Interception of 1,2-cyclohexadiene with TEMPO radical

Abstract

Transient strained cyclic intermediates, such as strained cyclic allenes, are useful building blocks for the synthesis of structurally and stereochemically complex scaffolds. Trappings of strained cyclic allenes are thought to occur primarily through either two or one electron processes. Regarding the latter, diradical intermediates have been invoked in (2 + 2) cycloadditions and (3 + 2) nitrone cycloadditions. The present study questions if a monoradical pathway could exist for strained cyclic allene reactivity, as examined in the reaction of 1,2-cyclohexadiene and TEMPO radical. Our findings suggest the viability of this monoradical pathway.

Introduction

The structure and reactivity of transient strained cyclic intermediates, such as arynes, cyclic alkynes, and cyclic allenes (e.g., 1–5, Fig. 1A) have drawn interest from the scientific community over the last century [1,2,3]. Of particular interest to our laboratory are strained cyclic allenes, such as 1,2-cyclohexadiene (5), which were first experimentally validated in the 1960s [4,5]. Recently, a number of synthetic transformations using strained cyclic allenes have been reported [6–20,30], with cycloaddition reactions providing an opportunity to generate structural and stereochemical complexity (Fig. 1B). Notably, strained cyclic allenes are thought to react through both two and one electron pathways [21]. For example, the (4 + 2) cycloaddition between 5 and furan (6) to afford 7 is suggested to be a concerted and asynchronous process [16,22], whereas the (3 + 2) cycloaddition between 5 and nitrone 8 to provide 9 is believed to involve competing two and one electron pathways [15]. In the case of the (2 + 2) cycloaddition between styrene (10) and 5 to furnish 11, one electron chemistry is thought to occur [23]. This latter example indicates the presence of diradical intermediate 12. In the present study, we questioned if a strained cyclic allene could be intercepted to give monoradical intermediates 13 [24,25].

Fig. 1.

Selected in situ generated strained intermediates and the reactivity of 1,2-cyclohexadiene (5) in cycloaddition reactions.

Results and discussion

To investigate the reactivity of strained cyclic allenes with monoradicals, we pursued the transformation shown in Figure 2. Silyl triflate 14, a well-established precursor to 1,2-cyclohexadiene, was treated with (2,2,6,6-tetramethylpiperidin-1-yl)oxyl (TEMPO, 15) in the presence of CsF at 80 °C. To our surprise, ketone 16 formed as the major product (>20:1 dr). Compound 16 had previously been prepared by Studer and co-workers, which provided validation of our structural assignment [26].

Fig. 2.

The reaction of in situ generated 1,2-cyclohexadiene and 15 to afford ketone 16.

A plausible mechanism for the formation of ketone 16 is shown in Figure 3. Generation of 5 would occur via standard fluoride-mediated 1,2-elimination of silyl triflate 14 [6]. Subsequently, one molecule of TEMPO (15) would react with 5 in situ to afford monoradical intermediate 17. An additional equivalent of 15 would then combine with 17, providing bis(alkoxyamine) 18. Formation of α-carbonyl radical 19 would arise from homolytic fragmentation of 18 [27,31]. Trapping of radical 19 with a third equivalent of 15 would furnish the observed ketone 16 [28].

Fig. 3.

Potential mechanism for the formation of ketone 16.

To provide evidence for the proposed mechanism, an isotope labeling study was performed, as summarized in Figure 4. ¹⁸O-enriched 15, prepared from the corresponding N-oxoammonium salt [29], was combined with silyl triflate 14 and CsF at 80 °C to afford ketone 16. Analysis of the isotopic distribution of ketone 16 by high resolution mass spectrometry demonstrated incorporation of ¹⁸O at three sites in 16 (33% ¹⁸O/¹⁸O/¹⁸O). Additionally, the experimental isotopic distribution of 16 agreed with the expected distribution values when considering the proposed mechanism. We postulate that the observed labeling of the ketone oxygen arises from TEMPO (15) reacting with a strained cyclic allene intermediate to give monoradical 17. Of note, no reaction between 14 and 15 occurs in the absence of CsF, which is indicative of cyclic allene formation en route to 16. Moreover, exposure of non-labeled 16 to ${\text{H}}_2^{18}\text{O}$ under standard reaction conditions did not lead to ¹⁸O incorporation.

Fig. 4.

Isotope labeling study leveraging ¹⁸O enriched TEMPO (15).

Conclusion

In summary, we have demonstrated the trapping of 1,2-cyclohexadiene (5) with TEMPO radical (15). The transformation gives rise to ketone 16, which we surmise involves the consumption of three TEMPO (15) molecules. This suggests the possible trapping of a cyclic allene to give a monoradical intermediate and should prompt further investigation into one electron chemistry of strained cyclic allenes.