Corrigendum: Bright Near-Infrared Fluorophores Based on Squaraines by Unexpected Halogen Effects.

Abstract

During recent research carried out in our laboratory directed toward improved near infrared (NIR) emitters1 we became aware of the fact that some of our previously reported fluorescence quantum yield values are incorrect. Accordingly, the whole series of dyes described in this manuscript was re-investigated with our meanwhile improved instrumentation. These results are now compiled in the corrected Table 1 and the relevant parameters for the applications of these dyes in fluorescence imaging applications are illustrated in the corrected Figures 1 and 2. UV/Vis/NIR absorption spectra of squaraines 6 a (red dotted line), 6 b (dashed line), and 6 d (thin solid line) in CH2Cl2 (c=10−5 m, at 25 °C) and the fluorescence spectrum (λex=780 nm) of squaraine 6 d (thick solid line) in CH2Cl2 (at 25 °C). Influence of halogen atoms on the optical properties (squares: Φfl; circles: ϵmax) of squaraines 5 a–d (filled symbols) and 6 a–d (open symbols). Dye λmax/ nm ϵ/ M−1 cm−1 λem/ nm Φem ϵ Φem[c]/ M−1 cm−1 5 a[a] 684 209 000 701 0.36±0.02 75 200 5 b[a] 694 215 000 711 0.51±0.02 109 650 5 c[a] 695 218 000 712 0.54±0.02 117 700 5 d[a] 699 223 000 716 0.55±0.02 122 650 6 a[b] 870 254 000 885 0.009±0.001 2250 6 b[b] 889 259 000 901 0.007±0.001 1800 6 c[b] 893 271 000 905 0.008±0.001 2150 6 d[b] 897 271 000 911 0.008±0.001 2150 We like to note that meanwhile also a corrected value for the utilized fluorescence standard rhodamine 800 in ethanol has been suggested (Φfl=0.252) but that we decided to go on with the previously reported value to allow for a comparison of our old and new data for dyes 5 a–d whose values are reported relative to this fluorescence standard. For this series of dyes 5 a–d (filled symbols in Figure 2) there are only modest corrections compared to our original data. In contrast, for NIR emitting dyes 6 a–d (open symbols in Figure 2) we noticed severe deviations, i.e. the newly determined fluorescence quantum yields are more than one order of magnitude lower compared to the previously published ones. During our studies we also noticed that the commonly utilized NIR fluorescence standard indocyanine green in dimethylsulfoxide (Φfl=0.133) is rather unstable and should be not used for extended measurement times. For this reason values provided in Table 1 and in Figure 2 for dyes 6 a–d are now determined with an Edinburgh Instruments FLS980-D2D2-ST spectrometer equipped with an integrating sphere F-M01 as absolute fluorescence quantum yields and corrected for reabsorption.4 While the absolute method for Φfl determination suffers from lower sensitivity and intrinsic reabsorption, it enables a better choice for the selection of an appropriate excitation wavelength (λex) for narrow-band absorbers, which is beneficial for chromophores with low Φfl, especially in the NIR region, where the liquid-nitrogen-cooled NIR photomultiplier tube (PMT) exhibits intrinsically a much lower signal-to-noise ratio than common UV-Vis PMTs.5 We regret the erroneous fluorescence quantum yields published in our earlier work and apologize for all inconvenience caused to other researchers when reporting new NIR fluorophores. With regard to the main topic of our original Communication we note that the unusual halogen effect has been confirmed for the series of dyes 5 a–d (wavelength, extinction coefficient, fluorescence quantum yield as well as brightness all increase from chloride to iodide derivatives) but is less pronounced for the series 6 a–d (Figure 2). The authors thank Chia-An Shen and Dr. Matthias Stolte for their efforts with the re-investigation of the absorption and fluorescence properties.