Predicting a molecule's properties from first principles

What makes this fascinating

• Compute before you synthesize — Knowing what a new drug or material will do before ever making it would transform chemistry.

• The quantum bottleneck — Solving the Schrödinger equation exactly is intractable for all but the smallest molecules.

• Where AI and quantum meet — Machine learning and quantum computers are both bids to crack this — neither has finished the job.

Frequently asked questions

Can we predict a molecule's properties from first principles?

Partly. Quantum chemistry and simulations predict many properties of small molecules well, but accuracy drops for large or complex systems, and predicting behavior like reactivity or drug efficacy before synthesis remains unreliable.

Why is it so hard?

The exact equations (the many-electron Schrödinger equation) are intractable for all but the smallest systems, so methods rely on approximations that trade accuracy for feasibility.

Why does it matter?

Reliably computing what a new material or drug will do before making it would dramatically speed up discovery and cut the cost of trial-and-error experimentation.

More summits in Chemistry

• Artificial photosynthesis

  Plants turn sunlight, water and air into fuel. Can we engineer it to power the world cleanly?

• Room-temperature catalysis of nitrogen fixation

  Feeding humanity depends on a brutally energy-hungry reaction. Nature does it gently — can we?

• The origin of homochirality

  Life uses only left-handed amino acids. Why did nature pick a hand — and why that one?

• Designing a catalyst from scratch

  Could we design an enzyme atom by atom, on a computer, for any reaction we want?