Approximation and Semiclassical Methods

Most quantum systems cannot be solved exactly. This volume teaches the methods used to extract reliable predictions anyway: perturbation theory, transition rates, variational estimates, WKB and semiclassical methods, tunneling estimates, and effective Hamiltonians. Scattering-specific amplitudes, operators, thresholds, and resonances have their own home in Scattering Theory.

The guiding question is not “Which formula do I remember?” It is:

\text{What controls the approximation?}

Sometimes the control is a small dimensionless coupling. Sometimes it is a large action compared with $\hbar$ , a variational bound, a weak drive, a separated energy scale, or a boundary-condition definition of an observable such as a cross section.

What This Volume Covers

This volume is the canonical home for the working approximation methods of quantum mechanics:

time-independent perturbation theory for energy shifts and state corrections,
degenerate perturbation theory for level splitting and avoided crossings,
time-dependent perturbation theory for transitions and rates,
Fermi’s golden rule and density-of-states reasoning,
variational principles and Rayleigh-Ritz estimates,
WKB wavefunctions, turning points, tunneling, and quantization rules,
effective Hamiltonians and scale separation,
bridges from quantum-mechanical calculation to quantum field theory methods.

The point is not to replace exact solutions. Exact solvable models are the calibration targets. Approximation methods become trustworthy when they reduce to exact results in known limits, preserve the right symmetries, and expose their own failure modes.

The Method Map

Different physical questions call for different methods:

A weak static correction to a known Hamiltonian calls for time-independent perturbation theory.
A degenerate or nearly degenerate subspace calls for diagonalization inside the relevant subspace.
A weak time-dependent drive calls for transition amplitudes and rate formulas.
An unknown ground-state energy may call for a variational upper bound.
A large quantum number or slowly varying potential may call for WKB.
A barrier penetration probability, level splitting, or metastable decay rate may call for a semiclassical tunneling exponent, but each observable imposes different boundary conditions.
A collision experiment calls for scattering amplitudes, cross sections, and asymptotic-state methods.
A central scattering potential may call for partial waves and phase shifts.
Separated energy scales may call for an effective Hamiltonian.

The Volume Overview explains the standards shared by every method. The Approximation Map gives the high-level taxonomy, and Choosing a Method turns the taxonomy into a practical decision tree. For effects exponentially small in a semiclassical parameter, Instantons, Tunneling, and Nonperturbative Effects separates the relevant observables and methods.

For weak static changes to spectra and eigenstates, continue to Time-Independent Perturbation Theory. For weak driving, resonance, and the probability-to-rate limit, continue to Time-Dependent Perturbation Theory and Transitions. For reductions organized by subspaces, gaps, or timescales, continue to Effective Hamiltonians and Scale Separation. For collisions, asymptotic states, cross sections, and unitarity, continue to Scattering Theory.

The Discipline of Approximation

A controlled approximation should state:

the exactly solved starting point,
the small parameter or asymptotic regime,
the observable being approximated,
the leading result,
the expected error scale,
the failure modes.

For example, a first-order energy correction is meaningful only after one identifies the unperturbed Hamiltonian, the perturbing operator, the relevant energy gaps, and whether degeneracy invalidates the nondegenerate formula.

The page Small Parameters and Error Estimates is the conceptual backbone for this discipline.

Recommended Paths

For a first advanced-undergraduate pass, read the overview pages, then nondegenerate perturbation theory, variational principle, WKB approximation, and scattering amplitude/cross sections.

For graduate quantum mechanics, add degenerate perturbation theory, Fermi’s golden rule, Lippmann-Schwinger equation, Born approximation, partial waves, phase shifts, and the optical theorem.

For quantum field theory preparation, focus on transition amplitudes, scattering observables, unitarity, optical theorem logic, effective Hamiltonians, stationary phase, and the QFT bridge pages.

For AMO, chemistry, and quantum matter applications, this volume supplies the method pages; detailed applications belong to their own subject volumes.

Connections and scope

Exact solutions of square wells, harmonic oscillators, and simple barriers belong in Wave Mechanics and Model Systems. General time evolution and pictures belong in Quantum Dynamics. Angular momentum and selection rules belong in Symmetry, Angular Momentum, and Spin. Scattering-specific Born expansions and the exact S-matrix, optical theorem, phase shifts, and resonances belong to Scattering Theory. This volume owns the transferable approximation strategies used across those applications.

Cross-Links

References

R. Shankar, Principles of Quantum Mechanics, 2nd ed., Springer, 1994.
J. J. Sakurai and J. Napolitano, Modern Quantum Mechanics, 3rd ed., Cambridge University Press, 2020.
C. Cohen-Tannoudji, B. Diu, and F. Laloë, Quantum Mechanics, Wiley, 1977.
L. D. Landau and E. M. Lifshitz, Quantum Mechanics: Non-Relativistic Theory, 3rd ed., Butterworth-Heinemann, 1977.
C. J. Joachain, Quantum Collision Theory, 3rd ed., North-Holland, 1983.

Exercises

For each problem, name the most natural first method: a weak quartic correction to a harmonic oscillator, a nearly degenerate two-level system, a slowly varying one-dimensional potential, and a weak short-range scattering potential.

Solution

A weak quartic correction suggests time-independent perturbation theory. A nearly degenerate two-level system suggests degenerate or quasi-degenerate perturbation theory. A slowly varying one-dimensional potential suggests WKB. A weak short-range scattering potential suggests the Born approximation, at least as a first test.

Why is “small” not meaningful until the relevant dimensionless ratio is identified?

Solution

A quantity with units cannot be absolutely small; it is small only compared with another quantity of the same units. In perturbation theory, the relevant ratio is often a matrix element divided by an energy gap. In WKB, it is often a variation length compared with the local de Broglie wavelength. The control parameter must be dimensionless to diagnose error.