Anion identification using silsesquioxane cages† †Electronic supplementary information (ESI) available: Experiments, characterization and spectroscopic studies. See DOI: 10.1039/c8sc02959h

Anthracene-conjugated octameric silsesquioxane cages thermodynamically display intramolecular excimer formation, which can be used to identify anions through the change of fluorescence.


Chemicals and Instruments
9-Bromoanthracene, PPh3, triethylamine, tetrabutylammonium salts and Pd(OAc)2 were purchased from TCI Chemicals and Sigma Aldrich, while octavinylsilsesquioxane (OVS) was prepared according to a previous literature report. 1 The list of chemicals suppliers are provided table S1. Deionized (DI) water was obtained from Ultra Clear SIEMENS with ASTM type 2. The commercial solvents of acetone, DCM, were further distilled. The AR grade of THF, Toluene, DMF and DMSO were used without purification. The silica gel (No. 60) used for column chromatography were purchased from Merck&Co., Inc. 19 F NMR results were obtained using a Bruker-AV 400 high-resolution magnetic resonance spectrometer. FT-IR spectra were recorded using the attenuated total reflectance (ATR) technique on a Bruker model Alpha spectrometer.
High-performance MALDI-TOF analysis was performed on a Bruker autoflexTM series instrument. UV-Vis spectroscopy was performed on a UV-Vis spectrophotometer (Shimadzu UV-2600), while all fluorescence spectra were recorded using a spectrofluorometer (Horiba FluoroMax4+, integration time 0.1 s, slit width 2 nm). These conditions were also employed for measurement of quantum yields, although such measurements required the use of a BaO coated spherical cube.

Quantum efficiency measurement
The quantum efficiencies were measured by comparing between solvents as a blank and sample according to this equation (original from Horiba with sphere cuvette correction): P = ∆ Area under emission curve ∆ Area under absorption curve the default mode for quantum yield measurement was set at slitwidth 3 nm, integration time 1 sec and increment of emission 3 nm per step. After that parabolic spherical barium oxide was equipped to cover the quartz cell. This quantum efficiency measurement does not require reference, except for calibration using Rhodamine B. 100 equiv. of anions were added into 6 μM of AnSQ and wait

Quantitative Analysis of Anion Detection
A polar solvent, DMSO, was selected for anion association studies due to high complex stability, fast formation kinetics and strong excimer emission in this medium. Binding, or association, constants (Ka) were calculated using the Benesi-Hildebrand equation as below.
(a) for fluorescence studies and (b) for absorption studies In equations (a) and (b), I0 and I refer to fluorescence intensities prior to, and post-anion addition.
Likewise, A0 and A refer to absorbance values prior to, and post-anion addition.
Limits of detection (LOD) and quantitation (LOQ) were obtained from titration results and calculated from the linear region using the following equations, with Sb referring to the standard deviation.
The standard deviation (Sb) can be calculated by the below equation, where S is the slope of the calibration curve.
In this work, the x value refers to the concentration of added anions, with y reflecting the fluorescence intensity and absorbance values.    Figure S10c. Calibration curve from AnSQ fluorescence upon addition of OHand PO4 3for limit of detection (LOD) and limit of qualitative (LOQ) calculation in DMSO S65 Figure S10d. Calibration curve from AnSQ fluorescence upon addition of Fand CNfor limit of detection (LOD) and limit of qualitative (LOQ) calculation in DMF S66 Figure S10e. Calibration curve from AnSQ fluorescence upon addition of OHand PO4 3-for limit of detection (LOD) and limit of qualitative (LOQ) calculation in DMF S67 Figure S10f. Calibration curve from AnSQ fluorescence upon addition of Fand CNfor limit of detection (LOD) and limit of qualitative (LOQ) calculation in acetone S68 Figure S10g. Calibration curve from AnSQ fluorescence upon addition of OHand PO4 3-for limit of detection (LOD) and limit of qualitative (LOQ) calculation in acetone S69 Figure S10h. Calibration curve from AnSQ fluorescence upon addition of Flimit of detection (LOD) and limit of qualitative (LOQ) calculation in toluene S70 Figure S10i. Calibration curve from AnSQ absorbance upon addition of F -, OHand CNfor limit of detection (LOD) and limit of qualitative (LOQ) calculation in THF S71 Figure S11a. FTIR of solid state AnSQ. The peaks at 1100 cm -1 is assigned as Si-O-Si stretching, and the peaks at 732 cm -1 is assigned as Si-O-Si bending Figure S11b. FTIR of solid state AnSQ+TBAF. The peaks at 1072 cm -1 is assigned as Si-O-Si stretching, and the peaks at 734 cm -1 is assigned as Si-O-Si bending S72 Figure S11c. FTIR of solid state AnSQ+TBACN. The peaks at 1106 cm -1 is assigned as Si-O-Si stretching, and the peaks at 735 cm -1 is assigned as Si-O-Si bending, the peak at 2149 cm -1 is assigned as C ≡N stretching Figure S11d. FTIR of solid state AnSQ+TBAOH. The peaks at 1094 cm -1 is assigned as Si-O-Si stretching, and the peaks at 731 cm -1 is assigned as Si-O-Si bending, the peak at 952 cm -1 is assigned as O-H bending S73 Figure S11e. FTIR of solid state AnSQ+TBAPO4. The peaks at 1098 cm -1 is assigned as Si-O-Si stretching, and the peaks at 732 cm -1 is assigned as Si-O-Si bending, the characteristic peak of phosphate is dominated by Si-O-Si stretching Figure S11f. FTIR of solid state AnSQ+TBABr. The peaks at 1090 cm -1 is assigned as Si-O-Si stretching, and the peaks at 732 cm -1 is assigned as Si-O-Si bending. Figure S11g. FTIR of solid state AnSQ+TBACl. The peaks at 1092 cm -1 is assigned as Si-O-Si stretching, and the peaks at 734 cm -1 is assigned as Si-O-Si bending. Figure S11h. FTIR of solid state AnSQ+TBANO3. The peaks at 1094 cm -1 is assigned as Si-O-Si stretching, and the peaks at 734 cm -1 is assigned as Si-O-Si bending. The characteristic of nitrate peak is assigned at 1334 cm -1