To characterise the kinetics of the colouring and bleaching processes during illumination and bleaching in the dark, we needed to have an experimental set-up which measures the transmittance in the same way in both cases. For this purpose, we developed the set-up illustrated in fig. 2. A halogen lamp illuminates the photoelectrochromic sample, and the intensity of the transmitted light is detected by a silicon photodiode. The light intensity of the halogen lamp on the surface of the photoelectrochromic cell corresponds to 1 sun (1000W/m2), taking into account the mismatch factor of dye solar cells. The two electrodes of the photoelectrochromic device are connected via a variable shunt resistance and a switch. For all configurations, the TiO2 layer is always directed towards the lamp, so that the colouring of the Wo3 does not alter the light intensity on the TiO2.

The shunt resistance was chosen to be 10 Q, which is similar to the resistance of the TCO layer. With this construction, the voltage in the open circuit state and the current in the short circuit state were measured. For the dark state, an optical filter was placed between the lamp and the sample. This filter absorbs all the light with wavelengths below 715 nm. Above this wavelength, the dye is not sensitive and no electrons are excited, as we demonstrated by spectral response measurements. It was necessary to install several collimators in order to suppress scattered light from the environment. For the illuminated state, the filter was removed and placed between the sample and the photodiode detector. In this way, the optical signal is the same for both filter positions, and the transmittance signals for the dark and illuminated states are equivalent.

The measured value of the transmittance is a convolution of the spectra of the halogen lamp, the filter and the photodiode. In order to calculate the visible (or solar) transmittance from the measured value of the transmittance, the set-up was calibrated in the following way: A special PEC device was coloured to different extents by applying different voltages (0V, 0.3V, 0.4V, 0.5V, 0.6V, 0.7V) for 30 min, so that an equilibrium state was achieved. While holding this voltage with the potentiostat, the transmittance was measured with our set-up and immediately afterwards with the spectrometer which recorded the spectrum between 320 and 2000 nm. From these spectra, the visible (and solar) transmittance and
optical density were calculated. The visible (solar) optical density plotted versus the optical density determined with our set-up shows a linear dependence to a good approximation. A linear fit gave a simple equation to calculate the visible (solar) optical density from the values determined with our set-up.

These calibration measurements were made with a special PEC cell which was made without the dye. Coloured by an external voltage, this special cell has the same transmittance spectrum as the normal PEC cell coloured by illumination, because the amount of dye in the normal cell is very low. Using this special cell for the calibration had the advantage that the measured transmittance is not influenced by the different lighting conditions inside the spectrometer and in our set-up.

The condition for this calibration is that the form of the transmittance spectrum is independent of the depth of coloration. This condition is fulfilled because the dependence of the optical density on the charge is linear: The coloration efficiency is independent of the intercalation degree x [8]. On the other hand, the optical density OD represents the intercalation degree x (x: ratio of electrons and Li+ cations per W-atom in the Wo3):

Long-duration transmittance measurements of self-bleaching in the dark under open circuit conditions were made with a light-emitting diode with an intensity maximum at a wavelength of 655 nm without any filters. This light-emitting diode was switched on once every minute for a short measurement so that it did not influence the coloration of the photoelectrochromic device.

Transmittance spectra were measured with a Perkin-Elmer 330 Spectrometer.