retrosynthetic analysis
/ RET-ro-sin-THET-ic /
Retrosynthetic analysis is the chemist's way of planning how to build a complex molecule by reasoning backwards from it. Instead of starting with cheap materials and guessing which reactions will eventually reach the target, you start with the target and ask: what slightly simpler molecules could be combined to make this, in one reasonable step? Then you ask the same question of those, and again, peeling the molecule apart until you arrive at simple, buyable starting materials. It is solving a maze by tracing back from the exit.
The core move is the disconnection: you mentally break one bond in the target molecule, splitting it into two simpler fragments. A special arrow (a double-lined open arrow, drawn target on the left pointing to fragments on the right) marks this 'imagined-in-reverse' step. The fragments you get are usually idealized charged pieces called synthons — for example, breaking a C-C bond might give an imaginary positive synthon and a negative synthon. Synthons are not real bottles on a shelf; they are placeholders. The chemist's job is to recognize the real reagents (the synthetic equivalents) that behave like those synthons: a negative carbon synthon corresponds to a real Grignard reagent or enolate, a positive carbon synthon to a real alkyl halide or carbonyl. Disconnect at the bonds that a known, reliable forward reaction can actually make, and you turn an abstract plan into a real recipe.
Retrosynthesis matters because it converts the daunting question 'how on earth do I make this?' into a structured, almost chess-like search. Formalized by E. J. Corey (a Nobel Prize in 1990), it is the organizing logic behind total synthesis and the daily working method of medicinal chemists designing drugs. Good disconnections — choosing bonds next to functional groups, splitting the molecule into two roughly equal halves for a convergent route — are the art that separates an elegant short synthesis from a clumsy thirty-step one.
To make the alcohol Ph-CH(OH)-CH3, disconnect the C-C bond next to the -OH. That suggests two synthons: a 'CH3 minus' (real equivalent: a methyl Grignard, CH3MgBr) and a 'PhCHO plus' (real equivalent: benzaldehyde). The forward step is then obvious: add CH3MgBr to benzaldehyde, then work up.
Disconnect a bond into synthons, match each to a real reagent, then run the reaction forwards.
Synthons are imaginary idealized fragments, not real chemicals — the work is matching them to real synthetic equivalents. And a disconnection on paper is only valid if a real, reliable forward reaction can actually make that bond; retrosynthesis plans, it does not guarantee.