Surface defect engineering: gigantic enhancement in the optical and gas detection ability of metal oxide sensor

Abstract

In this work, the detection ability of nanosensors can be improved extraordinarily via surface defect engineering. A kinked SnO_2−X/SnO₂ nanostructure was fabricated by tuning the oxygen flow, and this kinked SnO_2−X/SnO₂ nanostructure was used to study the mechanism of surface defect (oxygen vacancy, V_O) effects via electric measurements. For UV light sensing, the response of the SnO_2−X NW device is always better than the SnO₂ NW device, and is two orders higher under pure O₂ surrounding conditions. The detection mechanism can be clarified by changing the detection environment (oxygen concentration) and the UV light detection sensitivity can be improved by increasing the surface V_O density. Furthermore, the SnO_2−X NW device is very sensitive to its surrounding environment due to the high surface V_O density. Hence, CO/O₂ alternate-detection was used to verify our hypothesis; the results show that the SnO_2−X NW device presents great detection abilities, compared with the SnO₂ NW device. The sensitivity of the SnO_2−X NW device is two orders higher and the reset/response time is faster, compared with the SnO₂ NW device. To verify this hypothesis, the polycrystalline structure was fabricated to prove that the detection ability of metal oxide nanosensors can be improved gigantically by increasing surface defect amounts.

---

Introduction

Recently, nanomaterials, such as nanowires,^1–4 nanobelts^5–7 and nanotubes,^8–10 and polymer materials,¹¹ have been extensively studied and utilized to form nano-devices for different applications.^12–18 Metal oxide nanomaterials, because of the high surface-to-volume ratio and surface defects of their nanostructures, are quite sensitive to variation in the environment.^19–26 Many articles reported that Schottky contact mechanisms can enhance the performance of metal oxide nanomaterials, such as in piezotronic devices,^27,28 nanogenerators^29–32 and nanosensors.^20,33–39

For sensor applications, using Schottky contacts as control gates can improve the sensitivity, and the response and reset time.^39–42 Besides, the defect amount of metal oxide nanomaterials is the main parameter for sensor applications.^43–49 The surface defects⁵⁰ and the oxygen vacancies⁵¹ of metal oxide nanomaterials have been reported and discussed widely. Many research works tried to control the defect amount and level of metal oxides in chemical or physical ways to create other possibilities or applications.^52–60 Besides, the signal output of Schottky contact device has strong relation to the surface defect.

Tin dioxide, SnO₂, has potential for ultraviolet (UV) light detection applications, due to its large band gap.⁶¹ Otherwise, SnO₂ is also a candidate for gas detection because of its oxygen vacancies.^{18,39,48,57,58} Besides, SnO₂ can also be used as a photocatalyst and in solar-cell applications.^55,62–64 In this research work, tin oxide nanomaterials were fabricated with different levels of defects, and the devices were formed as Schottky gate devices for mechanism investigation. The relationship between the detection ability and defect amount of the tin oxide nanosensor can be figured out by using the Schottky contact devices made with this kinked structure.

Results and discussion

The detailed analyses of the tin oxide nanostructure (NS) can be obtained from the TEM images, as shown in Fig. 1. The growth direction of the tin oxide NS can be identified from the diffraction pattern and high resolution TEM images, as shown in Fig. 1(a)–(c). Due to our control of the synthesis, the oxygen composition of both sides would be different; a lower oxygen composition was detected for the NS growth without oxygen flow, as shown in Fig. 1(d)–(f). The schematic and SEM images of the nanodevice can be seen in Fig. 1(g). The Pt was deposited via focused ion beam (FIB) to form ohmic contacts; the free contacts on both sides can form Schottky contacts as detection units.⁶⁵ The composition of each side of the nanodevice can be identified via EDS analysis of the system, as illustrated in Fig. 1(h).

Fig. 1 (a) Low magnification TEM image of a kinked SnO_2−X/SnO₂ nanostructure with the growth orientation changed from (200) to (002); the growth orientation is indicated by the selected area electron diffraction (SAED) pattern in the inset. (b) and (c) High resolution images and lattice constants of SnO₂ and SnO_2−X, respectively. (d) The Sn and O concentrations of a kinked SnO_2−X/SnO₂ nanowire analyzed via EDS element line scan. (e) and (f) The elemental mapping of Sn and O, respectively. (g) Schematic diagram and SEM image of the SnO_2−X/SnO₂ nanowire Schottky contacted device. (h) EDS analysis and the atomic percentage data of SnO_2−X/SnO₂ NWs.

For photodetection, the sensitivity of the SnO_2−X NW device for UV (254 nm) light detection is higher than the SnO₂ NW device, as shown in Fig. 2(a), because the dark current (I_D) of the SnO_2−X NW device is smaller than the I_D of the SnO₂ NW device and the photo current (I_P) of the SnO_2−X NW device is larger than the I_P of the SnO₂ NW device. The measurement was taken at room temperature, 25 °C. The UV detection ability of both of the devices is related to the oxygen concentration of the sensing environment, as illustrated in Fig. 2(b). The sensitivity of the SnO_2−X NW device is always higher than the SnO₂ NW device, no matter what the oxygen concentration percentage is. The UV detection sensitivity of both of the NW devices will be increased when the oxygen concentration is increasing. The sensing mechanism is illustrated in Fig. 2(c) and (d) for each device.

Fig. 2 (a) 254 nm UV detection performance in a 40% oxygen gas environment (with 40% oxygen and 60% nitrogen). (b) The sensitivities at different oxygen concentrations of the SnO_2−X and SnO₂ NW devices. (c) and (d) The mechanism diagrams of the SnO_2−X and SnO₂ NW devices and oxygen molecule interaction at the interface. Step (i) and (ii) represent the adsorption of oxygen molecules. Step (iii) shows the desorption of oxygen molecules via UV illumination. The SnO_2−X NW device has higher sensitivity because its V_O is higher. When the UV light was off, more oxygen molecules were trapped at the Schottky contact interface to form O₂⁻ and raise the Schottky barrier height (SBH) to reduce the current, so the I_D of SnO_2−X NW device was lower. But turn on the UV light, and electron–hole pairs would be generated and O₂⁻ would be desorbed by the hole. Because the hole would combine with the O₂⁻ to form O_2(g) and desorb, that would reduce the SBH to increase the I_P. (e) The variation of SBH from a vacuum environment to pure oxygen environment (with 100% oxygen) of the SnO_2−X and SnO₂ NW devices.

The SnO_2−X NW device has an impressive improvement in sensitivity because the density of surface oxygen vacancies (V_O) is higher. When the UV light is off, the SnO_2−X NW device would trap more oxygen molecules at the Schottky contact interface to form O₂⁻ and raise the Schottky barrier height (SBH) to reduce the current, so the I_D of the SnO_2−X NW device would be lower than the SnO₂ NW device. But when the UV light is on, electron–hole pairs would be generated and O₂⁻ would be desorbed by the hole. Because the hole would combine with the O₂⁻ to form O_2(g) and desorb, that would reduce the SBH to increase the I_P. Besides, the lone pair electrons will increase the I_P of the SnO_2−X NW device, as compared with the SnO₂ NW device, the amount of lone pair electrons in the SnO_2−X NW device is larger, as shown in Fig. 2(c) and (d). Due to the above-mentioned results, the response can be improved no matter what the oxygen concentration by using the SnO_2−X NW device, compared with the SnO₂ NW device. The SBH variation also can be analyzed from the electrical measurements, the ΔΦ (SBH variation) would be stable for the SnO₂ NW device after 40% oxygen concentration. But for the SnO_2−X NW device, the ΔΦ would increase with the increasing concentration of oxygen, as shown in Fig. 2(e). The SBH can be approximately presented by the following equation:⁶⁶

I_r = AA*T²exp[−qΦ_eff/kT]

where I_r is the reverse bias we set, A is the contact area between the SnO₂ NW and Pt electrode, A* is the effective Richardson constant, q is the electric charge, Φ_eff is the SBH, and k and T are the Boltzmann constant and system temperature, respectively.

The ΔΦ of both of the devices is almost the same for an oxygen concentration of around 20%; but when the oxygen concentration increases to above 40%, the ΔΦ would be different. This result symbolizes that the surface V_O density of the SnO_2−X NW device is higher than that of the SnO₂ NW device based on the mechanism illustrated in Fig. 2(c) and (d), which is also consistent with the TEM and SEM analyses. The SBH differences can be measured and analyzed between the SnO_2−X and SnO₂ NW devices via oxygen concentration variation, and the difference will be enlarged as the oxygen concentration increases, as shown in Table 1 (detailed measurements can be seen in Fig. S1†). The above result shows that the tin oxide based NW device would be very sensitive to its surrounding environment, especially for oxygen concentration. So we can presume that the tin oxide based NW device will have great detection ability for gas, especially for the SnO_2−X NW device.

Table 1 The variation and difference of the two Schottky barrier heights in different O₂ concentration environments

Oxygen concentration (y%)	ΔΦy–(y-20%)		ΦSnO2−X− ΦSnO2
	SnO2−X	SnO2
0%	—	—	—
20%	85 meV	83 meV	2 meV
40%	27 meV	21 meV	8 meV
60%	6 meV	3 meV	11 meV
80%	18 meV	5 meV	25 meV
100%	26 meV	4 meV	47 meV

---

From our data, the gas detection ability of the SnO_2−X NW device is better than the SnO₂ NW device, as shown in Fig. 3. For gas detection, the O₂-detection current density (J_O2) was used for the base current (J₀ = J_O2) and the CO-detection current density (J_CO) was used for the reaction current (J = J_CO). The sensitivity (S) can be defined as S = ΔJ/J₀, and ΔJ = J_CO − J_O2. For sensing, the CO detection signal output of the SnO_2−X NW device is larger than the SnO₂ NW device, and the J₀ of the SnO_2−X NW device is smaller than that of the SnO₂ NW device, as can be seen in Fig. 3(a). This means that the sensitivity of the SnO_2−X NW device is higher than the SnO₂ NW device for the detection of each individual CO concentration, as shown in Fig. 3(b). The repetition of both devices for 2 ppm CO detection can be seen in Fig. S3.† The detection mechanism is described in Fig. 3(c). Because of the high surface V_O density of the SnO_2−X NW device, the J_O2 can be reduced due to the SBH raising on oxygen adsorption. Considering CO/O₂ alternate-detection, the SBH variation of the SnO_2−X NW device is larger than the SnO₂ NW device. Based on this surface V_O density difference, the improvements in the gas detection abilities (the sensitivity, and the response and reset time) of the SnO_2−X NW device are gigantic, compared with the SnO₂ NW device. From Table 2, we can see higher sensitivity and a faster reset time on increasing the CO concentration by using the SnO_2−X NW device (detailed data can be seen in Fig. S2†). That is because a higher CO concentration can reduce the SBH more, leading to a higher current (J_CO); so when the O₂ flows in to raise the SBH, the current (J_O2) will be decreased immediately. But the response time is related to the influence of gas-surface interactions, and the interactions most depend on the temperature effect. So the response time will be unaffected by the different CO concentrations. The results can support our hypothesis that increasing surface defects can improve the response. So we synthesized another polycrystalline nanostructure to reconfirm our hypothesis. The polycrystalline structure we use is SnO₂ (p-SnO₂) and ZnO (p-ZnO) nanowire to compare with single-crystalline SnO₂ (s-SnO₂). The sensitivity enhancements of p-SnO₂ and p-ZnO are two orders larger than that of s-SnO₂ for UV light detection; the sensitivities of s-SnO₂, p-SnO₂ and p-ZnO are 9.4%, 132.3% and 130.7%, as seen in Fig. S4.†

Fig. 3 (a) Detection performance of O₂ for sensing different CO concentrations at 200 °C operation temperature. (b) The sensitivities of the SnO_2−X NW device are higher than the SnO₂ NW device for CO detection at different concentrations. (c) The detection mechanism of the SnO_2−X and SnO₂ NW devices. Considering CO/O₂ alternate-detection, the SBH variation of the SnO_2−X NW device is larger than the SnO₂ NW device (ΔΦ_SnO2−X > ΔΦ_SnO2), due to the high surface V_O density of the SnO_2−X NW device.

Table 2 The sensitivity, response time and reset time of CO gas sensing at different concentrations

CO concentration		Sensitivity	τresponse	τreset
2 ppm	SnO2−X	490 ± 43%	169 s	221 s
	SnO2	75 ± 7%	180 s	238 s
50 ppm	SnO2−X	27 549 ± 1521%	234 s	7 s
	SnO2	254 ± 53%	451 s	9 s
100 ppm	SnO2−X	43 657 ± 2232%	420 s	2 s
	SnO2	541 ± 124%	630 s	3 s
200 ppm	SnO2−X	46 726 ± 780%	225 s	<1 s
	SnO2	551 ± 76%	341 s	2 s

---

Experiment

Asymmetric kinked SnO₂/SnO_2−X nanowires

The kinked SnO_2−X/SnO₂ NWs were synthesized on silicon substrates at 850 °C via the general catalyst free thermal evaporation method using SnO₂ and carbon powders as sources in a horizontal quartz tube connected to a vacuum pump and a programmable mass flow controller (MFC). The source was placed in an alumina boat located at the high temperature zone of 1000 °C. The substrates were then positioned in the source downstream. After the tube had been sealed and evacuated to the base pressure, a carrier gas, Ar/O₂, was kept flowing through the tube to direct the deposition process. Initially, SnO₂ NWs were grown at 100 standard cubic centimeters per minute (sccm) Ar/O₂ mixture gas with the volume ratio of 5 : 1 for 15 min to create the stoichiometric SnO₂ segment, as shown in Fig. S5.† After reaction, the oxygen input was turned off, and the background pressure was kept at 4 torr to create the SnO_2−X segment. The growth plane of the kinked NW shifts from (200) to (002) owing to the external pressure perturbation.

Conclusions

In this research work, we have demonstrated that a kinked SnO_2−X/SnO₂ nanostructure can be formed by controlling the oxygen flow. We used this kinked SnO_2−X/SnO₂ nanostructure to form Schottky contact devices and to study the effects of the surface V_O density via optical and gas condition electric measurements. For UV light detection, the gigantic enhancement in the sensitivity of the SnO_2−X NW device is because the V_O can generate more electron–hole pairs to improve the photo response. We also studied the mechanism by controlling the detection environment (oxygen concentration) and we figured out that the ΔΦ is related to the oxygen concentration. Based on this result, we found that the SnO_2−X NW device is quite sensitive to its surrounding environment. So we used CO/O₂ alternate-detection to verify our hypothesis, and the results show that the SnO_2−X NW device presents great detection abilities, compared with the SnO₂ NW device. The sensitivity of the SnO_2−X NW device is two orders larger than that of the SnO₂ NW device when the CO concentration is over 50 ppm. The response and reset time both improved by using the SnO_2−X NW device. We proved that increasing the surface defects and using Schottky contacts can improve the detection ability of metal oxide nanosensors extraordinary. Based on the above idea we also designed another, polycrystalline, nanostructure to reconfirm the hypothesis. The sensitivities of p-SnO₂ and p-ZnO devices are 132.3% and 130.7%, and all perform better than s-SnO₂, at 9.4%. We can use the technologies of material science engineering and physics to design high-resolution and fast monitor speed nanosensors.

Acknowledgements

This research was supported by the Ministry of Science and Technology, Taiwan, under grants NSC-101-2112-M-032-004-MY3 and MOST-104-2112-M-032-003-006.

References

1. S. Xu and Z. Wang, Nano Res., 2011, 4, 1013–1098.
2. P. Gao and Z. Wang, in Scanning Microscopy for Nanotechnology, Springer New York, 2007, ch. 13, pp. 384–426.
3. Ü. Özgür, Y. I. Alivov, C. Liu, A. Teke, M. A. Reshchikov, S. Doğan, V. Avrutin, S. J. Cho and H. Morkoç, J. Appl. Phys., 2005, 98, 041301.
4. Z. Wang, Chin. Sci. Bull., 2009, 54, 4021–4034.
5. J. Wang, D. N. Tafen, J. P. Lewis, Z. Hong, A. Manivannan, M. Zhi, M. Li and N. Wu, J. Am. Chem. Soc., 2009, 131, 12290–12297.
6. K. Shankar, J. I. Basham, N. K. Allam, O. K. Varghese, G. K. Mor, X. Feng, M. Paulose, J. A. Seabold, K.-S. Choi and C. A. Grimes, J. Phys. Chem. C, 2009, 113, 6327–6359.
7. O. Lupan, T. Braniste, M. Deng, L. Ghimpu, I. Paulowicz, Y. K. Mishra, L. Kienle, R. Adelung and I. Tiginyanu, Sens. Actuators, B, 2015, 221, 544–555.
8. G. She, X. Huang, L. Jin, X. Qi, L. Mu and W. Shi, Small, 2014, 10, 4685–4692.
9. S. Y. Yang, W. Choi and H. Park, ACS Appl. Mater. Interfaces, 2015, 7, 1907–1914.
10. D. H. Kim, Y.-S. Shim, J.-M. Jeon, H. Y. Jeong, S. S. Park, Y.-W. Kim, J.-S. Kim, J.-H. Lee and H. W. Jang, ACS Appl. Mater. Interfaces, 2014, 6, 14779–14784.
11. Y. Gao, Y. Zhao, L. Qiu, Z. Guo, D. O’Hare and Q. Wang, Polym. Compos., 2015 DOI:10.1002/pc.23764.
12. X. Wang, N. Aroonyadet, Y. Zhang, M. Mecklenburg, X. Fang, H. Chen, E. Goo and C. Zhou, Nano Lett., 2014, 14, 3014–3022.
13. L. Peng, L. Hu and X. Fang, Adv. Funct. Mater., 2014, 24, 2591–2610.
14. S. Jeong, M. Choe, J.-W. Kang, M. W. Kim, W. G. Jung, Y.-C. Leem, J. Chun, B.-J. Kim and S.-J. Park, ACS Appl. Mater. Interfaces, 2014, 6, 6170–6176.
15. A. Pescaglini, A. Martín, D. Cammi, G. Juska, C. Ronning, E. Pelucchi and D. Iacopino, Nano Lett., 2014, 14, 6202–6209.
16. X. Yang, R. Liu, C. Du, P. Dai, Z. Zheng and D. Wang, ACS Appl. Mater. Interfaces, 2014, 6, 12005–12011.
17. N. A. Kyeremateng, ChemElectroChem, 2014, 1, 1442–1466.
18. M. Law, H. Kind, B. Messer, F. Kim and P. Yang, Angew. Chem., 2002, 114, 2511–2514.
19. G. F. Fine, L. M. Cavanagh, A. Afonja and R. Binions, Sensors, 2010, 10(6), 5469–5502.
20. Z. Yang, X. Dou, S. Zhang, L. Guo, B. Zu, Z. Wu and H. Zeng, Adv. Funct. Mater., 2015, 25, 4039–4048.
21. H. S. Woo, C. H. Kwak, J. H. Chung and J. H. Lee, ACS Appl. Mater. Interfaces, 2014, 6, 22553–22560.
22. S. W. Choi, A. Katoch, G. J. Sun, J. H. Kim, S. H. Kim and S. S. Kim, ACS Appl. Mater. Interfaces, 2014, 6, 8281–8287.
23. A. Katoch, J. H. Kim and S. S. Kim, ACS Appl. Mater. Interfaces, 2014, 6, 21494–21499.
24. H. Zhou, R. Deng, Y.-F. Li, B. Yao, Z.-H. Ding, Q.-X. Wang, Y. Han, T. Wu and L. Liu, J. Phys. Chem. C, 2014, 118, 6365–6371.
25. F. Hernández-Ramírez, J. Rodríguez, O. Casals, E. Russinyol, A. Vilà, A. Romano-Rodríguez, J. R. Morante and M. Abid, Sens. Actuators, B, 2006, 118, 198–203.
26. F. Hernández-Ramírez, A. Tarancón, O. Casals, J. Arbiol, A. Romano-Rodríguez and J. R. Morante, Sens. Actuators, B, 2007, 121, 3–17.
27. Z. L. Wang, in Piezotronics and Piezo-Phototronics, Springer, Berlin, Heidelberg, 2012, ch. 1, pp. 1–17.
28. R. Yu, C. Pan, J. Chen, G. Zhu and Z. L. Wang, Adv. Funct. Mater., 2013, 23, 5868–5874.
29. W. Seung, M. K. Gupta, K. Y. Lee, K.-S. Shin, J.-H. Lee, T. Y. Kim, S. Kim, J. Lin, J. H. Kim and S.-W. Kim, ACS Nano, 2015, 9, 3501–3509.
30. L. W. Yamin, in Piezoelectric ZnO Nanostructure for Energy Harvesting, John Wiley & Sons, Inc., 2015, ch. 4, pp. 65–103.
31. S. Garain, T. K. Sinha, P. Adhikary, K. Henkel, S. Sen, S. Ram, C. Sinha, D. Schmeißer and D. Mandal, ACS Appl. Mater. Interfaces, 2015, 7, 1298–1307.
32. Y. Wu, Q. Jing, J. Chen, P. Bai, J. Bai, G. Zhu, Y. Su and Z. L. Wang, Adv. Funct. Mater., 2015, 25, 2166–2174.
33. T.-Y. Wei, P.-H. Yeh, S.-Y. Lu and Z. L. Wang, J. Am. Chem. Soc., 2009, 131, 17690–17695.
34. C. Pan, R. Yu, S. Niu, G. Zhu and Z. L. Wang, ACS Nano, 2013, 7, 1803–1810.
35. Y. Hu, J. Zhou, P. H. Yeh, Z. Li, T. Y. Wei and Z. L. Wang, Adv. Mater., 2010, 22, 3327–3332.
36. P. Chinnamuthu, J. C. Dhar, A. Mondal, A. Bhattacharyya and N. K. Singh, J. Phys. D: Appl. Phys., 2012, 45, 135102.
37. S. Lenaerts, J. Roggen and G. Maes, Spectrochim. Acta, Part A, 1995, 51, 883–894.
38. E. R. Viana, J. C. González, G. M. Ribeiro and A. G. de Oliveira, J. Phys. Chem. C, 2013, 117, 7844–7849.
39. C.-M. Chang, C.-H. Hsu, Y.-W. Liu, T.-C. Chien, C.-H. Sung and P.-H. Yeh, Nanoscale, 2015, 7, 20126–20131.
40. Y.-H. Zhang, L.-F. Han, Y.-H. Xiao, D.-Z. Jia, Z.-H. Guo and F. Li, Comput. Mater. Sci., 2013, 69, 222–228.
41. C.-H. Sung, T.-C. Chien, C.-M. Chang, C.-M. Chang and P.-H. Yeh, RSC Adv., 2015, 5, 16769–16773.
42. J. K. Hsu, T. Y. Lin, C. Y. Lai, T. C. Chien, J. H. Song and P. H. Yeh, Appl. Phys. Lett., 2013, 103, 123507.
43. N. O. Savage, S. A. Akbar and P. K. Dutta, Sens. Actuators, B, 2001, 72, 239–248.
44. C. S. Moon, H.-R. Kim, G. Auchterlonie, J. Drennan and J.-H. Lee, Sens. Actuators, B, 2008, 131, 556–564.
45. Q. H. Li, T. Gao, Y. G. Wang and T. H. Wang, Appl. Phys. Lett., 2005, 86, 123117.
46. S. Mathur, S. Barth, H. Shen, J.-C. Pyun and U. Werner, Small, 2005, 1, 713–717.
47. M. K. Nowotny, L. R. Sheppard, T. Bak and J. Nowotny, J. Phys. Chem. C, 2008, 112, 5275–5300.
48. S. Das and V. Jayaraman, Prog. Mater. Sci., 2014, 66, 112–255.
49. J. Nisar, Z. Topalian, A. De Sarkar, L. Österlund and R. Ahuja, ACS Appl. Mater. Interfaces, 2013, 5, 8516–8522.
50. G. Pacchioni, Phys. Chem. Chem. Phys., 2013, 15, 1737–1757.
51. X. Pan, M.-Q. Yang, X. Fu, N. Zhang and Y.-J. Xu, Nanoscale, 2013, 5, 3601–3614.
52. C. Shoou-Jinn, H. Ting-Jen, I. C. Chen and H. Bohr-Ran, Nanotechnology, 2008, 19, 175502.
53. J. M. Wu, Y.-R. Chen and W. T. Kao, ACS Appl. Mater. Interfaces, 2014, 6, 487–494.
54. A. Kar, S. Kundu and A. Patra, J. Phys. Chem. C, 2010, 115, 118–124.
55. Y. Zhang, A. Kolmakov, S. Chretien, H. Metiu and M. Moskovits, Nano Lett., 2004, 4, 403–407.
56. S. Bhaumik, A. K. Sinha, S. K. Ray and A. K. Das, IEEE Trans. Magn., 2014, 50, 1–6.
57. J. Pan, R. Ganesan, H. Shen and S. Mathur, J. Phys. Chem. C, 2010, 114, 8245–8250.
58. F. Hernandez-Ramirez, J. D. Prades, A. Tarancon, S. Barth, O. Casals, R. Jimenez-Diaz, E. Pellicer, J. Rodriguez, J. R. Morante, M. A. Juli, S. Mathur and A. Romano-Rodriguez, Adv. Funct. Mater., 2008, 18, 2990–2994.
59. M. Epifani, J. D. Prades, E. Comini, E. Pellicer, M. Avella, P. Siciliano, G. Faglia, A. Cirera, R. Scotti, F. Morazzoni and J. R. Morante, J. Phys. Chem. C, 2008, 112, 19540–19546.
60. H. Zeng, G. Duan, Y. Li, S. Yang, X. Xu and W. Cai, Adv. Funct. Mater., 2010, 20, 561–572.
61. W. Zhou, Y. Liu, Y. Yang and P. Wu, J. Phys. Chem. C, 2014, 118, 6448–6453.
62. Y. Li, J. Zhu, Y. Huang, F. Liu, M. Lv, S. Chen, L. Hu, J. Tang, J. Yao and S. Dai, RSC Adv., 2015, 5, 28424–28429.
63. X. Li, Q. Yu, C. Yu, Y. Huang, R. Li, J. Wang, F. Guo, Y. Zhang, S. Gao and L. Zhao, J. Mater. Chem. A, 2015, 3, 8076–8082.
64. J. Song, E. Zheng, J. Bian, X.-F. Wang, W. Tian, Y. Sanehira and T. Miyasaka, J. Mater. Chem. A, 2015, 3, 10837–10844.
65. P. H. Yeh, Z. Li and Z. L. Wang, Adv. Mater, 2009, 21, 4975–4978.
66. J. M. Shannon, Solid-State Electron., 1976, 19, 537–543.

---

Footnotes

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c6ra09033h

‡ The first two authors contributed to this work equally.

---