Low-temperature MoO₃ film from a facile synthetic route for an efficient anode interfacial layer in organic optoelectronic devices

Abstract

A low temperature solution-processed MoO₃ (sMoO₃) thin film with a facile method for organic optoelectronic devices is developed. The film is extremely smooth with a root mean square (RMS) roughness of 0.318 nm, which is a significant advance for fabricating devices. An X-ray photoelectron spectroscopy (XPS) measurement shows that the sMoO₃ possesses few Mo⁵⁺ states with a Mo : O stoichiometry of 2.99, demonstrating a nearly ideal MoO₃ stoichiometry at a low annealing temperature. The sMoO₃ film exhibits a superior hole injection ability in both an organic light-emitting diode (OLED) and organic photovoltaic cell (OPV), compared to conventional hole injection material poly (3,4-ethylenedioxythiophene):poly(styrenesulfonate) (PEDOT:PSS). It is proved that the sMoO₃ fabricated by the simple and low-cost method is an excellent alternative to a PEDOT:PSS anode buffer layer for solution-processed organic optoelectronic devices.

---

Introduction

In the last two decades, wide attention has been paid to solution-processed optoelectronic devices such as organic light-emitting diodes (OLEDs)^1,2 and organic photovoltaic cells (OPVs),^3,4 owing to their potential for realizing low-cost, light-weight, large-area and flexible devices. In order to enhance device efficiency and stability, an effective strategy is to obtain a significant energetic match at the electrode–organic interface, by inserting an interfacial layer.^5,6 A commonly used hole-injection material is poly (3,4-ethylenedioxythiophene):poly(styrenesulfonate) (PEDOT:PSS). However, its acidic and hygroscopic nature is associated with the long-term instability.^7–10 Moreover, the low work function (5.1 eV) of PEDOT:PSS limits charge injection into the hole-transporting material with a large ionization potential.^11,12 One of the most promising alternatives is molybdenum oxide (MoO₃) that exhibits non-toxicity and deep-lying electronic states.^13–17 In particular, MoO₃ as the hole injection layer (HIL) can be solution processed in view of low-cost and large-area production. Several precursors for fabricating a solution-processed MoO₃ layer have been reported. Liu et al. spin coated an acidified aqueous ammonium heptamolybdate (NH₄)₆Mo₇O₂₄·H₂O precursor solution on indium tin oxide (ITO) and heat treated this at 160 °C, to obtain a discontinuous coverage of MoO₃ nanoparticles on the ITO surface.¹⁸ On the other hand, Murase et al. obtained a smooth MoO₃ film by pre-heating the same precursor in deionized water before spin coating.¹⁹ Jasieniak et al. exploited oxomolybdate MoO₂(acac)₂ (acac = bis(2,4-pentanedionate)) as a precursor to develop a MoO_x thin film, and the OPV using this as the modifying layer exhibits a comparable efficiency to that using a PEDOT:PSS modified anode. However, the current density–voltage (J–V) curve shows an s-shape, becoming more serious with increased annealing temperature, which requires a deeper understanding.²⁰ The sol–gel technique provided another precursor, MoO₂(OH)(OOH), by dissolving MoO₃ powder in H₂O₂, but a high annealing temperature of 275 °C was required for the full conversion of the precursor to MoO₃.²¹ A commercially available MoO₃ nanoparticle precursor has also been spin coated, followed by a low annealing temperature of 100 °C, however, such an approach required oxygen plasma treatment to remove the polymer stabilizers.²² Recently, an aqueous MoO₃ solution was prepared by dissolving MoO₃ powder directly into deionized water, but extra heating (70 °C) and prolonged stirring were required when the solution concentration was higher than 0.1 wt%, and the root mean square (RMS) roughness of the MoO₃ film was over 3 nm.²³ Similarly, a MoO₃ solution was obtained by dispersing molybdenum powder in ethanol and H₂O₂ mixed solvents, however, 18 hours reaction time was required and then the solution needed to be dried to solid molybdenum oxide, that was dispersed uniformly into ethanol.²⁴

Here, we present an ultra facile synthetic route for a low-temperature solution processed MoO₃ anode buffer layer from an aqueous solution. The solution was prepared by dissolving MoO₃ powder in ammonia–water to form a transparent ammonium molybdate solution without a further stirring or heating process. During the annealing process in air, the ammonium molybdate underwent rapid decomposition to MoO₃, NH₃ and H₂O, the latter two components being volatile. The remaining MoO₃ film completely covered the ITO surface with uniformly distributed nanoplates, and it possessed desirable surface electronic structures for hole injection. Such a solution-processed MoO₃ (sMoO₃) layer exhibits a superior hole injection property compared to evaporated MoO₃ (eMoO₃) and PEDOT:PSS, and its application in optoelectronic devices such as OLEDs and OPVs suggests that our simple technique provides an ideal approach to develop an effective MoO₃ film as a HIL.

Results and discussions

As shown in Fig. 1, it is hard to dissolve MoO₃ powder in deionized water, while it can be easily dissolved in ammonia–water at room temperature at ambient conditions. The aqueous transparent solution with MoO₃ in ammonia–water is formed by the chemical reaction²⁵ between MoO₃, H₂O and NH₃ as follows

MoO₃ + 2NH₃ + H₂O → (NH₄)₂MoO₄.

Fig. 1 Photographs of solutions with MoO₃ powder in deionized water (a) and ammonia–water (b).

The ammonium molybdate (NH₄)₂MoO₄ is soluble in H₂O, and the solution is extremely stable and the properties do not change after 2 months storage in a refrigerating chamber at a temperature of 8 °C. The desired concentration is achieved by diluting a saturated solution (∼175 mg ml⁻¹) into a certain amount of deionized water.

The aqueous MoO₃ solution was spin coated onto a quartz glass substrate and annealed at 150 °C for 10 min in air. Fig. 2 shows atomic force microscopy (AFM) images of (a) the bare quartz glass, and the MoO₃ films coated from solution with different concentrations: (b) 2, (c) 10 and (d) 20 mg ml⁻¹. The bare quartz glass substrate exhibits a RMS roughness of 0.628 nm, with clear evidence of several particles stemming from sample contamination. The deposition of a MoO₃ film on the quartz glass from a 2 mg ml⁻¹ solution results in uniformly discrete nanoplates, which are formed by the aggregation of MoO₃ nanoparticles. The diameter and height of the nanoplate are about 110 and 10 nm, respectively. However, the discontinuous coverage of the MoO₃ nanoplates on the substrate causes a larger RMS roughness of 1.887 nm. By increasing the concentration of MoO₃ in the ammonia–water, the density of the nanoplates increases, leading to a continuous and smooth sMoO₃ film. The sMoO₃ film coated from a 10 mg ml⁻¹ solution exhibits a decreased RMS of 0.415 nm with retained tips of nanoplates. When the concentration increases to 20 mg ml⁻¹, no single tip of MoO₃ is observed and the RMS of the film is only 0.318 nm, indicating a homogeneous film covering the entire substrate. Furthermore, we have compared the optical transparency of the sMoO₃ and PEDOT:PSS with a 10 nm thickness. As shown in Fig. S2,† the 10 nm sMoO₃ film exhibits an ultra high transparency, better than PEDOT:PSS in the entire visible region. The obtained smooth surface morphology and extreme high transparency of the sMoO₃ film are well suited for its employment in organic electronics.

Fig. 2 AFM images of bare quartz glass (a), and the sMoO₃ films on quartz glass with 2 (b), 10 (c) and 20 (d) mg ml⁻¹ solutions.

X-ray photoemission spectroscopy (XPS) was performed to investigate the surface electronic structure of the sMoO₃ films. Fig. 3a compares the full scan spectra of the eMoO₃ and sMoO₃ films. The characteristic peaks of the O 1s, Mo 3s, Mo 3p and Mo 3d levels in the full spectra of the eMoO₃ film and the sMoO₃ film are identical. Moreover, the peaks of the core levels of N are not found for the sMoO₃ film, suggesting that the synthetic precursor containing the N element completely decomposes to MoO₃ by annealing at the temperature of 150 °C. In Fig. 3b and c, the Mo 3d doublet has a spin–orbit splitting with 3.1 eV, between the Mo 3d_3/2 and 3d_5/2 bands. The Mo 3d_3/2 and 3d_5/2 peaks of the eMoO₃ film locate at 232.9 and 236.0 eV, respectively, while those of the sMoO₃ film locate at 233.0 and 236.1 eV, respectively. The higher binding energy of the solution processed sample is attributed to an increased adhesion between the sMoO₃ film and the silicon substrate. To observe molybdenum in two different oxidation states, the Mo 3d spectra of the eMoO₃ and sMoO₃ films are fitted by two 3d doublets in the form of a Gaussian function. The calculated Mo⁵⁺/Mo⁶⁺ ratios for eMoO₃ and sMoO₃ are 0.058 and 0.019, respectively. By using the Mo⁵⁺/Mo⁶⁺ obtained from the Mo 3d core level spectra, the estimated Mo:O stoichiometries of eMoO₃ and sMoO₃ are 2.97 and 2.99, respectively. It can be seen that the concentration of the Mo⁵⁺ component in the eMoO₃ film is significantly larger than that in the sMoO₃ film, which coincides with the fact that preparation of MoO₃ in the oxygen deficient environment leads to increased Mo⁵⁺ states, corresponding to the generation of oxygen vacancies. Therefore, fewer Mo⁵⁺ states are observed in the sMoO₃ film from annealing in the air environment.

Fig. 3 XPS spectra of the MoO₃ films. (a) Full scanned spectra, Mo 3d core level of (b) eMoO₃, and (c) sMoO₃ prepared from a 20 mg ml⁻¹ solution.

To elucidate the interfacial effect of sMoO₃ as a HIL, hole-only devices with various hole injection materials were analyzed by means of the space-charge-limited current (SCLC). Fig. 4a shows the J–V characteristics of the hole-only devices with configuration of ITO/HIL (9000 rpm)/N,N′-diphenyl-N,N′-bis(1-naphthyl)-1,1′-biphenyl-4,4′′-diamine (NPB) (345 nm)/MoO₃ (10 nm)/Al (100 nm). The theoretical J_SCLC calculated from eqn (1).

where μ₀ is the zero-field mobility of NPB (μ₀ = 2.7 × 10⁻⁴ cm² V⁻¹ s⁻¹), ε₀ is the free-space permittivity (ε₀ = 8.85 × 10⁻¹⁴ C V⁻¹ cm⁻¹), ε_r is the relative dielectric constant (assumed to be 3.0), β is the Poole–Frenkel factor (β = 1.3 × 10⁻³ (cm V⁻¹)^−1/2) and F is the applied electric field. We found that all of devices exhibit much lower current densities compared to J_SCLC, meaning that hole injection contacts are not ohmic. The hole injection efficiency (η_INJ) was calculated from eqn (2).

η_INJ = J_INJ/J_SCLC (2)

where J_INJ is the measured steady-state current density and J_SCLC is the calculated theoretical space-charge-limited current density. Fig. 4b shows the hole injection efficiencies, η_INJversus the electric field. The injection efficiency for the device with 1 nm eMoO₃ HIL is less than 0.1, which is similar to the case of a 10 nm eMoO₃ HIL, reported by So.⁶ This contradicts Lee's statement of ohmic contact in the ITO–eMoO₃–NPB interface.²⁶ It has been reported that the eMoO₃ film is strongly n-type, and hole injection at the oxide–hole transporting layer (HTL) interface proceeds via electron extraction from the HTL HOMO through the low-lying conduction band of the eMoO₃.²⁷ As a HIL for an organic optoelectronic device, it is important to have favorable energy level alignment at the modified anode and HTL interface. In order to obtain the work function of the buffer layer, ultraviolet photoelectron spectroscopy (UPS) measurements were conducted to investigate the electronic structure of the sMoO₃ and PEDOT:PSS on the ITO substrates. As shown in Fig. S3,† the photoemission onset of the sMoO₃ is found at 15.85 eV, which corresponds to a work function of 5.35 eV. This is 0.25 and 1.05 eV higher than that of PEDOT:PSS (5.10 eV) and ITO (4.30 eV) respectively, indicating a reduced hole injection barrier at the sMoO₃–HTL interface compared to that at PEDOT:PSS–HTL and ITO–HTL interfaces. According to the literature,²⁰ the work function of eMoO₃ is 6.9 eV. It is commonly believed that an anode buffer layer with a high work function favors hole injection. However, our experiment reveals the hole injection efficiency for eMoO₃ is less than 0.1, which is consistent with the report by So.⁶ On the contrary, sMoO₃ shows a significantly improved hole injection at the sMoO₃–NPB interface. Therefore, the large difference in the hole injection between sMoO₃ and eMoO₃ cannot be explained by the difference in the work functions. It may originate from the complicated MoO₃–NPB interfacial energetic situation, such as interfacial dipole, charge transfer and sub-gap states. A further understanding is needed, while this is outside the scope of current work.

Fig. 4 Current density versus electric field characteristics (a) and hole injection efficiency as a function of electric field (b) of hole-only devices with various HILs.

The AFM image of the sMoO₃ HIL from the 2 mg ml⁻¹ solution shows a discontinuous film, leading to an unfavorable hole injection contact at the ITO–NPB interface, resulting in a moderate improvement of hole injection compared to the eMoO₃ HIL. Moreover, a gradual damage of such a discontinuous film with an electric field²⁸ might cause a serious roll-off of the injection efficiency, as depicted in Fig. 4b. However, the complete coverage of the sMoO₃ HIL from the 10 mg ml⁻¹ solution on ITO ensures a favorable and complete contact at the sMoO₃–NPB interface, so that it has a dramatically enhanced hole injection efficiency, compared to the sMoO₃ HIL from the 2 mg ml⁻¹ solution and the PEDOT:PSS HIL, indicating the substitution of the sMoO₃ HIL with optimum thickness for PEDOT:PSS. For the sMoO₃ HIL from the 20 mg ml⁻¹ solution, the device exhibits a reduced injection efficiency, due to the insulating nature of the thick sMoO₃ layer with about a 10 nm thickness, which consumes voltages and leads to a slight roll-off of the injection efficiency with increasing electric field.

Five conventional green phosphorescent OLEDs using sMoO₃ from 3 different concentrations, eMoO₃ and PEDOT:PSS as HILs were examined, with the configuration ITO/HIL/NPB (35 nm)/CBP:Ir(ppy)₃ (8%, 20 nm)/TPBI (60 nm)/LiF (0.5 nm)/Al (100 nm). Fig. 5a and b show the J–V and luminance–voltage (L–V) characteristics of the OLEDs with various HILs. Similar to the performance of the hole-only devices, the OLED with the sMoO₃ HIL from the 10 mg ml⁻¹ solution shows a higher current density than that with the sMoO₃ HIL from the 2 and 20 mg ml⁻¹ solutions and the eMoO₃ HIL. Moreover, the OLED with the sMoO₃ HIL exhibits a small leakage current, nearly one order of magnitude lower than that of the OLED with the PEDOT:PSS HIL.

Fig. 5 Current density–voltage (a), luminance–voltage (b), EL spectrum (c) and current efficiency–current density (d) characteristics of OLEDs with various HILs.

Fig. 5c shows the electroluminescent (EL) spectra at a current density of 5 mA cm⁻². Devices with the sMoO₃ and eMoO₃ HILs exhibit almost identical EL shapes with a minor difference of the intensity at 550 nm, while that with the PEDOT:PSS HIL shows an obviously higher shoulder at 550 nm, which may be due to the microcavity effect in the device. In Fig. 5d we present the current efficiency of devices with various HILs as a function of current density. Using the sMoO₃ HIL from the 2 mg ml⁻¹ solution, a maximum current efficiency of 54.8 cd A⁻¹ was obtained. This efficiency is obviously better than the devices with the eMoO₃ and PEDOT:PSS HILs. The enhancement of the efficiency is attributed to a more favorable hole injection and better balance of holes and electrons within the device with the sMoO₃ HIL. However, further enhancing the hole injection will induce excessive hole carriers and therefore result in a carrier imbalance within the device. According to the analysis of hole-only devices, the sMoO₃ HIL from solution with a high concentration exhibits a better hole injection than that from the 2 mg ml⁻¹ solution. Therefore, the efficiency of the device with the sMoO₃ HIL from the 10 or 20 mg ml⁻¹ solution decreases.

In addition to the application of the sMoO₃ as HIL in OLEDs, the effect of this anode buffer layer as a HIL in an OPV with a poly(3-hexylthiophene):[6,6]-phenyl C₆₁ butyric acid methyl ester bulk hetero junction (P3HT:PC₆₁BM BHJ) system was also investigated. Fig. 6 depicts the J–V characteristics of the OPV devices using the sMoO₃ and PEDOT:PSS HILs, and Table 1 summarizes the detailed results. The device with the sMoO₃ HIL from the 2 mg ml⁻¹ solution shows the best power conversion efficiency (PCE) of 2.35%, with an open-circuit voltage (V_OC) of 0.54 V, a short-circuit current (J_SC) of 7.26 mA cm⁻², and a fill factor (FF) of 60.0%, which is better than that using the PEDOT:PSS HIL (PCE = 2.03%). According to the literature,²⁰ an increase in the Mo⁵⁺ states in the sMoO₃ induces the recombination of photogenerated holes and unfavorable interfacial dipoles, leading to a drop in J_SC and the PCE. Therefore, the enhancement of the PCE in our experiment using sMoO₃ from the 2 mg ml⁻¹ solution is attributed to the reduced recombination, due to fewer Mo⁵⁺ states (Fig. 3) and enhanced hole collection through the sMoO₃ layer. On the other hand, the sMoO₃ layer from the 20 mg ml⁻¹ solution is thick enough (about 10 nm) to reduce the J_SC due to the insulating nature of MoO₃, resulting in a decreased PCE to 1.84%. The application of the sMoO₃ layer from the 2 and 20 mg ml⁻¹ solutions in OPVs has the same trend in the variation of hole injection as that in the OLEDs. However, for the sMoO₃ layer from the 10 mg ml⁻¹ solution, its hole injection ability from ITO to NPB in the OLED is the best, while that from P3HT:PC₆₁BM to ITO in the OPV declines. This difference may arise from the complicated sMoO₃–NPB and sMoO₃–P3HT:PC₆₁BM interfacial energetic situations, such as interfacial dipole, charge transfer, sub-gap states. This issue requires a further understanding that is outside the scope of our work, and a detailed investigation is currently underway.

Fig. 6 Current density–voltage characteristic of P3HT:PC61BM OPV with sMoO₃ and PEDOT:PSS HILs.

Table 1 Detailed parameters of device performance with sMoO₃ HILs from solutions with various concentrations

MoO3 solutions	V OC (V)	J SC (mA cm^−2)	FF (%)	PCE (%)
2 mg ml^−1	0.54	7.26	60.0	2.35
10 mg ml^−1	0.54	6.62	59.0	2.11
20 mg ml^−1	0.54	5.77	59.1	1.84
PEDOT:PSS	0.54	6.03	62.2	2.03

---

Experimental

Preparation of the MoO₃ solution: The MoO₃ solution was prepared by rapidly dissolving MoO₃ (422.8 mg) powder into ammonia–water (2.4 ml) without further stirring or a heating process. The saturated solution had a concentration of 175 mg ml⁻¹, and it could be diluted by deionized water to the desired concentration.

Characterization

The AFM images were obtained in room-temperature air by a Digital Instruments Multimode Nanoscope AFM with the tapping mode. The XPS test was carried out in a Kratos AXIS Ultra DLD ultrahigh vacuum surface analysis system with a monochromatic aluminum Kα source (1486.6 eV). For the AFM and XPS measurements, the sMoO₃ films were deposited on quartz glass substrates and silicon substrates, respectively.

Fabrication of OLEDs

OLEDs were fabricated on ITO-coated glass substrates. Prior to film deposition, the ITO glass substrates were cleaned successively using deionized water, acetone and isopropanol under an ultrasonic bath, then dried with nitrogen and pretreated with oxygen plasma. The sMoO₃ layer was spin coated from the solution prepared above at 9000 rpm for 60 s and annealed at 150 °C for 20 min. The PEDOT:PSS (VP AI 4083 from H. C. Stark, PEDOT4083) layer was spin coated at 4000 rpm for 60 s and annealed at 150 °C for 10 min. The other organic layers were thermally deposited within a high vacuum deposition system at a base pressure of 10⁻⁶ Torr through a shadow mask. The active area defined by the overlap of the ITO anode and the Al cathode was 3 mm × 3 mm. The deposition rate for the organic materials was monitored by a quartz-crystal monitor. The doping of the organic layers was made by thermal co-evaporation, controlled via two separate quartz-crystal monitors. The J–V and L–V characteristics were measured with a computer controlled Keithley 2400 Source Meter and BM-7A Luminance Colorimeter under ambient conditions.

Fabrication of OPV devices: the OPV devices were also fabricated on patterned ITO-coated glass substrates. The sMoO₃ and PEDOT:PSS HILs were spin coated as above. The P3HT and PC₆₁BM (1 : 1 weight ratio) solution dissolved in dichlorobenzene (20 mg ml⁻¹) was spin coated at 600 rpm for 120 s and annealed at 150 °C for 10 min in a nitrogen glove box. Then, a 20 nm thick Ca and a 80 nm thick Al layer were thermally evaporated as a cathode under a vacuum at a base pressure of 1 × 10⁻⁶ Torr. The active area was 0.04 cm². The J–V characteristics were measured by the Keithley 2400 Source Meter. The OPV performance was carried out using an Air Mass 1.5 Global (AM 1.5G) solar simulator with an irradiation intensity of 100 mW cm⁻².

Conclusions

In summary, we have proposed a simple synthetic route of MoO₃ anode buffer layer spin coated from an aqueous solution, which was prepared by rapidly dissolving MoO₃ powder in ammonia–water without a long-time stirring or heating process. The solution enables an ultra-smooth MoO₃ thin film with the appropriate electronic properties for hole injection in organic optoelectronic devices. A hole-only device study confirms that the sMoO₃ from the optimum solution provides an excellent hole injection property compared to that using the conventional PEDOT:PSS and eMoO₃. OLED and OPV devices utilizing the sMoO₃-modified anodes have also been investigated. The device performance depends on the concentration of the precursor solution, and the highest current efficiency and PCE are realized with the sMoO₃ HIL from a 2 mg ml⁻¹ solution. In a word, the sMoO₃ layer developed by the facile and low-cost technique is a feasible candidate for replacing the PEDOT:PSS anode buffer layer in organic electronics.

Acknowledgements

This work was supported by the National Natural Science Foundation of China (61377030) and the Science and Technology Commission of Shanghai Municipal (11PJ1404900, 12JC1404900). The authors would like to thank Mrs Huiqin Li and Mrs Limin Sun from the Instrumental Analysis Center of the SJTU, for helping with the AFM and XPS measurements.

Notes and references

1. R. H. Friend, R. W. Gymer, A. B. Holmes, J. H. Burroughes, R. N. Marks, C. Taliani, D. D. C. Bradley, D. A. D. Santos, J. L. Bredas, M. Logdlund and W. R. Salaneck, Nature, 1999, 397, 121.
2. C. W. Tang and S. A. Van Slyke, Appl. Phys. Lett., 1987, 51, 913.
3. R. Steim, F. R. Kogler and C. J. Brabec, J. Mater. Chem., 2010, 20, 2499.
4. H. Chen, J. Hou, S. Zhang, Y. Liang, G. Yang, Y. Yang, L. Yu, Y. Wu and G. Li, Nat. Photonics, 2009, 3, 649.
5. V. S. Gevaerts, A. Furlan, M. M. Wienk, M. Turbiez and R. A. J. Janssen, Adv. Mater., 2012, 24, 2130.
6. C. E. Small, S.-W. Tsang, J. Kido, S. K. So and F. So, Adv. Funct. Mater., 2012, 22, 3261.
7. M. Jorgensen, K. Norrman and F. C. Kreb, Sol. Energy Mater. Sol. Cells, 2008, 92, 686.
8. M. P. de Jong, L. J. van Ijzendoorn and M. J. A. de Voigt, Appl. Phys. Lett., 2000, 77, 2255.
9. F. So and D. Kondakov, Adv. Mater., 2010, 22, 3762.
10. K. Norrman, M. V. Madsen, S. A. Gevorgyan and F. C. Krebs, J. Am. Chem. Soc., 2010, 132, 16883.
11. A. W. Hains and T. J. Marks, Appl. Phys. Lett., 2008, 92, 023504.
12. K. R. Choudhury, J. Lee, N. Chopra, A. Gupta, X. Jiang, F. Amy and F. So, Adv. Funct. Mater., 2009, 19, 491.
13. Y. Sun, C. J. Takacs, S. R. Cowan, J. H. Seo, X. Gong, A. Roy and A. J. Heeger, Adv. Mater., 2011, 23, 2226.
14. D. Y. Kim, J. Subbiah, G. Sarasquesta, F. So, H. Ding, Irfan and Y. Gao, Appl. Phys. Lett., 2009, 95, 093304.
15. Irfan, H. Ding, Y. Gao, C. Small, D. Y. Kim, J. Subbiah and F. So, Appl. Phys. Lett., 2010, 96, 243307.
16. J. Qiu, Z. B. Wang, M. G. Helander and Z. H. Lu, Appl. Phys. Lett., 2011, 99, 153305.
17. H. You, Y. Dai, Z. Zhang and D. Ma, J. Appl. Phys., 2007, 101, 026105.
18. F. Liu, S. Shao, X. Guo, Y. Zhao and Z. Xie, Sol. Energy Mater. Sol. Cells, 2010, 94, 842.
19. S. Murase and Y. Yang, Adv. Mater., 2012, 24, 2459.
20. J. J. Jasieniak, J. Seifter, J. Jo, T. Mates and A. J. Heeger, Adv. Funct. Mater., 2012, 22, 2594.
21. C. Girotto, E. Voroshazi, D. Cheyns, P. Heremans and B. P. Rand, ACS Appl. Mater. Interfaces, 2011, 3, 3244.
22. J. Meyer, R. Khalandovsky, P. Görrn and A. Kahn, Adv. Mater., 2011, 23, 70.
23. M.-F. Xu, L.-S. Cui, X.-Z. Zhu, C.-H. Gao, X.-B. Shi, Z.-M. Jin, Z.-K. Wang and L.-S. Liao, Org. Electron., 2013, 14, 657.
24. F. Xie, W. C. H. Choy, C. Wang, X. Li, S. Zhang and J. Hou, Adv. Mater., 2013, 25, 2051.
25. P. S. Patil and R. S. Patil, Bull. Mater. Sci., 1995, 18, 911.
26. H. Lee, S. W. Cho, K. Han, P. E. Jeon, C.-N. Whang, K. Jeong, K. Cho and Y. Yi, Appl. Phys. Lett., 2008, 93, 043308.
27. M. Kröger, S. Hamwi, J. Meyer, T. Riedl, W. Kowalsky and A. Kahn, Appl. Phys. Lett., 2009, 95, 123301.
28. C.-C. Chang, S.-W. Hwang, C. H. Chen and J.-F. Chen, Jpn. J. Appl. Phys., 2004, 43, 6418.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c3tc31580k

---