The Electronic Structure Problem

Chemistry is, at bottom, a quantum-mechanical many-body problem. Almost every property a chemist cares about — bond strengths, reaction barriers, spectra, magnetic moments — follows from the behaviour of electrons moving in the electric field of a set of atomic nuclei. Computing that behaviour from first principles is called the electronic structure problem, and it is the single most cited motivation for building quantum computers.

The molecular Hamiltonian

A molecule with $N$ electrons and $M$ nuclei is governed by the non-relativistic Hamiltonian (in atomic units, $\hbar = m_e = e = 4\pi\varepsilon_0 = 1$ ):

\hat{H} = -\sum_{i} \frac{\nabla_i^2}{2} - \sum_{A} \frac{\nabla_A^2}{2 M_A} - \sum_{i,A} \frac{Z_A}{|\mathbf{r}_i - \mathbf{R}_A|} + \sum_{i<j} \frac{1}{|\mathbf{r}_i - \mathbf{r}_j|} + \sum_{A<B} \frac{Z_A Z_B}{|\mathbf{R}_A - \mathbf{R}_B|}.

The five terms are, in order: electron kinetic energy, nuclear kinetic energy, electron–nucleus attraction, electron–electron repulsion, and nucleus–nucleus repulsion. The electron–electron repulsion term is what makes the problem hard: it couples the coordinates of every electron to every other, so the wavefunction does not factor into independent single-particle pieces.

The Born–Oppenheimer approximation

Nuclei are thousands of times heavier than electrons (a proton is about 1836 times an electron's mass), so they move far more slowly. The Born–Oppenheimer approximation freezes the nuclei at fixed positions $\{\mathbf{R}_A\}$ , drops their kinetic-energy term, and treats the nucleus–nucleus repulsion as a constant. What remains is the electronic Hamiltonian:

\hat{H}_{\text{el}} = -\sum_i \frac{\nabla_i^2}{2} - \sum_{i,A} \frac{Z_A}{|\mathbf{r}_i - \mathbf{R}_A|} + \sum_{i<j} \frac{1}{|\mathbf{r}_i - \mathbf{r}_j|}.

Solving $\hat{H}_{\text{el}}\,|\Psi\rangle = E_{\text{el}}\,|\Psi\rangle$ for the lowest eigenvalue $E_{\text{el}}$ gives the electronic ground-state energy at that nuclear geometry. Adding the constant nuclear repulsion back yields the total energy $E(\mathbf{R})$ . Repeating this over geometries traces out a potential energy surface, from which equilibrium bond lengths and reaction paths follow.

Why it is hard classically

The electronic wavefunction lives in a space whose dimension grows exponentially with the number of orbitals. A faithful "full configuration interaction" treatment of $n$ spin-orbitals must consider on the order of $\binom{n}{N}$ ways to place $N$ electrons, which quickly exceeds any classical memory. For example, a modest molecule with 50 spin-orbitals and 25 electrons already has more configurations than there are atoms on Earth.

What a quantum computer offers

A quantum computer can store a superposition over all $2^n$ configurations of $n$ qubits using only $n$ qubits, sidestepping the exponential memory wall. The plan for the rest of this module is:

Rewrite $\hat{H}_{\text{el}}$ in second quantization using fermionic creation and annihilation operators (next lesson).
Map those fermionic operators onto qubits with the Jordan–Wigner or Bravyi–Kitaev transform, producing a Hamiltonian that is a weighted sum of Pauli strings.
Prepare a trial state (starting from the Hartree–Fock reference) and use the variational quantum eigensolver (VQE) to estimate the ground-state energy.
Sweep the bond length to compute a dissociation curve, and apply error mitigation so noisy hardware still yields chemically useful numbers.

The target accuracy throughout is chemical accuracy, conventionally $1.6 \times 10^{-3}$ Hartree (about $1\ \text{kcal/mol}$ ) — the threshold below which computed reaction rates become reliable. Reaching it on real hardware is the central engineering challenge of quantum chemistry.