Smart materials

Off and on reflection

Article metrics

The architect's palette has been enriched by smart materials, which enable a building to adjust its environment to the prevailing conditions. At a meeting last monthFootnote 1, Peter Duine (Philips Research Labs, Eindhoven) described a tantalizing addition to the palette: windows that can be reversibly switched from a transparent to a mirror state.

The appeal of a switchable mirror might be obvious to gimmick addicts, but are there more practical attractions in such devices? For privacy purposes it beats drawing the curtains, but one can also imagine applications in energy-conscious architectural design — using switchable mirrors as elements of smart systems that control the amount of sunlight admitted through a window. And as a means to influence the path of light beams, the mirrors might be put to good use in display devices and interior lighting systems, particularly when the light sources are hard to reach and adjust.

The new work, which brings these technological goals closer to fruition, stems from a report in 1996 of switchable optical properties in yttrium and lanthanum1. The underlying principle is a transition between a reflective metallic state and a translucent insulating state, induced by converting a thin film of a metal to its ionic hydride. All of the alkali, alkaline-earth and rare-earth metals form hydrides on exposure to hydrogen gas, whereupon ionization of the metal can empty the conduction band and make the compound insulating. A thin enough (sub-micrometre) film may then be transparent.

For trivalent rare-earth metals like yttrium and lanthanum, however, the situation is slightly more complicated. In particular, the metals can exist in both a divalent and a trivalent state, corresponding to the hydrides MH2and MH3. Three different phases occur on exposure to hydrogen gas: the α-phase, a solution of a small amount of interstitial hydrogen in the metal; the β-phase (the dihydride); and the γ-phase (the trihydride). The two hydride phases (β and γ) can easily be interconverted by altering the hydrogen pressure, as the gas can readily diffuse in and out of the thin film. The dihydride retains a partially filled conduction band, and so provides the metallic mirror, whereas the trihydride is transparent.

To make a device, the thin metal films deposited on glass are enclosed in a sealed cell after being coating with a 20-nm film of palladium to prevent oxidation. Palladium is a renowned hydrogen ‘sponge’, and so hydrogen gas introduced into the cell has easy access to the rare-earth metal. Graded-diffusion experiments, in which the gas is permitted access to the films from one end only, show the formation of the γ-, β- and α-phases, successively further away from the hydrogen source region.

The early work1 on yttrium demonstrated the principle but left much to be desired in terms of optical properties. For one thing, yttrium dihydride is not fully reflective, but has a spectral transmission window: it is dark red in transmission, blue in reflection. The trihydride, meanwhile, is not perfectly transparent, but has a yellowish tinge, because its bandgap is not large enough to prevent absorption of high-energy blue photons.

In both respects, a magnesium-based system would do better. But the transition between metallic magnesium and the insulating dihydride (there is no trihydride in this case) is not readily reversible, and the diffusivity of hydrogen in magnesium is low (so the formation of the hydride is slow). But the Philips team have found a compromise, in which the good optical properties of magnesium are combined with the good switchability of rare-earth elements. Their favoured material at present is an alloy of magnesium and gadolinium2: this alloy can be switched in about one second, has no transmission window in the metallic state, and for magnesium contents of around 50% the insulating state is virtually colourless. The transparent phase forms at high (>1 atmosphere) hydrogen pressures, and the metallic reflective state at low pressures (for which the metal/hydrogen ratio in the film is around 0.8). In between these two extremes a third phase is found which is dark and has low reflectivity (Fig. 1).

Figure 1: Transparent, dark and reflective states of a Gd30Mg70 alloy.


The states can be switched by pumping hydrogen in and out of a cell that contains the alloy film.

But devices that pump hydrogen gas are not destined to find a thriving market. To make a user-friendly package, the hydrogen will need to be transported in a condensed medium. Duine also reported a ‘wet’ electrochemical cell, in which the hydrogen is shuttled back and forth between the gadolinium-magnesium alloy film at one electrode, and a liquid electrolyte of potassium hydroxide in contact with another electrode, through the reduction of water (ref. 3): H2O + e- → H + OH-. What is really needed, however, is a solid-state cell in which the hydrogen is kept in some ‘storage’ layer and is transported through a solid electrolyte. ‘Solid’ proton conductors are widely known already, such as the polymer Nafion that is used in fuel cells. But a solid electrolyte that can deliver hydrogen for uptake as anions may be harder to find. Until then, we will have to keep drawing the curtains.


  1. 1.

    *Fall Meeting of the Materials Research Society, Boston, 1-5 December 1997.


  1. 1

    Huiberts, J. al. Nature 380, 231–234 (1996).

  2. 2

    Van der Sluis, P., Ouwerkerk, M. & Duine, P. Appl. Phys. Lett. 70, 3356–3358 (1997).

  3. 3

    Notten, P., Kremers, M. & Griessen, R. J. Electrochem. Soc. 143, 3348–3353 (1996).

Download references

Author information

Rights and permissions

Reprints and Permissions

About this article

Further reading


By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.