Different interactions between a metal electrode and an organic layer and their different electrical bistability performances

Abstract

A new p–π conjugated small molecule, 2-[(pyridin-4-ylmethylene)-amino]-6-(N′-pyridin-4-ylmethylene-hydrazino)-benzo[de]isoquinoline-1,3-dione (2PyNI), was synthesized, characterized and fabricated into memory devices. The device ITO/2PyNI/Au (1) showed volatile memory effect, whereas the device ITO/2PyNI/Al (2) showed write once read many times (WORM) effect. According to theoretical calculations, 2PyNI had different interactions with Au and Al, which produced the different memory behaviors of devices 1 and 2, respectively. In addition, the I–V characteristics of devices 1 and 2 were analyzed in detail with various conduction models. The temperature dependence of the current for the ON state of devices 1 and 2 were measured. Under the irradiation of the UV light (365 nm), device 1 could be switched on automatically when the voltage was fixed at −3.0 V and switched off as the bias was removed.

---

Introduction

Organic materials are expected to be widely applied in light-emitting diodes, solar cells and high density data storage (HDDS) memory devices.^1–6 For memory devices, different memory types based on organic materials have been obtained through changing the active layer such as non-volatile write once read many times (WORM), flash (rewritable) and volatile dynamic random access memory (DRAM).^7–9 Tuning the suitable memory types is important for future practical applications. There are several reported methods such as adjustment of the component ratio of polymer composites and modification of molecular structure. For instance, Chen et al. once fabricated memory devices from Flash to WORM memory behavior by only tuning the ratio of polyimide and the additive, graphene oxide.¹⁰ In our previous study, an aromatic hydrazone, SNACA ([3-(N-butyl-4-carbaldehyde-1,8-naphthalimide)-9-hexyl ether-9H-carbazole]) showed volatile memory behavior (DRAM) when fabricated as a ITO or Pt/SNACA film/Al memory device. However, changing the hydrazone linker with a linear π-spacer, pyridyl acetylene, the organic compound CAPyNA based device showed a different WORM memory performance.¹¹ However, the factors affecting the performances of the organic devices are far from the organic layer itself. The interfaces between active layer and metal electrode layer also played an important role in the memory behavior.^12–16 For example, it is well known that an organic molecule has different interactions with various metal layers. Most studies showed that the ON/OFF ratio and switching voltage could be influenced by the electrodes; however, the memory types were seldom changed.^17–19

Here, we fabricated devices with different metal electrodes intended to study the relationship of memory behaviour and the interface with organic layer. A new p–π conjugated molecule 2-[(pyridin-4-ylmethylene)-amino]-6-(N′-pyridin-4-ylmethylene-hydrazino)-benzo[de]isoquinoline-1,3-dione (2PyNI) based on the well-known backbone, naphthalimide and hydrazone, has been widely studied for its excellent photoelectric properties.^20,21 Moreover, pyridines were introduced at the edge of the molecule, which easily interacted with metal due to its inherent flexibility in being able to bind through the nitrogen atom or the π-ring. According to theoretical simulations, 2PyNI had different interactions with Au and Al. When two devices, ITO/2PyNI/Au (1) and ITO/2PyNI/Al (2), were fabricated, their current–voltage characteristics showed expectedly different memory types, in which device 1 showed volatile memory performance and device 2 showed a non-volatile WORM effect. In addition, the OFF (low-conductivity) state of the I–V characteristics for device 1 and 2 could be fitted with various conduction models. Moreover, device 1 could be switched on by irradiation with UV light. We speculated the mechanism using theoretical calculation, experimental characterization and conduction models.

Experimental

Materials

4-Bromo-1,8-naphthalic anhydride (97%, Liaoning Liangang Dyes Chemical Co. Ltd., China), hydrazine hydrate (85%, Sinopharm Chemical Reagent), pyridine-4-carbaldehyde (97%, TCI), sodium hydroxide (96%, Shanghai Sinopharm), and acetic acid (99%, Shanghai Sinopharm) were purchased from commercial sources.

Apparatus

¹H NMR spectra were obtained with an INOVA 400 MHz FT-NMR spectrometer, using CDCl₃ or DMSO-d₆ as solvent and tetramethylsilane (TMS) as the internal standard at ambient temperature. The UV-Vis absorption spectra were recorded using a Perkin Elmer Lambda-17 spectrophotometer at room temperature. Thermogravimetric analysis (TGA) was conducted using a TA instrument Dynamic TGA 2950 at a heating rate of 20 °C min⁻¹ under a nitrogen flow rate of 100 mL min⁻¹. The fluorescence spectra were measured using an Edinburgh-920 fluorescence spectrophotometer (Edinburgh Co. UK) with a slit of 3 nm. The fluorescence quantum yield (QY) in the solution was determined using fluorescein (Φ_F = 79% in 0.1 mmol L⁻¹ NaOH) as standard. Cyclic voltammetry was performed at room temperature using an ITO working electrode, a reference electrode Ag/AgCl, and a counter electrode (Pt wire) at a sweep rate of 100 mV s⁻¹ (CorrTest CS Electrochemical Workstation analyzer) in a solution of tetra-butylammonium hexafluorophosphate (TBAP) in CH₃CN (0.1 mmol L⁻¹). The scanning electron microscopy (SEM) images were taken using a Hitachi S-4700 scanning electron microscope. The atomic force microscopy (AFM) measurements were performed using a MFP-3DTM (Digital Instruments/Asylum Research) AFM instrument in the tapping mode. The single crystal X-ray diffraction was made using a Rigaku Mercury CCD X-ray diffractometer (50 kV, sealed tube) at 223 K using graphite monochromated Mo Kα and a suitable single crystal mounted at the top of a glass fiber. Diffraction data was collected in ω mode and reduced using the program CrystalClear and the application of a semi-empirical absorption correction. The reflection data was further corrected for Lorentz and polarization effects. The elemental analysis (EA) of C, H and N was performed by the Elemental Analysis Service using an EA1110-CHNS elemental analyzer.

Synthesis of compounds

Synthesis of N-amino-4-hydrazine-1,8 naphthalimide (1). Synthesis of N-amino-4-hydrazine-1,8 naphthalimide (1) was according to the references.^22,23 The product was characterized by ¹H NMR spectroscopy (DMSO-d₆, 400 MHz) δ (ppm): 9.18 (s, 1H), 8.61 (d, J = 8.4 Hz, 1H), 8.41 (d, J = 7.2 Hz, 1H), 8.27 (d, J = 8.8 Hz, 1H), 7.62 (t, J = 7.6 Hz, 8.0 Hz, 1H), 7.22 (d, J = 8.4 Hz, 1H), 5.70 (s, 2H), 4.68 (s, 2H).

Synthesis of 2-[(pyridin-4-ylmethylene)-amino]-6-(N′-pyridin-4-ylmethylene-hydrazino)-benzo[de]isoquinoline-1,3-dione (2PyNI). Compound 1 (1.2 g, 5.0 mmol), pyridine-4-carbaldehyde (1.2 g, 12 mmol) and a catalytic amount of acetic acid were dissolved in 150 mL anhydrous ethanol and heated to reflux for 8 h. The precipitate was filtered off and washed three times with anhydrous ethanol (10 mL). Recrystallization from a mixed solvent of DMF (N,N-dimethyl-formamide) and acetonitrile (volume ratio is 1 : 1) gave 2PyNI. Yellow solid (1.6 g, 3.8 mmol, yield, 76%) was obtained. EA: calcd for C₂₄H₁₆N₆O₂ (%): C, 68.56; H, 3.84; N, 19.99; found: C, 68.41; H, 3.94; N, 19.78. ¹H NMR (DMSO-d₆, 400 MHz) δ (ppm): 11.81 (s, 1H), 8.87 (s, 1H), 8.84 (d, J = 8.9 Hz, 2H), 8.81 (s, 1H), 8.65 (d, J = 5.4 Hz, 2H), 8.53 (d, J = 7.2 Hz, 1H), 8.42 (d, J = 8.2 Hz, 2H), 7.89 (s, 1H), 7.88 (s, 1H), 7.84 (dd, J = 8.1, 4.3 Hz, 2H), 7.75 (d, J = 5.6 Hz, 2H). ¹³C NMR (400 MHz, CF₃COOD) δ (ppm): 189.66, 153.38, 153.16, 150.56, 148.51, 147.66, 146.52, 143.43, 141.80, 140.70, 140.63, 137.37, 136.72, 136.48, 136.38, 134.76, 134.53, 131.18, 129.55, 129.40, 126.59, 126.49, 123.51, 123.31. HRMS-ESI (m/z): [M + H]⁺ calcd for C₂₄H₁₇N₆O₂, 421.1408, found, 421.1407 (Scheme 1).

Scheme 1 The synthesis and chemical structure of 2PyNI.

Fabrication of memory devices

The indium tin oxide (ITO) glass was pre-cleaned sequentially with deionized water, acetone and ethanol in an ultrasonic bath, each for 20 min. The active organic film was deposited on the ITO-glass substrate under high vacuum (∼10⁻⁶ Torr). The film was annealed at 70 °C in a vacuum oven for 12 h. Al or Au layer (∼50 nm) was thermally evaporated and deposited onto the organic surface at ∼10⁻⁵ Torr through a shadow mask to form the top electrode. The active area of the fabricated device was 0.126 mm² (a nummular point with a radius of 0.2 mm). The MoO₃ layer added in the device was deposited on the 2PyNI film under high vacuum (∼10⁻⁶ Torr). All electrical measurements of the device were characterized under ambient conditions, without any encapsulation, using a HP4145B semiconductor parameter analyzer.

Results and discussion

The product 2PyNI was characterized by ¹H NMR and IR spectroscopy and CCD X-ray diffraction. The thermal stability of 2PyNI was evaluated by TGA under a nitrogen atmosphere. As shown in the TGA curves (Fig. S3†), the thermal decomposition temperature (the 5% weight-lost temperature) of 2PyNI was up to 297 °C. The good-thermal stability of 2PyNI could endure heat deterioration in the memory devices. The X-ray diffraction (XRD) measurement showed no evident peak indicating the amorphous structure of 2PyNI film deposited by thermal vacuum deposition.

Crystal structure of 2PyNI

Single crystals of 2PyNI·2H₂O were obtained by slow evaporation of DMF solution (the key parameters of crystal structure were listed in Table S1†). In its asymmetric unit, there were one 2PyNI and two water molecules. The naphalamide ring and two pyridine rings linked via two hydrazone bridges in different modes. The dihedral angle of naphalamide and pyridine in planar conformation was 5.536°, whereas that in distorted conformation is 83.459°. Fig. 1 showed the molecular packing view. There was π–π stacking (3.382 Å) between nearby naphalamide rings. Due to the existence of water molecules, hydrogen bondings enclose the molecules into a 3-D supramolecular structure.

Fig. 1 Crystal structure of 2PyNI and 3-D stacking patterns of 2PyNI viewed along a axis (hydrogen bonds are denoted by red dotted lines), and hydrogen atoms were omitted for clarity.

Optical properties of 2PyNI

The UV-Vis absorption spectrum of 2PyNI in dilute tetrahydrofuran (THF) solution displayed one major absorption peak at 438 nm, which was assigned to the π–π* transition in the naphthalimide moieties.^24,25 The absorption spectrum of 2PyNI film on quartz substrate showed a visible bathochromic-shift (from 438 to 461 nm) and broadened absorption band (Fig. 2a), suggesting the formation of molecular aggregates and/or increased polarity of the thin film.^26–28 This was beneficial for the improvement of the charge carrier mobility of the films.²⁹ The fluorescence emission spectrum of 2PyNI in THF solution was 525 nm with a quantum yield of 60%. The fluorescence quantum yield (QY) in the solution was determined using fluorescein (Φ_F = 79% in 0.1 M NaOH) as standard. The fluorescence of 2PyNI film on quartz substrate was quenched and also assigned to the molecule aggregation.

Fig. 2 (a) UV-Vis absorption spectra of 2PyNI in THF solution (black) and film (red); (b) the fluorescence emission spectra of 2PyNI in THF solution and film with the excitation wavelength 460 nm.

Electrochemical properties of 2PyNI

The highest occupied molecular orbital (HOMO) and lowest unoccupied molecular orbital (LUMO) energy levels could be calculated from the UV-Vis absorption spectrum and the cyclic voltammetry (CV) results via the following equations:

E_HOMO = −[E^(onset)_OX + 4.80 − E_Foc]

E_LOMO = E_HOMO + E_g

where E^(onset)_OX was the onset oxidation potential, E_Foc was the external standard potential of the ferrocene/ferrocenium ion couple and E_g was the band gap determined from UV-Vis absorption spectrum. The optical band gap of the 2PyNI film, estimated from absorption edges, was 2.54 eV (Fig. 2a). The onset oxidation E^(onset)_OX of 2PyNI film was 1.12 eV (Fig. 3). The E_Foc was 0.50 eV, as determined from CV measurement with bare ITO glass substrate without molecular film. Assuming that the HOMO level for the Fc/Fc⁺ standard was −4.80 eV with respect to the zero vacuum level, the HOMO level for 2PyNI film is determined to be −5.4 eV. Thus, the LUMO level was estimated to be −2.9 eV. The energy barrier between the work functions (Φ) of the ITO (−4.8 eV) and HOMO energy level (−5.4 eV) was 0.6 eV, which was considerably lower than the energy barrier between the Φ of Al (−4.3 eV) or Au (−5.2 eV) and the LUMO energy level (−2.9 eV). This indicated that hole injection from ITO into the HOMO of 2PyNI (corresponding to ITO as the anode) was easier than electron injection from Al or Au into the LUMO of 2PyNI (Fig. 6b). Thus, the hole-injection dominated the conduction process to the film of 2PyNI.³⁰

Fig. 3 Cyclic voltammograms of 2PyNI film was measured in 0.1 mmol L⁻¹ TBAP/CH₃CN solution with Ag/AgCl as reference electrode and Pt wire as counter electrode. A scan rate of 100 mV s⁻¹ was used.

Morphology of 2PyNI

Atomic force microscopy (AFM) was used to characterize the surface morphologies and roughness of the vacuum-deposited electroactive layers. The surface root-mean-square roughness of 2PyNI film was 0.821 nm (Fig. 4). The small surface roughness in the AFM image of the film led to good quality of the film, which could offer a strong guarantee for high performance data-storage devices.

Fig. 4 Tapping-mode AFM height images of thin film spin-coating onto ITO substrates annealed at 70 °C; (a) morphology for 2PyNI; (b) roughness of 2PyNI film.

Electrochemical properties of memory devices

The current–voltage (I–V) characteristics of 2PyNI in device 1 is shown in Fig. 5a. When a voltage sweep was applied from 0 V to −6 V (sweep 1), the device was initially in the low-conductivity (OFF) state. As the negative bias increased further, the memory device switched from the OFF state to the high-conductivity (ON) state at the voltage of −4.3 V, as was indicated by the abrupt increase of the current from 10⁻⁵ A to 10⁻¹ A. The transition from the OFF state to the ON state could serve as the “write” process. The device remained in ON state during the subsequent scan from 0 to −6 V (sweep 2) and 0 to 6 V (sweep 3). Subsequent application of the negative scan from 0 to −6 V (sweep 4) was performed after turning off power for about 2 min, and device 1 could be reprogrammed from the OFF state to the ON state at −4.2 V again. The OFF state could be further written to the ON state when the voltage was reapplied, indicating that the memory device was rewritable and volatile. The short retention time of the ON state suggested that device 1 showed the volatile memory characteristic.³¹

Fig. 5 (a and c) Current–voltage (I–V) characteristics of the 2PyNI memory device; (b and d) the effect of retention time of the memory device under a constant stress of −1.0 V.

The current–voltage characteristics of device 2 is shown in Fig. 5c. The device switched at a threshold voltage of −3.8 V with the current from 10⁻⁵ A to 10⁻¹ A (sweep 1). The ON state was still observed during the subsequent voltage sweep (sweep 2 and sweep 3), even though the voltage was withdrawn for longer time. Thus, the non-volatile memory behavior of device 2 showed the characteristics of WORM memory performance.³²

At a constant stress of −1.0 V, no evident degradation in the current for the OFF and ON states of devices 1 and 2 were observed during the long-term testing of about 10⁴ s (Fig. 5b and d). The results suggested the excellent device stability.

Calculation details and mechanisms

To further understand the different memory behaviours due to the different top electrode, we first utilized theoretical calculations to illustrate the interactions of the organic molecule and metal electrodes. All calculations based on density functional theory (DFT) with generalized gradient approximation (GGA) were implemented in DMol³ code available in Materials Studio 6.0.³³ For the calculation, Au13 cluster was chosen as a representative of the gold layer, which was seen as a gold “magic” number cluster and used as the adsorptive of organic molecules.³⁴ As a comparison, the Al13 cluster was also used as adsorptive of 2PyNI. The generalized gradient approximation functional by B88 exchange and LYP correlation (GGA-BLYP) along with a double numerical plus polarization (DNP) basis set was used for all the calculations. According to the result shown in Fig. 6 the optimized 2PyNI–Au13 system had no significant interaction and maintained a distance of Au–N about 0.509 nm, the original value predetermined between Au and pyridine group. However, the optimized 2PyNI–Al13 had a strong Al–N interaction, in which the distance of Al–N was close to 0.2 nm.

Fig. 6 (a) Energy level diagram of HOMO and LUMO for 2PyNI, the work functions of ITO, Al and Au electrodes; (b) simulated highest occupied molecular orbital (HOMO), lowest unoccupied molecular orbital (LUMO) and electrostatic potential (ESP) plots of 2PyNI.

The energy level diagram showed that as a negative bias was applied, the electron injection from the Al electrode to the LUMO of 2PyNI was easier than that with the Au electrode. This result corresponded to the test that the turn-on voltage of device 2 was lower than device 1.

According to the calculation, the 2PyNI molecule itself contained some negative ESP zone (blue) forming charge traps (acceptor side). As the voltage increased, the traps were filled and then the device turned to the ON-state, showing bistable conductivity states. To device 1, as the voltage was applied, the charge injected into the 2PyNI film from the ITO and the electron transmitted from the HOMO to the LUMO of 2PyNI and device 1 turned to the ON-state. As the 2PyNI had no interaction with the Au, the electron hopped to the Au electrode and then the device returned to OFF-state showing volatile memory behavior.

To device 2, as the voltage was applied, the electron transmitted to the LUMO from the HOMO level and then the device turned to the ON-state. As the organic layer had strong interaction with Al, the charge was trapped in the 2PyNI molecule and could not be recovered after removing the voltage. Thus, device 2 showed WORM characteristics.

To further research these memory behaviours, we analyzed the I–V characteristics in detail with various conduction models. The ohmic contact model was found to be matched with the I–V data for the ON-state of devices 1 and 2 (Fig. 7b and d).^35–37 The results indicated that ohmic conduction was dominant for all devices in the ON-state.

Fig. 7 Analysis of current–voltage characteristics of the (a) OFF-state with trap-limited space charge limited current (SCLC) model and (b) ON-state based on device 1; (c) OFF-state with the space-charge-free Frenkel Poole emission current model and (d) ON-state based on device 2.

To device 1, the I–V data of the OFF-state could be fitted by the trap-limited space-charge limited conduction (SCLC) model.^38–40As shown in Fig. 7a, the logarithmic plot of the I–V data for the OFF-state contained two linear regions for <0.46 V and >0.46 V, with slopes of 1.1 and 5.6, respectively. To device 2, the I–V data of the OFF-state can be fitted by the space-charge-free Frenkel Poole emission current model.⁴¹ The different conduction models showed different interaction between 2PyNI and Au or Al.

To block the interaction between the 2PyNI molecule and Al, we fabricated the device ITO/2PyNI/MoO₃/Al. With the blocking layer MoO₃ thickness of 5 nm, the device ITO/2PyNI/MoO₃/Al also showed WORM characteristics with lower turn-on voltage of −2.0 V compared with device 2 (Fig. 8a). To device ITO/2PyNI/MoO₃/Al, the charge was also trapped in the 2PyNI molecule and could not return. Thus, device ITO/2PyNI/MoO₃/Al also showed WORM characteristics. The MoO₃ was conducive to the electron injection and therefore decreased the switching voltage of device.^42–44 The approach of adding MoO₃ into the device could provide a new method for decreasing turn-on voltage of the electrical memory device.

Fig. 8 (a) Current–voltage characteristics of device ITO/2PyNI/MoO₃/Al; (b) the effect of retention time of the memory device ITO/2PyNI/MoO₃/Al under a constant stress of −1.0 V.

Moreover, the temperature dependence (between 250 and 320 K) of the current for the ON state of devices 1 and 2 were measured (Fig. S7†). It was found that the ON state of devices 1 and 2 was weakly affected by the temperature, which excluded the formation of the metal bridge during the ON state. As the temperature dependence was very weak and there was linear I–V characteristics, a tunneling effect might have dominated the ON-state.⁴⁵

The aromatic hydrazone group was photo sensitive as suggested by previous reports.^46,47 Device 1 showed no evident change under irradiation with incandescent lamp. However, under irradiation with the UV lamp (365 nm), the threshold voltage of device 1 was decreased significantly from −4.3 V to −2.2 V (Fig. 9a). When the voltage was fixed at −3.0 V (Fig. 9b), for the first 8 s, the device 1 was in the OFF state in the dark. After being exposed to UV light (365 nm) for about 20 s, device 1 switched to the ON state.

Fig. 9 (a) Current–voltage (I–V) characteristics of device 1 lighted by 365 nm UV; (b) the electro-optical performance of the device 1.

The UV-Vis absorption spectra of the 2PyNI film before and after irradiation had showed no evident change, which indicated an unchanged electronic transit. We tentatively assigned it to the polarity of hydrazone group. As the UV lamp (365 nm) was irradiated, the polar molecule 2PyNI adsorbed the energy, beneficial for the charge separation and formation of ion channel. Therefore, the device 1 was inclined to be switched on a considerably lower voltage under the UV light. The electro-optical performance of device 1 could be repeated after being placed in the dark for a while, which showed potential application in UV sensor.

Conclusion

In summary, the small molecule 2PyNI containing a hydrazone group was synthesized and fabricated as an electro-active layer of the sandwich-structure memory devices. By changing the top electrodes, we obtained different memory types. Device 1 with Au as the top electrode showed a volatile memory effect, whereas device 2 with Al as top electrode showed a non-volatile WORM effect. The difference was assigned to the different interactions between metal and organic layer. Therefore, the interface of electrode and active film might play an important role in the memory performance in this case. It would provide useful information for tuning memory types by molecular design.

Acknowledgements

The authors graciously thank Prof. Qing-Hua Xu from the National University of Singapore for helpful discussion. The authors graciously thank the Chinese Natural Science Foundation (21371128 and 21336005), Chinese-Singapore Joint Project (2012DFG41900), and Specialized Research Fund for the Doctoral Program of Higher Education of China (grant no. 20113201130003).

Notes and references

1. B. Cho, J. Yun, S. Song, Y. Ji, D. Kim and T. Lee, Adv. Funct. Mater., 2011, 21, 3976.
2. S. L. Lian, C. L. Liu and W. C. Chen, ACS Appl. Mater. Interfaces, 2011, 3, 4504.
3. J. P. Wang, Nat. Mater., 2005, 4, 191.
4. M. Colle, M. Buchel and D. M. Leeuw, Org. Electron., 2006, 7, 305.
5. J. C. Scott and L. D. Bozano, Adv. Mater., 2007, 19, 1452.
6. P. Y. Gu, F. Zhou, J. K. Gao, G. Li, C. Y. Wang, Q. F. Xu, Q. C. Zhang and J. M. Lu, J. Am. Chem. Soc., 2013, 135, 14086.
7. S. ChandraKishore and A. Pandurangan, RSC Adv., 2014, 4, 9905.
8. C. Jin, J. Lee, E. Lee, E. Hwang and H. Lee, Chem. Commun., 2012, 48, 4235.
9. I. A. Hummelgen, N. J. Coville, I. Cruz-Cruz and R. Rodrigues, J. Mater. Chem. C, 2014, 2, 7708.
10. A. D. Yu, T. Kurosawa, Y. H. Chou, K. Aoyagi, Y. Shoji, T. Higashihara, M. Ueda, C. L. Liu and W. C. Chen, ACS Appl. Mater. Interfaces, 2013, 5, 4921.
11. G. Wang, S. F. Miao, Q. J. Zhang, H. F. Liu, H. Li, N. J. Li, Q. F. Xu, J. M. Lu and L. H. Wang, Chem. Commun., 2013, 49, 9470.
12. A. D. Yu, C. L. Liu and W. C. Chen, Chem. Commun., 2012, 48, 383.
13. E. Kapetanakis, A. M. Douvas, D. Velessiotis, E. Makarona, P. Argitis, N. Glezos and P. Normand, Adv. Mater., 2008, 20, 4568.
14. Q. D. Ling, D. J. Liaw, C. X. Zhu, D. S. H. Chan, E. T. Kang and K. G. Neoh, Prog. Polym. Sci., 2008, 33, 917.
15. N. G. Kang, B. Cho, B. G. Kang, S. Song, T. Lee and J. S. Lee, Adv. Mater., 2012, 24, 385.
16. Q. D. Ling, D. J. Liaw, E. Y. H. Teo, C. Zhu, D. S. H. Chan, E. T. Kang and K. G. Neoh, Polymer, 2007, 48, 5182.
17. R. Zazpe, P. Stoliar, F. Golmar, R. Llopis, F. Casanova and L. E. Hueso, Appl. Phys. Lett., 2013, 103, 073114.
18. S. H. Hong, O. Kim, S. Choi and M. Ree, Appl. Phys. Lett., 2007, 91, 093517.
19. S. G. Hahm, T. J. Lee, D. M. Kim, W. Kwon, Y. G. Ko, T. Michinobu and M. Ree, J. Phys. Chem. C, 2011, 115, 21954.
20. R. M. Duke, E. B. Veale, F. M. Pfeffer, P. E. Krugerc and T. Gunnlaugsson, Chem. Soc. Rev., 2010, 39, 3936.
21. D. Kolosov, V. Adamovich, P. Djurovich, M. E. Thompson and C. Adachi, J. Am. Chem. Soc., 2002, 124, 9945.
22. Z. Ma, W. Sun, L. Z Chen, J. Li, Z. Z. Liu, H. X. Bai, M. Y. Zhu, L. P. Du, X. D. Shi and M. Y. Li, Chem. Commun., 2013, 49, 6295.
23. N. R. Chereddy, K. Saranraj, A. K. Barui, C. R. Patra, V. J. Rao and S. Thennarasu, RSC Adv., 2014, 4, 24324.
24. G. Wang, S. f. Miao, Q. J. Zhang, H. F. Liu, H. Li and N. J. Li, Chem. Commun., 2013, 49, 9470.
25. H. Kanno, R. J. Holmes, Y. Sun, S. Kena-Cohen and S. R. Forrest, Adv. Mater., 2006, 18, 339.
26. S. L. Lim, N. J. Li, J. M. Lu, Q. D. Lin, C. X. Zhu, E. T. Kang and K. G. Neoh, ACS Appl. Mater. Interfaces, 2009, 1, 60.
27. K. H. Kim, S. Y. Bae, Y. S. Kim, J. A. Hur, M. H. Hoang, T. W. Lee, M. J. Cho, Y. Kim, M. Kim, J. I. I. Jin, S. J. Kim, K. Lee, S. J. Lee and D. H. Choi, Adv. Mater., 2011, 23, 3095.
28. Y. Ma, X. Cao, G. Li, T. Wen, Y. Yang, J. Wang, S. Du, L. Yang, H. Gao and Y. Song, Adv. Funct. Mater., 2010, 20, 803.
29. S. F. Miao, H. Li, Q. F. Xu, N. J. Li, J. W. Zheng, R. Sun, J. M. Lu and C. M. Li, J. Mater. Chem., 2012, 22, 16582.
30. H. Li, N. J. Li, H. W. Gu, Q. F. Xu, F. Yan, J. M. Lu, X. W. Xia, J. F. Ge and L. H. Wang, J. Phys. Chem. C, 2010, 114, 6117.
31. C. J. Chen, H. J. Yen, Y. C. Hu and G. S. Liou, J. Mater. Chem. C, 2013, 1, 7623.
32. Y. Q. Li, R. C. Fang, A. M. Zheng, Y. Y. Chu, X. Tao, H. H. Xu, S. J. Ding and Y. Z. She, J. Mater. Chem., 2011, 21, 15643.
33. B. Delley, J. Chem. Phys., 2000, 113, 7756.
34. Y. Shichibu and K. Konishi, Small, 2010, 6, 1216.
35. J. Hsu, Y. Chen, T. Kakuchi and W. C. Chen, Macromolecules, 2011, 44, 5168.
36. B. Zhang, T. Cai, S. Li, X. Zhang, Y. Chen, K. Neoh, E. T. Kang and C. Wang, J. Mater. Chem. C, 2014, 2, 5189.
37. L. M. Chen, Z. Xu, Z. Honga and Y. Yang, J. Mater. Chem., 2010, 20, 2575.
38. T. J. Lee, S. Park, S. G. Hahm, D. M. Kim, K. Kim, J. Kim, W. Kwon, Y. Kim, T. Chang and M. Ree, J. Phys. Chem. C, 2009, 113, 3855.
39. W. Kwon, B. Ahn, D. Kim, Y. Ko, S. GyuHahm, Y. Kim, H. Kim and M. Ree, J. Phys. Chem. C, 2011, 115, 19355.
40. G. Tian, D. Wu, L. Shi, S. Qi and Z. Wu, RSC Adv., 2012, 2, 9846.
41. G. Liu, Q. Ling, E. Y. H. Teo, C. X. Zhu, D. S. Chan, K. Neoh and E. T. Kang, ACS Nano, 2009, 37, 1929.
42. H. Bronstein, J. M. Frost, A. Hadipour, Y. Kim, C. B. Nielsen, R. S. Ashraf, B. P. Rand, S. Watkins and I. McCulloch, Chem. Mater., 2013, 25, 277.
43. Y. L. Chen, C. Y. Chang, Y. J. Cheng and C. S. Hsu, Chem. Mater., 2012, 24, 3964.
44. S. L. Lim, N. J. Li, J. M. Lu, Q. D. Lin, C. X. Zhu, E. T. Kang and K. G. Neoh, ACS Appl. Mater. Interfaces, 2009, 1, 60.
45. W. L. Leong, N. Mathews, B. Tan, S. Vaidyanathan, F. Dötz and S. Mhaisalkar, J. Mater. Chem., 2011, 21, 5203.
46. H. Li, N. J. Li, R. Sun, H. W. Gu, J. F. Ge, J. M. Lu, Q. F. Xu, X. W. Xia and L. H. Wang, J. Phys. Chem. C, 2011, 115, 8288.
47. S. S. Razi, R. Ali, P. Srivastava, M. Shahid and A. Misra, RSC Adv., 2014, 4, 16999.

---

Footnote

† Electronic supplementary information (ESI) available. CCDC 978046. For ESI and crystallographic data in CIF or other electronic format see DOI: 10.1039/c4ra12893a

---