Self-doped tungsten oxide films induced by in situ carbothermal reduction for high performance electrochromic devices

Lijun Zhou a, Peng Wei a, Huajing Fang *a, Wenting Wu a, Liangliang Wu b and Hong Wang *bc
aState Key Laboratory for Mechanical Behavior of Materials, School of Material Science and Engineering, Xi’an Jiaotong University, Xi’an 710049, China. E-mail:
bSchool of Electronic and Information Engineering and State Key Laboratory for Mechanical Behavior of Materials, Xi’an Jiaotong University, Xi’an 710049, China
cDepartment of Materials Science and Engineering & Shenzhen Engineering Research Center for Novel Electronic Information Materials and Devices, Southern University of Science and Technology, Shenzhen 518055, China. E-mail:

Received 1st July 2020 , Accepted 12th August 2020

First published on 13th August 2020

As a promising electrochromic material, tungsten oxide is of great value for applications in smart windows and energy saving displays. Fabricating high performance tungsten oxide by the cost-effective way is immensely in demand for promoting the large-scale application. Herein, we demonstrate a green and economic route in developing self-doped tungsten oxide (WO3−x) films. Through the precursor design and carbonization process, amorphous carbon can be obtained and used as an in situ reducing agent. Then, oxygen vacancies and nanopores were generated in WO3−x films by carbothermal reduction during annealing. The self-doping WO3−x film exhibits a remarkable optical modulation (∼70% at 680 nm) in both visible and near infrared bands. The electrochromic switching behavior of the WO3−x film shows a fast response speed (7 s for coloring and 15 s for bleaching), a high coloration efficiency up to 62 cm2 C−1 and a good cycling stability. The enhanced electrochromic properties are attributed to the large specific surface area and oxygen vacancies which facilitate the ion insertion and extraction at the interface. This work offers a highly active electrochromic material as well as an ingenious synthesis strategy for numerous oxygen deficient functional oxides.

1. Introduction

With the development of industrial society, energy shortage has become a common problem faced by all countries around the world. Smart windows with the ability to dynamically adjust their color and transmittance can effectively alleviate the energy consumption of lighting and air conditioning in both buildings and automobiles.1–4 Over the past half century, many electrochromic materials including inorganic compounds (such as transition metal oxides, Prussian blue) and organic compounds (such as viologens and polyaniline) have been studied for the application of smart windows.5–7 Among them, tungsten oxide is one of the most promising candidates due to its large modulation range, high contrast and stable physical and chemical properties.8–10 However, the complex manufacturing processes and high cost hinder the popularity of this material in the civilian sector.11 Exploring the synthesis process of low-cost and high performance electrochromic materials is still the only way to break through the development bottleneck of smart windows.

Compared with vacuum deposition techniques such as magnetron sputtering and pulsed laser deposition, solution processing technology eliminates the need for expensive equipment and a rigorous vacuum environment, which can significantly reduce production costs.12,13 At the same time, the solution processing technology provides the possibility to design material components from the source and facilitate precise doping and modification of the electrochromic materials. For instance, Lee et al. have synthesized the Ti doped WO3 with reduced crystallite size and improved coloration efficiency using the wet bath method.14 The Co doped WO3 film with modified morphology and better long term stability has been prepared through a seed free hydrothermal method by Xu's group.15 Tu et al. demonstrated a smart electrochromic battery electrode based on Mo doped WO3 nanowire arrays fabricated via a sulfate-assisted hydrothermal method.16 Nevertheless, the introduction of metal elements is prone to form phase segregation or impurities, especially in the case of high doping content. Instead of adding foreign atoms, self-doping (introducing anion vacancies or cation vacancies) is another promising method to improve the semiconductor and electrochemical properties of metal oxides. Recently, Wang et al. have demonstrated the self-doped SnO2−x nanocrystals with a shifted absorption band and photocatalytic activity which enables a visible light responsive color switching system.17 V2O5 nanoflakes self-doped with V4+ as the cathode material in lithium ion batteries have been reported to exhibit superior rate capability and remarkable cycling performance.18 Self-doped WO3/TiO2 nanotubes were developed with an enhanced photocatalytic degradation of volatile organic compounds, resulting from the reduced electron–hole recombination in the composite film.19 However, to the best of our knowledge, improving the electrochromic performance of tungsten oxide using the self-doping strategy has not been reported. At the same time, a simple and scalable fabrication method to prepare the self-doped WO3−x film remains a big challenge.

In this work, we presented a unique aqueous synthesis of the self-doped WO3−x film by spin coating. The polyvinylpyrrolidone (PVP) polymer was added in the precursor both as a thickener and the carbon source. Via a facile carbonization step, the polymer can be in situ converted into amorphous carbon as a reducing agent embedded in the film. Inspired by the carbothermal reduction of WO3 for the industrial production of tungsten powder,20 WO3−x with oxygen vacancies is accurately produced by changing the amount of the carbon source. The partially reduced polycrystalline film is transparent and porous. Owing to the synergistic effects of the large specific surface area and oxygen vacancies, the self-doped WO3−x films displayed favorable electrochromic behaviors as well as a good cycling stability, making them a promising material for electrochromic applications.

2. Experimental section

2.1. Materials

Tungstic acid powder (H2WO4, AR) and hydrogen peroxide solution (H2O2, 30 wt% in water) were purchased from Sinopharm Chemical Reagent Co., Ltd. Polyvinylpyrrolidone (PVP K90, 98%, number averaged molecular weight ∼107[thin space (1/6-em)]000) was purchased from Hefei BOSF Biotechnology Co., Ltd. The 1.25 M LiClO4 electrolyte in a mixed solvent (1[thin space (1/6-em)]:[thin space (1/6-em)]1[thin space (1/6-em)]:[thin space (1/6-em)]1 Vol%) of ethylene carbonate (EC)[thin space (1/6-em)]:[thin space (1/6-em)]dimethyl carbonate (DMC)[thin space (1/6-em)]:[thin space (1/6-em)]ethyl methyl carbonate (EMC) was purchased from Dodochem Technology Co., Ltd. Ethanol (CH3CH2OH, 99.7%) was purchased from Tianjin Zhiyuan Chemical Reagent Co., Ltd. All chemicals and reagents were used without any further purification.

2.2. Fabrication of the self-doped WO 3−x electrochromic films

The self-doped WO3−x films were synthesized via a simple solution processing route as schematically shown in Fig. 1a. First, the H2WO4 powder and the H2O2 solution were mixed with deionized water in an Erlenmeyer flask. After stirring at 90 °C for 4 hours, the yellow H2WO4 powder has been completely dissolved to get a colorless and transparent solution. Then, PVP K90 was added to the previous solution with a mass fraction of 5%. A sticky precursor can be obtained after continuous stirring for 4 hours.
image file: d0tc03103h-f1.tif
Fig. 1 (a) Schematic of fabrication steps from the H2WO4 powder to the carbonized film. The photographs of (b) H2WO4 powder (c) precursor (d) carbonized film and (e) annealed WO3−x film.

Fluorine-doped tin oxide (FTO) glasses with a sheet resistance of 7 Ω sq−1 were cleaned using the ultrasonic bath with deionized water and ethanol before using. Then, the sticky precursor was spin-coated onto a clean FTO glass at a speed of 1500 rpm for 10 s and dried at 240 °C for 5 min. The spin-coating operations were repeated three times, and the sample was subsequently transferred to a hot plate at 350 °C. After carbonizing for 30 min, the black film was annealed at 500 °C for 2 h in a muffle furnace to form the self-doped WO3−x film.

2.3. Characterization

The surface and cross-sectional morphologies of the WO3−x film were observed using a scanning electron microscope (SEM, FEI Quanta 250 FEG). The crystalline structure and chemical composition of the WO3−x film were measured by X-ray diffraction (XRD, X’Pert Pro) using Cu Kα radiation and X-ray photoelectron spectroscopy (XPS, Escalab Xi+). The molecular structure of the film was obtained via Raman spectroscopy (LabRAM HR Evolution) and Fourier transform infrared spectroscopy (FTIR, Bruker VERTEX 70). The electrochromic properties of the device were measured using a spectrophotometer (Mapada V-1600PC) in the wavelength region between 325 and 1000 nm. The cyclic voltammetry (CV) curves of the WO3−x film were tested by an electrochemical workstation (Zennium Pro, Zahner, Germany) with a three-electrode system. The WO3−x film on the FTO substrate, the platinum (Pt) foil and the silver (Ag) wire were used as the work electrode, the counter electrode and the reference electrode, respectively. The CV curves were recorded within the potential range from −0.6 V to 0.9 V vs. the reference electrode.

3. Results and discussion

Fig. 1a schematically presents the key steps of preparing WO3−x films. The yellow H2WO4 powder (see Fig. 1b) was used as a raw material because of its high tungsten element content and good water solubility. After mixing with hydrogen peroxide at a mild temperature, the H2WO4 powder can be completely dissolved. In order to increase the viscosity, polyvinylpyrrolidone (PVP), a water soluble polymer, was added in the solution with a suitable concentration. The obtained transparent precursor is colorless and transparent. As shown in Fig. 1c, the significant Tyndall phenomenon can be observed, which indicated the colloidal property of the precursor.21 Such a precursor can be easily spin-coated on the FTO transparent electrode. Here, a special process step was introduced for carbonizing the sample. As thermal decomposition of the PVP polymer produces a large amount of carbon in the film, the sample turns black as shown in Fig. 1d. Finally, the highly transparent WO3−x films (Fig. 1e) are obtained after annealing. The entire preparation process is environmentally friendly and green, since no toxic or harmful organic solvents are used. Moreover, if we trace the evolution of the tungsten element throughout the process (Fig. 1b–e), it can be found that no tungsten-containing by-products are formed. Theoretically, the conversion rate of the W element from the H2WO4 powdered raw material to the WO3−x film product in this synthetic route reaches 100%, which is in line with the atomic economy.22

Fig. 2a shows the SEM image of the annealed WO3−x film. It can be clearly seen that many nanoscale pores are formed in the film, which are caused by the combustion of the carbonized sample. Since the electrochromic phenomenon of tungsten oxides is an electrochemical reaction involving the injection and extraction of ions,9 the interface between the WO3−x film and the electrolyte is critical for the coloring and bleaching process. Compared with compact films, the nanoscale pores facilitate the expansion of interface contact area between the WO3−x film and the electrolyte. This nanoporous structuring can provide a larger active surface and a faster ion diffusion path, leading to enhanced electrochromic performances.16 As shown in the cross-sectional SEM image (inset of Fig. 2a), the annealed WO3−x film tightly deposits on the FTO electrode and has a thickness of about 200 nm. X-ray diffraction (XRD) analysis confirmed the phase change of the spin-coated film before and after annealing (Fig. 2b). The carbonized sample shows only XRD patterns of the FTO substrate, since the carbonization temperature is lower than the crystallization temperature of tungsten oxides, whereas the diffraction peaks of monoclinic phase WO3 appeared in the annealed film, without any crystalline impurities.23,24 The Raman spectrum of the annealed film as shown in Fig. 2c is consistent with the result of XRD patterns. In addition to the intrinsic peaks of the FTO substrate, the strong peaks around 800, 690 and 265 cm−1 can be found, which are associated with the fundamental bands of the monoclinic WO3 phase.25,26 However, the Raman spectra of the carbonized film show only two weak peaks at 1589 and 1365 cm−1. According to the literature,27,28 these wide bands can be attributed to the carbon materials, which are the cause of black color in the carbonized film. Moreover, all the intrinsic peaks of the FTO substrate have not been observed in the carbonized film as the excitation light of the Raman spectra cannot penetrate the black carbonized layer. Comparing the three curves as shown in Fig. 2c, it can be found that the carbon material is burned out during the annealing process, according well with the above SEM image.

image file: d0tc03103h-f2.tif
Fig. 2 (a) SEM image of the annealed WO3−x film; the inset shows the cross-sectional SEM image. (b) XRD patterns of the carbonized film and the annealed film. (c) Raman spectra of the FTO substrate, carbonized film and the annealed film. (d) FTIR spectra of the FTO substrate, carbonized film and the annealed film.

To further identify the chemical structure of the materials, we have measured the FTIR spectra of the carbonized and annealed WO3−x film. As shown in Fig. 2d, the carbonized film shows five vibration peaks in the test wave number region. The peaks at 1714 and 1585 cm−1 arise from the vibration of C[double bond, length as m-dash]O and C[double bond, length as m-dash]C bonds, which fully prove the existence of carbon nanomaterials in the carbonized film.29 The three peaks between 500 and 1000 cm−1 may be assigned to the bands of some complex tungsten-containing compounds. In the annealed film, the peaks related to carbon materials are eliminated, indicating the complete combustion during annealing. The characteristic peak at 1039 cm−1 is due to the W–OH plane pending, and the peak at 1012 cm−1 arises from the symmetrical stretching mode of W[double bond, length as m-dash]O bonds. The strong peaks at 890, 790 and 698 cm−1 are associated with the W–O–W bonds.30–32 Thus it can be seen that the complex tungsten-containing compounds have converted to monoclinic tungsten oxide after annealing.

Next, the annealed WO3−x film was assembled into an electrochromic device to test its functionality. The device structure is shown in Fig. 3a, in which the FTO glass is used as the transparent electrode. The electrolyte with 1.25 M LiClO4 in a mixed solvent was used in the electrochromic device to achieve the insertion and extraction of Li+ ions in the WO3−x film. The device in the bleached state shows a high transmittance in both the visible and infrared wavelength range (Fig. 3b), while in the colored state, the transmittance of the electrochromic device decreases significantly. At the wavelength of 680 nm, the transmittance difference between the two curves is around 70%. In the wavelength range of 700–1000 nm, the transmittance of the colored state is even less than 3%, indicating the excellent near infrared light shielding ability. Hence, in the scorching summer, smart windows made of this electrochromic WO3−x film can keep the room cool by blocking heat radiation. Fig. 3c shows the photographs of the packaged electrochromic device in the coloring and bleaching process. A series of different color states from light to deep blue can be easily obtained by adjusting the applied bias from −3.5 V to 3.5 V, which may have potential applications in the field of electrochromic displays.

image file: d0tc03103h-f3.tif
Fig. 3 (a) Schematic diagram of the electrochromic device. (b) Optical transmittance spectra of the electrochromic device in the bleached and colored state. (c) Photographs of the electrochromic device during bleaching and coloring processes; scale bar is 2 cm.

To investigate the dynamic characteristics of the annealed WO3−x film, we measured the real-time transmittance of the packaged electrochromic device at 680 nm during the bleaching and coloring process. By applying positive (3.5 V) and negative voltages (−3.5 V) periodically, the transmittance fluctuates accordingly as shown in Fig. 4a. The time step of coloring and bleaching procedures is 7 s and 53 s, respectively. In the coloring time of 7 s, Li+ ions and electrons were injected into the annealed WO3−x film, resulting in a 60% transmittance reduction. Upon switching the bias, the transmittance returns to its initial value (73%) in less than 15 s through the extraction of Li+ ions and electrons. Such a fast response speed is comparable to the results of previous works.33,34

image file: d0tc03103h-f4.tif
Fig. 4 (a) Dynamic change in transmittance at 680 nm during the coloring (7 s) and bleaching (53 s) process. (b) It curve of the annealed WO3−x film during the coloring (7 s) and bleaching (53 s) process. (c) CV curves of the annealed WO3−x film at different scan rates, the red arrow indicates the trend when scan rate increases from 20 to 100 mV s−1. The inset shows the relationship between peak current density and the square root of the scan rates. (d) CV cycling stability of the annealed WO3−x film.

The coloration efficiency (CE) is another important evaluation index of electrochromic materials, which can be defined as the change in optical density (ΔOD) per unit charge (ΔQ) injected into the film per unit area.35 Usually, the coloration efficiency is calculated as follows,

CE =ΔOD/ΔQ = log(Tb/Tc)/ΔQ
where Tb and Tc refer to the transmittance of the electrochromic film in the bleached and colored states. And ΔQ can be obtained by testing the area of the device and the chronoamperometry (It) curve during the bleaching and coloring process (Fig. 4b). The CE value of the annealed WO3−x film is found to be 62 cm2 C−1.

To evaluate the electrochemical behavior during the bleaching and coloring process, a three-electrode system was built to measure the cyclic voltammetry (CV) curves of the annealed WO3−x film. As shown in Fig. 4c, the CV curves show similar shape at different scan rates. The innermost curve is at 20 mV s−1 and the outmost curve is at 100 mV s−1. The red arrow in Fig. 4c indicates the change of the curves when the scan rate increases from 20 to 100 mV s−1. The closed area of each CV curve increases as the scan rate increases, indicating the ideal capacitance characteristics. The inset shows the relationship between peak current density (Jp) and the square root of the scan rates (v1/2). After linearly fitting the experimental data, the diffusion coefficient (D) of the Li+ ions can be calculated using the Randles-Sevcik equation.36,37

Jp = 2.72 × 105 × D1/2 × Co × v1/2
where Co is the concentration of active ions in the electrolyte. The Li+ ion diffusion coefficient in the annealed WO3−x film is about 1 × 10−10 cm2 s−1. The performance parameters of our self-doped WO3−x film is comparable to the values in previous works, as shown in Table 1.

Table 1 Performance summary of the tungsten oxide electrochromic materials
Materials Fabrication method Transmittance modulation Switching speed tc/tb [s] Coloration efficiency [cm2 C−1] Diffusion coefficient [cm2 s−1] Ref.
Self-doped WO3−x Spin coating 70%@680 nm 7/15 62 Li+ ion 1 × 10−10 This work
WO3 Reactive sputtering 74.4%@670 nm 5.7/2.2 76.45 2
WO3 Hydrothermal 78.1%@630 nm 5/6 56.5 9
Ti-doped WO3 Wet bath 67.6%@633 nm 15/5 106.6 H+ ion 4.4 × 10−9 14
Co-doped WO3 Hydrothermal 75.4%@680 nm 27.6/11.8 65.7 15
WO3/rGO Spin coating 64.2%@633 nm 9.5/4.5 43 Li+ ion 4.9 × 10−10 36
2D WO3 Solution-phase synthesis 62.57%@700 nm 10.7/7 Li+ ion 5.9 × 10−10 37

Moreover, cycling stability is a very important parameter of the electrochromic films in practical applications. We have measured the CV curves for more than 200 cycles from −0.6 to 0.9 V. In each cycle, the reversible process of coloring and bleaching can be observed in the annealed WO3−x film. Fig. 4d shows the CV curves for the first, 50th, 100th, 150th and 200th cycle. The CV curves in the first 50 cycles show slight promotion, and remain almost stable without any degradation even after 200 cycles. Such a high cycling stability may originate from the solid nanostructure of the annealed WO3−x film and the prevention of the harmful side reaction at the WO3−x film/electrolyte interface.

It must be highlighted that the satisfactory electrochromic performance was attributed to the self-doping effect via the solution processed method. The precursor we developed in this work does not contain any toxic organic solvents, which shows compliance with green chemistry principles.38 A key ingredient in the precursor is the water-soluble polymer. On the one hand, the right amount of the PVP polymer can significantly increase the viscosity of the precursor and facilitate film formation during spin coating. On the other hand, the carbonized PVP polymer provides a carbon source for the subsequent carbothermal reduction reaction. As shown in Fig. 5a, the carbothermal reduction reaction can realize defect engineering at two different scales. At the nanoscale, pores are evenly distributed in the polycrystalline film, leading to a large specific surface area for fast ion diffusion. At the atomic scale, oxygen vacancies are generated in the tungsten oxide lattice. The chemical valences of elements in the annealed WO3−x film were investigated via XPS. The W 4f spectrum as shown in Fig. 5b shows two main peaks at 37.8 and 35.7 eV, which are associated with the oxidation state of W6+. Furthermore, the weak peaks at 36.6 and 34.5 eV that originated from the oxidation state of W5+ can also be identified.39,40 For the O 1s spectrum shown in Fig. 5c, the dominated peak at 530.5 eV is related to the lattice oxygen. And a shoulder peak appears at 531.9 eV, which could be determined as the OH groups. The OH groups are usually bonded to metal cations to maintain the charge balance in sub-stoichiometric oxides. Hence, the peak intensity of OH groups can prove the existence of oxygen vacancies.41,42 These XPS results confirmed that self-doping of WO3−x has been successfully achieved through the carbothermal reduction reaction. It has been reported that oxygen vacancies can improve the electrochromic performance of the WO3 film by increasing its electrical conductivity.43 The self-doped WO3−x tends to be a degenerate n type semiconductor with gap states introduced by oxygen vacancies. Density functional theory (DFT) calculations in previous works demonstrated that oxygen vacancies can effectively decrease the ion insertion energy barrier in metal oxides.44,45 Hence, electrochemical reactions during coloring and bleaching processes will be easier to trigger. It is also worth mentioning that oxygen vacancies are conducive to relaxing the lattice strain during Li+ ion insertion and extraction processes, resulting in a good cycling stability in self-doped WO3−x films.

image file: d0tc03103h-f5.tif
Fig. 5 (a) Schematic diagram of the oxygen vacancy in the porous WO3−x film. The XPS spectra of the annealed film, (b) W4f core level spectrum and (c) O1s core level spectrum.

4. Conclusion

In summary, we successfully fabricated the self-doped tungsten oxide film using an environmentally friendly method. The aqueous solution processed method without any toxic solvents is economical and suitable for lab-to-fab translation. By introducing an appropriate amount of the PVP polymer in the precursor and a subsequent carbonization step, a black intermediate product containing amorphous carbon and tungsten compounds has been obtained. After annealing, the polycrystalline WO3−x film with a porous structure and oxygen vacancies was formed via the carbothermal reduction reaction. The large specific surface area and the oxygen vacancy defect together contributed to significant improvements in Li+ ion insertion and extraction processes. Therefore, the self-doped WO3−x film exhibits outstanding electrochromic performances, including the large optical modulation (70% at 680 nm), fast switching speeds and superior coloration efficiency (62 cm2 C−1). Meanwhile, the electrochromic switching behavior is stable in long-term cycling, with almost no attenuation even after two hundred cycles. These results indicate that adjusting the nanostructure and point defect in the WO3−x film by in situ carbothermal reduction is an effective way to refine electrochromic materials. Moreover, the demonstrated green chemistry route can be extended to the defect engineering of other metal oxides for tailoring properties towards various applications.

Conflicts of interest

There are no conflicts to declare.


This work was financially supported by the National Science Foundation of China (No. 61631166004 and 51902250), the Fundamental Research Funds for the Central Universities (xzy012019002), the Natural Science Basic Research Plan in Shaanxi Province of China (No. 2020JQ-035) and the National Postdoctoral Program for Innovative Talents (No. BX201700185). The instrument Analysis Center of Xi’an Jiaotong University is acknowledged for the great helps in measurements.


  1. S. Macher, M. Schott, M. Sassi, I. Facchinetti, R. Ruffo, G. Patriarca, L. Beverina, U. Posset, G. A. Giffin and P. Löbmann, Adv. Funct. Mater., 2020, 30, 1906254 CrossRef.
  2. L. Xie, S. Zhao, Y. Zhu, Q. Zhang, T. Chang, A. Huang, P. Jin, S. Ren and S. Bao, ACS Sustainable Chem. Eng., 2019, 7, 17390–17396 CrossRef.
  3. M. Wang, X. Xing, I. F. Perepichka, Y. Shi, D. Zhou, P. Wu and H. Meng, Adv. Energy Mater., 2019, 9, 1900433 CrossRef.
  4. S. M. Islam, T. S. Hernandez, M. D. McGehee and C. J. Barile, Nat. Energy, 2019, 4, 223–229 CrossRef CAS.
  5. L. L. Wu, H. J. Fang, C. Zheng, Q. Wang and H. Wang, J. Mater. Chem. C, 2019, 7, 10446–10453 RSC.
  6. S. L. Zhang, S. Cao, T. R. Zhang, Q. F. Yao, A. Fisher and J. Y. Lee, Mater. Horiz., 2018, 5, 291–297 RSC.
  7. Y. J. Ke, J. W. Chen, G. J. Lin, S. C. Wang, Y. Zhou, J. Yin, P. S. Lee and Y. Long, Adv. Energy Mater., 2019, 9, 1902066 CrossRef CAS.
  8. Y. D. Shi, Y. Zhang, K. Tang, J. W. Cui, X. Shu, Y. Wang, J. Q. Liu, Y. Jiang, H. H. Tan and Y. C. Wu, Chem. Eng. J., 2019, 355, 942–951 CrossRef CAS.
  9. J. B. Pan, Y. Wang, R. Z. Zheng, M. T. Wang, Z. Q. Wan, C. Y. Jia, X. L. Weng, J. L. Xie and L. J. Deng, J. Mater. Chem. A, 2019, 7, 13956–13967 RSC.
  10. H. J. Fang, P. Y. Zheng, R. Ma, C. Xu, G. Y. Yang, Q. Wang and H. Wang, Mater. Horiz., 2018, 5, 1000–1007 RSC.
  11. E. L. Runnerstrom, A. Llordes, S. D. Lounis and D. J. Milliron, Chem. Commun., 2014, 50, 10555–10572 RSC.
  12. D. Ouyang, Z. F. Huang and W. C. H. Choy, Adv. Funct. Mater., 2019, 29, 1804660 CrossRef.
  13. P. M. Sberna, M. Trifunovic and R. Ishihara, ACS Sustainable Chem. Eng., 2017, 5, 5642–5645 CrossRef CAS PubMed.
  14. Y. Zhan, M. R. J. Tan, X. Cheng, W. M. A. Tan, G. F. Cai, J. W. Chen, V. Kumar, S. Magdassi and P. S. Lee, J. Mater. Chem. C, 2017, 5, 9995–10000 RSC.
  15. K. Shen, K. Sheng, Z. T. Wang, J. M. Zheng and C. Y. Xu, Appl. Surf. Sci., 2020, 501, 144003 CrossRef CAS.
  16. D. Zhou, F. Shi, D. Xie, D. H. Wang, X. H. Xia, X. L. Wang, C. D. Gu and J. P. Tu, J. Colloid Interface Sci., 2016, 465, 112–120 CrossRef CAS PubMed.
  17. D. Han, B. L. Jiang, J. Feng, Y. D. Yin and W. S. Wang, Angew. Chem., Int. Ed., 2017, 56, 7792–7796 CrossRef CAS.
  18. H. Q. Song, C. F. Liu, C. K. Zhang and G. Z. Cao, Nano Energy, 2016, 22, 1–10 CrossRef CAS.
  19. X. G. Wang, M. H. Sun, M. Murugananthan, Y. R. Zhang and L. Z. Zhang, Appl. Catal., B, 2020, 260, 118205 CrossRef CAS.
  20. J. D. Chen, D. N. Yu, W. S. Liao, M. D. Zheng, L. F. Xiao, H. Zhu, M. Zhang, M. L. Du and J. M. Yao, ACS Appl. Mater. Interfaces, 2016, 8, 18132–18139 CrossRef CAS.
  21. Y. Jiang, Z. Wu, L. Jiang, Z. Pan, P. Yang, W. Tian and L. Hu, Nanoscale, 2018, 10, 12003–12010 RSC.
  22. R. A. Sheldon, Green Chem., 2016, 18, 3180–3183 RSC.
  23. W. Thongpan, D. Louloudakis, P. Pooseekheaw, T. Kumpika, E. Kantarak, W. sroila, A. Panthawan, W. Thongsuwan and P. Singjai, Mater. Lett., 2019, 257, 126747 CrossRef CAS.
  24. J. Shang, Z. L. Xiao, L. X. Yu, P. Aprea and S. Y. Hao, Nanotechnology, 2020, 31, 125603 CrossRef PubMed.
  25. R. Godbole, V. P. Godbole, P. S. Alegaonkar and S. Bhagwat, New J. Chem., 2017, 41, 11807–11816 RSC.
  26. R. F. Garcia-Sanchez, T. Ahmido, D. Casimir, S. Baliga and P. Misra, J. Phys. Chem. A, 2013, 117, 13825–13831 CrossRef CAS PubMed.
  27. K. T. Alali, J. Y. Liu, K. Aljebawi, Q. Liu, R. R. Chen, J. Yu, M. L. Zhang and J. Wang, J. Alloys Compd., 2019, 780, 680–689 CrossRef CAS.
  28. J. Feng, J. H. Zhang, Z. Zheng and T. Zhou, ACS Appl. Mater. Interfaces, 2019, 11, 41688–41700 CrossRef CAS PubMed.
  29. G. Xue, Y. Xu, T. P. Ding, J. Li, J. Yin, W. W. Fei, Y. Z. Cao, J. Yu, L. Y. Yuan, L. Gong, J. Chen, S. Z. Deng, J. Zhou and W. L. Guo, Nat. Nanotechnol., 2017, 12, 317–321 CrossRef CAS PubMed.
  30. R. Brahimi, K. Dib, M. Trari and Y. Bessekhouad, Mater. Chem. Phys., 2019, 223, 398–403 CrossRef CAS.
  31. S. Ghosh, S. S. Acharyya, M. Kumar and R. Bal, RSC Adv., 2015, 5, 37610–37616 RSC.
  32. K. Movlaee, P. Periasamy, T. Krishnakumar, M. R. Ganjali, S. G. Leonardi, G. Neri, M. Chavali, P. Felix Siril and V. P. Devarajan, J. Alloys Compd., 2018, 762, 745–753 CrossRef CAS.
  33. T. D. Nguyen, L. P. Yeo, T. C. Kei, D. Mandler, S. Magdassi and A. I. Y. Tok, Adv. Opt. Mater., 2019, 7, 1801389 CrossRef.
  34. Z. J. Bi, X. M. Li, X. L. He, Y. B. Chen, X. K. Xu and X. D. Gao, Sol. Energy Mater. Sol. Cells, 2018, 183, 59–65 CrossRef CAS.
  35. J. L. Zhou, Y. X. Wei, G. Luo, J. M. Zheng and C. Y. Xu, J. Mater. Chem. C, 2016, 4, 1613–1622 RSC.
  36. B. W. Zhao, S. J. Lu, X. Zhang, H. Wang, J. B. Liu and H. Yan, Ionics, 2016, 22, 261–267 CrossRef CAS.
  37. A. Azam, J. Kim, J. Park, T. G. Novak, A. P. Tiwari, S. H. Song, B. Kim and S. Jeon, Nano Lett., 2018, 18, 5646–5651 CrossRef CAS.
  38. A. H. Ribeiro, A. Fakih, B. Zee, L. Veith, G. Glaser, A. Kunz, K. Landfester, P. W. M. Blom and J. J. Michels, J. Mater. Chem. C, 2020, 8, 6528–6535 RSC.
  39. X. H. Xia, Z. L. Ku, D. Zhou, Y. Zhong, Y. Q. Zhang, Y. D. Wang, M. J. Huang, J. P. Tu and H. J. Fan, Mater. Horiz., 2016, 3, 588–595 RSC.
  40. Z. Xing, X. K. Zeng, A. Deletic, D. McCarthy, G. Wang and X. W. Zhang, Chem. Commun., 2016, 52, 6985–6988 RSC.
  41. S. Rahimnejad, J. H. He, W. Chen, K. Wu and G. Q. Xu, RSC Adv., 2014, 4, 62423–62429 RSC.
  42. S. Zhuiykov, E. Kats, B. Carey and S. Balendhran, Nanoscale, 2014, 6, 15029–15036 RSC.
  43. K. Bon-Ryul, K.-H. Kim and H.-J. Ahn, Nanoscale, 2019, 11, 3318–3325 RSC.
  44. Z. T. Li, Y. F. Dong, J. Z. Feng, T. Xu, H. Ren, C. Gao, Y. R. Li, M. J. Cheng, W. T. Wu and M. B. Wu, ACS Nano, 2019, 13, 9227–9236 CrossRef CAS PubMed.
  45. Y. Zhang, Z. Y. Ding, C. W. Foster, C. E. Banks, X. Q. Qiu and X. B. Ji, Adv. Funct. Mater., 2017, 27, 1700856 CrossRef.


These authors contributed equally to this work.

This journal is © The Royal Society of Chemistry 2020