JOVANA
Library Glossary Getting Started Three Levels Fields How it works Mission
Join the mission
All guides

Thinking Backwards: Retrosynthesis

The single most useful planning skill in synthesis is working backwards from the target. Learn what a disconnection is, how to spot the bonds worth breaking, and why medicinal chemists plan routes differently from total-synthesis chemists.

What retrosynthesis really is

When you look at a molecule you want to make, you cannot buy it — but you can usually buy, or already have, simpler pieces. Retrosynthesis is the discipline of mentally breaking your target apart, one bond at a time, until every piece is something you can purchase or make easily. Each imaginary bond-break is called a disconnection, and the pieces it leaves behind are called synthons. The clean trick is that you reason in the *reverse* direction of the actual chemistry: you plan backwards, then run the reactions forwards.

A good disconnection corresponds to a *real* forward reaction you trust. There is no point breaking a bond if you have no robust way to form it again. So retrosynthetic arrows (drawn as ⇒) are really shorthand for 'I know a reaction that builds this bond.' The art is choosing disconnections that lead back to cheap, stable, available building blocks.

How a medicinal chemist plans differently

A total-synthesis chemist making one complex natural product will accept a long, clever route. A medicinal chemist has a different goal: make *many* analogues of one scaffold quickly so the team can read the structure–activity relationship. That changes everything about route choice. You want a convergent plan where the part you will vary — the R-group — is added in the *last* step from a diverse menu of building blocks. Then a single late intermediate can branch into dozens of final compounds.

  1. Draw the target and identify the fixed core versus the variable region you want to explore.
  2. Place your first disconnection so the variable group comes off — ideally as an amide, amine, or coupling bond.
  3. Keep disconnecting the core until each fragment is a catalog building block or one short step away from one.
  4. Sanity-check each forward reaction for functional-group compatibility — will an earlier group survive a later step?
Target:  Ar–C(=O)–NH–R    (an amide)

Disconnection 1 (amide bond):
  Ar–C(=O)–NH–R  ⇒  Ar–C(=O)–OH  +  H2N–R
  forward = amide coupling (last step; R varied from a menu of amines)

Disconnection 2 (biaryl C–C, on the acid piece):
  Ar–C(=O)–OH    ⇒  Ar'–Br  +  (HO)2B–Ar''(CO2H)
  forward = Suzuki coupling

Both Ar'–Br and the boronic acid are catalog building blocks.
A two-disconnection plan that puts the variable amine in the final step — ideal for analogue series.