Towards full-colour tunability of inorganic electrochromic devices using ultracompact fabry-perot nanocavities

Intercalation-based inorganic materials that change their colours upon ion insertion/extraction lay an important foundation for existing electrochromic technology. However, using only such inorganic electrochromic materials, it is very difficult to achieve the utmost goal of full-colour tunability for future electrochromic technology mainly due to the absence of structural flexibility. Herein, we demonstrate an ultracompact asymmetric Fabry-Perot (F-P) nanocavity-type electrochromic device formed by using partially reflective metal tungsten as the current collector and reflector layer simultaneously; this approach enables fairly close matching of the reflections at both interfaces of the WO3 thin layer in device form, inducing a strong interference. Such an interference-enhanced device that is optically manipulated at the nanoscale displays various structural colours before coloration and, further, can change to other colours including blue, red, and yellow by changing the optical indexes (n, k) of the tungsten oxide layer through ion insertion.

and 47.1 cm 2 C -1 , respectively. The F-P nanocavity exhibits a CE of 48.6 cm 2 ·C -1 , which is slightly higher than those of the normal current collector-covered electrochromic electrodes. Figure 13. Optical switching of different electrochromic electrodes, which determines the colouration (tc) and bleaching (tb) times. For traditional FTOand ITO-covered electrochromic electrodes, the coloration times are found to be 2.9 and 7.5 s, and the bleaching times are 4.2 and 8.7 s, respectively. For the metal W-covered F-P nanocavity-type electrochromic electrode, the coloration time is 3.3 s, and the bleaching time is 3.1 s. Obviously, the switching speed of the metal W-covered F-P nanocavity-type electrochromic electrode is faster than that of the ITO-covered electrochromic electrode but is comparable to that of the FTO-covered electrochromic electrode. Figure 14. Normalized charge capacity profiles of a W-covered F-P nanocavity-type electrochromic electrode over 1000 voltammetric cycles at 20 mV s -1 between -0.8 and 0.5 V (vs. Ag/AgCl). Good cycling stability is demonstrated by a high capacity retention of 81% after 1000 cycles.  As can be seen from the table, the electrical conductivity of the substrates has a substantial effect on the electrochromic performances of the devices (particularly on switching times). Thus, given the same substrate surface area, the switching speed of the metal W-based F-P nanocavity-type electrochromic electrode is faster than the ITO-based electrochromic electrode but comparable to the FTO-based electrochromic electrode.

Characterizations.
Field emission scanning electron microscopy (FE-SEM) analysis was performed on a FEI Quanta FEG field emission scanning electron microscope. XRD patterns of the prepared samples were recorded on a Bruker AXS D8 Advance X-ray diffractometer with a Cu Ka radiation target (40 V, 40 A). Tapping-mode atomic force microscopy (AFM; Bruker Instruments Dimension ion) with a silicon-tip cantilever (0.24 N m −1 ) was used to characterize the top surface of the electrode.
Calculation of coloration efficiency.

CE = ΔOD/ΔQ = log(Tb/Tc)/ΔQ, or CE = ΔOD/ΔQ = log(Rb/Rc)/ΔQ
Here ΔQ is the inserted charge that promotes the change (ΔOD) in the optical absorbance and Tb (Rb) and Tc (Rc) refer to the bleached and coloured transmittances (reflectance) at a certain wavelength, respectively. The CE can be accordingly evaluated from the slope of the plots of ΔOD versus ΔQ.

Calculation of optical constants for sputtering WO3 and W.
The single-layer WO3 and W deposited on silicon (100)  Computational modelling of transmittance/reflectance spectra.
We use the characteristic matrix method derived from the basic principles of ], (S1) where B and C is the normalised electric and magnetic fields at the front interface and δ1 is the phase thickness. This last property is defined as: where N1 and d1 is the refractive index and the thickness of the first layer, respectively; θ1 is the angle obtained from Snell's law (light is considered to have normal incidence for our study); and λ is the wavelength. Optical admittance values Y1 and Y2 are given by: where Y0 is the optical admittance in free space. The electric and magnetic components of Y1 and Y2 are equal.
The transmittance/reflectance spectra of the F-P nanocavities are thus calculated by equations (S5)

Supplementary Note 1. The material selection of metal layer.
According to common understanding, one of the greatest weaknesses of inorganic electrochromic materials is their monotonous colour changes. Thus, the ultimate goal of full-colour tunability for future electrochromic technology has been difficult to achieve with devices based on these typical materials. Our present work aims to broaden the colour versatility of inorganic electrochromic materials by introducing ultracompact Fabry-Perot (F-P) nanocavities into relative electrochromic devices. For a proposed F-P nanocavity-type electrochromic device, the metal layer is the central component, as this layer acts as a reflecting mirror for the optical interference that allows the generation of different colours. Generally, the choice of metal layer in our work is based on the following criteria.
(1) Large dip depth. As is well known, the depth of the resonance dip (that is, the difference in intensity between the dip minimum and the threshold) in a reflectance spectrum is an important indicator of resonance efficiency. 2 The greater the depth of this dip, the better the efficiency of resonance that will be achieved. Supplementary Figure17 shows the dependence of the depth of resonance dip on refractive index (n) and extinction factor (k) for the metal layers analysed in our work (for these measurements, 200-nm-thick WO3 is set as the dielectric layer). The yellow region of the spectrum corresponds to a large dip depth, while the blue region indicates a small dip depth. As can be seen, the perceived (n, k) values for Al, Ag, Au and Cu are either fully or partly located in the blue region, indicating that the colour gamut was limited because a strong resonance could not be excited over the whole visible range. In contrast, the (n, k) values for Ni, Ti, Cr, V, Zn and W are all located in the yellow region, suggesting that a wide colour gamut is established by strong resonance behaviours operating across the entire visible spectrum.
(2) Good stability. We choose metal layers with good environmental and electrochemical stability.
(3) Strong adhesion characteristics. We choose metal layers that adhere well to the substrate and the electrochromic layer.
According to the above criteria, W is selected as the ideal choice for the proposed F-P nanocavity-type electrochromic device.

Supplementary Note 3.
For the experiments illustrated in Figure 2f, the reflectance valley of the F-P nanocavity-type electrode is red-shifted as the thickness of the WO3 layer increased.
These thickness-dependent shifts can be understood using the following equation: = 4 1 2 + 1 ( = 0, 1, 2 … ) where DIV is the location of the wavelength intensity minima in the reflective spectrum (which has also been described as the destructive interference valley, DIV); d is the film thickness; and k is the DIV order. As can be clearly inferred from this equation, the wavelength positions of the DIV will be gradually red-shifted as the thickness of the tungsten oxide layer increases, thus resulting in different structural colors.