Optical approaches to raise JSC, and therefore efficiencies, are based on the increase of optical absorption caused by either an enlargement of the photon path length through the working electrode or light trapping effects occurring within the TiO2 electrodes. JSC can be attained by integrating the product of the ratio between the solar spectral irradiance and the photon energy density, F (X), and the photon-to-current conversion efficiency, IPCE, of the cell over the wavelength of the incident light:
Jsc =f q4(X)F (X)IPCE (X) dX (3)
Here q is the electron charge and ^ (X) is a factor that accounts for the losses at the air- substrate interface. IPCE can be expressed as the product of light harvesting efficiency (LHE) and the electron-transfer yield Ф(Х), that is the product of the electron injection yield and the charge collection efficiency.
IPCE(X) = LHE(X)Q(X) (4)
The LHE or optical absorptance, A, at a certain wavelength is defined as the fraction of incident photons that are absorbed by the dyed electrode: A=Ia/Io, where Io is the incident intensity and IA the intensity absorbed. Therefore the relationship between LHE and JSC is given by the expression:
Jsc =J q^(X) F (X)LHE (X) Ф(Х) dX (5)
In a first approximation, absorptance for a standard dyed electrode is related with the extinction coefficient and the concentration of the dye. Since only a monolayer of dye molecules is attached to the surface of the TiO2 nanoparticles, its total amount is directly related to the oxide layer thickness. For example, a 7 pm thick dye sensitized (N719) mesoscopic electrode can absorb nearly the 80% of incident photons at the maximum absorption wavelength. However, photons still in the visible range but of lower energy are weakly absorbed. Looking for a dye absorption enhancement, devices having thicker electrodes (around 10 pm) have been previously reported (Ito et al, 2006). Nevertheless, the thickness of TiO2 layer cannot be increased at will without affecting its mechanical properties, reaching mass transport limitations in the electrolyte or/and reducing the photovoltage of the cell. In addition, electrons injected in the conduction band must travel a longer distance to reach the back contact, increasing the probability of recombination at grain boundaries and diminishing both current and voltage of the cell, as experimentally demonstrated (Ito et al., 2008). Finally, another disadvantage to scale-up the device would appear due to the high cost of sensitizer dyes (approx. 1000€/g). The thicker the electrodes, the higher the dye loads required, therefore raising the final cost of the DSSC. All these nondesirable features make it preferable to increase the absorption of the cell for a given dye and film thickness, modifying the optical path length within the film and improving the spectral response of the photoelectrode. Keeping in mind that a standard dye sensitized layer of around 7-8 pm thick will not absorb light strongly, these achievements can be obtained by reflecting light back to the dyed electrode. In what follows, we provide an overview of the different approaches taken towards the integration of optical passive components in order to increase the power conversion efficiency of DSSCs.
