A removable amide substituent decides whether light gives you a bicyclo[2.1.1] or a bicyclo[2.2.0]

The rule of five is a rate ratio before it is a heuristic, and rate ratios convert into energy. A better-than-50-fold advantage for 5-exo-trig closure over 6-endo-trig ^[3] works out to about 2.3 kcal/mol of activation free energy at room temperature, if the two pathways have comparable prefactors ^[15]. That is a small number by the standards of substituent effects, which is what makes the result plausible at all. It is also only the entry fee. A 6-endo closure that does happen yields a 1,4-biradical that must still form a C-C bond against transannular strain and bridgehead compression, a step the authors call equally unfavourable ^[5]. An electronic handle that merely cancels the first 2.3 kcal/mol leaves the second problem untouched, so whatever the amide N-substituent is doing, it has to overshoot.

What it is doing, on the authors' account, is reshaping the energy-transfer-driven first cyclization, with a computational study attributing the switch to an interplay of electronic effects ^[2][6]. The neat part is bookkeeping: the substituent that decides the skeleton sits on the nitrogen of an amide described as easily removable ^[2]. The cost is that the nitrogen is booked. Regiocontrol and N-substitution are one variable, so the frame you want and the auxiliary you tolerate are chosen together, then undone.

This is also the second system this group has pushed off the 5-exo default, after hex-5-enal oximes ^[10]. Two cases do not repeal anything, and related 1,6-dienes still return fused [3.2.0] products from 5-exo closure ^[9], but the bias looks softer than its reputation in these nitrogen-containing dienes.

Set that against the other visible-light route into rigid [2.1.1] chemistry. The HKU group's heteroleptic copper(I) photosensitizers target the older bicyclo[1.1.0]butane-plus-alkene reaction, where methods have been restricted in starting materials and prone to polymerization or catalyst instability ^[11][14]. By modulating triplet energy levels, Jian He reports suppressed polymerization and scope extended to electron-deficient alkenes, enynes, dienes and aliphatic alkenes, with gram-scale synthesis demonstrated ^[12][13]. That is diversity on the coupling partner. The aza-diene paper argues instead that preassembled strained frameworks intrinsically restrict substitution patterns and limit late-stage diversification ^[7], and answers with diversity on the skeleton. A group that wants both runs both.

One caution about the evidence. The material supplied carries no yields, substrate counts or selectivity ratios ^[16], so the width of the electronic window that flips the pathway is not knowable from it. And the medicinal case for these frames is at present a bicyclo[2.1.1]hexane case: reported BCH analogues of axitinib and conivaptan showed better solubility and metabolic stability with potency preserved or improved ^[8]. The fused [2.2.0] frame is, by the authors' own description, vastly underexplored ^[4], which makes it a synthetic opening and, so far, a biologically untested one.