. Olivier, S. Hecht, M. V. Nardi, D. Beljonne, E. Orgiu and P. Samori, Nat. Commun., 2018, 9, 2661.
15. S. H. Yu, S. Z. Hassan, G.-H. Nam, S. An, B. Kang and D. S. Chung, Chem. Mater., 2021, 33, 5991.
16. E. Orgiu, N. Crivillers, M. Herder, L. Grubert, M. Patzel, J. Frisch, E. Pavlica, D. T. Duong, G. Bratina, A. Salleo, N. Koch, S. Hecht and P. Samori, Nat. Chem., 2012, 4, 675.
17. T. Leydecker, M. Herder, E. Pavlica, G. Bratina, S. Hecht and E. Orgiu, Nat. Nanotechnol., 2016, 11, 769.
18. L. Hou, T. Leydecker, X. Zhang, W. Rekab, M. Herder, C. Cendra, S. Hecht, I. McCulloch, A. Salleo, E. Orgiu and P. Samori, J. Am. Chem. Soc., 2020, 142, 11050.
19. E. Orgiu and P. Samori, Adv. Mater., 2014, 26, 1827.
20. J. Zhang, Q. Zou and H. Tian, Adv. Mater., 2022, 34, 2200526.
21. M. Kang, S. Z. Hassan, S.-M. Ko, C. Choi, J. Kim, S. K. R. Parumala, Y.-H. Kim, Y. H. Jang, J. Yoon, D.-W. Jee and D. S. Chung, Adv. Mater., 2022, 34, 2200526.
22. S. Z. Hassan, J. Kwon, J. Lee, H. R. Sim, S. An, S. Lee and D. S. Chung, Adv. Sci., 2024, 11, 2401482.
23. S.-J. Wang, M. Sawatzki, G. Darbandy, F. Talnack, J. Vahland, M. Malfois, A. Kloes, S. Mannsfeld, H. Kleemann and K. Leo, Nature, 2022, 606, 700.
24. W. Huang, J. Chen, Y. Yao, D. Zheng, X. Ji, L.-W. Feng, D. Moore, N. R. Glavin, M. Xie, Y. Chen, R. M. Pankow, A. Surendran, Z. Wang, Y. Xia, L. Bai, J. Rivnay, J. Ping, X. Guo, Y. Cheng, T. J. Marks and A. Facchetti, Nature, 2023, 613, 496.
25. H. Kleemann, K. Krechan, A. Fischer and K. Leo, Adv. Funct. Mater., 2020, 30, 1907113.
26. Y. Chen, Y. Yao, N. Turetta and P. Samori, J. Mater. Chem. C, 2022, 10, 2494.
27. M. A. Mccarthy, B. Liu, E. P. Donoghue, I. Kravchenko, D. Y. Kim, F. So and A. G. Rinzler, Science, 2011, 332, 570.
28. E. Li, C. Gao, R. Yu, X. Wang, L. He, Y. Hu, H. Chen, H. Chen and T. Guo, Nat. Commun., 2022, 13, 2898.
29. H. R. Sim, S. Lee, J. Lee, S. Z. Hassan, G.-H. Nam, C. So, K. M. Sim and D. S. Chung, ACS Nano, 2023, 17, 24374.
30. M. Kiguchi, M. Nakayama, K. Fujiwara, K. Ueno, T. Shimada and K. Saiki, Jpn. J. Appl. Phys., 2003, 42, L1408.
31. S. Yadav and S. Ghosh, Phys. Scr., 2020, 95, 095801.
32. S. H. Yu, S. Z. Hassan, S. Lee, B. Lim and D. S. Chung, J. Mater. Chem. C, 2023, 11, 1560.
33. S. Z. Hassan, J. Song, S. H. Yu and D. S. Chung, Chem. Mater., 2021, 33, 7546.

---