Effect of molecular filtering and electrolyte composition on the spatial variation in performance of dye solar cells

Abstract

It is demonstrated that the molecular filtering effect of TiO2 has a significant influence on dye solar cell (DSC) performance. As electrolyte is injected to a DSC, some of the electrolyte components adsorb to the surface TiO2 (here 4-tert-butylpyridine and 1-methylbenzimidazole) and accumulate near the electrolyte filling hole resulting in varying electrolyte composition and performance across the cell. The spatial performance distribution was investigated with a new method, the segment cell method. Not only is the segmented cell method simple and cheap when compared to the only other method for examining spatial variation (photocurrent mapping), it also has the major advantage of allowing the spatial variation in all other operating parameters to be assessed. Here the molecular filtering effect was to influence the cell performance in case of all the five studied electrolytes causing up to 35 % losses in efficiency. Raman spectra indicated that the loss in photocurrent in the electrolyte filling was in correlation with the loss of thiocyanate ligands suggesting that dye regeneration may also be a significant factor in addition to electron injection in some of the cells. There were also shifts in the absorption spectra the photoelectrodes which further supported changes in the thiocyanate ligands.Besides absorption changes, there were additional shifts in the IPCE spectra which may relate to deprotonation of the dye. The efficiency losses were reduced to ~10 % with contemporary electrolyte compositions.

The effect of electrolyte filling method on the performance of dye-sensitized solar cells

“…Besides changes in the electron injection, [26][27][28][29][30][31][32] there are also other reactions which have been reported to affect photocurrent when changing the concentration of tBP, BI and Li + such as regeneration of dye, light harvesting and electron collection. 5,25,33,34 In our previous study, the relation between the thiocyanate (SCN) in the Raman spectra and the JSC suggests that some of segmented cells may have suffered from dye regeneration problems. 5 However in the present study this is not likely to be an issue since the presence of GuSCN in both electrolytes significantly increases the amount of SCN surrounding the dyes.…”

“…In our previous work, insitu Raman measurements were conducted on the segmented cells to investigate the surface composition. 5 Unfortunately, the key electrolyte components such as tBP or Nmethyl-benzimidazole (NMBI) did not show a characteristic response in the Raman spectra of a complete DSSC and thus their spatial distribution could not be directly verified. 5 In this contribution the focus is on the understanding the mechanisms causing photovoltage and photocurrent variation.…”

“…5 Unfortunately, the key electrolyte components such as tBP or Nmethyl-benzimidazole (NMBI) did not show a characteristic response in the Raman spectra of a complete DSSC and thus their spatial distribution could not be directly verified. 5 In this contribution the focus is on the understanding the mechanisms causing photovoltage and photocurrent variation. The electrolyte components affect the electrical performance of the cell in a particular manner and from that it is possible to evaluate which electrolyte components are dominating the effect.…”

“…4 Our later study indicated that effect could be reduced but not avoided by selection of suitable electrolyte composition. 5 Since the spatial variations are related to electrolyte filling (as well as electrolyte composition), it suggests that the selection of electrolyte application method could affect the cell performance in particular at larger scale. The purpose of this study is to examine the effects of two most commonly used electrolyte filling methods: simple injection and repeated pumping.…”

“…6 The latter one of these is similar to the electrolyte that resulted in one of smallest spatial performance distribution and the highest overall efficiency in our previous work. 5 In this investigation we apply the segment method 4 which was developed in our initial evaluation of the spatial distribution losses and discussed in detail in our previous study. 5 The main advantage of the technique compared to the only other approaches for detecting spatial variations (such as photocurrent imaging) [7][8][9][10] is that segmented cell method allows the examination of all the photovoltaic parameters separately in the different parts of the cell.…”