Vacuum energy is the energy associated with the lowest-energy state of a system of quantum fields. A quantum vacuum is not simply empty space: even when no particles are present, fields retain correlations and fluctuations, and their ground state can contribute to the system’s energy. In nongravitational physics, differences in vacuum energy can affect measurable forces and transitions. In gravitational physics, a uniform vacuum energy has the same mathematical effect as a cosmological constant.
The subject connects established quantum phenomena with one of the largest unresolved problems in fundamental physics. Laboratory measurements support predictions involving quantum fields in their ground state, but they do not directly measure a universal absolute vacuum energy. Meanwhile, observations of cosmic expansion are consistent with a small, positive, approximately constant energy density. Explaining why this density is so small compared with characteristic contributions from particle physics is the cosmological constant problem.
Quantum vacuum and zero-point motion
In quantum mechanics, the ground state is the state with the lowest eigenvalue of the Hamiltonian operator. For a harmonic oscillator of angular frequency , the allowed energies are
where and is the reduced Planck constant. The ground-state energy is therefore , rather than zero. This zero-point energy follows from the oscillator’s quantum dynamics: a state with both exactly zero displacement and exactly zero momentum would be incompatible with the uncertainty principle.
A free quantum field can be decomposed into independent modes, each behaving mathematically like an oscillator. For a free bosonic field, the formal vacuum energy is consequently a sum,
with additional labels for polarization or other internal degrees of freedom when needed. Interacting fields require a more general treatment, but the oscillator picture explains why vacuum energy appears naturally in field quantization rather than as an optional addition to classical physics.
Vacuum fluctuations are described by expectation values and correlation functions. A field can have a vanishing mean, , while its two-point correlation is nonzero. Such correlations affect how detectors, atoms, and boundaries interact with the field. They should not be interpreted as a classical noise signal that could be monitored without disturbing the system. Nor does the vacuum consist literally of observable particles continually appearing and disappearing. “Virtual particles” belong to particular methods of calculating interactions; they are not particles detected in the vacuum.
The word “vacuum” also depends on the physical theory and setting. The vacuum of an interacting theory generally differs from the no-particle state of a corresponding free theory. A field may possess a nonzero vacuum expectation value without representing a gas of particles. The Higgs field provides a central example: its vacuum expectation value participates in the mechanism that gives masses to elementary particles.
Zero-point energy is broader than relativistic vacuum energy. Vibrations in molecules and phonons in solids also have ground-state energies. These can influence bond lengths, isotope effects, and phase stability, but they are properties of material systems. Their existence does not by itself determine the energy density of the relativistic vacuum that enters cosmology.
Divergences, subtraction, and renormalization
The oscillator sum is divergent when arbitrarily short wavelengths are included. For one free, massless scalar field in three spatial dimensions, imposing a maximum wave number gives
where is energy per unit volume and is the speed of light. This fourth-power dependence illustrates the strong sensitivity of a naive estimate to short-distance physics. It is not, by itself, a prediction for the observed vacuum density.
The divergence must be treated using renormalization. A regulator first makes expressions mathematically manageable; counterterms then absorb regulator-dependent contributions into parameters whose physical values must be specified. Different regulators organize the calculation differently. A simple spatial momentum cutoff, for example, does not manifestly preserve Lorentz symmetry. Conclusions about vacuum pressure therefore cannot safely be drawn from its unrenormalized expressions alone. A consistent calculation must respect the relevant symmetries and distinguish intermediate quantities from physical results.
In ordinary nongravitational quantum theory, adding a constant to the Hamiltonian multiplies a state by an overall time-dependent phase. Transition probabilities are unchanged. This makes it possible to choose a convenient energy reference. For free fields, normal ordering places annihilation operators to the right and removes the vacuum constant from the normally ordered Hamiltonian. Such a prescription is useful, but it does not establish that gravity is insensitive to vacuum energy.
Interactions introduce additional vacuum contributions, represented in perturbation theory by vacuum diagrams. In normalized flat-spacetime correlation functions, disconnected vacuum factors cancel. When gravity is included, however, a constant term in the action affects spacetime dynamics. It is then associated with a gravitational coupling rather than merely an arbitrary reference for laboratory transition energies.
In effective field theory, vacuum energy is an allowed parameter that summarizes both known contributions and effects of higher-energy physics. The renormalized total may be expressed schematically as
The separation depends on the renormalization convention; the total does not. Thus, quantum field theory does not simply predict an infinite physical vacuum density. The unresolved issue is why the finite, gravitationally relevant result is so small despite contributions associated with much larger particle-physics scales.
Boundary effects and laboratory evidence
The best-known laboratory phenomenon associated with vacuum energy is the Casimir effect, predicted by Hendrik Casimir in 1948. Two ideal, parallel, perfectly conducting plates separated by a distance alter the allowed electromagnetic field modes. After the separation-independent contributions are removed consistently, their interaction energy per unit area is
The corresponding pressure is
where the negative sign denotes attraction. The result concerns the change in energy with plate separation, not a measurement of the absolute energy of empty space.
For ideal plates at a separation of , the pressure magnitude is approximately pascals. The rapid dependence makes the effect much stronger at smaller separations, but real materials then require careful treatment. Conductivity, surface roughness, temperature, geometry, and residual electrostatic potentials all affect measurements. Most precision experiments use a sphere opposite a plate, avoiding the difficulty of maintaining two macroscopic surfaces exactly parallel.
Steven Lamoreaux reported a precision torsion-pendulum measurement in 1997, and Umar Mohideen and Anushree Roy reported atomic-force-microscope measurements in 1998. These experiments helped establish quantitative agreement with electromagnetic fluctuation-force calculations. Their interpretation depends on suitable material models and experimental corrections; the ideal perfect-conductor expression is not an exact description of actual metal surfaces.
The Casimir force can be derived using field zero-point energies, stress tensors, or interactions between fluctuating charges and currents in matter. Lifshitz theory, developed in the 1950s, incorporates dielectric response and temperature and connects these forces with van der Waals forces. These approaches are complementary descriptions of interacting matter and electromagnetic fields. Consequently, observing a Casimir force does not select a unique interpretation in which an independently measurable reservoir of empty-space energy pushes the plates together.
Other quantum-electrodynamic phenomena also involve vacuum correlations. The Lamb shift, measured by Willis Lamb and Robert Retherford in 1947, includes radiative corrections that separate hydrogen energy levels coincident in the simplest relativistic treatment. Spontaneous emission can be analyzed through coupling between an atom and field modes, with vacuum fluctuations and radiation reaction contributing in ways that depend on the formulation. Such phenomena test quantum electrodynamics, but none directly determines the homogeneous vacuum density responsible for a cosmological constant.
Gravitational coupling and negative pressure
In general relativity, gravity responds to the stress-energy tensor, which includes energy density, momentum flow, and stresses. Using metric signature , a Lorentz-invariant vacuum has the form
This tensor has the same form in every inertial frame. Comparing it with the stress-energy tensor of an isotropic perfect fluid gives
when both quantities are expressed in energy-density units. Vacuum energy therefore has negative pressure if its energy density is positive.
The same relation follows thermodynamically for a constant density. If a region of volume contains vacuum energy , then
Unlike ordinary matter, whose density decreases as a fixed number of particles occupies an increasing volume, a cosmological-constant component retains the same density as space expands. Its total energy in an expanding comoving region consequently increases. This does not violate local stress-energy conservation: in an expanding universe, there need not be a globally conserved energy defined in the same way as in a stationary system.
The Einstein field equations can be written
Moving a vacuum contribution between the two sides shows that it shifts the cosmological constant by
A geometric cosmological constant and a constant Lorentz-invariant vacuum contribution are therefore indistinguishable through their gravitational effects alone. Only their combined effective value can be inferred from spacetime geometry.
Negative pressure matters because cosmic acceleration depends on both density and pressure. For a homogeneous and isotropic universe, the acceleration equation contains
with any cosmological constant included here as a fluid component. Positive vacuum energy gives , producing accelerated expansion when it dominates. This is not a general statement that vacuum energy acts as an ordinary repulsive force between objects.
The description assumes an exactly Lorentz-invariant vacuum contribution. Boundaries, thermal states, background fields, and curved spacetime can introduce additional structure, so their stress-energy need not have the simple form . In semiclassical gravity, the gravitational source includes a renormalized expectation value of the quantum stress-energy tensor. Defining this quantity consistently is more demanding than summing oscillator energies.
Cosmic expansion and observational constraints
Albert Einstein introduced the cosmological constant in 1917 while constructing a static cosmological model. Its later interpretation as a possible contribution from quantum vacuum energy developed as quantum field theory matured. The observational role of a positive cosmological constant changed decisively in 1998–1999, when two teams studying distant Type Ia supernovae reported evidence that cosmic expansion was accelerating. Saul Perlmutter, Brian Schmidt, and Adam Riess received the 2011 Nobel Prize in Physics for this discovery.
In the standard Lambda–Cold Dark Matter model, the accelerating component is represented by a positive cosmological constant. Its density is constrained jointly using supernova distances, the cosmic microwave background, galaxy clustering, gravitational lensing, and other observations. These measurements establish the expansion history and matter distribution rather than directly detecting vacuum energy in a laboratory apparatus.
The Planck collaboration’s 2018 results, interpreted within a spatially flat six-parameter Lambda–Cold Dark Matter model, place roughly 68–69 percent of the present total cosmic energy density in the cosmological-constant component. A representative corresponding energy density is about
Its mass-equivalent density, , is approximately . These values depend on the cosmological model and fitted parameters; they are not model-independent measurements of a separately identifiable quantum contribution.
A useful parameterization is the equation-of-state parameter
A cosmological constant has exactly . For a separately conserved homogeneous component with constant , its density changes with scale factor according to
Thus, nonrelativistic matter scales as , radiation as , and a cosmological constant remains constant. Their different dilution rates explain why a component negligible during much of early cosmic history can dominate later.
Dark energy is the broader name for the component or mechanism used to explain the observed acceleration. Vacuum energy is one candidate, not a synonym for every dark-energy model. A slowly evolving scalar field, for example, can have negative pressure without being a permanent vacuum contribution. Observations constrain departures from , time variation, and interactions with matter, but agreement with a cosmological constant does not reveal its microscopic origin. Likewise, evidence for evolving dark energy would require interpretation within the particular model and data combination used.
The cosmological constant problem
The cosmological constant problem concerns the mismatch between the small observed gravitational vacuum density and the much larger scales naturally associated with quantum fields. Expressed in units with , the observed density is approximately
The fourth root is only a few millielectronvolts. This is extraordinarily small compared with characteristic electroweak energies near electronvolts, or the still higher scales associated with possible unification and quantum gravity.
A cutoff estimate at the Planck scale gives a density roughly times larger than observed, with the precise order depending on conventions. This famous comparison illustrates the scale of the difficulty, but it is not a regulator-independent prediction that the vacuum must have that density. A more robust formulation is that changes in particle content, masses, interactions, and symmetry-breaking states generally shift vacuum-energy contributions by amounts much larger than the measured total.
For example, characteristic fourth-power estimates at an electroweak scale of exceed the observed density by about 55 orders of magnitude. Strong-interaction scales near give estimates exceeding it by roughly 43 orders. The coefficients and signs depend on the physical contribution and renormalization treatment. Nevertheless, the hierarchy persists even without invoking physics at the Planck scale.
The difficulty is often described as a problem of radiative stability or naturalness. A parameter can be selected so that the total vacuum density matches observations at one stage of a calculation. Additional quantum corrections then require the parameter to compensate further large contributions unless some protecting mechanism exists. Renormalization permits the measured value to be accommodated, but it does not by itself explain why that value is so small.
Supersymmetry illustrates a possible cancellation principle. In an exactly supersymmetric vacuum with unbroken global supersymmetry, bosonic and fermionic contributions cancel in a way associated with zero vacuum energy. Realistic applications require supersymmetry breaking, and the resulting contributions are generally much too large unless additional structure is supplied. Local supersymmetry introduces further complications rather than an automatic solution.
Other proposals involve dynamical adjustment, modifications of gravity, special treatments of vacuum contributions, or selection among many possible vacua. None has become an experimentally established explanation. A separate “coincidence problem” asks why matter and dark-energy densities are comparable during the present cosmic era despite their different evolution. Solving the smallness problem would not necessarily solve this timing question, and a model of late-time acceleration need not explain the vacuum contributions of known particle physics.
Symmetry breaking, phase transitions, and metastable vacua
Vacuum energy is not limited to oscillator zero-point terms. It also includes the value of field potentials in their vacuum configurations. For a scalar field with effective potential , a constant field at a minimum contributes an energy density . Different minima may have different energies. In nongravitational calculations, an overall constant in leaves the field equations unchanged, but gravity responds to that constant as well as to energy differences.
Spontaneous symmetry breaking occurs when the state of lowest energy does not share all the symmetries of the underlying equations. The electroweak theory provides an important example through the Higgs field. As the early universe cooled, its equilibrium field configuration changed. For the measured Higgs mass, the Standard Model electroweak change is a crossover rather than a first-order phase transition. Extensions of the model can have different thermal histories.
Quantum chromodynamics supplies another source of nontrivial vacuum structure. Strongly interacting quark and gluon fields have condensates and nonperturbative correlations. The change from a high-temperature quark–gluon regime to hadronic matter near temperatures of order 150 megaelectronvolts is a crossover for physical quark masses at low baryon chemical potential. Its vacuum and thermal energy contributions cannot be represented adequately by a simple sum of free-particle zero-point energies.
A false vacuum is a local energy minimum that is not the lowest available state. It can be long-lived because a barrier separates it from a lower-energy configuration. Decay can occur through quantum tunneling, producing a bubble of a lower-energy phase. Sidney Coleman developed a semiclassical treatment of false-vacuum decay in 1977; Coleman and Frank De Luccia included gravitational effects in 1980. Bubble formation reflects a competition between the energy cost of the bubble wall and the energy gained by converting its interior.
Potential energy also plays a central role in cosmic inflation, a proposed early period of accelerated expansion. When a scalar field changes sufficiently slowly, its potential energy dominates its kinetic energy and its pressure approaches the negative of its density. This resembles vacuum domination without requiring the field to sit permanently at a minimum. As the field evolves or decays, energy can be transferred into particles through reheating.
The possibility that the present electroweak vacuum is metastable has been studied by extrapolating Standard Model parameters to very high energies. Results depend sensitively on measured parameters, especially the top-quark mass, and on assumptions about unknown high-energy physics. Such calculations concern the shape and stability of an effective potential; they do not establish that a vacuum transition has occurred or determine the observed cosmological constant.
Observer dependence and quantum fields in curved spacetime
In flat spacetime without boundaries, the usual vacuum is selected by time-translation symmetry and positive frequency with respect to inertial time. In curved or time-dependent spacetime, an equally preferred global time coordinate may not exist. The division of a field into positive- and negative-frequency modes can then be nonunique. Consequently, different observers or different epochs may use different definitions of particles while describing the same underlying quantum field.
This does not make vacuum energy merely subjective. Observable detector responses and stress-energy measurements must be calculated for a specified state, trajectory, and spacetime geometry. What fails is the assumption that a universally preferred particle count always exists. Correlation functions provide a more general description than statements about how many particles occupy a region.
The Unruh effect, developed through work by Stephen Fulling, Paul Davies, and William Unruh in the 1970s, provides a precise example. An idealized detector undergoing uniform proper acceleration through the ordinary Minkowski vacuum has a thermal response characterized by
where is the Boltzmann constant. An acceleration of approximately corresponds to one kelvin. The effect does not imply that an inertial observer sees a surrounding thermal gas.
Hawking radiation, predicted by Stephen Hawking in 1974, is another consequence of quantum fields in a gravitational setting. A nonrotating, uncharged black hole has temperature
For a solar-mass black hole, this is about kelvin. The outgoing radiation is associated with quantum field evolution in black-hole geometry and a corresponding decrease in black-hole mass. The popular image of a particle pair separating at the horizon is a heuristic, not the complete derivation.
Curved-spacetime calculations also require a carefully renormalized stress-energy tensor. Its expectation value can depend on both the quantum state and local curvature. Renormalization introduces gravitational terms beyond the cosmological constant, including curvature-squared terms in the effective action. Even fields that are classically scale invariant can acquire a nonzero stress-energy trace through quantum effects.
These results delimit the simple picture of a fixed vacuum density filling space. An exact cosmological-constant contribution remains proportional to the metric, but the full quantum stress-energy can contain state-dependent radiation, polarization effects, and geometric contributions. Treating all of these as a single constant would obscure physically distinct phenomena.
Dynamical effects, usable energy, and experimental limits
The dynamical Casimir effect occurs when time-dependent boundaries or material properties convert an initial field state into real excitations. A sufficiently rapid change mixes positive- and negative-frequency modes, so a state initially regarded as vacuum need not remain a no-particle state. For mechanical mirrors, achieving the required motion at electromagnetic frequencies is exceptionally difficult. Experiments therefore use systems in which an effective boundary condition can be changed electronically.
In 2011, Christopher Wilson and collaborators reported microwave photon production in a superconducting circuit with a rapidly modulated effective boundary. The experiment used a superconducting quantum interference device to change the electrical boundary condition of a transmission line. It demonstrated a dynamical quantum-field effect without requiring a material mirror to oscillate at relativistic speed. The emitted photons obtained their energy from the external modulation.
The same energy accounting applies to static Casimir arrangements. Plates can perform work while attracting one another, but restoring the apparatus requires energy. A cycle that returns all components and fields to their initial conditions does not yield unlimited net work from the vacuum. For an ideal ground state, no cyclic operation can reduce the system below its lowest energy. Nonequilibrium driving, stored material energy, or coupling to another reservoir can produce useful output, but those resources must be included in the accounting.
Negative Casimir interaction energy also does not establish that unrestricted quantities of negative total energy can be manufactured. The energy is defined relative to an appropriate reference configuration, and the complete apparatus includes positive material and mechanical contributions. Quantum field theory allows some locally negative renormalized energy densities, but their magnitude and duration are constrained in many settings by quantum energy inequalities. A negative local value is not equivalent to a freely available macroscopic substance with negative mass.
Existing laboratory experiments therefore test particular vacuum correlations, radiative corrections, boundary forces, and driven field dynamics. They do not provide a direct measurement of the homogeneous gravitational vacuum density. Conversely, cosmological observations constrain an effective large-scale stress-energy component without isolating every microscopic contribution to it. The two approaches address different observables and cannot be substituted for one another.
Current research connects these domains through precision force measurements, superconducting quantum circuits, quantum sensors, effective field theory, and semiclassical gravity. Important distinctions remain between an energy difference and an absolute gravitational contribution, between particles and field correlations, and between a time-independent vacuum and a driven quantum system. Any claimed extraction of vacuum energy must specify the initial and final states of the complete apparatus, the external control work, and the measured energy flow.