Chemists break cycloaddition ‘rule-of-five’ by simply tweaking electronic properties of molecule

Tuning the substituent attached to a dialkene helps control the type of bicyclic ring formed during radical cyclisation reactions.¹ This breaks a long-standing rule of thumb, otherwise known as the ‘rule-of-five’, which chemists use to predict the outcome of such reactions. The researchers behind the work say that accessing different bicyclic rings from a common starting material could be useful to quickly screen the impact of these ring systems in drug molecules.

‘The rule-of-five is the observation that, in radical cyclisations of non-conjugated dienes, forming a five-membered ring is generally favoured over other ring sizes,’ says Varinder Aggarwal at the University of Bristol, UK. Exposing a 1,5-diene to light and a photocatalyst, for example, generates a diradical that intramolecularly attacks the terminal alkene. Better orbital overlap between the radical and alkene, and less ring strain, means that a five-membered bridged compound forms faster than the alternative product, which has two fused four-membered rings, also known as a ladderane.

Aggarwal notes that the ladderane structure is an attractive motif for synthetic chemists as it is similar to the beta-lactam rings found in many antimicrobial drugs. Yet, there are limited routes to make these structures.

Aggarwal and computational chemist Robert Paton at Colorado State University, US, have now found a way to both structures by tweaking the substituent on the backbone of an aza-1,5-diene. ‘This is the first time that I’m aware of that anyone’s been able to switch between these two reactive pathways without changing the core of the structure,’ says Johannes Walker at the Georg August University of Göttingen in Germany, who was not involved in the work.

The team found that molecules with an electron-withdrawing group on the nitrogen atom, such as an acyl group, favoured the ladderane structure, while electron-donating alkyl groups led to the expected bridged product. Using simple reactions then allowed the team to remove the protecting groups and functionalise with a wide variety of substituents, including amides, thioureas and sulfonyl groups.

Computational analysis revealed that the electronic properties of the substituents affect the stability of the radicals and transition states during cyclisation, and ultimately which product forms. Paton explains that one of the radicals is not involved in bond-formation, acting as a ‘spectator’ radical that ‘experiences more stabilisation in the five-ring pathway’. ‘But acylating the nitrogen, or adding electron-donating groups to the adjacent aromatic ring [on the alkene], takes the stabilisation away, to the point that the five-membered pathway becomes unfavourable,’ he says. Certain substituents, such as an acyl group, can also influence the planarity of the amide’s carbonyl group, which help stabilises the radical next to this group. ‘How do these rules hold when you don’t have the carbonyl group anymore?’ asks Walker, adding that ‘there is a bit of a question mark about how broadly applicable [these observations] are going to be’.

‘The rule-of-five is not wrong per se, but it assumes certain things are held constant,’ Paton says. Using computational tools to understand the underlying mechanisms behind these reactions allows chemists to know ‘which substituent effects can be tuned to make such long-standing selectivity rules “negotiable” rather than absolute’, he adds.

Walker notes that Aggarwal’s team previously broke the rule-of-five using aza-1,6-dienes and there are a handful of reactions that defy predictions.² ‘[But these reactions] are starting to become applied and more useful,’ he says.

Walker also points out that the selectivity of these reactions is not complete. ‘[However], in practice, pharma probably won’t care if they get two compounds at the end of the reaction – they’re actually happier … they’ve got two things to test rather than just one,’ he says.

The team now wants to explore the scope of these reactions further. Aggarwal explains that he would like to develop an asymmetric version of the reaction to produce products that give a single enantiomer. ‘Ultimately, we want the choice of molecular scaffold to become a design decision rather than a chance discovery.’