The general procedure for the synthesis of 9-bromofluorene from fluorene was as follows: free radical bromination of fluorene using N-bromosuccinimide (NBS) at room temperature was used as a baseline reaction for the evaluation of the initiator system of the present invention (Scheme 4). This was done as follows: fluorene (83 mg, 0.5 mmol, 1 eq.) and N-bromosuccinimide (98 mg, 0.55 mmol, 1.1 eq.) were dissolved in dichloromethane (5 mL) in a pre-oven dried Schlenk flask. The desired peroxide (0.025 mmol, 5 mol%) was added and the resulting mixture was degassed by freeze-pump-thaw cycle method (3 cycles). After recovery to room temperature, the acid catalyst was added under argon protection. After the reaction reached a predetermined time, the reaction mixture was quenched with triethylamine (250 μL) and dibromomethane (0.5 mmol) was added as an internal standard for analysis. Aliquots were taken for direct 1H NMR analysis, and the yield was determined by integrating the reference peak (5.9 ppm, s, 1H; determined from the actual sample) relative to the peak of dibromomethane. The reaction results are detailed in Table 1.The bromination reaction was efficiently carried out using peroxyacetone 1 (Trigonox 22, 50 wt% mineral oil solution) and different commercial solutions of Brownsted acid. Control experiments confirmed the necessity of the acid and peroxide, and no conversion was observed after 24 h if either component was omitted. The pKa values of the acid catalysts showed a clear trend: the stronger the acid, the faster the rate of conversion. The conversion of sulfuric acid and p-toluenesulfonic acid after 1 h was 72% and 28%, respectively (entries 1 and 2). Methanesulfonic acid had a slightly lower yield (45%; entry 3), whereas acids weaker than trifluoroacetic acid (22%, entry 5) or trichloroacetic acid (18%, entry 6) failed to initiate the reaction (entry 7). Nitric acid performed better than its pKa value predicted (96%; entry 4). Ultimately, high yields were obtained for all reactions (80-95% yield after 24-72 h) when the reaction time was extended to full conversion, indicating that the acid catalyst only affects the initiation rate. In addition, scandium(III) trifluoromethanesulfonate, a Lewis acid, was found to be catalytic (69%; entry 8). The efficiency of different commercial peroxydione solutions was evaluated using methanesulfonic acid as a standard catalyst of moderate reactivity. Peroxystrictone 2 (Trigonox D; 50 wt%) was less efficient than 1 (45%; entry 3), with 1% product after 1 h and 76% after 48 h (entry 11). Peroxiredoxin 3 (Trigonox 301; 41 wt%) showed low conversion after two days of reaction (8%, entry 15). Peroxiredoxin 4 (Luperox DHD-9, 32 wt%) was more reactive than 3, yielding 12% product after 48 h (entry 16). A series of structurally different peroxides were evaluated based on the significant effect of the peroxydione structure on its reactivity. Compounds 11a and 11b demonstrated the effect of group X (formula I). 11b was less reactive than 1 (21%, entry 9), whereas 11a was more efficient, with 28% conversion after 1 h (entry 10). The aromatic substituents around the peroxide portion had a significant effect: 5 was more efficient than 2 (20%, entry 12 vs. entry 11), whereas 6 was much less efficient (33% after 48 h, entry 13). 9 was slightly more reactive than 1 (50%, entry 17), whereas 10 was the most efficient of the evaluated structures, with a conversion of 74% after 1 h of reaction (entry 28).
