A major challenge for energy-efficient smart window technology is to ensure the durability of electrochromic (EC) devices over a service life of more than 20 years. Hence, their degradation under operating conditions must be better understood and preferably avoided. In this paper we review our recent results from a fundamental study of the ageing kinetics of EC tungsten oxide and nickel oxide thin films, as well as electrochemical rejuvenation mechanisms that are able to restore the films to their initial state. The ageing of tungsten oxide films upon cycling in the range 2-4 V vs. Li/Li⁺ displays a power-law decrease of the charge capacity as a function of cycle number. In the case of severe ageing at potentials below 2.0 V there is evidence for a superposition of two processes. In the case of nickel oxide the ageing process seems to be more complex and it can be described by either a power-law or a stretched exponential function. The ageing of nickel oxide was also observed to decrease and even stop after a few thousand cycles just prior to final breakdown of the film. The decay of charge capacity can be understood in terms of models built on so-called dispersive chemical kinetics.

The term “rejuvenation” describes electrochemical processes that may restore aged films to full EC functionality. During severe ageing of tungsten oxide coatings, Li ions are trapped in the film, and subsequently these ions can be released by application of a high electrochemical potential for a few hours. The trapping/detrapping of Li has been verified by Elastic Recoil Detection Analysis (ERDA) measurements. Very recently we have found that a similar process can be applied to EC nickel oxide films, although using a lower potential than for the case of tungsten oxide. For nickel oxide, the rejuvenation mechanism is not yet clear, but ERDA measurements are in progress and are expected to give detailed information.

An increased understanding of ageing and rejuvenation will facilitate large-scale practical application of electrochromic materials. Our phenomenological models of ageing may make it possible to predict the service life of EC devices using empirical relationships. An understanding of rejuvenation processes will be of major importance in order to prevent or minimize performance decrease under extended cycling. The next step will be to investigate whether rejuvenation treatments can be applied successfully to full EC devices.