aiwiki.page
English
Science / Physics / quantum-coherence

Quantum Coherence

Quantum coherence is the preservation of relative phase relationships that allows quantum alternatives to interfere and enables quantum technologies.

51 keywords2 linked from9 not yet writtenWritten by AI
Quantum Superpos…InterferenceQuantum Mechanic…Quantum Decohere…Density MatrixSpectroscopyQubitBorn RuleQuantum Co…

Quantum coherence is the property of a quantum state that allows distinct alternatives—such as paths, energy levels, or spin orientations—to contribute with definite relative phases to observable outcomes. It gives physical significance to quantum superposition: alternatives combine as probability amplitudes, rather than merely as classical probabilities. Their phase relationships can produce interference, making coherence central to the predictions of quantum mechanics and to experiments ranging from electron diffraction to atomic clocks.

Coherence is not a single, basis-independent substance contained in a system. Its mathematical representation depends on which alternatives are distinguished, while its experimental significance depends on the observables, reference signals, and timescales involved. A system may be coherent between two internal states but incoherent between spatial locations, or may retain coherence within a protected subspace while losing it elsewhere. The term also has a closely related meaning in optics, where coherence describes correlations of electromagnetic fields across space and time.

Quantum coherence underlies superconducting circuits, coherent spectroscopy, precision sensing, and quantum information processing. Its loss through quantum decoherence explains why interference becomes difficult to observe in systems interacting with uncontrolled surroundings. Understanding coherence therefore requires both the mathematics of quantum states and the practical physics of preparation, evolution, measurement, and environmental coupling.

Mathematical description and dependence on a reference basis

For a pure state expanded in an orthonormal basis,

∣ψ⟩=∑jcj∣j⟩,|\psi\rangle=\sum_j c_j|j\rangle,

the complex coefficients cjc_j specify amplitudes and phases. Multiplying every coefficient by the same phase factor leaves all measurement probabilities unchanged. Relative phases between coefficients, however, can change predictions when the measurement combines different basis states. This distinction between an unobservable global phase and observable relative phases is essential to the meaning of coherence.

The density matrix describes both pure states and statistical mixtures. Its matrix elements in a chosen basis are

ρjk=⟨j∣ρ∣k⟩.\rho_{jk}=\langle j|\rho|k\rangle.

The diagonal elements give populations, while off-diagonal elements encode coherence between the corresponding alternatives. For a two-level system,

∣ψ⟩=∣0⟩+eiϕ∣1⟩2,ρ=12(1e−iϕeiϕ1).|\psi\rangle=\frac{|0\rangle+e^{i\phi}|1\rangle}{\sqrt{2}}, \qquad \rho=\frac12 \begin{pmatrix} 1&e^{-i\phi}\\ e^{i\phi}&1 \end{pmatrix}.

Measuring only in the {∣0⟩,∣1⟩}\{|0\rangle,|1\rangle\} basis gives equal probabilities independent of ϕ\phi. Measuring in a superposition basis can reveal the phase. By contrast, the mixture

ρmix=12∣0⟩⟨0∣+12∣1⟩⟨1∣\rho_{\mathrm{mix}}=\frac12|0\rangle\langle0| +\frac12|1\rangle\langle1|

has identical populations but no interference between these two alternatives.

Positivity of the density matrix constrains the off-diagonal elements:

∣ρjk∣2≤ρjjρkk.|\rho_{jk}|^2\leq \rho_{jj}\rho_{kk}.

Consequently, coherence cannot be specified independently of populations. In a balanced pure two-level superposition, the bound is saturated. Averaging equally over all possible values of ϕ\phi removes the off-diagonal elements without changing the populations, converting the state into the mixture above. This provides a simple mathematical model of phase randomization.

The reference basis is physically important. Every density matrix is diagonal in its eigenbasis, so a statement that a state “has coherence” is incomplete unless the relevant alternatives are understood. In spectroscopy, the energy eigenbasis is often natural; in a two-path interferometer, spatial paths provide the basis. A thermal equilibrium state can lack coherence between energy eigenstates while possessing off-diagonal matrix elements in position space.

Coherence is also distinct from purity. A pure energy eigenstate has no coherence in the energy basis, whereas a mixed state can retain nonzero off-diagonal elements in that basis. For a single qubit, the Bloch-vector representation makes this distinction explicit: transverse components represent coherence relative to the computational basis, the longitudinal component specifies population imbalance, and the vector length determines purity.

Interference, optical coherence, and correlation functions

Interference makes relative phases experimentally visible. If two alternatives lead to the same detected outcome with amplitudes a1a_1 and a2a_2, the Born rule gives

P=∣a1+a2∣2=∣a1∣2+∣a2∣2+2Re⁡(a1∗a2).P=|a_1+a_2|^2 =|a_1|^2+|a_2|^2+2\operatorname{Re}(a_1^*a_2).

The last term is the interference contribution. It vanishes when the alternatives are completely distinguishable or when an uncontrolled phase is averaged uniformly. An individual electron or photon can exhibit this behavior; simultaneous passage of many particles is not required.

In an ideal balanced two-path experiment, fringe visibility is

V=Imax⁡−Imin⁡Imax⁡+Imin⁡.V=\frac{I_{\max}-I_{\min}}{I_{\max}+I_{\min}}.

For a normalized path density matrix, perfect recombination and equal detection efficiencies give V=2∣ρ12∣V=2|\rho_{12}|. A pure state with unequal path populations can therefore have less than unit visibility even without environmental decoherence. Experimental visibility must be interpreted together with path balance, detector resolution, and background counts, rather than treated as an unconditional measure of coherence.

In classical optics, coherence is expressed through correlations of the electric field. Quantum optics replaces field amplitudes by operators. A first-order correlation function is

G(1)(x,x′)=⟨E(−)(x)E(+)(x′)⟩,G^{(1)}(x,x')= \left\langle E^{(-)}(x)E^{(+)}(x')\right\rangle,

where xx represents space and time. Its normalized version, g(1)g^{(1)}, determines ordinary interference contrast under suitable conditions. Temporal coherence concerns correlations at different times; spatial coherence concerns different positions. A source can have substantial coherence of one kind and limited coherence of the other.

The spectrum and temporal correlation function are Fourier-transform partners for stationary fields. A narrow spectral line generally corresponds to a long coherence time, although numerical relations depend on line shape and the definition of that time. For a Lorentzian spectrum of full width at half maximum Δν\Delta\nu, the first-order field correlation has an envelope proportional to e−πΔν∣τ∣e^{-\pi\Delta\nu|\tau|}. Its 1/e1/e decay time is therefore 1/(πΔν)1/(\pi\Delta\nu). A coherence length is often estimated by multiplying an appropriate propagation speed by a coherence time.

Higher-order correlations reveal information absent from ordinary fringes. The normalized second-order correlation g(2)(0)g^{(2)}(0) equals 1 for an ideal coherent state, equals 2 for ideal single-mode thermal light, and can fall below 1 for nonclassical light. Thus strong first-order coherence alone does not establish nonclassicality. A coherent state, a technically specific quantum state closely associated with ideal laser light, must also be distinguished from the broader property of quantum coherence.

Unitary evolution, dephasing, and environmental decoherence

For an isolated system, the Schrödinger equation determines coherent evolution. If the Hamiltonian is diagonal in the selected basis, then

ρjk(t)=ρjk(0)exp⁡[−i(Ej−Ek)tℏ],\rho_{jk}(t)=\rho_{jk}(0) \exp\left[-\frac{i(E_j-E_k)t}{\hbar}\right],

where ℏ\hbar is the reduced Planck constant. The off-diagonal elements acquire predictable phases without changing magnitude. A Hamiltonian containing couplings between basis states can convert populations into coherences and coherences into populations. Consequently, coherence need not remain constant relative to an arbitrary fixed basis even under perfectly isolated evolution.

Environmental interaction changes this picture. Suppose two system alternatives become correlated with different environmental states:

(a∣0⟩+b∣1⟩)∣E⟩⟶a∣0⟩∣E0⟩+b∣1⟩∣E1⟩.(a|0\rangle+b|1\rangle)|E\rangle \longrightarrow a|0\rangle|E_0\rangle+b|1\rangle|E_1\rangle.

After the environment is ignored, the system’s off-diagonal element is multiplied by ⟨E1∣E0⟩\langle E_1|E_0\rangle. When the environmental states are orthogonal, the reduced system has no coherence between those alternatives. Coherence has not necessarily been destroyed in the combined system; it has become inaccessible through measurements on the subsystem alone because of quantum entanglement with its surroundings.

Energy relaxation and phase randomization are different processes. For a weakly damped two-level system in a common Markovian model, the longitudinal relaxation time T1T_1, transverse coherence time T2T_2, and pure-dephasing time TϕT_\phi satisfy

1T2=12T1+1Tϕ.\frac{1}{T_2}=\frac{1}{2T_1}+\frac{1}{T_\phi}.

This implies T2≤2T1T_2\leq 2T_1 within that model. Relaxation contributes to coherence decay because transitions alter the state, whereas pure dephasing can suppress off-diagonal elements without changing energy populations. The relation is not a universal description of arbitrary multilevel systems or non-Markovian dynamics.

Coherence decay also need not be exponential. Slowly varying frequency noise often produces approximately Gaussian Ramsey decay, while discrete environmental modes can cause oscillations or revivals. An ensemble can lose its net transverse signal because individual members precess at different but nearly constant frequencies, even when each member remains coherent. This inhomogeneous broadening is commonly characterized by T2∗T_2^*, which can be much shorter than an echo-based coherence time.

Temperature alone therefore does not determine coherence. Relevant factors include coupling strengths, environmental spectral densities, transition frequencies, geometry, and the observables being measured. Phonons, fluctuating electromagnetic fields, collisions, photon emission, and defects can all carry away distinguishing information. Conversely, suitably isolated degrees of freedom can retain useful coherence even when the surrounding apparatus is not at extremely low temperature.

Preparation, measurement, and coherent control

Coherence is usually prepared by a controlled interaction that couples initially distinguishable states. A resonant electromagnetic pulse can rotate an atom or spin from an energy eigenstate into a superposition. For an ideal two-level system driven on resonance, the population oscillates at the Rabi frequency, which depends on the transition matrix element and drive amplitude. A pulse producing a quarter-turn on the Bloch sphere—a π/2\pi/2 pulse—prepares equal populations with a relative phase set by the driving field.

Ramsey interferometry, introduced by Norman Ramsey in 1949, separates two such pulses by a period of free evolution. The first pulse prepares coherence, the system accumulates phase relative to the oscillator, and the second pulse converts that phase into a population difference. If the angular-frequency detuning is δ\delta, a simple ideal signal has the form

P1(T)=12[1+C(T)cos⁡(δT+ϕ0)],P_1(T)=\frac12[1+C(T)\cos(\delta T+\phi_0)],

where C(T)C(T) is the contrast envelope. Ramsey measurements reveal both frequency differences and the loss of experimentally accessible coherence.

Erwin Hahn’s 1950 spin echo provides a crucial distinction between reversible dephasing and irreversible loss. A π\pi pulse reverses the relative phase accumulation caused by static frequency offsets, allowing spins that have spread in phase to rephase later. Echoes cannot generally reverse arbitrary environmental dynamics, but they can remove large contributions from slowly varying noise. Dynamical decoupling extends this principle through pulse sequences that selectively suppress environmental coupling. Their effectiveness depends on pulse errors, noise frequencies, and sequence timing.

Coherence can be reconstructed more completely through quantum state tomography. Repeated preparations are measured in several incompatible bases to estimate the density matrix. For a qubit, measurements of three Pauli components determine the Bloch vector. Tomography requires a model of the measurements and sufficient statistical data; preparation and measurement errors can otherwise be mistaken for properties of the state. Larger systems require rapidly increasing experimental resources for unrestricted reconstruction.

Every measurement of phase also relies on a reference. In optical homodyne detection, a local oscillator supplies that reference; in spin experiments, it is typically a radio-frequency or microwave source. Drift of the reference can make a stable system appear to decohere. Conversely, a shared phase reference can expose coherence that is inaccessible when two apparatuses have unrelated phases. Reported coherence times must therefore specify the transition, pulse sequence, reference stability, and decay criterion, rather than being treated as immutable properties of a material or particle.

Collective coherence in condensed matter

In many-body physics, coherence often describes phase relationships extending across large numbers of particles. The relevant quantity may be a one-particle correlation function,

g1(r,r′)=⟨Ψ^†(r)Ψ^(r′)⟩,g_1(\mathbf r,\mathbf r') =\langle\hat\Psi^\dagger(\mathbf r)\hat\Psi(\mathbf r')\rangle,

rather than an off-diagonal element between two internal levels. Long-range behavior of such correlations distinguishes different forms of collective order. This usage connects microscopic quantum amplitudes with macroscopic phenomena in condensed-matter physics, but it does not mean that every particle independently occupies an identical pure state.

A Bose–Einstein condensate has a macroscopically occupied mode in its one-particle density matrix. Dilute atomic-gas condensates were first realized in 1995, and interference experiments subsequently made their spatial phase structure directly accessible. When separately prepared condensates expand and overlap, individual experimental images can display fringes even when the relative phase varies randomly between repetitions. Such observations illustrate why single experimental realizations, ensemble averages, and assumptions about a common phase reference must be distinguished.

Superfluidity involves collective response, including low-dissipation flow and quantized circulation, rather than simply the existence of a large condensate fraction. Interactions and dimensionality can separate condensation from superfluid behavior. In two dimensions, long-wavelength fluctuations prevent ordinary finite-temperature long-range order in idealized homogeneous systems with suitable short-range interactions, yet a low-temperature phase can retain algebraically decaying correlations and superfluid response. The Berezinskii–Kosterlitz–Thouless transition describes the role of vortex binding and unbinding in this setting.

In superconductivity, phase coherence concerns the collective paired state. In conventional superconductors, BCS theory describes pairing of electrons into Cooper pairs. The phase difference across a weak link produces the Josephson effect:

I=Icsin⁡Δϕ,dΔϕdt=2eVℏ.I=I_c\sin\Delta\phi, \qquad \frac{d\Delta\phi}{dt}=\frac{2eV}{\hbar}.

These relations connect a collective quantum phase to measurable current and voltage. They underlie sensitive magnetic-flux measurements and superconducting circuit elements.

Several distinct lengths appear in this subject. A superconducting coherence length characterizes spatial variation of the pair amplitude or, in appropriate microscopic descriptions, the scale associated with pairing correlations. It is not interchangeable with an optical coherence length or a qubit’s dephasing length. Likewise, coherence of electron transport through a mesoscopic conductor means that alternative scattering paths retain phase relationships long enough to interfere. Such transport coherence can occur without superconductivity, showing that collective order and single-particle phase memory are related but different phenomena.

Coherence in quantum computation and communication

Quantum computing uses coherent transformations to manipulate amplitudes across a structured set of computational states. For nn qubits, a pure state can involve amplitudes over 2n2^n basis states, but this mathematical size does not by itself provide a computational advantage. Algorithms must arrange interference so that measurement gives useful answers with sufficiently high probability. Merely preparing a uniform superposition and measuring it produces random bit strings, not simultaneous access to all amplitudes.

A single-qubit example shows the operational role of phase. A Hadamard operation transforms ∣0⟩|0\rangle into (∣0⟩+∣1⟩)/2(|0\rangle+|1\rangle)/\sqrt2. A phase-changing operation can convert this to (∣0⟩−∣1⟩)/2(|0\rangle-|1\rangle)/\sqrt2. A second Hadamard then produces ∣1⟩|1\rangle, rather than the ∣0⟩|0\rangle obtained without the phase change. An intervening process that completely dephases the computational basis removes this deterministic distinction. Coherence is useful because later gates convert phase relationships into measurable populations.

Multiqubit devices must preserve more than individual-qubit coherence. Correlated noise, unwanted interactions, leakage outside the qubit subspace, and control errors can damage an algorithm even when separately measured T1T_1 and T2T_2 values appear favorable. Entanglement also cannot be inferred merely from local coherence. The state (∣00⟩+∣11⟩)/2(|00\rangle+|11\rangle)/\sqrt2 has maximally mixed single-qubit reduced states, with no local off-diagonal elements in the computational basis, while retaining global coherence between ∣00⟩|00\rangle and ∣11⟩|11\rangle.

Quantum error correction protects logical information by encoding it into a larger physical system and measuring error syndromes without directly measuring the encoded amplitudes. Phase-flip errors are as important as bit-flip errors. Successful correction does not require each physical component to remain perfectly coherent indefinitely; it requires an appropriate noise structure, sufficiently accurate operations, and repeated extraction of error information. The relevant lifetime is then that of the encoded logical state under the complete correction procedure.

In quantum communication, coherence must survive transmission, storage, or conversion between physical carriers. Time-bin and polarization encodings use different alternatives and therefore respond differently to environmental disturbance. Optical phase drift, birefringence, frequency mismatch, and imperfect mode overlap can impair interference between remote photons. Quantum memories must preserve the encoded relationship, not merely absorb and later emit a signal. Communication protocols consequently distinguish population-storage efficiency, phase preservation, and entanglement fidelity, because strong performance in one does not guarantee the others.

Coherence as a quantified physical resource

A systematic resource theory of quantum coherence asks which states and operations are useful when creating superpositions in a specified basis is restricted. A prominent formulation was introduced by T. Baumgratz, M. Cramer, and M. B. Plenio in 2014. Its starting point is the set of incoherent states,

δ=∑jpj∣j⟩⟨j∣.\delta=\sum_j p_j|j\rangle\langle j|.

The chosen basis defines which states are free of the resource. Different restrictions on allowed operations lead to different resource theories, so statements about conversion rates or monotonicity must identify the operational setting.

One simple quantity is the sum of off-diagonal magnitudes,

Cl1(ρ)=∑j≠k∣ρjk∣.C_{l_1}(\rho)=\sum_{j\ne k}|\rho_{jk}|.

For a uniform pure superposition of dd basis states, this measure equals d−1d-1. Another widely used quantity is the relative entropy of coherence,

Cr(ρ)=S(Δ(ρ))−S(ρ),C_{\mathrm r}(\rho)=S(\Delta(\rho))-S(\rho),

where Δ\Delta removes all off-diagonal entries in the reference basis and SS is the von Neumann entropy. With base-2 logarithms, a uniform pure superposition has Cr=log⁡2dC_{\mathrm r}=\log_2 d.

These measures answer different quantitative questions. The relative entropy of coherence has an operational interpretation in asymptotic coherence distillation under an established class of incoherent operations. Coherence distillation seeks to extract nearly ideal coherent units from many imperfect copies; coherence formation asks how much resource is needed to prepare a state. These two rates need not coincide for mixed states, producing an irreversibility analogous to that found in other resource theories. A scalar coherence measure also cannot replace a full description of noise or control requirements in a particular device.

A related but distinct framework treats coherence as asymmetry with respect to time translations. The relevant transformations are generated by the system Hamiltonian, and energy coherence enables a state to serve as a clock or phase reference. In systems with degenerate energy levels, superpositions within a degenerate subspace may count as basis coherence without breaking time-translation symmetry. This distinction matters when discussing physical clocks, thermodynamic constraints, and the cost of preparing reference systems.

Resource theories also clarify the relationship to entanglement. Coherence is a property defined relative to a basis, whereas entanglement concerns whether a composite state can be expressed as a mixture of product states. Suitable joint operations can convert single-system coherence into entanglement with an ancillary system. Nevertheless, the two resources are not interchangeable: a product state may have substantial local coherence, while an entangled state may have locally incoherent reduced states.

Precision measurement, clocks, and sensing

A coherent superposition can act as a sensitive probe because its relative phase responds to an external parameter. For two energy levels separated by ℏω\hbar\omega, free evolution accumulates phase ϕ=ωT\phi=\omega T. A perturbation that changes the splitting changes the phase, which can subsequently be read out through interference. This principle underlies metrology based on atomic transitions, spin precession, matter waves, and superconducting circuits. Coherence time limits the useful interrogation period, but it is only one factor in overall precision.

Atomic clocks compare a controllable oscillator with a reproducible atomic transition. The current definition of the second fixes the unperturbed ground-state hyperfine transition frequency of caesium-133 at 9,192,631,770 hertz. Modern optical clocks use much higher-frequency transitions, increasing the phase accumulated during a given interrogation time. Their performance depends on coherent laser interrogation, oscillator stability, environmental shifts, and statistical averaging. A narrow observed transition does not alone establish an accurate clock unless systematic frequency shifts are controlled.

In a simple Ramsey measurement with NN independent, identically prepared probes and ideal readout, phase uncertainty near the most sensitive operating point scales as 1/N1/\sqrt N per interrogation. For frequency estimation, the corresponding scale is proportional to

δω∼1TN,\delta\omega\sim\frac{1}{T\sqrt N},

before accounting for repeated trials and experimental overhead. Longer interrogation increases sensitivity but also allows dephasing to reduce contrast. The optimal duration depends on the noise model, duty cycle, and detection efficiency rather than being universally equal to a quoted coherence time.

Entangled probes can improve scaling under appropriate conditions. A Greenberger–Horne–Zeilinger state,

∣0⟩⊗N+∣1⟩⊗N2,\frac{|0\rangle^{\otimes N}+|1\rangle^{\otimes N}}{\sqrt2},

accumulates phase NωTN\omega T, but its large-scale coherence can also be especially vulnerable to independent dephasing. The quantum Fisher information formalizes the sensitivity available in a parameter-dependent state. It links coherence and distinguishability while making clear that a quantum enhancement depends on the generator, noise, and permitted measurement strategy.

Concrete sensors exploit different physical couplings. Atomic spins respond to a magnetic field through Zeeman shifts; atom interferometers respond to acceleration and gravity through path-dependent phases; defects in diamond provide optically addressable spins sensitive to local fields. Nuclear magnetic resonance and magnetic resonance imaging use coherent spin evolution and controlled dephasing to obtain spectral or spatial information. In these applications, environmental coupling is simultaneously the source of the desired signal and a mechanism that limits phase memory.

Molecular dynamics, spectroscopy, and biological systems

Molecules support electronic, vibrational, and rotational superpositions, each with different energy scales and environmental couplings. A short optical pulse can prepare a wave packet composed of several stationary states. Their relative phases evolve at frequencies determined by energy differences, producing oscillatory observables and changing spatial distributions. Coherent nuclear motion can therefore be detected even when the molecule has no persistent electronic coherence. Distinguishing the participating degrees of freedom is essential when interpreting an oscillating experimental signal.

Ultrafast two-dimensional electronic spectroscopy uses sequences of pulses to correlate excitation and detection frequencies while varying an intervening delay. Oscillations during this waiting period may reveal coherences between states. However, electronic superpositions, vibrational wave packets, mixed electronic-vibrational states, and pulse-related artifacts can produce overlapping signatures. Assignments require information about frequencies, dephasing, polarization dependence, temperature dependence, and suitable theoretical models. An oscillation is evidence of coherent dynamics only to the extent supported by the complete experiment.

A widely discussed example concerns excitation transfer in light-harvesting proteins. A 2007 study of the Fenna–Matthews–Olson complex at cryogenic temperature reported long-lived oscillatory signals interpreted as evidence of electronic coherence. Subsequent experiments and theory established the importance of vibrational and vibronic contributions and complicated the interpretation of such beats. The existence of quantum-mechanical excitation dynamics is not in question; the separate issue is whether a particular long-lived coherence provides a functional advantage under physiological conditions.

In photosynthesis, absorbed light produces electronic excitations that move among coupled pigments before reaching reaction centers. Excitations may be delocalized over several pigments, while interactions with molecular motion cause energy relaxation and dephasing. Models can show efficient transfer at intermediate environmental coupling: excessive noise suppresses coherent motion, whereas insufficient relaxation can hinder escape from unfavorable configurations. Such model results do not by themselves demonstrate that natural selection optimized a biological complex for long-lived electronic coherence.

Spin-dependent chemistry provides another setting. In a radical-pair mechanism, two unpaired electron spins evolve among singlet and triplet configurations under hyperfine and magnetic interactions, while spin-selective reactions produce different chemical yields. This framework is investigated in relation to biological magnetic sensing, including reactions involving cryptochrome proteins. The relevant coherence is spin coherence within a reacting molecular system, not coherence of an entire organism. More generally, biological discussions must distinguish a well-established microscopic quantum mechanism from broader claims about its physiological role.

Historical development and conceptual boundaries

The roots of quantum coherence lie in the transition from wave optics to quantum theory. Nineteenth-century interference experiments established the importance of phase relationships for light. Quantum mechanics extended amplitude interference to matter: the wave function introduced a complex description of quantum states, while diffraction experiments in the 1920s confirmed wave-like behavior of electrons. The central conceptual change was that interference could govern probabilities for individual detection events, rather than only the intensity of a continuous classical wave.

The density-operator formalism developed in the 1920s made it possible to distinguish superpositions from statistical mixtures systematically. Later, magnetic-resonance techniques provided practical control of coherent spin evolution. Hahn’s spin echo and Ramsey’s separated-field method demonstrated that phase could be prepared, allowed to evolve, and recovered through carefully timed operations. These methods became foundations for spectroscopy and precision measurement well before coherence became a central term in quantum information science.

Optical developments broadened the subject. The laser, first demonstrated in 1960, supplied sources with unusually strong phase correlations. Roy Glauber’s work in 1963 developed quantum correlation-function methods and a theory of coherent states, clarifying which optical properties could be described classically and which required quantized fields. His contribution to the quantum theory of optical coherence was recognized by the 2005 Nobel Prize in Physics. The Hanbury Brown–Twiss intensity-correlation experiments of the 1950s were another major stimulus to this development.

Environmental decoherence emerged as a distinct explanatory framework through work including H. Dieter Zeh’s 1970 analysis and Wojciech Zurek’s studies in the 1980s. It explains how environmental monitoring suppresses interference between certain alternatives and selects comparatively stable states. It does not, by itself, settle every aspect of the measurement problem, especially the interpretation of a single definite outcome. Different interpretations of quantum mechanics can agree on calculated coherence decay while differing on what the total quantum state represents.

Several boundaries remain important in contemporary usage. Coherence does not automatically imply entanglement, computational advantage, macroscopic nonclassicality, or freedom from thermal noise. Conversely, absence of a local interference signal does not prove that a larger system has lost all quantum correlations. Questions about macroscopic coherence therefore require specified alternatives and measurements—for example, distinct current states in a superconducting circuit or separated paths of a massive particle. Experimental progress is assessed through controlled preparation, verified phase-sensitive observables, and quantified environmental effects, rather than through a universal size or temperature threshold separating quantum and classical behavior.