Binding energy is the energy required to separate a bound physical system into specified constituents in specified final states. Equivalently, it is the energy released when those constituents assemble into that bound state, provided the initial and final conditions are reversed consistently. The concept applies to electrons in atoms, atoms in molecules and solids, protons and neutrons in nuclei, and matter held together by gravity.
Binding energy connects the structure of matter with measurable quantities such as masses, radiation frequencies, reaction energies, and dissociation thresholds. Its magnitude ranges from fractions of an electronvolt for weak molecular associations to millions of electronvolts per constituent in atomic nuclei. Although the underlying interactions differ, the central comparison is always between the energy of a composite system and that of a specified separated reference state.
A binding energy is therefore not fully defined by naming the object alone. The separation products, electronic or nuclear states, temperature, and surrounding medium can matter. In chemistry and biology, it is particularly important to distinguish a microscopic binding energy from the thermodynamic free-energy change that determines an equilibrium population.
Definition, reference states, and sign conventions
For an isolated system at rest, a positive binding energy can be defined as
Here the separated constituents are sufficiently far apart that their mutual interaction is negligible, and their internal states are specified. A stable bound state lies below the corresponding dissociation threshold, so . Supplying this amount of energy makes separation energetically possible, although the actual process can require a particular transfer of momentum or angular momentum.
The opposite sign convention is common when reporting interaction energies or energies of association. If the energy of separated constituents is chosen as zero, the bound state has energy . Thus a “binding energy of electronvolts” in one computational convention may describe the same stabilization as a positive separation energy of electronvolts in another. The physical meaning follows from the stated energy difference, not from the sign alone. In the International System of Units, binding energy is expressed in joules, but the electronvolt is usually more convenient at atomic and nuclear scales. One electronvolt is exactly joules.
A composite object may have several distinct binding energies. The energy required to remove one outer electron from an atom differs from that required to remove all its electrons. Likewise, the energy required to separate a nucleus into individual nucleons differs from the threshold for ejecting one neutron or an alpha particle. Each quantity refers to a different channel. An excited state generally has a smaller binding energy relative to the same threshold than the ground state does.
Through mass–energy equivalence, binding also changes the rest mass of a system. For constituents initially at rest and separated,
where are constituent rest masses, is the rest mass of the bound system, and is the speed of light. This relation compares the same constituent content and specified internal states. A system that emits radiation while binding loses both energy and rest mass. If the released radiation remains trapped inside a closed container, however, that radiation still contributes to the container’s total mass.
Binding energy is not identical to an energy barrier. A system may be energetically able to break apart but remain long-lived because the transition is inhibited. Conversely, an energetically favorable association may not occur unless excess energy is carried away by a photon, another particle, or the surrounding material. Energy differences determine which states are accessible; transition mechanisms determine how rapidly they are reached.
Quantum bound states and atomic electrons
In quantum mechanics, a bound state is usually represented by a spatially localized, normalizable state whose energy lies below an appropriate continuum threshold. For a particle moving in an attractive potential that approaches zero at large distance, negative-energy eigenstates are bound, whereas positive-energy states can describe unbound motion. The Schrödinger equation determines the allowed energies and wave functions for a specified Hamiltonian.
Binding does not mean that the particle has ceased moving. Localization generally entails a positive kinetic-energy contribution, while the attractive interaction provides a sufficiently negative potential-energy contribution. For a stationary nonrelativistic system governed solely by Coulomb interactions, the virial theorem gives . Consequently, the total energy is negative even though the kinetic energy is positive. This balance is central to the size and stability of atoms.
The simplest example is hydrogen. Neglecting relativistic, radiative, and finite-nuclear-size corrections, the energy levels of a hydrogen-like ion are
where is the electron–nucleus reduced mass, is the fine-structure constant, is the nuclear charge number, and is the principal quantum number. The ground-state binding energy of hydrogen is approximately eV. Its higher levels become progressively less tightly bound, approaching the ionization limit as increases.
For many-electron atoms, electron–electron repulsion, screening, and quantum statistics complicate this picture. The first ionization energy is the total-energy difference between the neutral atom and its singly charged positive ion, with the detached electron at rest at infinity. Successive ionization energies refer to further removals. They are not generally equal to the negatives of orbital energies, because the remaining electrons reorganize after removal and their correlated motion changes.
Electron binding energies also depend on the atom’s chemical environment. Inner-shell energies shift when oxidation state, neighboring atoms, or local electrostatic conditions change. Such shifts underpin important forms of spectroscopy. A core electron in a metal and one in an oxide of the same element can have measurably different removal energies even though both belong to the same nominal atomic shell.
A discrete energy or long lifetime alone does not establish a true bound state. A resonance can temporarily localize a particle while lying above a dissociation threshold. Such a state can decay into the continuum and has a finite spectral width. Distinguishing bound states from resonances is essential in atomic collisions, molecular dissociation, and unstable nuclear systems.
Molecular bonds and dissociation energies
A chemical bond stabilizes a molecular structure relative to specified separated fragments. Its binding arises from the quantum behavior of electrons and nuclei, rather than from a simple cancellation of classical attractive and repulsive forces. In a covalent bond, electron sharing and redistribution can lower the total energy; ionic bonding prominently involves electrostatic attraction between oppositely charged species. These descriptions emphasize different contributions to a single electronic structure.
For a diatomic molecule, a potential-energy curve plots electronic energy against internuclear distance. At large separation, the curve approaches the energy of the chosen fragments. Near the equilibrium bond length, it has a minimum. The depth measured from that minimum to the dissociation limit is conventionally denoted . The nuclei cannot ordinarily occupy the exact minimum with zero vibrational motion, so the experimentally relevant ground-state dissociation energy is smaller:
where is the vibrational zero-point energy relative to the potential minimum. For molecular hydrogen, is approximately eV, whereas is about eV.
A bond dissociation energy must specify the fragments produced. Homolytic cleavage divides the bonding electrons between fragments, often producing radicals; heterolytic cleavage produces ions. These channels can have very different energies, especially when the fragments are placed in a polar solvent. A gas-phase bond dissociation energy therefore cannot automatically be used as the energy required to break the corresponding bond in solution.
In polyatomic molecules, individual bonds do not possess universally fixed energies independent of their surroundings. Removing a hydrogen atom from methane differs energetically from removing one from a methyl radical. Conjugation, electronic relaxation, ring strain, and the stability of the products all affect the result. Tabulated average bond energies are useful approximations for estimating reaction energies, but they average over distinct molecular environments.
The energy of a chemical reaction depends on all bonds broken and formed, together with changes in molecular state and environment. Breaking bonds requires energy; forming bonds releases it. Combustion releases energy because the complete set of products is lower in energy than the reactants, not because fuel bonds release energy merely by being broken. The same distinction applies to biochemical reactions involving adenosine triphosphate.
Bond strength also differs from activation energy. A strongly exothermic reaction can proceed slowly if it requires passage through a high-energy transition state. A catalyst changes the accessible reaction pathway and its barriers, but it does not change the energy difference between the same initial and final states. Molecular binding energies therefore characterize energetic stability, while reaction rates require additional kinetic information.
Cohesion in solids and binding at surfaces
In a solid, binding is distributed across a many-particle structure rather than confined to an isolated pair of atoms. The cohesive energy is commonly defined as the energy required to separate a crystal into isolated, neutral, ground-state atoms, usually expressed per atom or per formula unit. This reference distinguishes it from lattice energies defined using separated gaseous ions. Different conventions for lattice energy also differ in sign, depending on whether formation or dissociation is described.
The physical origin of cohesion depends on the material. Metals are stabilized by electronic states extending through the crystal. Covalent solids such as diamond contain extended directional bonding networks. Ionic crystals combine long-range electrostatic attraction with short-range repulsion and other quantum contributions. Molecular crystals may be held together largely by van der Waals forces, hydrogen bonding, or both. These categories overlap, and quantitative calculations treat the full electronic and nuclear energy.
Typical cohesive energies illustrate the scale involved. Silicon has a cohesive energy of roughly eV per atom, while diamond is bound by about eV per atom relative to isolated carbon atoms. These values are not energies for breaking one arbitrarily selected bond: separating an entire crystal changes the coordination of every atom and allows extensive electronic reorganization. Measured sublimation enthalpies can help determine cohesion, but thermal and zero-point contributions must be accounted for when comparing them with calculated static-lattice energies.
At a surface, adsorption binds an atom or molecule to a substrate. Weak physisorption is often dominated by dispersion interactions, whereas chemisorption involves substantial electronic rearrangement or bond formation. Adsorption energies depend on surface orientation, defects, adsorbate coverage, and the state chosen for the free adsorbate. Neighboring adsorbates can attract or repel one another, so the energy of adding the next molecule need not equal the average adsorption energy.
Electronic binding in solids introduces additional reference choices. The work function is the energy needed to take an electron from the Fermi level to the vacuum level. In photoelectron measurements, an electron’s binding energy is often quoted relative to the Fermi level rather than to vacuum. These quantities are related but not identical. Surface dipoles can alter the work function without producing an equivalent shift in every internal electronic energy.
A further example is an exciton, a bound state of an electron and a hole. Its binding energy is measured relative to the energy of a separated electron–hole pair in the material. Dielectric screening and effective particle masses determine how tightly it is bound. In many bulk semiconductors, exciton binding energies are only a few to a few tens of millielectronvolts; in some atomically thin semiconductors they reach hundreds of millielectronvolts. Thus the surrounding medium can radically alter binding even when the interacting charges remain the same.
Association free energy in chemistry and biology
For molecules in solution, the tendency to associate is governed by thermodynamic free energy rather than by an isolated interaction energy alone. When a ligand binds to a protein, both partners can change conformation, water molecules can be displaced, and ions or protons can be exchanged with the surrounding solution. The energetic difference between two optimized molecular structures captures only part of this process.
For the association reaction , the standard binding Gibbs free energy is
where is the gas constant, is absolute temperature, and is a dimensionless association constant defined using activities. In dilute solutions, experimental results are often reported as a dissociation constant with concentration units. With a standard concentration , the corresponding expression is
The concentration ratio is necessary because the argument of a logarithm must be dimensionless. At K, a of nanomolar corresponds to a standard association free energy of approximately kJ mol, using a mol L standard state.
The decomposition
shows why an attractive molecular interaction does not by itself determine affinity. Favorable enthalpy can arise from electrostatic interactions, hydrogen bonds, and dispersion, but binding can restrict molecular translation, rotation, or internal motion. Release of solvent molecules can contribute favorably to entropy. The hydrophobic effect is consequently not a single attractive bond between nonpolar groups; it involves the free-energy consequences of exposing or burying nonpolar surfaces in water.
Measured affinity depends on experimental conditions. Temperature, pH, salt concentration, and competing molecules can change protonation, screening, conformational populations, or solvent organization. A reported binding free energy is therefore associated with a defined chemical environment. Coupled proton uptake or release can also make the observed binding enthalpy depend on the buffer used.
Affinity must further be distinguished from residence time and binding mechanism. For an elementary reversible association, , but the same equilibrium constant can result from different association and dissociation rates. Cooperative systems require additional care: oxygen binding to hemoglobin, for example, involves interacting sites and changing protein conformations. A single microscopic binding energy cannot replace the thermodynamic description of all populated states. In this setting, “binding energy” is sometimes used loosely for an interaction score, whereas binding free energy has a specific equilibrium meaning.
Nuclear mass defects and the binding-energy curve
Nuclear binding energy is the energy required to separate an atomic nucleus into its constituent protons and neutrons. If a nucleus contains protons and neutrons, with mass number , its total binding energy is
The difference between the summed free-nucleon masses and the nuclear mass is the mass defect. When neutral atomic masses are used instead, electron masses largely cancel if hydrogen-atom masses replace proton masses, but precise work must also account for electronic binding-energy differences.
Nuclear binding results primarily from the residual strong interaction between nucleons. It is short-ranged and effectively saturating: adding a nucleon does not give it equally strong attraction to every other nucleon in a large nucleus. The electromagnetic repulsion between protons, by contrast, acts over long distances and grows increasingly important in heavy nuclei. The interplay produces a characteristic binding-energy-per-nucleon curve.
The deuteron, containing one proton and one neutron, has a total binding energy of about MeV. Helium-4 is much more tightly bound per nucleon, with a total near MeV. The average rises across light nuclei, reaches roughly MeV near iron and nickel, and then declines gradually for heavier nuclei. Nickel-62 has the highest measured binding energy per nucleon, approximately MeV; iron-56 is close to the maximum but is not its exact peak.
The broad pattern can be represented by the semi-empirical mass formula:
Its terms describe volume binding, a surface penalty, proton repulsion, a neutron–proton asymmetry penalty, and pairing effects. The coefficients are fitted to nuclear masses and depend on the chosen formulation. The expression explains broad trends but does not reproduce all local structure.
Quantum shell effects create additional stability near particular proton or neutron numbers. Pairing generally favors even proton and neutron counts. These effects appear strongly in one- and two-nucleon separation energies, which compare neighboring nuclei. Average binding energy alone therefore cannot establish whether a particular isotope is stable against every decay channel. Stability against beta decay, for instance, depends on mass differences between nuclei with the same mass number but different proton numbers, together with the relevant lepton masses.
Energy release in nuclear reactions and stars
A nuclear reaction releases energy when its products have a smaller total rest mass than its reactants. The reaction energy, conventionally called the -value, is
A positive indicates an exoergic reaction. When the reactants and products contain the same total numbers of protons and neutrons, the -value can also be expressed as the increase in total nuclear binding energy. For reactions involving changes of particle identity, including weak processes, all particle masses and emitted particles must be included consistently.
The binding-energy curve explains why both nuclear fusion of light nuclei and nuclear fission of very heavy nuclei can release energy. Fusion can move matter toward more strongly bound nuclei from the light side of the curve; fission can do so from the heavy side. Neither process releases energy for every possible choice of reactants and products. The specific masses determine the result.
A widely studied fusion reaction is
which releases about MeV. Most of this energy initially appears as kinetic energy of the neutron, with the remainder carried by the helium nucleus. The favorable energy balance does not remove the Coulomb barrier between the reacting nuclei. High temperatures and quantum tunneling determine whether fusion occurs at an appreciable rate.
Fission of uranium-235 following neutron absorption typically releases about MeV per fission, depending on the products and on which delayed contributions are included. The energy appears mainly as fragment kinetic energy, with additional contributions from neutrons, gamma rays, and subsequent radioactive decays. Antineutrinos carry away part of the decay energy and generally escape without depositing it locally. Fission barriers and neutron-induced reaction probabilities are separate from the overall mass-energy balance.
In the Sun, hydrogen is converted into helium mainly through the proton–proton chain. The net conversion of four hydrogen atoms into a helium atom releases approximately MeV, including energy carried by neutrinos. Because some protons become neutrons through weak interactions, this total is not simply the binding energy of helium-4. Stellar energy accounting must include the associated particle conversions and annihilation processes.
The iron-group region marks an important limit for energy-producing stellar fusion. Further fusion toward heavier nuclei generally does not supply net energy under ordinary stellar burning conditions. Nevertheless, nuclei heavier than iron form through neutron captures and other pathways in nucleosynthesis. Their production depends on available particles, reaction rates, radioactive decay, and astrophysical conditions, not on a rule that all nuclear growth must increase binding energy per nucleon.
Gravitational binding of planets, stars, and compact objects
Gravitational binding energy is the energy needed to disperse a self-gravitating object to infinity, with the final material states specified. In Newtonian theory, a spherical body can be imagined as assembled by bringing successive shells inward from infinity. Its gravitational potential energy is
where is the mass inside radius . For a sphere of uniform density,
The magnitude of this expression is often called its gravitational binding energy. Real objects have nonuniform density, so the numerical coefficient depends on internal structure.
Applied to Earth, the uniform-density estimate is about J. Earth’s concentration of denser material toward its center makes its gravitational potential energy somewhat more negative than this simple model predicts. The scale is enormous compared with everyday energies, although small relative to the planet’s total rest energy. The formula describes collective gravitational cohesion, not the chemical or nuclear energy required to separate matter into elementary constituents.
Gravitational potential energy should not always be equated with the net energy required to unbind a hot object. A star also contains thermal energy, radiation, and sometimes substantial rotational or magnetic energy. The virial theorem relates these contributions to gravity. For an idealized, nonrotating, self-gravitating system supported by nonrelativistic ideal-gas pressure, , so the total energy is . The magnitude of the negative total energy is then only half the magnitude of the gravitational potential energy.
This distinction explains how contraction can heat a star while allowing it to radiate. As gravitational potential energy decreases, part of the released energy increases internal motion and part escapes. Before sustained hydrogen fusion begins, gravitational contraction is a major energy source for a forming star. Accretion onto compact objects can likewise convert a substantial fraction of infalling matter’s gravitational energy into radiation.
For a neutron star, Newtonian formulas are inadequate for precise binding calculations. In general relativity, one distinguishes the gravitational mass measured from afar from a baryonic mass constructed from the conserved baryon number and an adopted reference mass per baryon. Their difference characterizes a commonly used binding-energy measure. For typical neutron-star models, the binding energy is of order J and can represent roughly a tenth of the baryonic rest energy, with strong dependence on mass and the dense-matter equation of state.
During neutron-star formation, much of the released energy is carried away by neutrinos. Supernova neutrino observations therefore probe gravitational energy release on a scale far larger than the visible-light output. In strong gravity, however, local gravitational energy cannot generally be assigned a unique coordinate-independent density. Global masses and carefully defined reference states are essential to a meaningful binding-energy statement.
Measurement, historical development, and theoretical limits
Binding energies are determined by comparing well-defined states, but the appropriate experimental method depends on the scale. Atomic and molecular spectroscopy identifies transitions or thresholds from measured radiation frequencies. With the Planck constant , a photon has energy . A series of spectral lines approaching an ionization or dissociation limit can therefore establish a binding energy without physically separating every constituent in each measurement.
In photoelectron spectroscopy, a photon ejects an electron and an analyzer measures its kinetic energy. Energy conservation then relates the photon energy, electron kinetic energy, and removal energy. For conducting samples referenced to the analyzer Fermi level, a common expression is
where is the analyzer work function. Charging, surface contamination, final-state relaxation, and calibration affect the interpretation. The measured feature represents a transition between many-electron states, not a direct photograph of a pre-existing orbital energy.
Nuclear binding energies are obtained especially precisely through mass spectrometry and reaction measurements. Penning traps determine charged-particle mass ratios through cyclotron frequencies in a magnetic field. Reaction thresholds and emitted-particle energies constrain mass differences, while gamma-ray measurements establish excitation energies. Combining these measurements produces internally connected nuclear mass evaluations. Uncertainties and correlations matter because separation energies can be small differences between much larger measured quantities.
Historically, Albert Einstein’s 1905 mass–energy relation supplied the framework connecting mass deficits with released energy. Early twentieth-century mass spectrograph measurements, notably those of Francis Aston, established precise isotopic masses and revealed systematic departures from integer mass values. Nuclear experiments and the discovery of the neutron in 1932 made it possible to interpret those masses in terms of proton–neutron nuclei. In 1935, Carl Friedrich von Weizsäcker introduced a semi-empirical mass formula that captured major nuclear binding trends.
Theoretical calculations face different limitations in different fields. Atomic and molecular calculations must approximate electron correlation and, where necessary, relativistic effects. Solid-state predictions depend on electronic-structure methods and treatment of dispersion, screening, and temperature. Nuclear calculations must describe many interacting nucleons using interactions and approximations with controlled uncertainties. Agreement on a total energy does not automatically guarantee equally accurate predictions for every small energy difference.
An additional limit appears inside hadrons. Most of the proton’s mass arises from the dynamics of quarks and gluons in quantum chromodynamics, rather than from the sum of the quarks’ small mass parameters. Yet isolated quarks are not available as physical separated reference states because of color confinement. A proton therefore cannot be assigned a directly measurable quark-separation binding energy in the same way that a nucleus can be assigned an energy for separation into free nucleons.