Spectral conversion properties were investigated by preparing SCFs under various conditions by the casting method onto glass (25 mm × 25 mm) and a copper–indium–gallium–selenide solar cell (CIGS, 25 mm × 25 mm) (Figure S5, Supporting Information). Power-conversion efficiency was then calculated from current–voltage curves (Figure 7). Table 3 summarizes the increase in rate of conversion efficiency (Δη) and the short-circuit current (ΔJSC) for the SCF-covered cell, as well as for the bare cell.
In our previous work, we reported the increase in power conversion efficiency of an amorphous silicone solar cell using a g-Pyr/polystyrene film.[9b] In the case of the as-prepared CIGS solar cell, we also observed an increase in power-conversion efficiency using the g-Pyr/EVA film (ΔJSC = 2.68%, Δη = 3.57%). The highest power-conversion efficiency was observed for the 1 mol% C6-doped g-Pyr/EVA film-covered cell (η = 15.77%). This film exhibited an improvement in power-conversion efficiency of 4.85% compared to the bare cell (η = 15.04%). It was found that only JSC, a property essential to power-conversion efficiency, showed a significant increase (ΔJSC = 4.57%), while all other factors (open voltage (VOC), and fill factor (FF)) remained relatively unchanged. This finding strongly supports the presumption that the increase in power-conversion efficiency was due to the spectral conversion effect of SCF from UV to visible light. Figure 7b exhibits a C6 concentration-dependent Δη change. A decrease in the Δη value was observed when increasing the concentration of C6 (0–10 mol%); this might be caused by the strong C6 absorption band in the visible region (400–500 nm). Incident photon-to-current conversion (IPCE) value of the CIGS solar cell, an indicator of spectral sensitivity character, was quite low in the pyrene absorption region (300–350 nm, average IPCE <10%). However, in the C6 absorption region (400–500 nm), the average IPCE value was found to be more than 80%, a likely reason why the strong absorption band in the visible region decreases the power-conversion efficiency. Transmission and fluorescence spectra of the C6-doped g-Pyr/EVA film support this assumption. For example, 1 mol% C6-doped film showed high transparency in the visible region (T% = 80%). In contrast, 10 mol% C6-doped film showed low transparency (T% = 50%) in the visible region due to the strong absorption of C6 (Figure S6a,b, Supporting Information). For the same reason, NR-doped g-Pyr/EVA film did not exhibit an effective increase in power-conversion efficiency. In the absence of g-Pyr, no enhancement in the power-conversion efficiency was observed by addition of each dye (C6, NR and Rub) as an additive into an EVA film because of the extremely low concentration. Therefore, the significant increase of the power-conversion efficiency is not delivered by these additives but realized by cooperation of g-Pyr with the additives. On the other hand, the modified cells with Pyr/EVA, C6-doped Pyr/EVA, and NR-doped Pyr/EVA films showed slight increases in power-conversion efficiencies (Δη = 1.24%, 0.49% and 0.29%, respectively), while the cell coated with only EVA showed no change.
The advantages of the phase-separated system were explained based on comparison of transmission and fluorescence spectra between C6-doped g-Pyr/EVA and Pyr/EVA films (Figure S6c, Supporting Information). A small amount (1 mol%) of C6-doped g-Pyr/EVA film showed strong fluorescence at 492 nm as C6 emission, due to efficient FRET in the phase-separated nano-domain. In contrast, in the case of Pyr/EVA film as non-assembling system, 4-fold C6 (4 mol%) dopant loadings were required to obtain a similar fluorescence emission. However, as indicated in the above discussion it is expected that a large amount of C6 act as visible-light absorbents in decreasing the power-conversion efficiency. From these results, we may state that our phase-separated system, which provides control over emission wavelength by efficient FRET with addition of small amounts of acceptor molecules, is suitable for SCF.