A broad-spectral-response perovskite photodetector with a high on/off ratio and high detectivity

Abstract

Given their high photoconversion efficiency, organometal halide perovskites have attracted much interest in ultraviolet-visible-near infrared photodetectors (PDs). However, recombination of carriers in perovskites, which leads to low photocurrent, has always been an adverse factor for device performance. Here, we demonstrate a way to overcome this problem by incorporating technologically-mature Si. Meanwhile, the dark current of the device can be maintained at a low level after inserting a TiO₂ layer between the CH₃NH₃PbI₃ film and the Si substrate. The fabricated CH₃NH₃PbI₃/TiO₂/Si tri-component PD displays a high on/off ratio of 6 × 10³ and a high detectivity of 4.85 × 10¹³ Jones. In addition, the PD shows broad photoresponse with a long cut-off wavelength of 1100 nm. The design here provides an intriguing way for improving perovskite-based PDs.

---

Introduction

Due to their high absorption coefficients, long diffusion lengths and high defect tolerances,^1–5 organometal halide perovskites (e.g., CH₃NH₃PbI₃) have attracted great attention in photodetector (PD) applications.^6–28 Currently, perovskite-only PDs usually suffer from severe carrier recombination.^6–9 Integrating perovskites with high-conductivity 2-dimensional materials,^10–13 transition metal sulfides (selenides),^14–17 or solution-processed ZnO, TiO₂^18–20 and organics^21–23 has always been used to address this issue via interfacial charge separation. However, high mobility materials would induce high dark current, while low mobility materials would cause low photoconductive gain. For example, typical perovskite/graphene hybrid PDs (generally composed of a phototransistor structure) exhibit outstanding responsivities of 1.8 × 10² to 2.6 × 10⁶ A W⁻¹, but are accompanied by extremely low on/off ratios lower than 5; CH₃NH₃PbI₃/WS₂ hybrid PDs show a high on/off ratio of about 10⁵ but a relatively low R of 17 A W⁻¹.¹⁴ For practical applications, it is very important to design perovskite-based PDs with both high photocurrent and low dark current.

Si is a mature semiconductor with a high electron (hole) mobility of 1450 (500) cm² V⁻¹ s⁻¹ and a narrow bandgap of 1.12 eV.^29–32 Integrating perovskites with Si could offer a unique way for achieving high-performance hybrid PDs: perovskites have strong light absorption abilities and hence are excellent photocarrier generators; Si can function as a very efficient transporter for the photocarriers generated in the perovskites. Meanwhile, the narrow bandgap of Si, which corresponds to an absorption edge with a wavelength of 1100 nm, could also largely expand the absorption and hence light detection ranges of perovskite-based PDs.

Here we design and fabricate a CH₃NH₃PbI₃/TiO₂/Si tri-component hybrid PD. Among them, Si is used to increase the photocurrent and the thin TiO₂ layer dramatically reduces the dark current. This device shows a high on/off ratio of 6 × 10³ as well as a high detectivity (D*) of 4.85 × 10¹³ Jones. In addition, the PD exhibits broad photoresponse reaching 1100 nm. The elaborate combination of CH₃NH₃PbI₃ and Si here opens up an efficient way for developing advanced high-performance perovskite-based PDs.

Results and discussion

The schematic structure of the CH₃NH₃PbI₃/TiO₂/Si hybrid PD is presented in Fig. 1a. A TiO₂ film with an optimized thickness of 20 nm was spin-coated on a Si substrate with a low n-doping concentration of about 10¹⁵ cm⁻³. Then two Ag electrodes with a thickness of 80 nm were deposited on the TiO₂ film by thermal evaporation with the help of a shadow mask, resulting in a channel length of 60 μm and a channel width of 2000 μm. After that, a 180 nm-thick CH₃NH₃PbI₃ film was spin-coated on top of the TiO₂ film by a simple one-step method.³³ For reference, the corresponding bare Si PD, bare CH₃NH₃PbI₃/glass PD and CH₃NH₃PbI₃/Si hybrid bi-component PD were also fabricated, and their schematic structures are given in Fig. S1 (ESI†), in which an optical microscopy image of the CH₃NH₃PbI₃/TiO₂/Si PD is also provided.

Fig. 1 (a) Schematic structure of the CH₃NH₃PbI₃/TiO₂/Si PD. (b) Top-view SEM image and AFM image of the TiO₂/Si film. (c) XPS of the TiO₂/Si film. (d) Top-view SEM image of the CH₃NH₃PbI₃/TiO₂/Si film. (e) Cross-sectional SEM image of the CH₃NH₃PbI₃/TiO₂/Si film. (f) XRD spectra of the CH₃NH₃PbI₃/glass, CH₃NH₃PbI₃/Si and CH₃NH₃PbI₃/TiO₂/Si films.

The surface morphology of the TiO₂ film deposited on the Si substrate was acquired by top-view scanning electron microscopy (SEM) and atomic force microscopy (AFM), and the images are given in Fig. 1b (with the AFM image as an inset). Analysis of the images shows that the TiO₂ film has a roughness of about 2.5 nm. A Ti 2p X-ray photoelectron spectroscopy spectrum of the TiO₂ film is illustrated in Fig. 1c. The Ti_1/2 and Ti_3/2 spin-orbital splitting photoelectrons are located at 465.2 and 459.3 eV, respectively. The peak separation of 5.9 eV between the Ti 2P_1/2 and 2P_3/2 signals is consistent with those reported.^34,35 The top-view SEM image of the CH₃NH₃PbI₃/TiO₂/Si film is shown in Fig. 1d. It is seen that the CH₃NH₃PbI₃ film is almost pinhole-free, and consists of dense crystals with sizes in the 50–250 nm range. The cross-sectional SEM image given in Fig. 1e shows that the thickness of the perovskite film is about 180 nm. Huang reported that the nucleation and crystallization of the perovskite film heavily depend on the substrate.³⁵ We compared the SEM images of the CH₃NH₃PbI₃/glass, CH₃NH₃PbI₃/Si and CH₃NH₃PbI₃/TiO₂/Si samples, as presented in Fig. S2 (ESI†). There is no obvious difference among the three samples. The X-ray diffraction (XRD) patterns of the three films are presented in Fig. 1f. The three typical peaks at 14.1°, 28.4°, and 31.9° can be assigned to the (110), (220), and (310) crystallographic planes of CH₃NH₃PbI₃.^8,14 It should be noted that no obvious signals from the TiO₂ film can be observed due to its weak crystallization, which is in agreement with a previous report.³⁶ A very strong Si(100) peak at 69.1° is observed and is not shown here. For the main (110) crystallographic planes of both the CH₃NH₃PbI₃/Si and CH₃NH₃PbI₃/TiO₂/Si samples, the full width at half-maximum (FWHM) of the peak only increases slightly compared with that of the CH₃NH₃PbI₃/glass sample. Therefore, these substrates in our experiment have negligible influence on the quality of the CH₃NH₃PbI₃ film. As the improvement of the device parameters followed by the change in the order of magnitude (shown below), the slight differences in the perovskite film could not be the major reason for this.

We also performed photoluminescence (PL) measurements for the three samples. As shown in Fig. 2a, strong PL quenching is observed for both the CH₃NH₃PbI₃/Si and CH₃NH₃PbI₃/TiO₂/Si samples (compared with the CH₃NH₃PbI₃/glass sample). The time-resolved PL spectra for the three samples are given in Fig. 2b; it is seen that the CH₃NH₃PbI₃/glass sample shows a PL lifetime of over 100 ns, which is about tenfold longer than those of the CH₃NH₃PbI₃/Si and CH₃NH₃PbI₃/TiO₂/Si samples, which can be attributed to efficient interface-involved carrier transfer. As shown in Fig. 2c, the band alignments of both the CH₃NH₃PbI₃/Si and CH₃NH₃PbI₃/TiO₂/Si samples enable them to form a heterojunction. For the CH₃NH₃PbI₃/Si sample, photoholes generated in the CH₃NH₃PbI₃ film can be injected into the Si, while photoelectrons are obstructed by the elevated interface barrier. For the CH₃NH₃PbI₃/TiO₂/Si sample, the TiO₂ layer can efficiently facilitate (block) the photoelectron (photohole) transfer from the CH₃NH₃PbI₃ into the Si. Thus the photoelectron and photoholes are largely spatially separated, and hence an enormous increase in photoresponse could be realized for both the CH₃NH₃PbI₃/Si and CH₃NH₃PbI₃/TiO₂/Si PDs.

Fig. 2 (a) Steady-state PL spectra. (b) Time-resolved PL spectra of the CH₃NH₃PbI₃/glass, CH₃NH₃PbI₃/Si and CH₃NH₃PbI₃/TiO₂/Si samples (under an excitation wavelength of 460 nm). (c) Band diagrams of the CH₃NH₃PbI₃/Si and CH₃NH₃PbI₃/TiO₂/Si samples.

The dark and light currents of PDs with different constructions (Si, CH₃NH₃PbI₃/glass, CH₃NH₃PbI₃/Si and CH₃NH₃PbI₃/TiO₂/Si) as a function of applied voltage are given in Fig. 3a and b, respectively. The PD parameters extracted under a bias of 10 V are summarized in Table 1. It is seen that the Si PD only presents a very low on/off ratio of 1.3. This can be explained as follows: Si possesses an intrinsically high conductance due to its high carrier mobility, and this leads to high dark current in the order of 10⁻⁵ A; meanwhile, Si is an indirect bandgap semiconductor and hence has weak light absorption, leading to low photocurrent (defined as the difference between light current and dark current). In comparison, the CH₃NH₃PbI₃/glass PD has a much lower dark current in the order of 10⁻⁹ A, but still a relatively low on/off ratio of 116. The unsatisfactory light current is due to severe carrier recombination inside the CH₃NH₃PbI₃ film, in which both electrons and holes are transported.^6,8 By combining the CH₃NH₃PbI₃ film with the Si substrate, the light current is enhanced significantly compared with the pure perovskite device. It should be attributed to the heterojunction formed at the CH₃NH₃PbI₃/Si interface, which helps holes to flow into Si under illumination. This greatly suppresses the photogenerated carrier recombination and improves the photocurrent.^36,37 However, although the dark current is lower than that for the Si device due to the depletion region, it is still in the order of 10⁻⁶ A, and such high dark current is detrimental to the light detecting capability of PDs. Encouragingly, by inserting a TiO₂ layer, the CH₃NH₃PbI₃/TiO₂/Si PD exhibits a much lower dark current of 10⁻⁹ A as well as a comparable light current compared to that of the CH₃NH₃PbI₃/Si PD. The high barrier for electron injection at the Ag/TiO₂ interface effectively reduces the dark current from the Si substrate without any impact on the photocurrent in the CH₃NH₃PbI₃ film.

Fig. 3 The dark current and light current (under 0.5 mW cm⁻² white light illumination) as a function of applied voltage: (a) the Si and CH₃NH₃PbI₃/glass PDs and (b) the CH₃NH₃PbI₃/Si and CH₃NH₃PbI₃/TiO₂/Si PDs.

Table 1 Parameters of the PDs (under a white light illumination power of 0.5 mW cm⁻² at a bias of 10 V)

Structure	Dark current (A)	Light current (A)	On/off ratio	R (A W^−1)	D* (Jones)
Si	2.62 × 10^−5	3.34 × 10^−5	1.3	12.0	1.44 × 10^11
CH3NH3PbI3/glass	3.15 × 10^−10	3.65 × 10^−8	116	0.057	1.95 × 10^11
CH3NH3PbI3/Si	2.02 × 10^−6	9.12 × 10^−5	45	148.4	6.40 × 10^12
CH3NH3PbI3/TiO2/Si	6.23 × 10^−9	3.75 × 10^−5	6019	62.5	4.85 × 10^13

---

Benefiting from the high photocurrent and low dark current, the CH₃NH₃PbI₃/TiO₂/Si PD exhibits a fairly high on/off ratio of 6.0 × 10³ and D* of 4.85 × 10¹³ Jones, with a decent R of 62.5 A W⁻¹ (D* and R are two important parameters characterizing a PD device; their definitions and calculations are provided in the ESI†). Table S2 (ESI†) lists the device parameters of current typical perovskite-based PDs. It is seen that the PDs with high R (usually use highly conductive graphene or its derivatives) have very low on/off ratios, and the ones with high on/off ratios have very low R. In comparison, the CH₃NH₃PbI₃/TiO₂/Si PD we designed here shows high overall performance. In addition, a device simulation was made to verify our analysis, which is presented in the ESI,† Fig. S5. In order to confirm the reliability and reproducibility of our results, the experiments were conducted repeatedly. The histograms of dark and light currents at a bias of 10 V for the CH₃NH₃PbI₃/TiO₂/Si PD were collected from 20 independent products.

Fig. 4a shows the photocurrent and R dependences on light power for the CH₃NH₃PbI₃/TiO₂/Si PD. It is seen that the photocurrent increases, while R decreases with increasing light illumination. The decrease of R is a trap state induced effect.³⁸Fig. 4b shows the spectral response of the CH₃NH₃PbI₃/TiO₂/Si PD. It is seen that, benefiting from the light absorption of narrow-bandgap Si, the photoresponse wavelength reaches 1100 nm. In comparison, the response of the CH₃NH₃PbI₃/glass PD is cut off at 780 nm.²⁰ The on–off switching properties of the CH₃NH₃PbI₃/TiO₂/Si PD are presented in Fig. 4c and d. It is seen that the PD can be switched on and off repeatedly and stably, and the rise time and decay time are 0.89 s and 0.42 s, respectively (the rise time is defined as the time for the current to increase to 90% of the peak value and the decay time is defined as the time for the current to decay to 10% of the peak value^38–40).

Fig. 4 (a) Photocurrent and responsivity of the CH₃NH₃PbI₃/TiO₂/Si PD versus white light illumination power at a bias of 10 V. (b) Responsivity of the CH₃NH₃PbI₃/TiO₂/Si PD under 10.6 μW cm⁻² illumination with various wavelengths at a bias of 10 V. (c and d) On–off switching properties of the CH₃NH₃PbI₃/TiO₂/Si PD at a bias of 10 V.

Conclusions

In summary, CH₃NH₃PbI₃-Si hybrid PDs are fabricated and investigated. The reduced recombination due to the spatial separation of photocarriers significantly enhances the photocurrent. Further, by inserting a thin TiO₂ layer in-between the CH₃NH₃PbI₃ film and the Si substrate, the dark current is greatly suppressed. The designed CH₃NH₃PbI₃/TiO₂/Si PD shows excellent overall performance with a high on/off ratio of 6 × 10³ and a high D* of 4.85 × 10¹³ Jones. Moreover, the PD shows broad photoresponse reaching 1100 nm, which is helped by light absorption from the Si. Our work here provides a feasible strategy for developing high-performance perovskite-based PDs.

Experimental

Material synthesis and device fabrication

A TiO₂ solution was made using previously reported sol–gel methods:^2,40 1 mL TiCl₄ (99.9%, Alfa) was added into 4 mL anhydrous ethanol. The mixed solution was then stirred, then 20 mL benzyl alcohol was added. The obtained solution is yellow in color and was stored at 85 °C for one whole night. The solution was then mixed with 400 mL diethyl ether and centrifuged at 6000 rpm for 2 min to obtain TiO₂ nanocrystal precipitates. Finally, the TiO₂ nanocrystals were dispersed into anhydrous chloroform and anhydrous methanol (1 : 1 volume ratio) with a concentration of about 5 mg mL⁻¹. A perovskite precursor (0.5 M) was prepared by dissolving CH₃NH₃I and PbI₂ in a 1 : 1 molar ratio in a 4 : 1 (v : v) mixed solvent of N,N-dimethylformamide (DMF) and dimethylsulfoxide (DMSO). The solution was then heated at 60 °C for 12 hours for sufficient dissolution. A Si substrate (2–5 Ω cm) was sequentially cleaned using detergent, acetone, and isopropanol with ultrasonication for 10 min each and then was immersed into 10% HF solution for 1 min to remove the native oxide layer. After that, the Si substrate was dried and treated with O₂ plasma for 5 min. The TiO₂ solution was spin-coated on the substrate at 6000 rpm for 30 s, and annealed in ambient air at 150 °C for 30 min. After that, 80 nm-thick Ag electrodes were thermally deposited on the TiO₂ film through a shadow mask, resulting in a channel length of 60 μm and a channel width of 2000 μm. The perovskite film was then fabricated on top using an anti-solvent spin coating procedure.

Characterization

Electrical characterization data were recorded with a Keithley 4200 in ambient air. Monochromatic light and white light were from a Newport Oriel 200 and a XZ-150WA halogen lamp, respectively. Prior to the utilization of the light, the spectral and light intensity were calibrated using a monosilicon detector. In detail, the I–V curves were measured in reverse (10 V to −10 V) or forward (−10 V to 10 V) mode. The step voltage and decay time were fixed at 200 mV and 10 ms, respectively. The optical images were obtained using an Olympus BX51 optical microscope. The morphology of the formed film and the cross-sectional image were characterized by SEM (Hitachi S-4800). An AFM image was obtained using a Dimension Icon microscope. Visible spectra were recorded using a JASCO V-570 spectrophotometer and PL spectra were characterized using a JASCO FP-6600 spectrophotometer with a Xenon flash lamp. XPS spectra were obtained using a KRATOS ULTRA AXIS DLD photoelectron spectroscopy system with an Al X-ray source. XRD patterns were obtained using a Rigaku D/MAX-2004 XRD with Cu Kα radiation (λ = 1.54178 Å).

Author contributions

The most experiments (including simulation) and data analysis were performed by Xiaohui Yi. The AFM was performed with the help of Yisen Wang. Zhengwei Ren helped in synthesising the TiO₂ NCs. The XPS and PL were performed with the help of Yisen Wang. Zhiwei Huang and Hui Li participated in discussion and provided some useful advice. The manuscript was written by Xiaohui Yi, Tao Lin, Cheng Li and Jizheng Wang. All authors reviewed the manuscript.

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

This research was supported by the National Natural Science Foundation of China (61474094 and 61405208), the National Key Research Program of China (2016YFA0200104), the 973 Program (2014CB643600 and 2014CB643503), and the Strategic Priority Research Program of the Chinese Academy of Sciences (XDB12030200).

Notes and references

1. H. Zhou, Q. Chen, G. Li, S. Luo, T. Song, H. Duan, Z. Hong, J. You, Y. Liu and Y. Yang, Science, 2014, 345, 6196.
2. W. Nie, H. Tsai, R. Asadpour, J.-C. Blancon, A. J. Neukirch, G. Gupta, J. J. Crochet, M. Chhowalla, S. Tretiak, M. A. Alam, H.-L. Wang and A. D. Mohite, Science, 2015, 347, 6221.
3. D.-Y. Son, J.-W. Lee, Y. J. Choi, I.-H. Jang, S. Lee, P. J. Yoo, H. Shin, N. Ahn, M. Choi, D. Kim and N.-G. Park, Nat. Energy, 2016, 1, 16081.
4. Z. Zhu, Y. Bai, X. Liu, C.-C. Chueh, S. Yang and Al. K.-Y. Jen, Adv. Mater., 2016, 28, 6478.
5. W. S. Yang, B.-W. Park, E. H. Jung, N. J. Jeon, Y. C. Kim, D. U. Lee, S. S. Shin, J. Seo, E. K. Kim, J. H. Noh and S. I. Seok, Science, 2017, 356, 1376.
6. Y. Zhang, J. Du, X. Wu, G. Zhang, Y. Chu, D. Liu, Y. Zhao, Z. Liang and J. Huang, ACS Appl. Mater. Interfaces, 2015, 7, 21634.
7. X. Hu, X. Zhang, L. Liang, J. Bao, S. Li, W. Yang and Y. Xie, Adv. Mater., 2014, 24, 7373.
8. H. Lu, W. Tian, F. Cao, Y. Ma, B. Gu and L. Li, Adv. Func. Mater., 2016, 26, 1296.
9. K. C. Kwon, K. Hong, Q. V. Le, S. Y. Lee, J. Choi, K.-B. Kim, S. Y. Kim and H. W. Jang, Adv. Funct. Mater., 2016, 26, 4213.
10. Y. Lee, J. Kwon, E. Hwang, C.-H. Ra, W. J. Yoo, J.-H. Ahn, J. H. Park and J. H. Cho, Adv. Mater., 2015, 27, 41.
11. Y. Wang, Y. Zhang, Y. Lu, W. Xu, H. Mu, C. Chen and H. Qiao, Adv. Opt. Mater., 2015, 3, 1389.
12. Z. Sun, L. Aigouy and Z. Chen, Nanoscale, 2016, 8, 7377.
13. M. Spina, M. Lehmann, B. Náfrádi, L. Bernard, E. Bonvin, R. Gaál, A. Magrez, L. Forró and E. Horváth, Small, 2015, 11, 4824.
14. C. Ma, Y. Shi, W. Hu, M.-H. Chiu, Z. Liu, A. Bera, F. Li, H. Wang, L.-J. Li and T. Wu, Adv. Mater., 2016, 28, 3683.
15. D.-H. Kang, S. R. Pae, J. Shim, G. Yoo, J. Jeon, J. W. Leem, J. S. Yu, S. Lee, B. Shin and J.-H. Park, Adv. Mater., 2016, 28, 7799.
16. Y. Wang, R. Fullon, M. Acerce, C. E. Petoukhoff, J. Yang, C. Chen, S. Du, S. Ki Lai, S. P. Lau, D. Voiry, D. O’Carroll, G. Gupta, A. D. Mohite, S. Zhang, H. Zhou and M. Chhowalla, Adv. Mater., 2017, 29, 1603995.
17. J. Lu, A. Carvalho, H. Liu, S. X. Lim, A. H. C. Neto and C. H. Sow, Angew. Chem., 2016, 128, 12124.
18. L. Zhou, K. Yu, F. Yang, J. Zheng, Y. Zuo, C. Li, B. Cheng and Q. Wang, Dalton Trans., 2017, 46, 1766.
19. J. Yu, X. Chen, Y. Wang, H. Zhou, M. Xue, Y. Xu, Z. Li, C. Ye, J. Zhang, P. A. van Aken, P. D. Lundac and H. Wang, J. Mater. Chem. C, 2016, 4, 7302.
20. X. Yi, Z. Ren, N. Chen, C. Li, X. Zhong, S. Yang and J. Wang, Adv. Electron. Mater., 2017, 1700251.
21. S. Chen, C. Teng, M. Zhang, Y. Li, D. Xie and G. Shi, Adv. Mater., 2016, 28, 5969.
22. Y. Guo, C. Liu, H. Tanaka and E. Nakamura, J. Phys. Chem. Lett., 2015, 6, 535.
23. S. Tong, J. Sun, C. Wang, Y. Huang, C. Zhang, J. Shen, H. Xie, D. Niu, S. Xiao, Y. Yuan, J. He, J. Yang and Y. Gao, Adv. Electron. Mater., 2017, 1700058.
24. H. Wang and D. H. Kim, Chem. Soc. Rec., 2017, 46, 5204.
25. L. Li, Y. Deng, C. Bao, Y. Fang, H. Wei, S. Tang, F. Zhang and J. Huang, Adv. Opt. Mater., 2017, 5, 1700672.
26. W. Wang, D. Zhao, F. Zhang, L. Li, M. Du, C. Wang, Y. Yu, Q. Huang, M. Zhang, L. Li, J. Miao, Z. Lou, G. Shen, Y. Fang and Y. Yan, Adv. Funct. Mater., 2017, 27, 1703953.
27. W. Wang, F. Zhang, M. Du, L. Li, M. Zhang, K. Wang, Y. Wang, B. Hu, Y. Fang and J. Huang, Nano Lett., 2017, 17, 1995.
28. L. Qian, Y. Sun, M. Wu, D. Xie, L. Ding and G. Shi, Adv. Mater., 2017, 29, 1606175.
29. S. Ke, S. Lin, X. Li, J. Li, J. Xu, C. Li and S. Chen, Opt. Express, 2016, 24, 1943.
30. G. Lin, X. Yi, C. Li, N. Chen, L. Zhang, S. Chen, W. Huang, J. Wang, X. Xiong and J. Sun, Appl. Phys. Lett., 2016, 109, 141104.
31. X. Li, J. E. Carey, J. W. Sickler, M. U. Pralle, C. Palsule and C. J. Vineis, Opt. Express, 2012, 20, 5518.
32. L.-D. Haret, X. Checoury, Z. Han, P. Boucaud, S. Combrié and A. D. Rossi, Opt. Express, 2010, 18, 23965.
33. K.-F. Lin, S. H. Chang, K.-H. Wang, H.-M. Cheng, K. Y. Chiu, K.-M. Lee, S.-H. Chen and C.-G. Wu, Sol. Energy Mater. Sol. Cells, 2015, 141, 309.
34. Z. Song, J. Hrbek and R. Osgood, Nano Lett., 2005, 5, 1327.
35. B. Erdem, R. A. Hunsicker, G. W. Simmons, E. D. Sudol, V. L. Dimonie and M. S. El-Aasser, Langmuir, 2001, 17, 2664.
36. (a) H. Tan, A. Jain, O. Voznyy, X. Lan, F. P. G. de Arquer, J. Z. Fan, R. Q. Bermudez, M. Yuan, B. Zhang, Y. Zhao, F. Fan, P. Li, L. N. Quan, Y. Zhao, Z.-H. Lu, Z. Yang, S. Hoogland and E. H. Sargent, Science, 2017, 355, 722; (b) J. Zhao, Y. Deng, H. Wei, X. Zheng, Z. Yu, Y. Shao, J. E. Shield and J. Huang, Sci. Adv., 2017, 3, 5616.
37. Z. Jin and J. Wang, J. Mater. Chem. C, 2013, 1, 7996.
38. Z. Qi, J. Cao, H. Li, L. Ding and J. Wang, Adv. Funct. Mater., 2015, 25, 3138.
39. Z. Jin, Q. Zhou, Y. Chen, P. Mao, H. Li, H. Liu, J. Wang and Y. Li, Adv. Mater., 2016, 28, 3697.
40. J. Zhang, Q. Wang, L. Wang, X. Li and W. Huang, Nanoscale, 2015, 7, 10391.

---

Footnote

† Electronic supplementary information (ESI) available: Optical microscopy image, schematic structures, SEM images, I–V curve, on–off switching properties, and device simulation. See DOI: 10.1039/c8qm00303c

---