Atomic, Molecular, and Optical Physics
Atomic, Molecular, and Optical Physics is the canonical home for applying quantum mechanics to atomic structure, molecular structure, spectroscopy, coherent light–matter interaction, quantum optical fields, lasers, AMO platforms, and precision measurement. Its central question is not merely how to solve another Hamiltonian, but how microscopic degrees of freedom become preparation procedures, spectra, transition rates, time-domain signals, correlations, and controlled quantum devices.
The volume begins after the basic formalism and canonical models are known. It uses the hydrogen atom, oscillator, rotor, two-level system, angular-momentum algebra, perturbation theory, identical-particle structure, density operators, and open-system dynamics without duplicating their canonical derivations. The work here is to assemble those ingredients into experimentally interpretable models of real matter and radiation.
Why This Volume Matters
Section titled “Why This Volume Matters”Atomic, molecular, and optical physics is one of the places where the abstract structure of quantum mechanics can be tested with exceptional precision. Discrete energy levels become spectral lines. Matrix elements become intensities and lifetimes. Relative phases become Ramsey fringes. Field commutators become photon statistics. Weak environmental couplings become linewidths, shifts, and decoherence rates.
The same subject also supplies much of the working language of quantum chemistry and quantum technology:
- atomic orbitals and antisymmetry organize multi-electron structure;
- molecular potential-energy surfaces organize bonding, vibrations, rotations, and reactions;
- selection rules organize what radiation can prepare or detect;
- driven two-level systems organize coherent control, clocks, and qubits;
- quantized modes organize photon counting, squeezing, cavity QED, and optical networks;
- laser cooling, traps, and optical lattices turn Hamiltonian parameters into laboratory controls;
- precision spectroscopy converts a frequency comparison into a test of theory, constants, and possible new physics.
These are not separate applications pasted onto quantum mechanics. They are a connected hierarchy of models, approximations, and observables.
The Organizing Hamiltonian
Section titled “The Organizing Hamiltonian”A useful top-level decomposition is
The four terms have different roles:
| Term | Degrees of freedom | Typical questions |
|---|---|---|
| electrons, nuclei, internal spin | What are the bound states, symmetries, and intrinsic energy scales? | |
| classical drives or quantized radiation modes | What frequencies, polarizations, mode functions, and quantum states illuminate the system? | |
| charge–field or multipole coupling | Which transitions occur, at what rates, and with what coherent dynamics? | |
| unobserved modes, collisions, technical noise | What shifts, broadening, dissipation, and decoherence remain after degrees of freedom are ignored? |
No single approximation is implied by this decomposition. A semiclassical laser drive treats part of the field as prescribed. Quantum optics retains selected field modes as quantum degrees of freedom. A master equation may eliminate a broadband reservoir. Molecular structure may first separate electronic and nuclear motion. The page must always state which degrees of freedom remain explicit.
Matter before approximations
Section titled “Matter before approximations”For nonrelativistic electrons at positions and nuclei at positions , the Coulomb Hamiltonian is
This expression already contains the central difficulty of atomic and molecular structure: several particles, several masses, attractive and repulsive interactions, antisymmetry for electrons, and usually no exact closed-form solution. Spin-dependent, relativistic, radiative, external-field, and finite-nuclear-size terms are added at the accuracy demanded by the observable.
The Hamiltonian is therefore a starting point, not a complete model specification. One must also state:
- whether the nuclear center of mass has been removed;
- whether nuclei are fixed, dynamical, or treated through Born–Oppenheimer separation;
- whether relativistic and radiative corrections are retained;
- whether the electromagnetic field is classical, quantized, or split into drive and reservoir sectors;
- which symmetries and quantum numbers survive external fields;
- whether the target is an isolated level, a transition, a line shape, or a time-dependent signal.
Radiation and light–matter coupling
Section titled “Radiation and light–matter coupling”For a set of quantized electromagnetic modes,
The zero-point term, mode normalization, polarization basis, and boundary conditions depend on the field geometry. In a cavity or waveguide, the mode labels and density of states differ from free space.
At wavelengths long compared with the matter system, the leading electric-dipole interaction is
where is the electric dipole operator. This is an approximation to the underlying minimal-coupling theory, not a universal identity. Magnetic-dipole, electric-quadrupole, retardation, strong-field, and gauge-consistency issues matter outside its domain.
Scale Hierarchies
Section titled “Scale Hierarchies”The most productive habit in AMO physics is to identify scales before diagonalizing anything. Atomic units expose the gross Coulomb scale:
where is the fine-structure constant. Species-dependent coefficients and quantum numbers matter, but several useful order-of-magnitude relations follow.
| Structure | Characteristic scale | Expansion or control parameter |
|---|---|---|
| Electronic Coulomb structure | correlation strength and nuclear charge | |
| Fine structure | roughly | and relativistic velocity |
| Hyperfine structure | often below fine structure | nuclear moments and |
| Molecular vibration | roughly | electron–nuclear mass ratio |
| Molecular rotation | roughly | moment of inertia and rotational quantum number |
| Zeeman shift | field strength relative to internal splittings | |
| Stark shift | or | field strength and nearby-level spacing |
| Natural linewidth | radiative coupling and photonic density of states | |
| Coherent drive | drive amplitude, detuning, and saturation |
The molecular scalings assume a bound geometry and comparable electronic force constants; they are not universal formulas for every molecule. Hyperfine intervals can also violate a naive hierarchy because nuclear spin, magnetic moments, near-degeneracies, and electronic angular momentum vary substantially.
Length, frequency, and time
Section titled “Length, frequency, and time”Energy is only one part of the hierarchy. A complete problem also compares
where is the system size, the optical wave number, an internal transition frequency, a drive scale, a decay rate, and the reservoir-to-system time-scale ratio.
These ratios control, respectively, the dipole approximation, the rotating-wave approximation, spectral resolution, and common Markovian reductions. Writing the ratio is more informative than merely naming the approximation.
From a Hamiltonian to an Observable
Section titled “From a Hamiltonian to an Observable”An AMO calculation is incomplete until it identifies how a result becomes data. The usual chain is:
| Layer | Mathematical object | Experimental meaning |
|---|---|---|
| Model | Hilbert space and Hamiltonian | prepared species, fields, geometry, and controls |
| Symmetry | commuting operators and transformation rules | labels, degeneracies, and forbidden couplings |
| Approximation | projected or effective Hamiltonian | declared accuracy and neglected processes |
| Spectrum | eigenvalues or quasienergies | resonance positions and thresholds |
| Matrix element | $\langle f | O |
| Dynamics | propagator, density operator, or correlator | populations, coherences, counts, and fringes |
| Environment | self-energy, channel, or master equation | shifts, widths, loss, heating, and noise |
| Instrument | response function and calibration | the reported signal and its uncertainty |
For a transition between stationary levels,
The resonance position alone does not determine what is observed. In a weak-coupling continuum treatment, a transition rate has the schematic golden-rule form
where the normalization of final states and the definition of must be stated together. The canonical derivation belongs in Fermi’s Golden Rule. This volume specializes the formula to atomic, molecular, and optical transitions.
Three Physical Anchors
Section titled “Three Physical Anchors”Hydrogenic spectroscopy
Section titled “Hydrogenic spectroscopy”In the nonrelativistic Coulomb model with an infinitely heavy point nucleus,
The to energy difference is
For hydrogen, this gross-structure result is approximately , corresponding to a vacuum wavelength near . A measured line requires more: the orbital and polarization selection rules, reduced-mass correction, fine and hyperfine structure, radiative corrections, recoil, Doppler profile, natural width, and instrumental response may all matter at the relevant resolution.
The exact Coulomb solution remains in The Hydrogen Atom. Atomic physics begins when that prototype is embedded in a hierarchy of corrections and measurement procedures.
A diatomic molecule
Section titled “A diatomic molecule”After separating center-of-mass motion and applying an adiabatic electronic description, a low-lying diatomic spectrum is often organized as
with
Here is the nuclear reduced mass, a local curvature of the potential-energy curve, and the moment of inertia. The expression is an effective low-energy model. Anharmonicity, vibration–rotation coupling, electronic degeneracy, nonadiabatic coupling, nuclear-spin statistics, and dissociation all require refinements.
The oscillator and rotor solutions remain in Quantum Harmonic Oscillator and Rigid Rotor. Their molecular meaning belongs here.
A coherently driven two-level transition
Section titled “A coherently driven two-level transition”For two relevant internal levels with transition frequency , a near-resonant drive can lead, after a rotating-frame transformation and rotating-wave approximation, to
where is the detuning and is the on-resonance Rabi frequency in the chosen convention. Starting in the lower state and neglecting decoherence,
This compact result links a Hamiltonian parameter to a measured excited-state population. Its validity depends on level isolation, drive strength, pulse envelope, phase convention, and environmental time scales. The canonical finite-dimensional treatment begins with Two-Level Systems.
What to Know First
Section titled “What to Know First”No reader needs every prerequisite at once. The following map identifies the minimum useful entry points.
| Needed idea | Canonical preparation | Why it is used here |
|---|---|---|
| States, observables, probabilities | Core Formalism | Every spectrum and measurement starts from a state and an observable. |
| Coulomb bound states | Hydrogen Atom | Atomic orbitals, scales, and spectroscopic notation build from this prototype. |
| Oscillators and rotors | Harmonic Oscillator and Rigid Rotor | Molecular vibration, rotation, and field modes reuse these models. |
| Angular momentum | Angular Momentum Algebra | Atomic terms, polarization, coupling schemes, and selection rules require it. |
| Tensor operators | Wigner–Eckart Theorem | It separates geometry from reduced transition strength. |
| Controlled approximations | Time-Independent Perturbation Theory | Fine, hyperfine, Zeeman, Stark, and effective interactions are organized perturbatively. |
| Molecular scale separation | Born–Oppenheimer Scale Separation | It explains electronic surfaces and nuclear motion without pretending the separation is exact. |
| Identical particles | Identical Particles | Electronic antisymmetry controls shells, terms, and bonding. |
| Occupation-number language | Fock Space | Photon modes and many AMO platforms use variable occupation. |
| Reduced dynamics | Open Quantum Systems | Decay, pumping, linewidths, and detector backaction require more than unitary dynamics. |
Choose a Path
Section titled “Choose a Path”Atomic physics
Section titled “Atomic physics”Start with the Hydrogen Atom, then review Angular Momentum Algebra and Time-Independent Perturbation Theory. The atomic sequence then moves from atomic units and the hydrogenic prototype to fine structure, hyperfine structure, external-field splittings, selection rules, multi-electron atoms, and precision tests.
Quantum chemistry and molecular physics
Section titled “Quantum chemistry and molecular physics”Begin with Born–Oppenheimer Scale Separation, the Harmonic Oscillator, the Rigid Rotor, and Identical Particles. Then follow molecular Hamiltonians, potential-energy surfaces, molecular orbitals, bonding, vibration, rotation, and nonadiabatic effects. The Quantum Chemistry Roadmap supplies the broader route.
Spectroscopy
Section titled “Spectroscopy”Begin with Spectroscopy, which organizes the measurement chain from energy differences and transition amplitudes to intensities, lifetimes, line shapes, and time-domain response. Review Fermi’s Golden Rule and the Wigner–Eckart Theorem when the rate and symmetry derivations are needed.
Quantum optics
Section titled “Quantum optics”Start with Quantum Optics, using the oscillator, Fock Space, and Open Quantum Systems as foundations. Continue from Quantized Electromagnetic Modes and Photon Number States to Coherent Light, Thermal Light, Squeezed Light, and Phase-Space Distributions, then continue through Beam Splitters and Interferometers, then Photon Counting, Correlation Functions, Hanbury Brown–Twiss Interferometry, cavity QED, and input–output descriptions.
AMO platforms and control
Section titled “AMO platforms and control”Use the AMO Physics Roadmap together with Two-Level Systems and the Many-Body and Quantum Statistical Mechanics overview. The platform path emphasizes which degrees of freedom are controlled, how cooling and trapping work, how states are read out, and which noise sources bound fidelity.
Computational AMO
Section titled “Computational AMO”Begin with Computational AMO and Quantum Chemistry and the Computational Quantum Mechanics Roadmap. The computational path separates reproducible model problems from production electronic-structure packages: basis convergence, radial solvers, matrix elements, time-dependent two-level dynamics, master equations, and benchmark comparisons each require different validation tests.
Bridge to QFT
Section titled “Bridge to QFT”Follow the Bridge to QFT Roadmap and Fock-Space Bridge. The AMO bridge asks when a prescribed classical field must become a quantum field, how spontaneous emission depends on vacuum modes, why radiative shifts are not contained in a bare Schrödinger Hamiltonian, and how effective cavity models arise from field theory.
Main Sections
Section titled “Main Sections”Each section owns a distinct layer of the subject.
| Section | Canonical responsibility |
|---|---|
| Overview | scales, approximation hierarchy, notation, and navigation |
| Atomic Physics | one-electron structure, relativistic and spin corrections, external-field shifts, and atomic selection rules |
| Multi-Electron Atoms | antisymmetry, shell structure, term symbols, coupling schemes, and electron correlation |
| Molecular Quantum Mechanics | molecular Hamiltonians, electronic surfaces, bonding, rotations, vibrations, and nonadiabatic structure |
| Spectroscopy | transitions, intensities, line shapes, resolution, and inference from spectra |
| Light–Matter Interaction | dipole coupling, coherent driving, optical Bloch dynamics, emission, and controlled approximations |
| Quantum Optics | quantized modes, states of light, correlations, photodetection, interference, and cavity models |
| Lasers | gain, inversion, threshold, coherence, linewidth, and laser architectures |
| AMO Platforms and Quantum Control | cooling, trapping, ultracold matter, ions, Rydberg systems, cavities, and engineered Hamiltonians |
| Precision Measurement and Metrology | clocks, interferometry, standards, systematic shifts, and uncertainty budgets |
| Computational AMO and Quantum Chemistry | validated solvers, basis choices, convergence, and reproducible benchmarks |
| Frontiers and Open Problems | carefully dated research questions, evidence levels, and representative platforms |
| Reference and Data | constants, units, term symbols, selection rules, line shapes, model cards, and authoritative data sources |
Approximation Registry
Section titled “Approximation Registry”Names such as “dipole approximation” or “two-level atom” are not sufficient documentation. A mature page states the retained space, the neglected terms, a control parameter, and an observable-level failure test.
| Approximation | Small or separating parameter | Typical use | Warning sign |
|---|---|---|---|
| Fixed nucleus | atomic electronic structure | isotope shifts or recoil matter | |
| Nonrelativistic electrons | and | light-atom gross structure | high or fine precision |
| Independent particles | residual correlations are perturbative | shells and central-field models | near-degeneracy or strong configuration mixing |
| Born–Oppenheimer | electronic and nuclear time scales separate | molecular surfaces | avoided crossings or conical intersections |
| Electric dipole | optical transitions | forbidden lines or short wavelengths | |
| Two-level truncation | off-resonant levels remain far away | coherent control | strong drive, short pulses, or leakage |
| Rotating-wave approximation | counter-rotating effects are small | near-resonant dynamics | ultrastrong coupling or large bandwidth |
| Weak excitation | saturation is negligible | linear spectroscopy | power broadening or population redistribution |
| Markov reduction | optical decay and pumping | structured reservoirs or memory effects | |
| Secular approximation | distinct Bohr frequencies are well resolved | Lindblad-form generators | near-degenerate transitions |
Exact, effective, and phenomenological statements
Section titled “Exact, effective, and phenomenological statements”The volume uses four labels in a precise way:
- Exact within a model means a result follows without further approximation from the stated Hamiltonian and domain.
- Controlled approximation means an expansion or error estimate identifies a regime of validity.
- Effective model means eliminated degrees of freedom have been absorbed into fitted or derived parameters over a specified scale range.
- Phenomenological model means the form is motivated by observed behavior and must be validated empirically.
An exactly solved two-level Hamiltonian can still be only an effective description of a multilevel atom. “Exact solution” and “exact physical model” are different claims.
Canonical-Home Boundaries
Section titled “Canonical-Home Boundaries”This volume applies foundational tools but does not duplicate them.
| Topic | Canonical home | What this volume adds |
|---|---|---|
| Hydrogenic Coulomb solution | Wave Mechanics and Model Systems | atomic corrections, spectra, isotope effects, and experimental interpretation |
| Harmonic oscillator | Wave Mechanics and Model Systems | molecular vibration and radiation-mode applications |
| Rigid rotor | Wave Mechanics and Model Systems | molecular rotational structure and spectroscopy |
| Angular-momentum algebra | Symmetry, Angular Momentum, and Spin | atomic terms, coupling schemes, polarization, and line strengths |
| Perturbation theory | Approximation and Semiclassical Methods | fine, hyperfine, Zeeman, Stark, and effective AMO Hamiltonians |
| Born–Oppenheimer method | Approximation and Semiclassical Methods | molecular surfaces, spectra, chemistry, and breakdown mechanisms |
| Density operators and channels | Core Formalism and Open Systems | optical pumping, fluorescence, linewidths, and platform noise |
| Identical-particle and Fock-space formalism | Composite Systems and Entanglement | electrons in atoms, photons, and ultracold AMO implementations |
| Generic quantum gases and lattice models | Many-Body and Quantum Statistical Mechanics | preparation, trapping, imaging, and AMO realization |
| Quantum algorithms and error correction | Quantum Information and Computation | trapped-ion, neutral-atom, and photonic hardware implementations |
| Full field quantization and QED | QFT.org | AMO motivation, low-energy models, and observable bridges |
What Counts as a Complete AMO Result
Section titled “What Counts as a Complete AMO Result”A result intended for comparison or reuse should report:
- the species, isotope, charge state, and internal level labels;
- the Hamiltonian and unit convention;
- the retained degrees of freedom and basis;
- the approximation hierarchy and parameter regime;
- the field geometry, polarization, frequency, and intensity convention;
- whether quoted frequencies are angular frequencies or ordinary frequencies;
- the state-preparation and measurement model;
- uncertainties, convergence checks, and relevant systematic shifts;
- the data source or primary reference;
- the date and version for live databases or recommended constants.
For spectroscopy, a bare wavelength without medium, vacuum/air convention, uncertainty, and transition assignment is not a reproducible datum. For simulations, a population curve without basis truncation, solver tolerance, pulse definition, and conservation checks is not a reproducible benchmark.
Common Mistakes
Section titled “Common Mistakes”Treating orbitals as particle trajectories
Section titled “Treating orbitals as particle trajectories”An orbital is a one-particle state or basis function, not a classical path followed by an electron. Orbital pictures are useful representations of amplitudes and symmetry, but they do not license a planetary interpretation.
Reporting only level energies
Section titled “Reporting only level energies”Spectroscopy depends on energy differences, matrix elements, populations, selection rules, widths, shifts, and detector response. Correct eigenvalues can still predict no visible line if the relevant transition matrix element vanishes.
Calling selection rules absolute prohibitions
Section titled “Calling selection rules absolute prohibitions”A transition forbidden in the electric-dipole approximation may occur through magnetic-dipole, electric-quadrupole, hyperfine-induced, collision-induced, or multiphoton processes. A selection rule is always tied to an operator and symmetry assumptions.
Treating Born–Oppenheimer separation as exact
Section titled “Treating Born–Oppenheimer separation as exact”Electronic and nuclear coordinates remain part of one quantum system. The approximation can fail near degeneracies, avoided crossings, conical intersections, dissociation thresholds, and sufficiently rapid nuclear motion.
Confusing a model’s exact solution with exact physics
Section titled “Confusing a model’s exact solution with exact physics”Rabi oscillations may be exact for the chosen two-level rotating-frame Hamiltonian while the two-level truncation and rotating-wave approximation remain imperfect descriptions of the laboratory system.
Deriving spontaneous emission from matter alone
Section titled “Deriving spontaneous emission from matter alone”A closed time-independent atomic Hamiltonian does not by itself produce irreversible spontaneous decay. The electromagnetic field and the treatment of its modes, or an appropriate open-system reduction, are essential.
Equating laser light with one universal coherent state
Section titled “Equating laser light with one universal coherent state”An ideal single-mode coherent state is a powerful model. Real lasers have phase diffusion, technical noise, multimode structure, finite linewidth, and measurement-dependent descriptions.
Mixing frequency conventions
Section titled “Mixing frequency conventions”with . Missing factors of propagate directly into detunings, Rabi frequencies, linewidths, and spectral densities.
Using database values without provenance
Section titled “Using database values without provenance”Evaluated data can change as measurements and analyses improve. Record the database name, version or access date, units, uncertainty, and transition assignment.
Exercises
Section titled “Exercises”1. Route a question to its canonical home
Section titled “1. Route a question to its canonical home”For each question, identify the canonical home and the AMO-specific layer:
- Derive the spectrum of the ideal rigid rotor.
- Explain microwave rotational line intensities of a polar diatomic molecule.
- Derive the Wigner–Eckart theorem.
- Apply it to polarization-dependent atomic transition strengths.
Solution
The ideal rotor derivation belongs in Wave Mechanics and Model Systems. Molecular line intensities belong here because they combine the rotor spectrum with a molecular dipole, thermal populations, polarization, and a spectroscopic measurement model.
The Wigner–Eckart theorem belongs in Symmetry, Angular Momentum, and Spin. Its use for concrete atomic line strengths belongs here because the reduced matrix elements, level labels, polarization geometry, and experimental observables are system specific.
2. Compare gross and fine-structure scales
Section titled “2. Compare gross and fine-structure scales”Suppose an atomic gross-structure interval is of order . Use to estimate the relative size and energy scale of a correction of order . Take .
Solution
The relative size is
The corresponding energy is
This is only a scale estimate. Actual fine-structure intervals contain powers of , quantum-number dependence, angular coefficients, reduced-mass effects, and sometimes cancellations.
3. Use parity to test electric-dipole transitions
Section titled “3. Use parity to test electric-dipole transitions”The electric dipole operator is odd under parity. In the central-field approximation, determine whether and transitions can have nonzero electric-dipole matrix elements.
Solution
For a nonzero matrix element , the initial and final states must have opposite parity because is parity odd. Both and have even parity, so the electric-dipole matrix element vanishes. The state has odd parity, so parity permits .
Parity permission is necessary, not sufficient. Angular-momentum and polarization rules must also be satisfied, and the conclusion assumes the central-field parity labels remain valid.
4. Predict isotope scaling
Section titled “4. Predict isotope scaling”Within the harmonic-vibration and rigid-rotor approximations, hold the electronic potential curve and equilibrium bond length fixed while changing the nuclear reduced mass from to . Find the scaling of the vibrational spacing and rotational constant.
Solution
Since
the vibrational spacing scales as
At fixed bond length, , so
The result illustrates why rotational spectra are generally more sensitive to isotope substitution than vibrational frequencies in relative scaling. Real isotopic substitution also produces smaller adiabatic and nonadiabatic corrections.
5. Interpret detuned Rabi oscillations
Section titled “5. Interpret detuned Rabi oscillations”For the ideal driven two-level result above, find the largest possible excited-state population and explain what detuning changes besides the oscillation frequency.
Solution
The sine-squared factor can reach one, so
Detuning increases the generalized Rabi frequency to but reduces the oscillation amplitude. Thus faster population oscillation does not imply more efficient inversion. Decoherence, pulse shape, and multilevel leakage further modify the laboratory signal.
Reference and Data Practice
Section titled “Reference and Data Practice”Theory references and live data resources serve different purposes.
- Textbooks establish notation, approximation hierarchies, and durable derivations.
- Primary papers establish original results and experimental methods.
- Review articles organize a mature subfield or a rapidly developing platform.
- Evaluated databases provide recommended energies, wavelengths, transition probabilities, and constants.
- A live database entry must be cited with enough metadata to reproduce the lookup.
For atomic levels and radiative transitions, use the NIST Atomic Spectroscopy Databases as an evaluated-data starting point, then follow its source references when a result requires primary provenance. For fundamental constants, use the stated release of the NIST/CODATA Fundamental Physical Constants rather than silently mixing values from different adjustments.
The AMO Bibliography and Reading Guide organizes textbooks, monographs, reviews, primary papers, standards, databases, and software sources by audience and task.
Continue
Section titled “Continue”The most direct preparation is:
- Conceptual Overview
- Atomic Physics
- Multi-Electron Atoms
- AMO Physics Roadmap
- Hydrogen Atom
- Angular Momentum Algebra
- Time-Independent Perturbation Theory
- Fermi’s Golden Rule
Readers oriented toward molecules should continue to Molecular Quantum Mechanics, then use the Quantum Chemistry Roadmap and Born–Oppenheimer Scale Separation. Readers oriented toward photons and cavities should add Fock Space and Open Quantum Systems.
References
Section titled “References”- C. J. Foot, Atomic Physics, Oxford University Press, 2005.
- B. H. Bransden and C. J. Joachain, Physics of Atoms and Molecules, 2nd ed., Pearson, 2003.
- I. I. Sobelman, Atomic Spectra and Radiative Transitions, 2nd ed., Springer, 1992.
- C. Cohen-Tannoudji, J. Dupont-Roc, and G. Grynberg, Atom–Photon Interactions: Basic Processes and Applications, Wiley, 1992.
- P. W. Atkins and R. S. Friedman, Molecular Quantum Mechanics, 5th ed., Oxford University Press, 2011.
- J. M. Brown and A. Carrington, Rotational Spectroscopy of Diatomic Molecules, Cambridge University Press, 2003.
- R. Loudon, The Quantum Theory of Light, 3rd ed., Oxford University Press, 2000.
- M. O. Scully and M. S. Zubairy, Quantum Optics, Cambridge University Press, 1997.
- D. F. Walls and G. J. Milburn, Quantum Optics, 2nd ed., Springer, 2008.
- H. J. Metcalf and P. van der Straten, Laser Cooling and Trapping, Springer, 1999.
- C. Cohen-Tannoudji and D. Guéry-Odelin, Advances in Atomic Physics: An Overview, World Scientific, 2011.
- A. Kramida, Yu. Ralchenko, J. Reader, and the NIST ASD Team, NIST Atomic Spectra Database, National Institute of Standards and Technology, current evaluated release.
- P. J. Mohr, D. B. Newell, B. N. Taylor, and E. Tiesinga, “CODATA recommended values of the fundamental physical constants: 2022,” Reviews of Modern Physics 97, 025002 (2025).