Limitation of room temperature phosphorescence efficiency in metal organic frameworks due to triplet-triplet annihilation

Abstract

The effect of triplet-triplet annihilation (TTA) on the room-temperature phosphorescence (RTP) in metal-organic frameworks (MOFs) is studied in benchmark RTP MOFs based on Zn metal centers and isophthalic or terephthalic acid linkers (ZnIPA and ZnTPA). The ratio of RTP to singlet fluorescence is observed to decrease with increasing excitation power density. Explicitly, in ZnIPA the ratio of the RTP to fluorescence is 0.58 at 1.04 mW cm−2, but only 0.42 at (the still modest) 52.6 mW cm−2. The decrease in ratio is due to the reduction of RTP efficiency at higher excitation due to TTA. The density of triplet states increases at higher excitation power densities, allowing triplets to diffuse far enough during their long lifetime to meet another triplet and annihilate. On the other hand, the shorter-lived singlet species can never meet an annihilate. Therefore, the singlet fluorescence scales linearly with excitation power density whereas the RTP scales sub-linearly. Equivalently, the efficiency of fluorescence is unaffected by excitation power density but the efficiency of RTP is significantly reduced at higher excitation power density due to TTA. Interestingly, in time-resolved measurements, the fraction of fast decay increases but the lifetime of long tail of the RTP remains unaffected by excitation power density. This may be due to the confinement of triplets to individual grains, leading decay to be faster until there is only one triplet per grain left. Subsequently, the remaining “lone triplets” decay with the unchanging rate expressed by the long tail. These results increase the understanding of RTP in MOFs by explicitly showing the importance of TTA in determining the (excitation power density dependent) efficiency of RTP. Also, for applications in optical sensing, these results suggest that a method based on long tail lifetime of the RTP is preferable to a ratiometric approach as the former will not be affected by variation in excitation power density whereas the latter will be.