Functionalising a Carbon-Hydrogen Bond Without a Handle

Many cross-couplings start from a carbon-halogen bond installed beforehand. Direct C-H functionalisation can avoid that operation, but not all classical reactions require such a handle. Reactivity and site selectivity both matter. Several C-H sites may compete, and their electronic and steric environments can differ. State the catalyst and conditions before predicting a site.

Use an O-bound ortho-activation model for N,N-dimethylbenzamide. Number its arene carbons C1 to C6 around the ring; C1 bears C(=O)N(CH3)2. Oxygen coordination gives the five-membered path M-O-C(carbonyl)-C1-C2-M. Activation at C6 gives an equivalent path. The model predicts activation at accessible ortho sites C2 and C6, not at the equivalent meta sites C3 and C5. This is not a rule for every amide or catalyst.

Compare complete routes, including directing-group installation and removal. A four-step route saves two operations compared with a six-step route. Fewer steps do not prove less waste. Compare measured waste per amount of product using the same boundary, including solvent, reagents, catalysts and work-up. Waste of 6 versus 4 kg per kilogram of product favours the latter on that metric.

Many nondirected iridium-catalysed arene borylations favour less crowded C-H sites. In a stated steric-control model, compare bulky neighbours around candidate sites. This differs from coordination-directed activation. Blocking directed ortho sites may suppress reaction or change selectivity; it does not automatically cause reaction at a distant site.

Test a directing-group hypothesis with controls. Compare the original substrate with a nonbinding analogue under matched conditions, measuring conversion and site ratios. Loss of ortho preference supports a role for coordination. Replacement also changes electronics and sterics, so additional controls or mechanistic evidence are needed before assigning the effect exclusively to binding.