The impact of post-deposition annealing on the performance of solution-processed single layer In₂O₃ and isotype In₂O₃/ZnO heterojunction transistors

Abstract

We study the influence of post-deposition annealing temperature on the morphology, chemical state and electrical properties of solution-processed single layer In₂O₃ and isotype In₂O₃/ZnO heterojunction transistors. Through careful optimisation of the material deposition and annealing conditions we demonstrate remarkable enhancement in the electron mobility of In₂O₃/ZnO heterojunction transistors, as compared to single layer In₂O₃ devices, with a maximum value of 48 cm² V⁻¹ s⁻¹.

---

The use of metal oxide semiconductors has gained significant interest in recent years due to their high charge carrier mobility and long-term reliability, transparency, and low processing temperature.^1,2 This unique combination of physical properties is paving the way for novel applications in the field of next generation optoelectronic devices by improving the performance and functionality of optical displays,³ solar cells,⁴ and light-emitting diodes.⁵ To date, several oxide semiconductor materials, such as ZnO,⁶ In₂O₃,⁷ Ga₂O₃⁸ or SnO₂⁹ have shown superior performance compared to the conventional hydrogenated amorphous silicon. Among these, In₂O₃ has received tremendous attention based on its unique ability to combine high electron mobilities with high optical transparency. Various deposition methods have been used for the fabrication of high performance In₂O₃ thin-film transistors, spanning from sputtering,¹⁰ reactive thermal evaporation,¹¹ and atomic-layer deposition,¹² to solution-based deposition techniques such as inkjet-printing¹³ or spraying¹⁴ for high-throughput, large-area fabrication. However, the maximum mobility values achieved through these methods are still far below the data reported using single crystal devices with 160 cm² V⁻¹ s⁻¹.¹⁵ Current work continues to address this issue by selective post-deposition treatments,¹⁶ or optimization of the dielectric/semiconductor interaction.¹⁷

Interestingly, the electronic structure of In₂O₃ has received significant interest recently¹⁸ and surface electron accumulation has now been observed in In₂O₃.^19,20 In addition, a novel approach using solution-processed bilayer heterostructures of In₂O₃ and ZnO with film thicknesses of a few nanometers has shown to remarkably enhance the charge carrier mobility in thin-film transistors due to the formation of a quasi-2D-electron-gas-like system at the oxide heterointerface.²¹ Here, the significant difference in the energy levels between ZnO and In₂O₃ in combination with the low-dimensionality of the individual layers appear to induce charge carrier migration towards the ultra-thin In₂O₃ layer, which not only serves as an excellent charge-transporting layer, but also exhibits clear energy quantization.²²

Here, we study the effects of different post-deposition annealing temperatures on the electronic properties of solution-processed isotype In₂O₃/ZnO heterojunctions and compare them to single layer In₂O₃ devices. Combination of field-effect electrical measurements with a detailed characterisation of the surface properties of the different layers by atomic force microscopy (AFM) and X-ray photoelectron spectroscopy (XPS), allows elucidation of the key physical properties required to achieve exceptional electron transport.

In₂O₃ and ZnO precursor solutions were prepared by dissolving 20 mg mL⁻¹ indium nitrate hydrate (In(NO₃)₃·xH₂O, 99.99%, Indium Corporation) in 2-methoxyethanol (99.8%, Sigma-Aldrich), and 10 mg mL⁻¹ zinc oxide (ZnO, 99.99%, Sigma-Aldrich) nanoparticles in aqueous ammonium hydroxide solution (NH₄OH, 50% v/v aqn. soln., Alfar Aesar) under constant stirring at room temperature. As-prepared In₂O₃ solutions were deposited by spin-coating at 4000 rpm for 30 s onto highly doped Si (Si⁺⁺) substrates with a 100 nm thermally grown SiO₂ layer acting as the common gate and dielectric. Prior to spin-coating, the substrates were ultrasonically cleaned in deionized water, acetone, and 2-propanol for 10 minutes each, and exposed to UV-ozone for 15 minutes to remove any unwanted organic residues and improve the surface wettability. The as-deposited In₂O₃ films were dried for 5 minutes at 100 °C to remove the solvent, and finally annealed at 200, 300, 400 and 600 °C for 1 hour. Bilayer structures were realized by subsequently depositing ZnO solutions using the same spin-coating and annealing steps. Finally, 40 nm-thick Al source/drain contacts were deposited through a shadow mask via thermal evaporation in high vacuum (<10⁻⁵ mbar). The channel length and width of the resulting transistors were 100 and 1000 μm, respectively.

The surface morphology of the semiconductor layers were investigated by AFM in intermittent contact mode using an Agilent 5500AFM system and respective images along with the height distributions for In₂O₃ single layer and In₂O₃/ZnO heterojunctions are shown in Fig. 1. The AFM measurements reveal a strong dependence of the surface morphology of In₂O₃ layers with annealing temperature. Specifically, the root-mean-square (rms) surface roughness increases 6-fold between 200 and 600 °C, and the height distribution shifts to higher values, indicating an increased grain growth in the In₂O₃ film [Fig. 1(b)]. A similar trend is observed for the In₂O₃/ZnO heterojunctions where the rms surface roughness is found to double for layers annealed at 600 °C as compared to those annealed at 200 °C.

Fig. 1 (a) AFM topography images of In₂O₃ single layers and In₂O₃/ZnO heterojunctions (deposited onto Si⁺⁺/SiO₂ substrates) annealed at various temperatures. The z value is the maximum height feature in the scan represented by (lighter shade). The corresponding height distributions including root-mean-square (rms) values for (b) In₂O₃ single layers and (c) In₂O₃/ZnO bilayers are shown.

To gain more insight into the impact of post-deposition annealing on the chemical properties, the surface chemistry and electronic structure of metal oxide layers was characterised using XPS. The spectra were recorded on a Thermo Scientific K-Alpha⁺ X-ray photoelectron spectrometer operating at 2 × 10⁻⁹ mbar base pressure. This system incorporates a monochromated, microfocused Al Kα X-ray source (hν = 1486.6 eV) and a 180° double focusing hemispherical analyser with a 2D detector. The X-ray source was operated at 6 mA emission current and 12 kV anode bias. Data were collected at 200 eV pass energy for survey, 20 eV pass energy for core level, and 15 eV pass energy for valence band spectra using an X-ray spot size of 400 μm². A flood gun was used to minimize sample charging.

All data were analysed using the Avantage software package.

Fig. 2(a) shows the In 3d core level of the In₂O₃ single layers which displays no significant changes with annealing temperature. All samples show a binding energy position of 444.5 eV for In 3d_5/2 and a spin–orbit-splitting (SOS) of 7.6 eV, values typical for In₂O₃.^23,24 The inelastic mean free path (IMFP) is a frequently used measure of the surface sensitivity of XPS. Using the Tanuma, Powell and Penn (TPP-2M) method incorporated within the QUASES software package,²⁵ the IMFP was calculated to be only 2.0 nm in the case of ZnO. Due to this fact in combination with the low-dimensional structure of the devices used in these experiments with thicknesses of the semiconductor layers in the range of 7 nm as measured by AFM and ellipsometry (see ESI,† Table S1) we assume moreover a more bulk-related analysis of the materials using XPS rather than only the surface. Since In₂O₃ is fully covered by a continuous layer of ZnO with a thickness of approximately 7 nm in the case of the bilayer devices one would not expect to detect indium in the XPS measurements. Nevertheless, indium was found to be present on all sample surfaces with its intensity increasing with the annealing temperature, while the position of the In 3d core level shifted to higher binding energy (see Fig. 2(c)). Fig. 3 shows the Zn 2p_3/2 and In 3d_5/2 core levels normalised to the Zn intensity and plotted over the same intensity range enabling a direct comparison of the peak height. In the samples annealed at 200 and 300 °C a comparable amount (4 at%) of indium is found which increases to 9 and 19 at% at 400 and 600 °C, respectively.

Fig. 2 In 3d and O 1s core level XPS spectra for (a) and (b) In₂O₃ single layer, and (c) and (d) In₂O₃/ZnO bilayer samples annealed at varying temperatures T_A. A higher binding energy shoulder observed on the O 1s spectra is marked with an asterisk.

Fig. 3 Zn 2p_3/2 and In 3d_5/2 core level XPS spectra of the In₂O₃/ZnO bilayer samples annealed at varying temperatures T_A plotted over the same intensity range. The table inset shows the annealing temperature and the observed Zn : In ratios in atomic percent (at%).

Using a simple mathematical model which takes into account the layer thickness, surface and interface roughness, indium and zinc interdiffusion, and the IMFP of ZnO, it can be shown that the increase of the In 3d core level intensity is caused by the increase of both surface roughness and interdiffusion at higher temperatures (see ESI†). The shift in energy position suggests a change in the direct chemical environment of In, e.g. differences in oxygen coordination of In caused by varying concentration of oxygen vacancies.

With an increase in the annealing temperature the Zn 2p_3/2 core level shifts consecutively to higher binding energy from 1021.5 to 1021.8 eV as shown in Fig. 3. It is not possible to distinguish between different Zn compounds using the core level alone as it shows very minute chemical shifts which cannot be resolved in XPS. Nevertheless, an increase in binding energy could indicate increased oxidation of the Zn. The Zn L₃M_4,5M_4,5 Auger line (see ESI,† Fig. S2) confirms the presence of ZnO, showing structural features and energy positions that agree well with data reported for ZnO in the literature.^26,27 Three features of the Auger line can be resolved: ³F at 495 eV, a combination of ¹G, ³P, ¹D at 498.4 eV and ¹S at 503.0 eV.

The O 1s core level for both single and heterojunction samples shows two resolved components (see Fig. 2(b) and (d)). The lower binding energy peaks at 530.0 eV for the single layer and 530.3 eV for the bilayer can be assigned to In₂O₃ and ZnO, respectively.^24,28 The contribution towards higher binding energy (labelled with an asterisk in the figures) which can be assigned to hydroxide groups in the layers decreases in intensity, with an increase in annealing temperature in both groups, and is more pronounced in the single layers. The ratios of the main peak to this higher binding energy contribution range from 50 : 50 (In₂O₃ single layer) and 74 : 26 (In₂O₃/ZnO bilayer) at 200 °C, to 72 : 28 (In₂O₃ single layer) and 86 : 14 (In₂O₃/ZnO heterojunction) at 600 °C. This result indicates that with increasing temperature a higher amount of unreacted hydroxide-related precursor material is converted into pure metal oxide material through condensation reactions. Hence the quality of the semiconductor in terms of purity and composition is improved as observed in previous investigations.^7,29 In addition, small contributions from the oxidised Si substrate are found in survey spectra (see ESI,† Fig. S3), which also contribute to the overall oxygen signal. In addition to the analysis of core and Auger levels, XPS was used to study the electronic structure of the samples by measuring valence band spectra as depicted in Fig. 4. The In₂O₃ single layer shows a valence band structure typical for In₂O₃ with 5 resolved features, denoted by Roman numerals I–V.^23,30 Whilst feature I is dominated by O 2p states, features II–IV involve mixing between the O 2p and In 5p (II and III) and 5s (IV) states. Feature V below the main VB is due to water adsorbed to the sample surface.³¹

Fig. 4 XPS valence band spectra of (a) the In₂O₃ single layer and (b) the In₂O₃/ZnO bilayer samples annealed at varying temperatures T_A. Roman numerals indicate major features of the valence bands. The position of the Fermi level E_F is indicated in the figures.

The spectra for the In₂O₃/ZnO heterostructures show, besides the shallow Zn 3d core level at the bottom of the valence band (BE = 10.5 eV), two features typical for ZnO.³² From all valence band spectra the separation Δ between the valence band maximum VB_max and the Fermi energy E_F was extracted by linear fits (see ESI,† Fig. S4 for exemplary fits for samples annealed at 600 °C). Table 1 summarises the values. For both sample groups the separation increases with annealing temperature with values being generally larger for the double layers.

Table 1 Values for the separation Δ between the valence band maximum V_Bmax and the Fermi energy E_F at different annealing temperatures (T_A) as extracted from XPS valence band spectra

T A (°C)	Δ In2O3 single layer (eV)	Δ In2O3/ZnO heterojunction (eV)
200	2.57	2.67
300	2.74	2.74
400	2.67	2.78
600	2.77	2.93

---

The impact of the morphological and chemical changes observed in In₂O₃ and In₂O₃/ZnO systems, on the electron transport properties, was studied using electrical field-effect measurements. Fig. 5 shows representative transfer characteristics measured for In₂O₃ and In₂O₃/ZnO transistors annealed at different temperatures in the range of 200–600 °C. Key transistor parameters including, electron mobility (μ_e), channel current on/off ratio (on/off), threshold voltage (V_Th) and switch-on voltage (V_on), are summarized in Table 2. The most important outcome of these measurements is the improved performance observed for all isotype In₂O₃/ZnO heterostructure devices when compared to single layer In₂O₃ transistors. For instance, hysteresis effects observed in In₂O₃ transistors annealed at 200 °C (Fig. 5(a)) are almost absent in heterojunction devices (Fig. 5(b)). Additionally, the V_Th for all In₂O₃/ZnO transistors shifts towards more negative voltages, indicating the presence of a higher concentration of mobile electrons. More importantly, the electron mobility increases to a maximum value of 47.9 cm² V⁻¹ s⁻¹ for transistors annealed at 400 °C, followed by a small reduction at higher annealing temperatures (Fig. 6). For comparison, mobility values of ZnO single layer devices annealed at different temperatures are also presented in Fig. 6.

Fig. 5 Transfer characteristics of thin-film transistors based on: (a) single layer In₂O₃, and (b) In₂O₃/ZnO heterojunction channels annealed at various temperatures. The insets show the respective device architecture employed.

Table 2 Electrical parameters of transistor devices based on In₂O₃ single layers and In₂O₃/ZnO bilayers annealed at various temperatures

Semiconductor	μ sat (cm^2 V^−1 s^−1)	On/off ratio	V Th (V)	V on (V)
In2O3 (200 °C)	0.3 (±0.1)	1 × 10^5	23.8	22.4
In2O3 (300 °C)	12.7 (±1.2)	2 × 10^6	10.3	1.2
In2O3 (400 °C)	30.1 (±1.3)	4 × 10^5	3.2	−5.3
In2O3 (600 °C)	37.9 (±2.1)	1 × 10^4	−0.3	−8.6
In2O3/ZnO (200 °C)	2.9 (±0.2)	5 × 10^6	9.4	3.2
In2O3/ZnO (300 °C)	17.8 (±0.4)	6 × 10^3	1.5	−6.2
In2O3/ZnO (400 °C)	47.9 (±0.9)	3 × 10^4	−3.1	−10.4
In2O3/ZnO (600 °C)	33.1 (±1.0)	2 × 10^3	1.8	−6.4

---

Fig. 6 Evolution of the field-effect electron mobility (μ_e) measured in saturation for transistor based on single layers of In₂O₃ and ZnO as well as In₂O₃/ZnO heterojunctions as a function of annealing temperature.

The clockwise hysteresis observed in the transfer curves of transistors annealed at lower temperatures is attributed to charge carrier trapping at the dielectric/semiconductor interface due to the presence of chemical defects.³³ To this end, the increased amount of hydroxyl groups in the low temperature annealed layers observed in the XPS measurements is believed to act as electron trapping sites causing the hysteresis.³⁴ With increasing annealing temperature hysteresis becomes less pronounced due to the reduction of the hydroxyl group concentration. In case of the heterojunction devices annealed at 200 °C we believe that the presence of a higher concentration of mobile electrons results in the filling of the trap states which in turn significantly reduces the operating hysteresis effects as compared to single layer In₂O₃ devices.

Single layer In₂O₃ transistors exhibit a steady increase in μ_e with increasing temperature. This is attributed to an improved conversion of the precursor material to In₂O₃ at higher temperatures, as indicated by the XPS measurements. The same holds for In₂O₃/ZnO heterojunctions for annealing temperatures in the range from 200 °C to 400 °C. However, in the latter type of devices the μ_e drops below that of single layer In₂O₃ transistors when annealed at 600 °C (Fig. 6), and V_Th shifts back to more positive gate voltages (Fig. 5(b)). The deterioration of electrical performance is thought to originate from the increase in heterointerface roughness and the possible interdiffusion of indium and zinc atoms, as indicated by the AFM and XPS analysis. The interdiffusion processes most likely alters the local electronic properties of the semiconductors, whereas the increased In₂O₃ layer roughness degrades the electron transport at the vicinity of the all-important oxide heterointerface. To this end, Lin et al., has recently argued that the formation of a high quality In₂O₃/ZnO heterointerface is at the heart of the enhanced μ_e observed.²¹ In the latter study, it was shown that the large conduction band offset between In₂O₃ and ZnO leads to the formation of a quasi-2D electron gas in a process similar to that observed in high electron mobility transistors based on III–V heterostructure channels.²² The confined charge shifts the Fermi energy closer to the conduction band edge which in turn allows access to extended high mobility states. However, further dedicated effort would be required to elucidate the exact nature of the In₂O₃/ZnO heterojunctions, especially those subjected to high temperature post-deposition annealing steps.

Finally, for all types of devices higher annealing temperatures lead to higher channel conductivity due to increased mobility and as a consequence to higher lateral leakage currents which reduce the channel on/off current ratio. It should be noted that the active semiconductor channels in the devices discussed in Fig. 5 and 6 were not patterned and as such the relatively large lateral gate leakage currents observed directly influence the off-state current of these devices. However, this issue can be rectified by patterning the transistor channel using standard photolithography. The latter step reduces the channel off-current by nearly 2 orders of magnitude leading to a dramatic improvement in the on/off current ratio (see ESI,† Fig. S5).

In summary, single layer In₂O₃ and isotype In₂O₃/ZnO heterojunction transistors were successfully grown via spin coating in air and the influence of post-deposition annealing temperature was investigated. In the case of In₂O₃ transistors, it was shown that the electron mobility increases continuously with increasing annealing temperature up to 600 °C reaching a maximum value of 37.9 cm² V⁻¹ s⁻¹. The electron mobility could be further increased to 47.9 cm² V⁻¹ s⁻¹ by sequentially depositing In₂O₃ and ZnO to form low-dimensional (<14 nm-thick) In₂O₃/ZnO heterostructure channel transistors. Unlike single layer In₂O₃ transistors, however, the improved performance of the latter devices was obtained at the significantly lower annealing temperature of 400 °C. Analysis of the heterojunction channel morphology following post-deposition thermal annealing, reveals that the surface roughness and possible elemental interdiffusion/alloying/recrystallization processes may all play critical roles. On the basis of these results, we conclude that improvements in deposition methods coupled with an improved understanding of their chemical and physical characteristics, could enable further improvement in the performance of heterojunction metal oxide transistors to levels on par with, or beyond, incumbent TFT technologies but with the added advantage of low-cost, large-area manufacturing.

Acknowledgements

This work was funded by the People Programme (Marie Curie Actions) of the European Union's Framework Programme Horizon2020: “Flexible Complementary Hybrid Integrated Circuits” (FlexCHIC), grant agreement no. 658563. A. R. and D. J. P. acknowledge support from the EPSRC (EP/K004913/1 and EP/J021199/1). D. J. P. acknowledges support from the Royal Society (UF100105).

References

1. E. Fortunato, P. Barquinha and R. Martins, Adv. Mater., 2012, 24, 2945.
2. S. R. Thomas, P. Pattanasattayavong and T. D. Anthopoulos, Chem. Soc. Rev., 2013, 42, 6910.
3. Y. Fukui, M. Shibata, Y. Tanaka, K. Okumoto, K. Morita, K. Otake, A. K. Tripathi, B. van der Putten, J.-L. van der Steen, K. Tempelaars, L. van Leuken, F. Li, I. Yakimets, G. Gelinck, K. Myny, S. Smout, M. Willegems, S. Schols, S. Steudel, J. Genoe and P. Heremans, SID Symp. Dig. Tech. Pap., 2013, 44, 2168.
4. Y. S. Lee, J. Heo, S. C. Siah, J. P. Mailoa, R. E. Brandt, S. B. Kim, R. G. Gordon and T. Buonassisi, Energy Environ. Sci., 2013, 6, 2112.
5. W. Z. Xu, Z. Z. Ye, Y. J. Zeng, L. P. Zhu, B. H. Zhao, L. Jiang, J. G. Lu, H. P. He and S. B. Zhang, Appl. Phys. Lett., 2006, 88, 173506.
6. H. Bong, W. H. Lee, D. Y. Lee, B. J. Kim, J. H. Cho and K. Cho, Appl. Phys. Lett., 2010, 96, 192115.
7. Y. H. Hwang, J.-S. Seo, J. M. Yun, H. J. Park, S. Yang, S.-H. Ko Park and B.-S. Bae, NPG Asia Mater., 2013, 5, e45.
8. S. R. Thomas, G. Adamopoulos, Y.-H. Lin, H. Faber, L. Sygellou, E. Stratakis, N. Pliatsikas, P. A. Patsalas and T. D. Anthopoulos, Appl. Phys. Lett., 2014, 105, 092105.
9. G. Huang, L. Duan, G. Dong, D. Zhang and Y. Qiu, ACS Appl. Mater. Interfaces, 2014, 6, 20786.
10. J. H. Noh, S. Y. Ryu, S. J. Jo, C. S. Kim, S.-W. Sohn, P. D. Rack, D.-J. Kim and H. K. Baik, IEEE Electron Device Lett., 2010, 31, 567.
11. Dhananjay and C.-W. Chu, Appl. Phys. Lett., 2007, 91, 132111.
12. H. I. Yeom, J. B. Ko, C.-S. Hwang, S. Cho and S.-H. K. Park, SID Symp. Dig. Tech. Pap., 2015, 46, 52.
13. J. S. Lee, Y.-J. Kwack and W.-S. Choi, ACS Appl. Mater. Interfaces, 2013, 5, 11578.
14. H. Faber, Y.-H. Lin, S. R. Thomas, K. Zhao, N. Pliatsikas, M. A. McLachlan, A. Amassian, P. A. Patsalas and T. D. Anthopoulos, ACS Appl. Mater. Interfaces, 2015, 7, 782.
15. R. L. Weiher, J. Appl. Phys., 1962, 33, 2834.
16. S.-Y. Han, G. S. Herman and C.-h. Chang, J. Am. Chem. Soc., 2011, 133, 5166.
17. H. S. Kim, P. D. Byrne, A. Facchetti and T. J. Marks, J. Am. Chem. Soc., 2008, 130, 12580.
18. A. Walsh, J. L. F. Da Silva, S.-H. Wei, C. Körber, A. Klein, L. F. J. Piper, A. DeMasi, K. E. Smith, G. Panaccione, P. Torelli, D. J. Payne, A. Bourlange and R. G. Egdell, Phys. Rev. Lett., 2008, 100, 167402.
19. P. D. C. King, T. D. Veal, D. J. Payne, A. Bourlange, R. G. Egdell and C. F. McConville, Phys. Rev. Lett., 2008, 101, 116808.
20. K. H. L. Zhang, R. G. Egdell, F. Offi, S. Iacobucci, L. Petaccia, S. Gorovikov and P. D. C. King, Phys. Rev. Lett., 2013, 110, 056803.
21. Y.-H. Lin, H. Faber, J. G. Labram, E. Stratakis, L. Sygellou, E. Kymakis, N. A. Hastas, R. Li, K. Zhao, A. Amassian, N. D. Treat, M. McLachlan and T. D. Anthopoulos, Adv. Sci., 2015, 2, 2198.
22. J. G. Labram, Y.-H. Lin and T. D. Anthopoulos, Small, 2015, 11, 1613.
23. C. Körber, V. Krishnakumar, A. Klein, G. Panaccione, P. Torelli, A. Walsh, J. L. F. Da Silva, S.-H. Wei, R. G. Egdell and D. J. Payne, Phys. Rev. B: Condens. Matter Mater. Phys., 2010, 81, 165207.
24. R. Henríquez, E. Muñoz, E. A. Dalchiele, R. E. Marotti, F. Martín, D. Leinen, J. R. Ramos-Barrado and H. Gómez, Phys. Status Solidi A, 2013, 210, 297.
25. S. Tanuma, C. J. Powell and D. R. Penn, Surf. Interface Anal., 1994, 21, 165.
26. S. P. Kowalczyk, R. A. Pollack, F. R. McFeely, L. Ley and D. A. Shirley, Phys. Rev. B: Solid State, 1973, 8, 2387.
27. G. Schön, J. Electron Spectrosc. Relat. Phenom., 1973, 2, 75.
28. V. Dimitrov and T. Komatsu, J. Solid State Chem., 2002, 163, 100.
29. C.-H. Choi, S.-Y. Han, Y.-W. Su, Z. Fang, L.-Y. Lin, C.-C. Cheng and C.-H. Chang, J. Mater. Chem. C, 2015, 3, 854.
30. P. D. C. King, T. D. Veal, F. Fuchs, C. Y. Wang, D. J. Payne, A. Bourlange, H. Zhang, G. R. Bell, V. Cimalla, O. Ambacher, R. G. Egdell, F. Bechstedt and C. F. McConville, Phys. Rev. B: Condens. Matter Mater. Phys., 2009, 79, 205211.
31. S. Eriksen, P. D. Naylor and R. G. Egdell, Spectrochim. Acta, Part A, 1987, 43A, 1535.
32. P. D. C. King, T. D. Veal, A. Schleife, J. Zúñiga-Pérez, B. Martel, P. H. Jefferson, F. Fuchs, V. Muñoz-Sanjosé, F. Bechstedt and C. F. McConville, Phys. Rev. B: Condens. Matter Mater. Phys., 2009, 79, 205205.
33. C. Zhu, A. Liu, G. Liu, G. Jiang, Y. Meng, E. Fortunato, R. Martins and F. Shan, J. Mater. Chem. C, 2016, 4, 10715.
34. J.-S. Seo, J.-H. Jeon, Y. H. Hwang, H. Park, M. Ryu, S.-H. K. Park and B.-S. Bae, Sci. Rep., 2013, 3, 2085.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c6tc04907a

---